2_5_5_Topic_Modeling_for_Research_Articles.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Project Overview: Topic Modeling for Research Articles**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Objective**</span>\n",
        "Discover the abstract topics that occur in a collection of research articles using Latent Dirichlet Allocation (LDA).\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Key Libraries Used**</span>\n",
        "- **Pandas**: For data manipulation.\n",
        "- **NumPy**: For numerical operations.\n",
        "- **NLTK**: For natural language processing tasks.\n",
        "- **Gensim**: For implementing LDA modeling.\n",
        "- **pyLDAvis**: For visualizing the results of the LDA model.\n",
        "- **Flask**: For creating a web application.\n",
        "- **Matplotlib & Seaborn**: For data visualization.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Project Steps**</span>\n",
        "\n",
        "### <span style=\"color:green;\">**1. Environment Setup**</span>\n",
        "- Install necessary libraries using pip.\n",
        "\n",
        "### <span style=\"color:green;\">**2. Data Loading and Inspection**</span>\n",
        "- Load the dataset into a DataFrame.\n",
        "- Inspect the first few rows, dataset information, summary statistics, and check for missing values.\n",
        "\n",
        "### <span style=\"color:green;\">**3. Data Visualization**</span>\n",
        "- Visualize missing values using a heatmap.\n",
        "- Create plots to analyze the distribution of papers by year and the length of abstracts.\n",
        "\n",
        "### <span style=\"color:green;\">**4. Text Preprocessing**</span>\n",
        "- Tokenize abstracts and remove stop words and non-alphabetic tokens.\n",
        "\n",
        "### <span style=\"color:green;\">**5. LDA Model Training**</span>\n",
        "- Prepare data for LDA and train the model.\n",
        "- Display and visualize the identified topics.\n",
        "\n",
        "### <span style=\"color:green;\">**6. Model Evaluation**</span>\n",
        "- Evaluate the model using perplexity and coherence scores.\n",
        "\n",
        "### <span style=\"color:green;\">**7. Model Persistence**</span>\n",
        "- Save the trained LDA model and dictionary for future use.\n",
        "\n",
        "### <span style=\"color:green;\">**8. Inference Function**</span>\n",
        "- Create a function to infer topics from new abstracts using the trained model.\n",
        "\n",
        "### <span style=\"color:green;\">**9. Web Application**</span>\n",
        "- Develop a Flask app to expose the topic inference functionality via an API endpoint.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Conclusion**</span>\n",
        "This project demonstrates the application of LDA for topic modeling in research articles, enabling the discovery of underlying themes in a collection of documents."
      ],
      "metadata": {
        "id": "M01BeBONiI60"
      }
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "NdieNHVfiMVM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "8mLHr2WEiMYV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "cXImxsfgiMq7"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import warnings  # Import the warnings library to manage warning messages"
      ],
      "metadata": {
        "id": "Ly7U_y1OjMxf"
      },
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "warnings.filterwarnings('ignore')  # Suppress all warning messages"
      ],
      "metadata": {
        "id": "eVkxtWuMjM1E"
      },
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "KlfIw_ymjYyp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Install Required Libraries"
      ],
      "metadata": {
        "id": "rC5VNjJHd-go"
      },
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!nohup python script.py &  # Run the Python script in the background and ignore hangups"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wObOFpYCd-jf",
        "outputId": "e6045870-4249-4b1d-9bd3-2369c3f58cb4"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "nohup: appending output to 'nohup.out'\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "aSKVq3U-d-mQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install pandas numpy nltk gensim pyLDAvis Flask  # Install necessary libraries for data manipulation, NLP, topic modeling, and web application\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "90nbZOi3d-v6",
        "outputId": "21c9d2e6-258c-49a7-eb48-80cc617ee2fc"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: pandas in /usr/local/lib/python3.10/dist-packages (2.2.2)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (1.26.4)\n",
            "Requirement already satisfied: nltk in /usr/local/lib/python3.10/dist-packages (3.8.1)\n",
            "Requirement already satisfied: gensim in /usr/local/lib/python3.10/dist-packages (4.3.3)\n",
            "Requirement already satisfied: pyLDAvis in /usr/local/lib/python3.10/dist-packages (3.4.1)\n",
            "Requirement already satisfied: Flask in /usr/local/lib/python3.10/dist-packages (2.2.5)\n",
            "Requirement already satisfied: python-dateutil>=2.8.2 in /usr/local/lib/python3.10/dist-packages (from pandas) (2.8.2)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas) (2024.2)\n",
            "Requirement already satisfied: tzdata>=2022.7 in /usr/local/lib/python3.10/dist-packages (from pandas) (2024.2)\n",
            "Requirement already satisfied: click in /usr/local/lib/python3.10/dist-packages (from nltk) (8.1.7)\n",
            "Requirement already satisfied: joblib in /usr/local/lib/python3.10/dist-packages (from nltk) (1.4.2)\n",
            "Requirement already satisfied: regex>=2021.8.3 in /usr/local/lib/python3.10/dist-packages (from nltk) (2024.9.11)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (from nltk) (4.66.5)\n",
            "Requirement already satisfied: scipy<1.14.0,>=1.7.0 in /usr/local/lib/python3.10/dist-packages (from gensim) (1.13.1)\n",
            "Requirement already satisfied: smart-open>=1.8.1 in /usr/local/lib/python3.10/dist-packages (from gensim) (7.0.4)\n",
            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from pyLDAvis) (3.1.4)\n",
            "Requirement already satisfied: numexpr in /usr/local/lib/python3.10/dist-packages (from pyLDAvis) (2.10.1)\n",
            "Requirement already satisfied: funcy in /usr/local/lib/python3.10/dist-packages (from pyLDAvis) (2.0)\n",
            "Requirement already satisfied: scikit-learn>=1.0.0 in /usr/local/lib/python3.10/dist-packages (from pyLDAvis) (1.5.2)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.10/dist-packages (from pyLDAvis) (71.0.4)\n",
            "Requirement already satisfied: Werkzeug>=2.2.2 in /usr/local/lib/python3.10/dist-packages (from Flask) (3.0.4)\n",
            "Requirement already satisfied: itsdangerous>=2.0 in /usr/local/lib/python3.10/dist-packages (from Flask) (2.2.0)\n",
            "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->pyLDAvis) (2.1.5)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.8.2->pandas) (1.16.0)\n",
            "Requirement already satisfied: threadpoolctl>=3.1.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn>=1.0.0->pyLDAvis) (3.5.0)\n",
            "Requirement already satisfied: wrapt in /usr/local/lib/python3.10/dist-packages (from smart-open>=1.8.1->gensim) (1.16.0)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "jYC02_jbd-0p"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "TwJIlMlod-5n"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "U-7OR28zd-7_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 1: Import Libraries**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "In this step, we import the necessary libraries required for data manipulation, natural language processing (NLP), topic modeling, visualization, and web application development.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Key Libraries**</span>\n",
        "- **Pandas**: Essential for data manipulation and analysis, allowing us to work with data in DataFrame format.\n",
        "- **NumPy**: Provides support for numerical operations and array handling.\n",
        "- **NLTK**: A powerful toolkit for NLP tasks, enabling us to preprocess text data.\n",
        "- **Gensim**: A robust library for topic modeling and handling large text corpora efficiently.\n",
        "- **pyLDAvis**: Used for visualizing the results of the LDA model, helping to interpret the identified topics.\n",
        "- **Matplotlib & Seaborn**: Libraries for data visualization, enhancing our ability to present findings graphically.\n",
        "- **Flask**: A web framework used to create a web application for topic inference.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Suppressing Warnings**</span>\n",
        "- The warnings library is imported to suppress warning messages that may arise during the execution of the code, ensuring a cleaner output."
      ],
      "metadata": {
        "id": "VCHGiM-aiQpA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Import Libraries"
      ],
      "metadata": {
        "id": "PDEq3ALod--h"
      },
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd  # For data manipulation"
      ],
      "metadata": {
        "id": "IWQOBRXHd_BM"
      },
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np  # For numerical operations"
      ],
      "metadata": {
        "id": "tppGrFSXd_Do"
      },
      "execution_count": 29,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import nltk  # For natural language processing"
      ],
      "metadata": {
        "id": "K0023WYhd_IN"
      },
      "execution_count": 30,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from nltk.corpus import stopwords  # For English stop words"
      ],
      "metadata": {
        "id": "fyR_FeZid_Kh"
      },
      "execution_count": 31,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from nltk.tokenize import word_tokenize  # For tokenizing text"
      ],
      "metadata": {
        "id": "pJxQc-f2d_MF"
      },
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from gensim import corpora  # For creating a dictionary and corpus for LDA"
      ],
      "metadata": {
        "id": "0CwtwD_ed_Ov"
      },
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from gensim.models import LdaModel  # For LDA modeling"
      ],
      "metadata": {
        "id": "IMY1qLUOd_Qt"
      },
      "execution_count": 34,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import pyLDAvis  # For visualizing LDA results"
      ],
      "metadata": {
        "id": "0kMc4j-qd_TM"
      },
      "execution_count": 35,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import pyLDAvis.gensim_models  # For Gensim integration with pyLDAvis"
      ],
      "metadata": {
        "id": "g_C5mZoPd_Xi"
      },
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt  # For plotting graphs"
      ],
      "metadata": {
        "id": "HUItYD9Hd_bw"
      },
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import seaborn as sns  # For enhanced visualizations"
      ],
      "metadata": {
        "id": "f7YqjnJEd_ga"
      },
      "execution_count": 38,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from flask import Flask, request, jsonify  # For creating the web application"
      ],
      "metadata": {
        "id": "I5H-TvFqd_kY"
      },
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "zfE4POkSfv_q"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "7xGEXwwaf3ta"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Download stopwords"
      ],
      "metadata": {
        "id": "2zZkSTI6f3wU"
      },
      "execution_count": 40,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "nltk.download('punkt')  # Download tokenizer"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Nh0ayadVf3zl",
        "outputId": "0cf84613-7013-4c0f-e03a-e8d3484028cd"
      },
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Package punkt is already up-to-date!\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {},
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "ddeidMdqf31u"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "nltk.download('stopwords')  # Download stopwords"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Xecm6o45f34H",
        "outputId": "a976fce0-7010-4c48-83b5-12b487d48ee6"
      },
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Package stopwords is already up-to-date!\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {},
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "d5hcDXafgAEx"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "viN9m5ltgAOQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "TFB3Y95YgAQ5"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 2: Load Data**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "In this step, we load the dataset into a Pandas DataFrame for further inspection and analysis.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Dataset**</span>\n",
        "- The dataset `papers.csv` contains research articles along with their abstracts and metadata.\n",
        "- Using `pd.read_csv()`, we load the dataset, which allows us to perform various data manipulation tasks.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Inspecting the Dataset**</span>\n",
        "- We inspect the first few rows of the dataset using `head()`, which provides a glimpse of the data structure and content.\n",
        "- The `info()` method gives us information about data types and non-null counts, while `describe()` provides summary statistics.\n",
        "- We check for missing values using `isnull().sum()` to understand data quality."
      ],
      "metadata": {
        "id": "SdlaZtWNiVxV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Load Data"
      ],
      "metadata": {
        "id": "akaMINgbgAWA"
      },
      "execution_count": 43,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "papers_df = pd.read_csv('papers.csv')  # Load the dataset into a DataFrame"
      ],
      "metadata": {
        "id": "oDTr6wDxgAY7"
      },
      "execution_count": 44,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "rney3nglgGqQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "4PE8UB3ggGsa"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Inspect the Dataset"
      ],
      "metadata": {
        "id": "8xmiiXgdgGz5"
      },
      "execution_count": 45,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(papers_df.head())  # Display the first 5 rows of the dataset"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tIFDQT_RgG6J",
        "outputId": "d477128c-9ba6-47ce-cc2c-655817f27afc"
      },
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "     id  year                                              title event_type  \\\n",
            "0     1  1987  Self-Organization of Associative Database and ...        NaN   \n",
            "1    10  1987  A Mean Field Theory of Layer IV of Visual Cort...        NaN   \n",
            "2   100  1988  Storing Covariance by the Associative Long-Ter...        NaN   \n",
            "3  1000  1994  Bayesian Query Construction for Neural Network...        NaN   \n",
            "4  1001  1994  Neural Network Ensembles, Cross Validation, an...        NaN   \n",
            "\n",
            "                                            pdf_name          abstract  \\\n",
            "0  1-self-organization-of-associative-database-an...  Abstract Missing   \n",
            "1  10-a-mean-field-theory-of-layer-iv-of-visual-c...  Abstract Missing   \n",
            "2  100-storing-covariance-by-the-associative-long...  Abstract Missing   \n",
            "3  1000-bayesian-query-construction-for-neural-ne...  Abstract Missing   \n",
            "4  1001-neural-network-ensembles-cross-validation...  Abstract Missing   \n",
            "\n",
            "                                          paper_text  \n",
            "0  767\\n\\nSELF-ORGANIZATION OF ASSOCIATIVE DATABA...  \n",
            "1  683\\n\\nA MEAN FIELD THEORY OF LAYER IV OF VISU...  \n",
            "2  394\\n\\nSTORING COVARIANCE BY THE ASSOCIATIVE\\n...  \n",
            "3  Bayesian Query Construction for Neural\\nNetwor...  \n",
            "4  Neural Network Ensembles, Cross\\nValidation, a...  \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "Yw_2vaZ_gQAk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "Bs1IoJIdgQDk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(papers_df.info())  # Display information about the dataset (types and non-null counts)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wsTpYRvlgG9a",
        "outputId": "921b7eb6-e29c-4792-80b9-cd223c4066ee"
      },
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 7241 entries, 0 to 7240\n",
            "Data columns (total 7 columns):\n",
            " #   Column      Non-Null Count  Dtype \n",
            "---  ------      --------------  ----- \n",
            " 0   id          7241 non-null   int64 \n",
            " 1   year        7241 non-null   int64 \n",
            " 2   title       7241 non-null   object\n",
            " 3   event_type  2422 non-null   object\n",
            " 4   pdf_name    7241 non-null   object\n",
            " 5   abstract    7241 non-null   object\n",
            " 6   paper_text  7241 non-null   object\n",
            "dtypes: int64(2), object(5)\n",
            "memory usage: 396.1+ KB\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "0Z89kpCggHAV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "2fV8Vlz_gHDB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(papers_df.describe())  # Display summary statistics for numerical columns"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mo9QSbrggHGT",
        "outputId": "5572d280-3b8e-49f6-9247-08c8c9508358"
      },
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                id         year\n",
            "count  7241.000000  7241.000000\n",
            "mean   3655.912167  2006.439718\n",
            "std    2098.435219     8.759919\n",
            "min       1.000000  1987.000000\n",
            "25%    1849.000000  2000.000000\n",
            "50%    3659.000000  2009.000000\n",
            "75%    5473.000000  2014.000000\n",
            "max    7284.000000  2017.000000\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "lyseL79XgRFq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "gzrjALxAgRIi"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(papers_df.isnull().sum())  # Check for missing values in each column"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JFra6-jugRLF",
        "outputId": "f6314f7f-8255-4a41-ff08-6afbaca75424"
      },
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "id               0\n",
            "year             0\n",
            "title            0\n",
            "event_type    4819\n",
            "pdf_name         0\n",
            "abstract         0\n",
            "paper_text       0\n",
            "dtype: int64\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "2M2WDbgCgRNr"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "IGBUtjhOgRQc"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "CclElKjNgHJb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 3: Visualize Missing Values**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "This step involves visualizing the presence of missing values in the dataset to identify potential data quality issues.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Heatmap Visualization**</span>\n",
        "- A heatmap is created using Seaborn’s `heatmap()` function to visually represent missing values in the dataset.\n",
        "- The `cbar` parameter is set to `False` to disable the color bar, and `cmap` is set to 'viridis' for better visual distinction.\n",
        "- The title 'Missing Values in Dataset' is added to provide context to the visualization."
      ],
      "metadata": {
        "id": "NKyYsDAiiZO-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Visualize Missing Values"
      ],
      "metadata": {
        "id": "pfu7gBzhgAb-"
      },
      "execution_count": 50,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(10, 6))  # Set the figure size\n",
        "sns.heatmap(papers_df.isnull(), cbar=False, cmap='viridis')  # Visualize missing values\n",
        "plt.title('Missing Values in Dataset')  # Title for the heatmap\n",
        "plt.show()  # Show the heatmap"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 545
        },
        "id": "9V1YFcGAgVn0",
        "outputId": "8c1f3adf-187e-4fdf-a83a-b579ed2c6ff4"
      },
      "execution_count": 51,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "RvUY1aUogVq7"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "hNvv8eiDgVtp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "o4ZIMbSUibtt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 4: Sample Data**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "To expedite the processing and analysis, we sample a smaller subset of the dataset.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Sampling Method**</span>\n",
        "- We use the `sample()` method with `frac=0.1` to select 10% of the dataset randomly.\n",
        "- The `random_state` parameter ensures reproducibility of the results."
      ],
      "metadata": {
        "id": "8_IBTD07icHq"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Sample a smaller subset of the data for testing"
      ],
      "metadata": {
        "id": "khXEjbtFgVwM"
      },
      "execution_count": 52,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "papers_df = papers_df.sample(frac=0.1, random_state=42)  # Use 10% of the dataset for faster processing"
      ],
      "metadata": {
        "id": "MOcHxGQJgVyZ"
      },
      "execution_count": 53,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "Gi2lt7Q7gdIz"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "d91tL1LyigBn"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "GSUdqMt2igT2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 5: Distribution of Year**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "This step visualizes the distribution of research papers across different years to identify trends over time.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Count Plot**</span>\n",
        "- A count plot is created using Seaborn's `countplot()` function to display the number of papers published each year.\n",
        "- The x-axis is labeled 'year', and the title 'Distribution of Papers by Year' provides context for the visualization.\n",
        "- The x-axis labels are rotated for improved readability."
      ],
      "metadata": {
        "id": "mJfdn07yigll"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Distribution of Year"
      ],
      "metadata": {
        "id": "5eSi73V_gdOR"
      },
      "execution_count": 54,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(10, 6))  # Set the figure size\n",
        "sns.countplot(data=papers_df, x='year', palette='viridis')  # Plot the distribution of articles by year\n",
        "plt.title('Distribution of Papers by Year')  # Title for the plot\n",
        "plt.xticks(rotation=45)  # Rotate x-axis labels for better readability\n",
        "plt.show()  # Show the distribution plot"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 585
        },
        "id": "PcPIBAKegdRF",
        "outputId": "1a1f81a0-6af5-45ef-c47a-b870273bd225"
      },
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "yGsbmTs6gdVa"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "W-T1Ue71gjwb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "cZqdzkcPijEV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 6: Word Count Analysis in Abstracts**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "In this step, we analyze the word count in the abstracts of the research papers to understand their length distribution.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Abstract Length Calculation**</span>\n",
        "- A new column `abstract_length` is created to store the word count of each abstract using a lambda function.\n",
        "- We then plot a histogram using Seaborn's `histplot()` to visualize the distribution of abstract lengths.\n",
        "- The histogram includes a kernel density estimate (KDE) for smoother representation, with titles and axis labels for clarity."
      ],
      "metadata": {
        "id": "k7LLtkhOijRP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Word Count Analysis in Abstracts"
      ],
      "metadata": {
        "id": "mBv2dN6Lgjz4"
      },
      "execution_count": 56,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "papers_df['abstract_length'] = papers_df['abstract'].fillna('').apply(lambda x: len(x.split()))  # Calculate word count in abstracts\n",
        "plt.figure(figsize=(10, 6))  # Set the figure size\n",
        "sns.histplot(papers_df['abstract_length'], bins=30, kde=True)  # Plot the distribution of abstract lengths\n",
        "plt.title('Distribution of Abstract Lengths')  # Title for the plot\n",
        "plt.xlabel('Number of Words in Abstract')  # X-axis label\n",
        "plt.ylabel('Frequency')  # Y-axis label\n",
        "plt.show()  # Show the histogram"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        },
        "id": "fxtNN4Vpgj50",
        "outputId": "9b863c73-4db6-462a-bbd1-c7016fc1e765"
      },
      "execution_count": 57,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "YlXjvVlWgj8d"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "XyyQW__JgkBr"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "-z0JnA51gdZG"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 7: Preprocessing Text**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "This step involves cleaning and preparing the abstract text data for topic modeling.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Text Preprocessing Function**</span>\n",
        "- We define a function `preprocess_text()` to handle the preprocessing tasks.\n",
        "- The function converts the text to lowercase, tokenizes it, and removes stopwords and non-alphabetic tokens.\n",
        "- The processed tokens are then returned for further analysis, ensuring that the data is clean and relevant for LDA modeling."
      ],
      "metadata": {
        "id": "-RjUKc4Xil1V"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Preprocessing"
      ],
      "metadata": {
        "id": "VVflVVnJgV8y"
      },
      "execution_count": 58,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def preprocess_text(text):  # Define a function to preprocess text\n",
        "    text = text.lower()  # Convert text to lowercase\n",
        "    tokens = word_tokenize(text)  # Tokenize the text\n",
        "    tokens = [word for word in tokens if word.isalpha() and word not in stopwords.words('english')]  # Remove stopwords and non-alphabetic tokens\n",
        "    return tokens  # Return the processed tokens"
      ],
      "metadata": {
        "id": "faK0MAFJgV_8"
      },
      "execution_count": 59,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "UwWGdEhXgWGt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "papers_df['processed_abstract'] = papers_df['abstract'].fillna('').apply(preprocess_text)  # Apply preprocessing to the abstracts\n"
      ],
      "metadata": {
        "id": "92jGMOgBf36u"
      },
      "execution_count": 60,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "xrnW6DUkgw1x"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "gX6RmJzPgw49"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "J13GVqd5iomt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 8: Prepare Data for LDA**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "In this step, we prepare the data for training the LDA model by creating a dictionary and corpus.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Dictionary and Corpus Creation**</span>\n",
        "- We create a dictionary using Gensim's `corpora.Dictionary()` method, which maps each word to a unique ID based on the processed abstracts.\n",
        "- A bag-of-words corpus is generated, which represents each document as a collection of word IDs and their corresponding counts, enabling LDA to process the text data effectively."
      ],
      "metadata": {
        "id": "biNBtgm7io20"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Prepare Data for LDA"
      ],
      "metadata": {
        "id": "kioew9BMgw8C"
      },
      "execution_count": 61,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "dictionary = corpora.Dictionary(papers_df['processed_abstract'])  # Create a dictionary from processed abstracts"
      ],
      "metadata": {
        "id": "zx8mefmwgw-2"
      },
      "execution_count": 62,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "corpus = [dictionary.doc2bow(text) for text in papers_df['processed_abstract']]  # Create a bag-of-words corpus"
      ],
      "metadata": {
        "id": "mZkiIC6dgxBz"
      },
      "execution_count": 63,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "9lUYZ7SHgxEq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "DhOwvwi6f38k"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "oQyB-vQ6irMw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 9: Train LDA Model**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "This step involves training the Latent Dirichlet Allocation (LDA) model on the prepared corpus to discover topics.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Model Training**</span>\n",
        "- The number of topics is set (e.g., `num_topics = 3`), and the number of passes over the corpus is defined (e.g., `passes = 5`).\n",
        "- The LDA model is trained using Gensim's `LdaModel()` method, which applies the LDA algorithm to identify latent topics in the dataset.\n",
        "- After training, the topics and their associated keywords are printed for review."
      ],
      "metadata": {
        "id": "WfXvsA4Kirfn"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Train LDA Model"
      ],
      "metadata": {
        "id": "P4a3bDEAfwCR"
      },
      "execution_count": 64,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "num_topics = 3  # Number of topics"
      ],
      "metadata": {
        "id": "OgvIucYhg34q"
      },
      "execution_count": 65,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "passes = 5  # Number of passes over the corpus"
      ],
      "metadata": {
        "id": "9uUcR5KXg37r"
      },
      "execution_count": 66,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "lda_model = LdaModel(corpus, num_topics=num_topics, id2word=dictionary, passes=passes)  # Train the LDA model"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0HrSaTm-g3-m",
        "outputId": "fe08ddef-c160-4af3-abd5-6ce04893fbb7"
      },
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:gensim.models.ldamodel:too few updates, training might not converge; consider increasing the number of passes or iterations to improve accuracy\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "nx70RoCTg4A6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "FvWvU0CMhEZE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Display the topics"
      ],
      "metadata": {
        "id": "SrmwBzAhg4Da"
      },
      "execution_count": 68,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for idx, topic in lda_model.print_topics(-1):  # Loop through the topics\n",
        "    print(f\"Topic {idx}: {topic}\")  # Print each topic and its words"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4fwcTCWmhCos",
        "outputId": "e59e62d0-d55a-430b-b77a-b838a4d91e9b"
      },
      "execution_count": 69,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Topic 0: 0.010*\"learning\" + 0.007*\"model\" + 0.007*\"data\" + 0.006*\"show\" + 0.006*\"methods\" + 0.005*\"approach\" + 0.005*\"paper\" + 0.005*\"algorithm\" + 0.005*\"networks\" + 0.005*\"algorithms\"\n",
            "Topic 1: 0.037*\"missing\" + 0.037*\"abstract\" + 0.009*\"model\" + 0.007*\"algorithm\" + 0.005*\"data\" + 0.005*\"show\" + 0.005*\"problem\" + 0.005*\"models\" + 0.005*\"method\" + 0.004*\"learning\"\n",
            "Topic 2: 0.011*\"learning\" + 0.010*\"algorithm\" + 0.008*\"data\" + 0.008*\"problem\" + 0.007*\"model\" + 0.006*\"show\" + 0.005*\"models\" + 0.005*\"method\" + 0.005*\"algorithms\" + 0.005*\"propose\"\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "NiKbA_OLhCqW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "F_zK4-9VhCs1"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "sgwTOV8viukP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 10: Visualize Topics**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "In this step, we visualize the results of the LDA model to interpret the identified topics.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Visualization Preparation**</span>\n",
        "- The `pyLDAvis` library is used to prepare the data for visualization, allowing us to interactively explore the topics.\n",
        "- The visualization is displayed inline, providing insights into the relationships between topics and the distribution of words associated with each topic."
      ],
      "metadata": {
        "id": "cWNdF8YdiuzW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Visualize Topics"
      ],
      "metadata": {
        "id": "LcCxFt3FhCvo"
      },
      "execution_count": 70,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "pyLDAvis.enable_notebook()  # Enable notebook visualization"
      ],
      "metadata": {
        "id": "5EwQOA2OhC1I"
      },
      "execution_count": 75,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "vis = pyLDAvis.gensim_models.prepare(lda_model, corpus, dictionary)  # Prepare visualization data"
      ],
      "metadata": {
        "id": "C6NPjNejhC3u"
      },
      "execution_count": 76,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "pyLDAvis.display(vis)  # Display the visualization"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 881
        },
        "id": "UdnmJwC6hO7-",
        "outputId": "92d783b6-8397-4226-d9a6-787392fbd52b"
      },
      "execution_count": 99,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "\n",
              "<link rel=\"stylesheet\" type=\"text/css\" href=\"https://cdn.jsdelivr.net/gh/bmabey/pyLDAvis@3.4.0/pyLDAvis/js/ldavis.v1.0.0.css\">\n",
              "\n",
              "\n",
              "<div id=\"ldavis_el205139081472960192344760793\" style=\"background-color:white;\"></div>\n",
              "<script type=\"text/javascript\">\n",
              "\n",
              "var ldavis_el205139081472960192344760793_data = {\"mdsDat\": {\"x\": [-0.04174254633555557, -0.035027361087352535, 0.0767699074229081], \"y\": [0.0469856556394259, -0.049807883640444195, 0.0028222280010183408], \"topics\": [1, 2, 3], \"cluster\": [1, 1, 1], \"Freq\": [44.594596321218575, 34.226275880926295, 21.17912779785513]}, \"tinfo\": {\"Term\": [\"abstract\", \"missing\", \"model\", \"convergence\", \"stochastic\", \"clustering\", \"gradient\", \"bandit\", \"methods\", \"algorithm\", \"networks\", \"neurons\", \"guarantees\", \"margin\", \"game\", \"network\", \"games\", \"submodular\", \"dynamical\", \"setting\", \"information\", \"variational\", \"p\", \"estimators\", \"bayesian\", \"proposed\", \"rates\", \"plasticity\", \"strategies\", \"feedback\", \"entries\", \"deliberation\", \"motion\", \"unitary\", \"mmd\", \"sources\", \"lstm\", \"pmd\", \"rational\", \"might\", \"sentence\", \"cirl\", \"hashing\", \"domain\", \"independence\", \"lp\", \"translation\", \"discriminators\", \"rl\", \"spn\", \"false\", \"easier\", \"hierarchy\", \"text\", \"augmentation\", \"evaluations\", \"softmax\", \"mccfr\", \"selected\", \"databases\", \"clustering\", \"successfully\", \"node\", \"tests\", \"utility\", \"human\", \"gan\", \"x\", \"image\", \"constraints\", \"hard\", \"testing\", \"classification\", \"sequence\", \"tree\", \"algorithm\", \"problem\", \"k\", \"real\", \"learning\", \"adversarial\", \"first\", \"metrics\", \"data\", \"test\", \"tasks\", \"decision\", \"one\", \"two\", \"also\", \"propose\", \"based\", \"features\", \"algorithms\", \"models\", \"show\", \"using\", \"results\", \"method\", \"model\", \"new\", \"structure\", \"performance\", \"demonstrate\", \"training\", \"paper\", \"matrix\", \"optimal\", \"approach\", \"network\", \"problems\", \"methods\", \"g\", \"lbp\", \"margin\", \"dynamical\", \"connectivity\", \"std\", \"workers\", \"spore\", \"bco\", \"dha\", \"jacobian\", \"bethe\", \"tensor\", \"established\", \"de\", \"states\", \"identification\", \"replica\", \"link\", \"protein\", \"cation\", \"q\", \"classi\", \"eigenfeatures\", \"hots\", \"principle\", \"grfs\", \"patients\", \"stale\", \"samplers\", \"social\", \"relational\", \"mcmc\", \"color\", \"brain\", \"gradient\", \"questions\", \"methods\", \"quantization\", \"networks\", \"bayesian\", \"programs\", \"stochastic\", \"output\", \"network\", \"ranking\", \"gradients\", \"spectral\", \"risk\", \"approach\", \"prediction\", \"learning\", \"properties\", \"neural\", \"variational\", \"inference\", \"paper\", \"n\", \"function\", \"show\", \"large\", \"number\", \"model\", \"data\", \"algorithms\", \"deep\", \"estimation\", \"error\", \"propose\", \"method\", \"based\", \"efficient\", \"analysis\", \"models\", \"new\", \"algorithm\", \"problem\", \"time\", \"performance\", \"using\", \"proposed\", \"also\", \"abstract\", \"missing\", \"arms\", \"memristive\", \"dendritic\", \"curvature\", \"net\", \"luce\", \"side\", \"sgns\", \"trustvi\", \"axiom\", \"deflation\", \"minsky\", \"rf\", \"preferred\", \"geometry\", \"waveforms\", \"ode\", \"relies\", \"disparity\", \"voc\", \"copula\", \"sussman\", \"influence\", \"lgmd\", \"shifted\", \"nonlinearities\", \"euclidean\", \"collapsed\", \"goss\", \"implementations\", \"bandit\", \"feedback\", \"neurons\", \"convergence\", \"outliers\", \"strategies\", \"estimators\", \"submodular\", \"model\", \"constraint\", \"guarantees\", \"rates\", \"algorithm\", \"collaborative\", \"information\", \"models\", \"game\", \"setting\", \"method\", \"proposed\", \"problem\", \"show\", \"games\", \"data\", \"distributions\", \"distribution\", \"stochastic\", \"optimization\", \"used\", \"new\", \"matrix\", \"performance\", \"time\", \"demonstrate\", \"algorithms\", \"however\", \"learning\", \"approach\", \"also\", \"function\", \"using\", \"number\", \"paper\", \"results\"], \"Freq\": [262.0, 265.0, 237.0, 67.0, 60.0, 49.0, 69.0, 19.0, 125.0, 251.0, 91.0, 22.0, 38.0, 16.0, 21.0, 96.0, 19.0, 18.0, 15.0, 54.0, 80.0, 45.0, 18.0, 16.0, 39.0, 84.0, 17.0, 13.0, 14.0, 12.0, 14.636443208383227, 7.656959647950735, 9.190888699095648, 8.33837368277035, 6.659069315212273, 6.643890559914276, 9.13968595504848, 5.782654456117464, 6.615913904483619, 7.425249937256845, 6.583812433551396, 5.748109554888647, 6.565810939798151, 13.186055386242266, 14.748601845018047, 5.742290496278428, 5.721591414332452, 5.714369987027188, 6.527493529628689, 4.886461140980394, 7.303310721887017, 5.685818154705102, 5.68631661277122, 8.893029359919678, 4.868922103162954, 5.679794655426321, 5.673144547641379, 4.851718228034857, 5.659371161215954, 4.850707596805359, 43.83331712734903, 7.273483814116349, 10.396684905439328, 10.2534144278591, 8.76459176002837, 21.078172377461144, 10.944993921330639, 13.035684774248185, 48.14051599230479, 16.79317996935981, 12.168066007063041, 10.625607879808284, 48.1118994418675, 26.392215390683553, 14.878005502072751, 144.4467871846509, 115.08763280610607, 29.15667839766501, 24.30575135930346, 163.6382542079667, 13.034630426146164, 37.783327138205884, 14.44639139123569, 123.24609955773491, 28.545226788366328, 37.51741883248904, 18.66393123661505, 41.8448577522293, 50.16768622944668, 62.47731518448082, 66.87189979164182, 61.62598480968153, 47.67677769490567, 76.10757085487883, 78.97286039938233, 95.31184777903741, 60.27126922007282, 55.374660621505654, 76.96005329784813, 98.71179981765647, 61.83371367281271, 39.70039600330227, 54.8668488590549, 46.76074291999829, 41.21964818598961, 61.689197692962374, 39.72712672786602, 38.22776143620866, 48.62988094893808, 41.17698051840734, 39.17113725750796, 39.91598704593892, 9.160778954771201, 7.362772761750205, 14.689962636401125, 13.582611558294916, 5.615556689406499, 8.011691701412774, 6.387094023802729, 4.78549403862485, 4.781344368462909, 4.773088778129637, 4.7574045709698, 4.743354573985509, 11.089298935568022, 4.728431879820114, 4.725643064629235, 5.513248487681853, 5.512954737744904, 4.7098858311692835, 7.056311643318889, 7.764254254523067, 4.674419062466127, 4.675240443868772, 4.663495051652975, 4.663576630582834, 4.659405578846535, 5.441960016141198, 3.858637000724537, 3.8565054222346937, 3.8460557534249924, 3.835905755603601, 9.113499359656196, 8.28530940584101, 9.493616383330545, 8.207502185699548, 12.623799921990223, 45.01263770685034, 12.100480297474117, 70.73299534409428, 9.566851236032374, 53.29961487549127, 25.234207741591785, 8.29103572040138, 35.16264771580176, 13.57968392734436, 50.2552734236695, 12.210363168625557, 16.177792580209278, 13.955491741417328, 18.23145924467077, 58.68172424658415, 21.953678182570993, 115.89762775516239, 22.791624055344787, 42.4534856245711, 25.68376553983496, 37.87496555467899, 56.13746163479323, 25.87985497143028, 42.73553556403154, 71.02014791403843, 36.37924745466922, 42.03704097711521, 78.03963588105067, 74.98251548460337, 52.44269926994706, 26.52568378017717, 31.913561571084543, 25.23338932071482, 43.855540586790106, 50.568210198605605, 40.594914970865524, 32.41164041154468, 30.11883564051736, 44.71854829628986, 39.67243358992076, 54.95182504717454, 48.21918965017083, 30.67165657449074, 35.10473983349434, 34.92925039747741, 31.32659064444116, 33.43127177163706, 259.8523757548077, 260.8499810394478, 5.364652853203017, 4.643244436353133, 3.945450460065826, 3.9451455396096424, 3.9182571734013703, 3.893348958192932, 3.891323618714278, 3.238274785630662, 3.8818971391910138, 3.8720094115010535, 3.216214055243017, 3.851556128938829, 3.847734649202924, 3.834328326716178, 5.095791988911412, 3.185199492394363, 3.1822310978599697, 3.18028290447804, 4.444676897038571, 3.8032476095629444, 3.1677779089472216, 3.799294635424707, 3.1638493610137517, 3.154205488268055, 3.1397258376334753, 3.112389657671129, 3.1133785118608075, 2.4761344510204184, 3.080726579085893, 5.95194145781106, 12.746387499435125, 7.643540001276566, 12.078126433125606, 27.166150779372234, 5.394239907934706, 8.342378604921105, 8.987418492564808, 9.648989221388458, 60.32892831555754, 5.346496448095781, 15.649688640743129, 9.00370832852438, 52.0597167786044, 6.670646300896573, 22.795291501555273, 34.335290731098524, 9.9935012986729, 17.473080400446353, 33.367876235551336, 22.006855950409197, 35.80403516695237, 35.96084163107902, 8.890027143121905, 36.94936362437279, 15.817801180849923, 17.55805265317854, 16.52205100305733, 16.065629454422897, 17.484707273915774, 22.484595062151218, 16.80236287214403, 19.94720905295193, 16.568008093120913, 17.62675216829449, 22.025714608903506, 15.882454088042694, 28.645880493716934, 19.696863251874234, 18.149485977632246, 17.02041560210519, 17.201299790478185, 16.76218013688753, 17.191902215297226, 16.498951955415354], \"Total\": [262.0, 265.0, 237.0, 67.0, 60.0, 49.0, 69.0, 19.0, 125.0, 251.0, 91.0, 22.0, 38.0, 16.0, 21.0, 96.0, 19.0, 18.0, 15.0, 54.0, 80.0, 45.0, 18.0, 16.0, 39.0, 84.0, 17.0, 13.0, 14.0, 12.0, 15.580284026548107, 8.229972370413654, 10.061573141898108, 9.133207262512528, 7.301208625409307, 7.2947085497779325, 10.040345723773823, 6.380390496348948, 7.299898776037754, 8.210344810401404, 7.2948969663430585, 6.375115339794154, 7.293085910931215, 14.651959884514024, 16.401159305607088, 6.386316513438928, 6.379376449240009, 6.378007650949709, 7.296587465942979, 5.462554917973804, 8.169938201539738, 6.370441371728033, 6.373430492758508, 9.968097643233705, 5.465953502732091, 6.38195533305451, 6.377863983799761, 5.455060344862739, 6.369416554813603, 5.460889623836713, 49.7259924407442, 8.190414995435647, 11.877251106281179, 11.771644422997058, 10.024941465783746, 25.24520616680737, 12.719148017024391, 15.475401189448888, 63.66134091676659, 20.593243473011796, 14.590267838701815, 12.725926805253337, 69.93947303505612, 35.39573085516372, 18.592871858901184, 251.45832901042985, 199.11085762322926, 40.65284805180293, 32.91874612204602, 308.18176245684606, 16.119320829528785, 57.25069862452271, 18.248703018367657, 235.17797866671106, 41.689430173851356, 58.224619037376584, 24.924968273571903, 66.87542952472957, 84.67236170200056, 114.05807293375014, 125.65272594962066, 114.37380541115529, 82.56705498510263, 150.5759847337294, 158.0266994267707, 202.29283732415485, 112.40181940802842, 102.50894284341939, 160.89613973200508, 237.0803640142647, 123.99074232488468, 67.25768396279759, 109.91879774550117, 86.7438481844331, 71.88505761095652, 135.01856154305284, 73.21975575443187, 70.90994582319006, 127.00846844739647, 96.4603198001323, 77.53233253137998, 125.93478734469745, 9.871888000321047, 8.069827739133302, 16.211078397100525, 15.15138959735282, 6.269517251607185, 8.981065028319277, 7.166233801985371, 5.37257554597101, 5.369957771263105, 5.3716941337707445, 5.370514225399344, 5.364891833755458, 12.55825490725245, 5.366662225644656, 5.364877329997607, 6.2619240389792425, 6.262580005541469, 5.359939301073283, 8.049855744947623, 8.894678254015176, 5.362221477621214, 5.3723169012145044, 5.362121681759092, 5.3677269192825605, 5.36593346896336, 6.275918910603398, 4.468968327771153, 4.466562607938424, 4.468001862108456, 4.465731626820853, 10.64453137365291, 9.78301385853195, 11.497728366573936, 9.89321574653624, 16.23709534810637, 69.04924855254959, 16.086039788314203, 125.93478734469745, 12.315427663890627, 91.56164725104333, 39.47821576450438, 10.48686222424718, 60.91556949805438, 19.107322536431433, 96.4603198001323, 17.07063641415883, 24.258575060670616, 20.29983897651048, 28.414539384355052, 127.00846844739647, 36.18894469545682, 308.18176245684606, 38.26015428712508, 86.46369501225753, 45.86592810422198, 78.01636573131603, 135.01856154305284, 47.404615238230306, 96.60632252818816, 202.29283732415485, 78.31703276644461, 96.68005573619325, 237.0803640142647, 235.17797866671106, 150.5759847337294, 53.47705763708746, 71.27613076816762, 50.304276568027404, 125.65272594962066, 160.89613973200508, 114.37380541115529, 78.21179904764242, 70.85960885475811, 158.0266994267707, 123.99074232488468, 251.45832901042985, 199.11085762322926, 73.90605513888676, 109.91879774550117, 112.40181940802842, 84.62300091303027, 114.05807293375014, 262.5648512428456, 265.98456618579286, 6.12618666830923, 5.370975462726836, 4.61163748379257, 4.611534463849805, 4.617083037497168, 4.6229787808775935, 4.623297701058478, 3.853916702721413, 4.6248441009380725, 4.627903477934661, 3.8584161294245307, 4.631212047795733, 4.632833411196915, 4.636677145279849, 6.180926641164488, 3.86497542776355, 3.865941604651347, 3.866561818778737, 5.417034609536362, 4.64302597100772, 3.8689878221389513, 4.641891801856529, 3.8694776234783577, 3.8711920966794726, 3.873558211417475, 3.880660977915406, 3.882012055861782, 3.1076024482561015, 3.889346805614249, 7.8325673426818785, 19.127897361206074, 12.01071441353427, 22.855370568616483, 67.93322768628234, 7.961411383850535, 14.601213326702691, 16.27444171671385, 18.022269054056288, 237.0803640142647, 7.970646938910738, 38.53448170442027, 17.18741262908194, 251.45832901042985, 11.315618296502413, 80.0593915243749, 158.0266994267707, 21.641694139586953, 54.56700882325664, 160.89613973200508, 84.62300091303027, 199.11085762322926, 202.29283732415485, 19.13869572317251, 235.17797866671106, 54.14154280243092, 65.54268358710236, 60.91556949805438, 61.23593061757955, 72.87153974051743, 123.99074232488468, 73.21975575443187, 109.91879774550117, 73.90605513888676, 86.7438481844331, 150.5759847337294, 68.80658274009585, 308.18176245684606, 127.00846844739647, 114.05807293375014, 96.60632252818816, 112.40181940802842, 96.68005573619325, 135.01856154305284, 102.50894284341939], \"Category\": [\"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\"], \"logprob\": [30.0, 29.0, 28.0, 27.0, 26.0, 25.0, 24.0, 23.0, 22.0, 21.0, 20.0, 19.0, 18.0, 17.0, 16.0, 15.0, 14.0, 13.0, 12.0, 11.0, 10.0, 9.0, 8.0, 7.0, 6.0, 5.0, 4.0, 3.0, 2.0, 1.0, -6.9139, -7.5618, -7.3792, -7.4765, -7.7014, -7.7037, -7.3848, -7.8425, -7.7079, -7.5925, -7.7128, -7.8485, -7.7155, -7.0182, -6.9063, -7.8495, -7.8532, -7.8544, -7.7214, -8.0109, -7.6091, -7.8594, -7.8593, -7.4121, -8.0145, -7.8605, -7.8617, -8.0181, -7.8641, -8.0183, -5.817, -7.6132, -7.2559, -7.2698, -7.4267, -6.5492, -7.2045, -7.0297, -5.7233, -6.7764, -7.0986, -7.2341, -5.7239, -6.3243, -6.8975, -4.6245, -4.8517, -6.2247, -6.4067, -4.4997, -7.0298, -5.9655, -6.927, -4.7832, -6.2459, -5.9726, -6.6708, -5.8634, -5.682, -5.4626, -5.3946, -5.4763, -5.733, -5.2653, -5.2283, -5.0402, -5.4985, -5.5833, -5.2541, -5.0052, -5.473, -5.916, -5.5925, -5.7524, -5.8785, -5.4753, -5.9154, -5.9538, -5.7132, -5.8795, -5.9295, -5.9106, -7.1179, -7.3363, -6.6456, -6.724, -7.6072, -7.2519, -7.4785, -7.7672, -7.7681, -7.7698, -7.7731, -7.776, -6.9268, -7.7792, -7.7798, -7.6256, -7.6257, -7.7831, -7.3789, -7.2833, -7.7907, -7.7905, -7.793, -7.793, -7.7939, -7.6386, -7.9825, -7.983, -7.9857, -7.9884, -7.123, -7.2183, -7.0822, -7.2277, -6.7972, -5.5258, -6.8395, -5.0739, -7.0745, -5.3569, -6.1046, -7.2176, -5.7728, -6.7242, -5.4157, -6.8305, -6.5491, -6.6969, -6.4296, -5.2607, -6.2438, -4.5801, -6.2064, -5.5844, -6.0869, -5.6985, -5.305, -6.0793, -5.5778, -5.0698, -5.7388, -5.5942, -4.9756, -5.0155, -5.3731, -6.0547, -5.8698, -6.1046, -5.5519, -5.4095, -5.6291, -5.8543, -5.9276, -5.5324, -5.6521, -5.3263, -5.457, -5.9094, -5.7744, -5.7795, -5.8883, -5.8233, -3.2927, -3.2889, -7.173, -7.3174, -7.4802, -7.4803, -7.4872, -7.4935, -7.4941, -7.6778, -7.4965, -7.499, -7.6846, -7.5043, -7.5053, -7.5088, -7.2244, -7.6943, -7.6952, -7.6958, -7.3611, -7.517, -7.6998, -7.518, -7.701, -7.7041, -7.7087, -7.7174, -7.7171, -7.9461, -7.7276, -7.0691, -6.3076, -6.8189, -6.3614, -5.5508, -7.1675, -6.7315, -6.657, -6.586, -4.753, -7.1764, -6.1024, -6.6552, -4.9004, -6.9551, -5.7263, -5.3166, -6.5509, -5.9921, -5.3452, -5.7615, -5.2747, -5.2704, -6.6679, -5.2433, -6.0917, -5.9873, -6.0481, -6.0761, -5.9915, -5.74, -6.0313, -5.8597, -6.0453, -5.9834, -5.7606, -6.0876, -5.4978, -5.8723, -5.9542, -6.0184, -6.0078, -6.0337, -6.0084, -6.0495], \"loglift\": [30.0, 29.0, 28.0, 27.0, 26.0, 25.0, 24.0, 23.0, 22.0, 21.0, 20.0, 19.0, 18.0, 17.0, 16.0, 15.0, 14.0, 13.0, 12.0, 11.0, 10.0, 9.0, 8.0, 7.0, 6.0, 5.0, 4.0, 3.0, 2.0, 1.0, 0.7451, 0.7354, 0.717, 0.7165, 0.7155, 0.7141, 0.7136, 0.7092, 0.7092, 0.707, 0.705, 0.704, 0.7025, 0.7021, 0.7014, 0.7013, 0.6987, 0.6977, 0.6962, 0.6961, 0.6954, 0.6939, 0.6935, 0.6934, 0.6919, 0.691, 0.6905, 0.6903, 0.6894, 0.6891, 0.6814, 0.6888, 0.6744, 0.6695, 0.6732, 0.6272, 0.6573, 0.636, 0.5281, 0.6036, 0.626, 0.6272, 0.4335, 0.514, 0.5847, 0.2532, 0.2594, 0.4752, 0.5042, 0.1745, 0.5951, 0.392, 0.5739, 0.1614, 0.4288, 0.3681, 0.5183, 0.3387, 0.2841, 0.2057, 0.1768, 0.1892, 0.2584, 0.1252, 0.1139, 0.055, 0.1843, 0.1917, 0.0701, -0.0686, 0.1118, 0.2804, 0.1127, 0.1896, 0.2514, 0.0243, 0.1961, 0.1897, -0.1525, -0.0437, 0.1248, -0.3414, 0.9974, 0.9805, 0.9736, 0.9629, 0.962, 0.958, 0.9571, 0.9565, 0.9561, 0.954, 0.951, 0.949, 0.9478, 0.9456, 0.9453, 0.9448, 0.9447, 0.9429, 0.9404, 0.9363, 0.9349, 0.9332, 0.9326, 0.9316, 0.931, 0.9296, 0.9253, 0.9253, 0.9223, 0.9201, 0.9169, 0.906, 0.8806, 0.8854, 0.8205, 0.6443, 0.7875, 0.4953, 0.8196, 0.5311, 0.6246, 0.8372, 0.5227, 0.7307, 0.4202, 0.7371, 0.667, 0.6974, 0.6284, 0.3001, 0.5724, 0.0942, 0.5542, 0.3609, 0.4923, 0.3495, 0.1946, 0.4669, 0.2566, 0.0254, 0.3054, 0.2393, -0.039, -0.0709, 0.0174, 0.371, 0.2686, 0.3823, 0.0196, -0.0853, 0.0363, 0.1913, 0.2166, -0.1902, -0.0674, -0.4486, -0.3459, 0.1927, -0.0692, -0.0966, 0.0784, -0.155, 1.5418, 1.5327, 1.4194, 1.4066, 1.3961, 1.3961, 1.388, 1.3804, 1.3798, 1.3781, 1.377, 1.3738, 1.3701, 1.3678, 1.3665, 1.3622, 1.3591, 1.3587, 1.3575, 1.3568, 1.3543, 1.3526, 1.3522, 1.3518, 1.3508, 1.3473, 1.3421, 1.3315, 1.3315, 1.325, 1.3191, 1.2776, 1.1463, 1.1002, 0.9144, 0.6356, 1.1629, 0.9924, 0.9584, 0.9274, 0.1836, 1.1528, 0.6511, 0.9056, -0.0227, 1.0237, 0.2959, 0.0256, 0.7795, 0.4134, -0.021, 0.2053, -0.1636, -0.1751, 0.7854, -0.2986, 0.3217, 0.235, 0.2474, 0.2141, 0.1248, -0.1552, 0.0802, -0.1545, 0.0568, -0.0414, -0.3701, 0.0861, -0.8235, -0.3116, -0.2859, -0.1841, -0.3249, -0.2001, -0.5088, -0.2745]}, \"token.table\": {\"Topic\": [1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 3, 1, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 2, 2, 1, 2, 2, 1, 2, 1, 2, 3, 1, 2, 3, 1, 2, 3, 3, 1, 2, 2, 1, 3, 1, 2, 3, 1, 2, 3, 3, 3, 1, 2, 3, 1, 2, 1, 2, 3, 1, 2, 3, 3, 1, 1, 2, 3, 3, 2, 1, 1, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1, 2, 3, 1, 1, 2, 3, 2, 1, 2, 1, 2, 3, 2, 1, 2, 3, 1, 2, 3, 3, 1, 1, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 2, 1, 2, 3, 1, 3, 1, 2, 3, 2, 3, 3, 1, 2, 3, 1, 2, 3, 2, 1, 2, 3, 1, 2, 3, 1, 1, 2, 1, 2, 3, 1, 2, 3, 2, 1, 2, 3, 1, 3, 1, 2, 3, 1, 2, 3, 3, 1, 2, 3, 2, 1, 2, 3, 1, 2, 3, 2, 1, 2, 3, 3, 1, 2, 1, 1, 2, 3, 1, 2, 1, 2, 3, 1, 1, 2, 3, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 3, 1, 2, 3, 1, 1, 2, 3, 1, 2, 3, 1, 2, 1, 2, 3, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 2, 1, 2, 3, 2, 3, 1, 1, 2, 3, 3, 1, 2, 1, 2, 3, 1, 2, 3, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 2, 3, 2, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 1, 2, 3, 1, 2, 3, 3, 2, 1, 2, 3, 3, 1, 2, 3, 1, 2, 1, 1, 1, 2, 3, 1, 2, 3, 3, 3, 1, 2, 3, 3, 1, 2, 3, 1, 1, 1, 2, 3, 1, 2, 2, 2, 1, 2, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 3, 1, 2, 3, 1, 2, 3, 1, 1, 3, 3, 1, 2, 3, 1, 1, 2, 3, 1, 2, 3, 1, 2, 1, 2, 3, 3, 3, 2, 1, 2, 3], \"Freq\": [0.0038085828901565444, 0.007617165780313089, 0.9902315514407016, 0.8064855918858231, 0.0620373532219864, 0.1240747064439728, 0.5726594961745222, 0.21872411312221335, 0.2067937069519108, 0.5047285603636886, 0.3453405939330501, 0.14610563589475198, 0.5435827417144973, 0.28932629800932924, 0.15781434436872502, 0.38103512616534846, 0.4233723624059428, 0.19757376912277327, 0.3858010461742911, 0.4645359535567995, 0.15746981476501679, 0.8161684047051659, 0.9147534821693617, 0.8643222614887204, 0.2091186461567142, 0.1045593230783571, 0.6796356000093211, 0.5420821645054132, 0.35847368943099905, 0.10491912861395095, 0.2026434033321474, 0.6332606354129606, 0.15198255249911055, 0.9311060185160293, 0.9319852393929756, 0.18476211019786068, 0.8006358108573963, 0.9324493627999301, 0.9411594426452736, 0.9324667168611708, 0.6863077160438528, 0.214471161263704, 0.10008654192306186, 0.8848491068817267, 0.0402204139491694, 0.0804408278983388, 0.17674685974677914, 0.17674685974677914, 0.6186140091137271, 0.6435829657433638, 0.10107936849048446, 0.8086349479238757, 0.957011482576574, 0.2509206611870477, 0.6273016529676192, 0.8255134759261757, 0.048559616230951515, 0.14567884869285455, 0.26496606466462236, 0.33856774929368416, 0.39744909699693354, 0.7753965992949196, 0.8673902431731405, 0.5230081519422907, 0.3189074097209089, 0.15732765546231506, 0.9156017323944919, 0.931987758982409, 0.7622878308794405, 0.12036123645464851, 0.160481648606198, 0.4674901930778999, 0.5048894085241319, 0.037399215446231995, 0.7775211121272809, 0.9720567263092653, 0.5418251666685285, 0.25362029078101334, 0.2075075106390109, 0.8673708664347214, 0.9308050450166216, 0.9407326438541632, 0.18460284492913528, 0.7384113797165411, 0.38143082693244434, 0.35091636077784877, 0.2746301953913599, 0.44328252867818935, 0.25858147506227713, 0.29552168578545956, 0.8872533164481281, 0.06825025511139447, 0.0660005469184632, 0.9240076568584847, 0.0660005469184632, 0.9418499676691087, 0.4602886065575699, 0.40914542805117327, 0.12785794626599165, 0.9314929904571766, 0.9627552344001349, 0.06418368229334233, 0.43733856246295466, 0.49697563916244847, 0.05963707669949382, 0.9316777896151995, 0.4349282104107256, 0.44895815268203937, 0.11223953817050984, 0.18433812060778715, 0.24578416081038287, 0.5530143618233615, 0.7727951270707786, 0.9401507354530951, 0.8567996265480629, 0.12239994664972327, 0.581345671208214, 0.23011599485325138, 0.193781890402738, 0.16651798811786755, 0.16651798811786755, 0.6660719524714702, 0.6637473587741184, 0.22707146484377733, 0.10480221454328184, 0.38299770689650253, 0.44510544314998945, 0.17597191938487955, 0.9116797110853879, 0.5082781379799106, 0.046207103452719146, 0.46207103452719145, 0.522501645077743, 0.4702514805699686, 0.8648378008713055, 0.07862161826102777, 0.07862161826102777, 0.16178803892284987, 0.8089401946142494, 0.7713377463972917, 0.23171866943379232, 0.6517087577825409, 0.11585933471689616, 0.12366761000994535, 0.6595605867197085, 0.20611268334990893, 0.8950611654916236, 0.2595078370770692, 0.31140940449248306, 0.4152125393233107, 0.8224660528965116, 0.13707767548275196, 0.06853883774137598, 0.9598131827170822, 0.9414082426751497, 0.9318043223830627, 0.5086722607952503, 0.2616028769804144, 0.23253589064925725, 0.8318410973252812, 0.07922296165002678, 0.07922296165002678, 0.9580715926488564, 0.7539897732087852, 0.1256649622014642, 0.10995684192628118, 0.1276720590132327, 0.7660323540793962, 0.9145694960033666, 0.06097129973355777, 0.06097129973355777, 0.3332634089819377, 0.4870772900505243, 0.17944952791335106, 0.7752984490199053, 0.46215689746698824, 0.249814539171345, 0.28728672004704675, 0.9310095439935656, 0.7133571542895595, 0.22138670305538055, 0.07379556768512685, 0.3702898204338663, 0.459670121917903, 0.16599198847035387, 0.8674286770775356, 0.5321534885535757, 0.3764012480013096, 0.0941003120003274, 0.774955084913833, 0.124225828596697, 0.869580800176879, 0.939508711692258, 0.8963834759882359, 0.09959816399869288, 0.865242993661474, 0.06168621084325011, 0.9252931626487516, 0.5463006477944836, 0.23217777531265552, 0.23217777531265552, 0.9165801446557253, 0.0869737019450893, 0.7827633175058037, 0.0869737019450893, 0.9309295927152695, 0.47856959233611335, 0.31697466505378935, 0.20510125385833428, 0.31762470754419564, 0.5637838558909473, 0.11910926532907337, 0.7671778090699783, 0.16439524480070966, 0.054798414933569885, 0.8525829501255465, 0.8637047836977872, 0.0075192332723658115, 0.011278849908548718, 0.9812599420437385, 0.9587453747916399, 0.41757992236777297, 0.3290023630776393, 0.2530787408289533, 0.49991552241846615, 0.28476200644089844, 0.2151535159775677, 0.894492329685749, 0.09938803663174989, 0.37970986389282146, 0.5484698034007421, 0.08437996975396032, 0.8663478580554884, 0.42504524228151863, 0.5183478564408763, 0.05183478564408764, 0.34949130952463686, 0.5788449814001798, 0.06552962053586942, 0.3585319826501202, 0.48575300875177574, 0.150352121756502, 0.13126017760216963, 0.35002714027245235, 0.5250407104086785, 0.500037332122309, 0.3226047304014897, 0.17743260172081934, 0.8419456581760395, 0.08419456581760394, 0.773064180837442, 0.3930490079948752, 0.43442258778380943, 0.17583771410297047, 0.7760075828332532, 0.6280333494451652, 0.254203974775424, 0.11962539989431717, 0.5358909749381399, 0.2538430933917505, 0.21153591115979206, 0.3266056349318944, 0.408257043664868, 0.2612845079455155, 0.25121173917189293, 0.6280293479297324, 0.10467191288505505, 0.7327033901953853, 0.2093438257701101, 0.1102106899025887, 0.4408427596103548, 0.4959481045616491, 0.4591961230473508, 0.4147577885588975, 0.12590861438395104, 0.8955432512892124, 0.5003693738294284, 0.3184168742550908, 0.1819524995743376, 0.4599205188108468, 0.5365739386126546, 0.940381314189685, 0.2486947346969713, 0.6079204625925965, 0.1381637414983174, 0.86268676353108, 0.1593392161760508, 0.796696080880254, 0.5775676995857785, 0.24107173547928148, 0.1808038016094611, 0.5030159512383479, 0.3224461225886845, 0.16767198374611594, 0.7628592641851263, 0.19071481604628157, 0.3136422271051353, 0.6011476019515093, 0.10454740903504509, 0.5332156504655781, 0.3501714719475438, 0.11937663816393539, 0.3663306626511589, 0.3663306626511589, 0.2599765993008224, 0.8994142083091868, 0.11242677603864835, 0.9306971446285427, 0.08119896663694788, 0.8119896663694788, 0.16239793327389576, 0.18649711423562235, 0.7459884569424894, 0.06216570474520745, 0.23432049649199346, 0.7029614894759804, 0.058580124122998366, 0.2327284557788416, 0.2327284557788416, 0.5236390255023936, 0.9589174062218253, 0.7290678664071885, 0.12151131106786475, 0.15188913883483093, 0.1022179886955676, 0.8177439095645408, 0.1022179886955676, 0.7758831076823588, 0.9328463848459612, 0.5365385543387325, 0.3024126397181947, 0.15608394308035856, 0.8634025109412644, 0.24635275290980702, 0.6334785074823609, 0.10557975124706015, 0.9593525785406796, 0.8957098935315093, 0.9420015080448112, 0.9595748963002981, 0.734551861816041, 0.08475598405569704, 0.16951196811139407, 0.32986964812937186, 0.34819573969211476, 0.311543556566629, 0.7784288637794309, 0.7744817132623371, 0.4696162318776103, 0.3509763417190561, 0.17795983523783127, 0.8651833082442912, 0.09394495303712252, 0.8455045773341028, 0.09394495303712252, 0.9407538347071115, 0.9595996813626085, 0.19704589798118663, 0.6896606429341532, 0.14778442348588997, 0.9153226054621749, 0.9306523393141655, 0.8952547746057552, 0.9581719552411021, 0.1113453690455174, 0.8907629523641392, 0.14774547909771174, 0.5745657520466567, 0.2790747938512333, 0.20546237719255847, 0.20546237719255847, 0.5478996725134893, 0.594727585655868, 0.3122319824693307, 0.0892091378483802, 0.38840836184412103, 0.055486908834874435, 0.5548690883487444, 0.854657548353894, 0.12209393547912771, 0.8617176295234189, 0.6526448885068078, 0.2576229823053189, 0.10304919292212755, 0.0796288980742456, 0.8759178788167016, 0.0796288980742456, 0.6956199659977487, 0.167908267654629, 0.167908267654629, 0.8643771230445185, 0.07857973845859259, 0.07857973845859259, 0.8494989859245172, 0.08494989859245172, 0.08494989859245172, 0.9028804012678543, 0.10032004458531714, 0.36532865878527393, 0.41945142304975896, 0.23002174812406137, 0.5703549717090425, 0.34777742177380644, 0.08346658122571354, 0.9405307944657811, 0.8067607905778619, 0.16135215811557238, 0.8648940186305235, 0.590511460823215, 0.23620458432928604, 0.17715343824696453, 0.8759244994730597, 0.41168335548864005, 0.35679224142348803, 0.23328723477689603, 0.5337991886251837, 0.3113828600313572, 0.15124310344380207, 0.8977608528406887, 0.09975120587118763, 0.21802676656355496, 0.5668695930652429, 0.23982944321991045, 0.8615071345663488, 0.7762015712829347, 0.8372598725899411, 0.8400428422407159, 0.12923736034472552, 0.06461868017236276], \"Term\": [\"abstract\", \"abstract\", \"abstract\", \"adversarial\", \"adversarial\", \"adversarial\", \"algorithm\", \"algorithm\", \"algorithm\", \"algorithms\", \"algorithms\", \"algorithms\", \"also\", \"also\", \"also\", \"analysis\", \"analysis\", \"analysis\", \"approach\", \"approach\", \"approach\", \"arms\", \"augmentation\", \"axiom\", \"bandit\", \"bandit\", \"bandit\", \"based\", \"based\", \"based\", \"bayesian\", \"bayesian\", \"bayesian\", \"bco\", \"bethe\", \"brain\", \"brain\", \"cation\", \"cirl\", \"classi\", \"classification\", \"classification\", \"classification\", \"clustering\", \"clustering\", \"clustering\", \"collaborative\", \"collaborative\", \"collaborative\", \"collapsed\", \"color\", \"color\", \"connectivity\", \"constraint\", \"constraint\", \"constraints\", \"constraints\", \"constraints\", \"convergence\", \"convergence\", \"convergence\", \"copula\", \"curvature\", \"data\", \"data\", \"data\", \"databases\", \"de\", \"decision\", \"decision\", \"decision\", \"deep\", \"deep\", \"deep\", \"deflation\", \"deliberation\", \"demonstrate\", \"demonstrate\", \"demonstrate\", \"dendritic\", \"dha\", \"discriminators\", \"disparity\", \"disparity\", \"distribution\", \"distribution\", \"distribution\", \"distributions\", \"distributions\", \"distributions\", \"domain\", \"domain\", \"dynamical\", \"dynamical\", \"dynamical\", \"easier\", \"efficient\", \"efficient\", \"efficient\", \"eigenfeatures\", \"entries\", \"entries\", \"error\", \"error\", \"error\", \"established\", \"estimation\", \"estimation\", \"estimation\", \"estimators\", \"estimators\", \"estimators\", \"euclidean\", \"evaluations\", \"false\", \"false\", \"features\", \"features\", \"features\", \"feedback\", \"feedback\", \"feedback\", \"first\", \"first\", \"first\", \"function\", \"function\", \"function\", \"g\", \"game\", \"game\", \"game\", \"games\", \"games\", \"gan\", \"gan\", \"gan\", \"geometry\", \"geometry\", \"goss\", \"gradient\", \"gradient\", \"gradient\", \"gradients\", \"gradients\", \"gradients\", \"grfs\", \"guarantees\", \"guarantees\", \"guarantees\", \"hard\", \"hard\", \"hard\", \"hashing\", \"hierarchy\", \"hots\", \"however\", \"however\", \"however\", \"human\", \"human\", \"human\", \"identification\", \"image\", \"image\", \"image\", \"implementations\", \"implementations\", \"independence\", \"independence\", \"independence\", \"inference\", \"inference\", \"inference\", \"influence\", \"information\", \"information\", \"information\", \"jacobian\", \"k\", \"k\", \"k\", \"large\", \"large\", \"large\", \"lbp\", \"learning\", \"learning\", \"learning\", \"lgmd\", \"link\", \"link\", \"lp\", \"lstm\", \"lstm\", \"luce\", \"margin\", \"margin\", \"matrix\", \"matrix\", \"matrix\", \"mccfr\", \"mcmc\", \"mcmc\", \"mcmc\", \"memristive\", \"method\", \"method\", \"method\", \"methods\", \"methods\", \"methods\", \"metrics\", \"metrics\", \"metrics\", \"might\", \"minsky\", \"missing\", \"missing\", \"missing\", \"mmd\", \"model\", \"model\", \"model\", \"models\", \"models\", \"models\", \"motion\", \"motion\", \"n\", \"n\", \"n\", \"net\", \"network\", \"network\", \"network\", \"networks\", \"networks\", \"networks\", \"neural\", \"neural\", \"neural\", \"neurons\", \"neurons\", \"neurons\", \"new\", \"new\", \"new\", \"node\", \"node\", \"nonlinearities\", \"number\", \"number\", \"number\", \"ode\", \"one\", \"one\", \"one\", \"optimal\", \"optimal\", \"optimal\", \"optimization\", \"optimization\", \"optimization\", \"outliers\", \"outliers\", \"output\", \"output\", \"output\", \"p\", \"p\", \"p\", \"paper\", \"paper\", \"paper\", \"patients\", \"performance\", \"performance\", \"performance\", \"plasticity\", \"plasticity\", \"pmd\", \"prediction\", \"prediction\", \"prediction\", \"preferred\", \"principle\", \"principle\", \"problem\", \"problem\", \"problem\", \"problems\", \"problems\", \"problems\", \"programs\", \"programs\", \"properties\", \"properties\", \"properties\", \"propose\", \"propose\", \"propose\", \"proposed\", \"proposed\", \"proposed\", \"protein\", \"protein\", \"q\", \"quantization\", \"quantization\", \"quantization\", \"questions\", \"questions\", \"questions\", \"ranking\", \"ranking\", \"ranking\", \"rates\", \"rates\", \"rates\", \"rational\", \"real\", \"real\", \"real\", \"relational\", \"relational\", \"relational\", \"relies\", \"replica\", \"results\", \"results\", \"results\", \"rf\", \"risk\", \"risk\", \"risk\", \"rl\", \"samplers\", \"selected\", \"sentence\", \"sequence\", \"sequence\", \"sequence\", \"setting\", \"setting\", \"setting\", \"sgns\", \"shifted\", \"show\", \"show\", \"show\", \"side\", \"social\", \"social\", \"social\", \"softmax\", \"sources\", \"spectral\", \"spectral\", \"spectral\", \"spn\", \"spore\", \"stale\", \"states\", \"std\", \"std\", \"stochastic\", \"stochastic\", \"stochastic\", \"strategies\", \"strategies\", \"strategies\", \"structure\", \"structure\", \"structure\", \"submodular\", \"submodular\", \"submodular\", \"successfully\", \"successfully\", \"sussman\", \"tasks\", \"tasks\", \"tasks\", \"tensor\", \"tensor\", \"tensor\", \"test\", \"test\", \"test\", \"testing\", \"testing\", \"testing\", \"tests\", \"tests\", \"tests\", \"text\", \"text\", \"time\", \"time\", \"time\", \"training\", \"training\", \"training\", \"translation\", \"tree\", \"tree\", \"trustvi\", \"two\", \"two\", \"two\", \"unitary\", \"used\", \"used\", \"used\", \"using\", \"using\", \"using\", \"utility\", \"utility\", \"variational\", \"variational\", \"variational\", \"voc\", \"waveforms\", \"workers\", \"x\", \"x\", \"x\"]}, \"R\": 30, \"lambda.step\": 0.01, \"plot.opts\": {\"xlab\": \"PC1\", \"ylab\": \"PC2\"}, \"topic.order\": [3, 1, 2]};\n",
              "\n",
              "function LDAvis_load_lib(url, callback){\n",
              "  var s = document.createElement('script');\n",
              "  s.src = url;\n",
              "  s.async = true;\n",
              "  s.onreadystatechange = s.onload = callback;\n",
              "  s.onerror = function(){console.warn(\"failed to load library \" + url);};\n",
              "  document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
              "}\n",
              "\n",
              "if(typeof(LDAvis) !== \"undefined\"){\n",
              "   // already loaded: just create the visualization\n",
              "   !function(LDAvis){\n",
              "       new LDAvis(\"#\" + \"ldavis_el205139081472960192344760793\", ldavis_el205139081472960192344760793_data);\n",
              "   }(LDAvis);\n",
              "}else if(typeof define === \"function\" && define.amd){\n",
              "   // require.js is available: use it to load d3/LDAvis\n",
              "   require.config({paths: {d3: \"https://d3js.org/d3.v5\"}});\n",
              "   require([\"d3\"], function(d3){\n",
              "      window.d3 = d3;\n",
              "      LDAvis_load_lib(\"https://cdn.jsdelivr.net/gh/bmabey/pyLDAvis@3.4.0/pyLDAvis/js/ldavis.v3.0.0.js\", function(){\n",
              "        new LDAvis(\"#\" + \"ldavis_el205139081472960192344760793\", ldavis_el205139081472960192344760793_data);\n",
              "      });\n",
              "    });\n",
              "}else{\n",
              "    // require.js not available: dynamically load d3 & LDAvis\n",
              "    LDAvis_load_lib(\"https://d3js.org/d3.v5.js\", function(){\n",
              "         LDAvis_load_lib(\"https://cdn.jsdelivr.net/gh/bmabey/pyLDAvis@3.4.0/pyLDAvis/js/ldavis.v3.0.0.js\", function(){\n",
              "                 new LDAvis(\"#\" + \"ldavis_el205139081472960192344760793\", ldavis_el205139081472960192344760793_data);\n",
              "            })\n",
              "         });\n",
              "}\n",
              "</script>"
            ]
          },
          "metadata": {},
          "execution_count": 99
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "R_LdHny4hO_K"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "3Wa8zTi0kV4w"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "bam1ljkvhPGl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 11: Model Evaluation**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "This step evaluates the performance of the LDA model using perplexity and coherence scores.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Evaluation Metrics**</span>\n",
        "- **Perplexity**: A measure of how well the model predicts a sample. A lower perplexity indicates a better model.\n",
        "- **Coherence Score**: Evaluates the degree of semantic similarity between high-scoring words in the topic. A higher coherence score indicates more interpretable topics.\n",
        "- Both metrics are printed to assess the effectiveness of the trained LDA model."
      ],
      "metadata": {
        "id": "KxVVujNBixpI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Model Evaluation"
      ],
      "metadata": {
        "id": "6GqQJd1ahPJG"
      },
      "execution_count": 78,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "perplexity = lda_model.log_perplexity(corpus)  # Calculate perplexity"
      ],
      "metadata": {
        "id": "pBkCbGoihTTS"
      },
      "execution_count": 79,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(f'Perplexity: {perplexity}')  # Print perplexity value"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PGeWLHqFhTY8",
        "outputId": "71317d39-426e-4840-a4e2-8ae538746f63"
      },
      "execution_count": 80,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Perplexity: -7.70936757329736\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "WUPw3BkYhTb-"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "wzNxm1LxhTgH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from gensim.models import CoherenceModel  # Import CoherenceModel for evaluation"
      ],
      "metadata": {
        "id": "lU8T__X_hTlT"
      },
      "execution_count": 81,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "coherence_model_lda = CoherenceModel(model=lda_model, texts=papers_df['processed_abstract'], dictionary=dictionary, coherence='c_v')  # Create coherence model\n"
      ],
      "metadata": {
        "id": "KD7YjRTghZjE"
      },
      "execution_count": 82,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "coherence_lda = coherence_model_lda.get_coherence()  # Get coherence score"
      ],
      "metadata": {
        "id": "et3HfZCjhZwq"
      },
      "execution_count": 83,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(f'Coherence Score: {coherence_lda}')  # Print coherence score"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2CvpKGgbhZzV",
        "outputId": "2f6df4f3-76bb-4b69-dc9e-43e851ab68ea"
      },
      "execution_count": 84,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Coherence Score: 0.5047528461835386\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "rv2u5VPZhZ35"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "9-eyoXmRhZ-c"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "9YD_t1s6i0H3"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 12: Save the Model**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "In this step, we save the trained LDA model and dictionary to disk for future use.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Model Persistence**</span>\n",
        "- The LDA model is saved using the `save()` method, allowing us to load it later without retraining.\n",
        "- The dictionary is also saved to enable consistent processing of new text data in the future."
      ],
      "metadata": {
        "id": "KNOGZnZBi0UX"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Save the Model"
      ],
      "metadata": {
        "id": "7dMhwwPehaCg"
      },
      "execution_count": 85,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "lda_model.save('lda_model.gensim')  # Save the LDA model"
      ],
      "metadata": {
        "id": "f8Nfx_0khaHV"
      },
      "execution_count": 86,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "dictionary.save('lda_dictionary.gensim')  # Save the dictionary"
      ],
      "metadata": {
        "id": "XkopVD9bhmTi"
      },
      "execution_count": 87,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "DUUa2VnRhmWp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "anLAg-xChmZP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Load and Use the Model"
      ],
      "metadata": {
        "id": "PiMUgVS_hmcC"
      },
      "execution_count": 88,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "lda_model = LdaModel.load('lda_model.gensim')  # Load the LDA model"
      ],
      "metadata": {
        "id": "cVr6Wvnihmgs"
      },
      "execution_count": 89,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "dictionary = corpora.Dictionary.load('lda_dictionary.gensim')  # Load the dictionary"
      ],
      "metadata": {
        "id": "lDVsCJYehToO"
      },
      "execution_count": 90,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "MquEkmduhPL1"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def infer_topics(new_text):  # Define a function to infer topics from new text\n",
        "    new_text_processed = preprocess_text(new_text)  # Preprocess the new text\n",
        "    new_corpus = dictionary.doc2bow(new_text_processed)  # Create a bag-of-words representation\n",
        "    topics = lda_model.get_document_topics(new_corpus)  # Get the topics for the new document\n",
        "    return topics  # Return the inferred topics"
      ],
      "metadata": {
        "id": "HXdPssnqhwgz"
      },
      "execution_count": 91,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "8OZ2i7oHhwj9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "new_abstract = \"This study investigates the effects of X on Y.\"  # Example new abstract"
      ],
      "metadata": {
        "id": "AU6aKR0Nhwoy"
      },
      "execution_count": 92,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "predicted_topics = infer_topics(new_abstract)  # Infer topics from the new abstract"
      ],
      "metadata": {
        "id": "oHEhrwE8hwuF"
      },
      "execution_count": 93,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(predicted_topics)  # Print the predicted topics"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4BzOQUqIhwxd",
        "outputId": "96c6e1f6-76fc-4588-e842-ee92d4e2f415"
      },
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[(0, 0.39351973), (1, 0.068457015), (2, 0.5380233)]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "GvgBYJ9gh4nj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "SNvmf106h4pz"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "n4QxwFpoh4sR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Step 13: Create Flask Application**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Overview**</span>\n",
        "In this final step, we create a Flask web application that allows users to interact with the LDA model.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Application Structure**</span>\n",
        "- The application includes routes for rendering the input form and displaying the inferred topics based on user input.\n",
        "- The model is loaded when the application starts, allowing users to submit abstracts for topic inference in real time.\n",
        "- The application is designed to provide a user-friendly interface for interacting with the LDA model, enabling broader access to the insights generated by our analysis."
      ],
      "metadata": {
        "id": "2CTX1iP1i7NN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a Flask App"
      ],
      "metadata": {
        "id": "7v32a4R2h4ue"
      },
      "execution_count": 95,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "app = Flask(__name__)  # Initialize the Flask application"
      ],
      "metadata": {
        "id": "n7av21kRh4xh"
      },
      "execution_count": 96,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "nvVNZ8bah9F6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "@app.route('/infer', methods=['POST'])  # Define the inference route\n",
        "def infer():  # Define the inference function\n",
        "    data = request.json  # Get the JSON data from the request\n",
        "    abstract = data.get('abstract', '')  # Extract the abstract from the JSON data\n",
        "    topics = infer_topics(abstract)  # Infer topics using the LDA model\n",
        "    return jsonify(topics)  # Return the inferred topics as JSON response"
      ],
      "metadata": {
        "id": "tODMhOM1h9I-"
      },
      "execution_count": 97,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "HZzfhK76h9Lx"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "if __name__ == '__main__':  # Check if the script is run directly\n",
        "    app.run(port=5000)  # Run the Flask app on port 5000"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZNx71feMh9Q-",
        "outputId": "30164016-e2a5-4df1-a6d7-c78245fba2e5"
      },
      "execution_count": 98,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            " * Serving Flask app '__main__'\n",
            " * Debug mode: off\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "INFO:werkzeug:\u001b[31m\u001b[1mWARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.\u001b[0m\n",
            " * Running on http://127.0.0.1:5000\n",
            "INFO:werkzeug:\u001b[33mPress CTRL+C to quit\u001b[0m\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "u8068jRfh9Tc"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "6zGsg9Bwh9cN"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "Q3eToLsThw0P"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <span style=\"color:blue;\">**Conclusion**</span>\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Summary**</span>\n",
        "In this project, we successfully implemented a topic modeling pipeline using Latent Dirichlet Allocation (LDA) on a dataset of research papers. The key steps included data loading, preprocessing, model training, and evaluation.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Key Achievements**</span>\n",
        "- **Data Preprocessing**: Effectively cleaned and prepared text data for analysis, ensuring the quality of input for the LDA model.\n",
        "- **Model Training**: Trained the LDA model to identify distinct topics present in the abstracts, enabling insights into the main themes of the research papers.\n",
        "- **Visualization**: Employed `pyLDAvis` for an interactive visualization of the topics, enhancing interpretability and understanding of the model outputs.\n",
        "- **Web Application**: Developed a Flask application to facilitate user interaction with the model, allowing for real-time topic inference based on new abstracts.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Future Work**</span>\n",
        "- Further refinement of the model can be achieved by exploring advanced NLP techniques, such as using pre-trained embeddings or fine-tuning transformer models.\n",
        "- Expanding the dataset to include more diverse research topics could improve the model's robustness and applicability.\n",
        "- Implementing additional evaluation metrics and user feedback mechanisms in the Flask application could enhance user experience and model accuracy.\n",
        "\n",
        "## <span style=\"color:darkblue;\">**Final Thoughts**</span>\n",
        "This project highlights the power of topic modeling in extracting meaningful insights from large text corpora, demonstrating the value of combining data science techniques with web technologies to make research accessible and interactive."
      ],
      "metadata": {
        "id": "y_YITzOji8nm"
      }
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "SEBXOqKei_12"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}