|
1 | | -欧洲语言资源协会（ELRA）和国际计算语言学委员会（ICCL）将联合主办 2024 年计算语言学、语言资源和评估联合国际会议（LREC-COLING 2024），该会议将于 2024 年 5 月 20-25 日在意大利都灵举行。 |
| 1 | +The European Language Resources Association (ELRA) and the International Council for Computational Linguistics (ICCL) are co-hosting the Joint International Conference on Computational Linguistics, Language Resources and Evaluation 2024 (LREC-COLING 2024), which will take place from 20-25 May 2024 in Turin, Italy. |
2 | 2 |
|
3 | | -此次联合会议将聚焦计算语言学、语音处理、多模态和自然语言处理领域的研究人员和从业者，重点关注评估和资源开发，以支持这些领域的工作。延续 COLING 和 LREC 的悠久传统，本次联合会议将突出重大挑战，提供口头报告和广泛的海报展示，为与会者提供充分交流的机会，并配以丰富的社交活动。 |
| 3 | +The joint conference will focus on researchers and practitioners in the fields of computational linguistics, speech processing, multimodality and natural language processing, with an emphasis on evaluation and resource development to support work in these areas. Continuing the long tradition of COLING and LREC, this joint conference will highlight major challenges, offer oral presentations and extensive poster presentations, and provide ample opportunities for participants to network, with a rich social programme. |
4 | 4 |
|
5 | | -此次上海大学自然语言处理与人机交互实验室（H!NTLAB）由李唐同学主要参与的工作1篇论文被LREC-COLING 2024接收。 |
6 | | -被录用论文的简要介绍如下： |
7 | | -Towards Human-Like Machine Comprehension: |
| 5 | +This time, one paper of the work of the Natural Language Processing and Human-Computer Interaction Laboratory (H!NTLAB) of Shanghai University, in which Tang Li was mainly involved, was accepted by LREC-COLING 2024. |
| 6 | +The brief introduction of the accepted paper is as follows: |
| 7 | +Towards Human-Like Machine Comprehension. |
8 | 8 | Few-Shot Relational Learning in Visually-Rich Documents |
9 | 9 |
|
10 | | -类型：Long Paper |
| 10 | +**Type:** |
| 11 | +Long Paper |
11 | 12 |
|
12 | | -作者：王昊 （讲师）、李唐 （2021级硕士研究生）等 |
| 13 | +**Author(s):** |
| 14 | +Hao Wang (Lecturer), Tang Li (M.S. Candidate, Class of 2021), et al. |
13 | 15 |
|
14 | | -简介：键值关系在视觉丰富文档（VRD）中很普遍，通常在不同的空间区域中描述，并伴有特定的颜色和字体样式。这些非文本是重要的特征，极大地增强了人类对这种关系三元组的理解。然而，当前的文档AI方法无法考虑与视觉和空间特征相关的这些有价值的先验信息，导致性能次优，尤其是在处理有限的示例时。为了解决这一局限性，我们的研究重点是少样本关系学习，特别是针对VRD中键值关系三元组的提取。鉴于缺乏适用于该任务的数据集，我们引入了两个新的基于现有监督基准数据集的少样本基准。此外，我们提出了一种结合关系二维空间先验和原型校正技术的变分方法。这种方法旨在生成关系表示以类似于人类感知的方式意识到空间上下文和看不见的关系。实验结果证明了我们提出的方法的有效性，展示了其优于现有方法的能力。这项研究也为实际应用开辟了新的可能性。123 |
| 16 | +**Introduction:** |
| 17 | +Key-value relationships are common in visually-rich documents (VRDs), which are usually described in different spatial regions accompanied by specific colours and font styles. These non-texts are important features that greatly enhance human understanding of this relational triad. However, current document AI approaches are unable to take into account this valuable a priori information related to visual and spatial features, resulting in sub-optimal performance, especially when dealing with a limited number of examples. To address this limitation, our research focuses on few-shot relational learning, especially for the extraction of key-value relation triples in VRD. In view of the lack of datasets suitable for this task, we have introduced two new few-shot datums based on the existing supervised datum datasets. In addition, we propose a variational method that combines relational two-dimensional space prior and prototype correction techniques. This approach aims to generate relational representations that are aware of spatial context and unseen relationships in a manner similar to human perception. Experimental results demonstrate the effectiveness of our proposed method and demonstrate its ability to outperform existing methods. This research also opens up new possibilities for practical applications. |
15 | 18 |
|
| 19 | +<div align="center"> |
| 20 | +<img src="../../assets/img/news/announcement_5_p1.png" width = "800" height = "500" alt="图1" align=center /> |
| 21 | +</div> |
| 22 | +<div style="text-align:center">Figure 1: Visual rich document few-shot dataset with multimodal features</div> |
| 23 | + |
| 24 | + |
| 25 | +<div align="center"> |
| 26 | +<img src="../../assets/img/news/announcement_5_p2.png" width = "800" height = "500" alt="图2" align=center /> |
| 27 | +</div> |
| 28 | +<div style="text-align:center">Figure 2: Diagram of the model architecture combining relational 2D space prior and prototype correction techniques</div> |
| 29 | +<div align="center"> |
| 30 | +<img src="../../assets/img/news/announcement_5_p3.png" width = "800" height = "500" alt="图3" align=center /> |
| 31 | +</div> |
| 32 | +<div style="text-align:center">Figure 3: Comparison of the performance of different methods on different datasets</div> |
0 commit comments