Parliament_Browser is an application designed to analyze parliamentary protocols from the German Bundestag using Natural Language Processing (NLP) techniques.
Key features include:
- Data ingestion: Scraping and processing XML protocol files.
- Custom storage: Extracted data is structured and stored in a MongoDB database.
- Web interface: A Java Spark Webserver provides an accessible web application with navigation and menus.
- Sentiment analysis: NLP models analyze speeches, discussions, and topics, providing insights into sentiment (positive, neutral, negative).
The sentiment analysis pipeline leverages standard NLP techniques such as tokenization, part-of-speech tagging, and sentiment classification. Results are stored alongside protocol data, enabling sentiment-based queries and exploration.
To start the application, simply run the main method (no arguments required).
The webserver will launch, and you can access the interface at:
http://localhost:4567- Or from another device in the same network using the server’s IP and port
4567.
- Webserver: Java Version 1.8+
- Database: MongoDB Server Version 4.4.0+
- Browser Support: Safari 15.4+, Chrome 98+, Edge 98+, Firefox 94+, Opera 84+
- A Minimum of 4 GB RAM
- Broadband internet connection
- 200 MB of Free Disk Space
- Kenan Khauto
- Maximilian Chen
- Simon Schütt
- Stanley Mathew
- Tim König
This project is licensed under the MIT License.