Summary of Enhancing Cross-document Event Coreference Resolution by Discourse Structure and Semantic Information, By Qiang Gao et al.
Enhancing Cross-Document Event Coreference Resolution by Discourse Structure and Semantic Information
by Qiang Gao, Bobo Li, Zixiang Meng, Yunlong Li, Jun Zhou, Fei Li, Chong Teng, Donghong Ji
First submitted to arxiv on: 23 Jun 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed approach addresses limitations in existing cross-document event coreference resolution models by incorporating document-level information. A novel combination of Rhetorical Structure Theory (RST) trees and Lexical Chains is introduced to model structural and semantic information within documents. This framework enables the construction of heterogeneous graphs, which are then used to learn event representations via Graph Attention Networks (GAT). The similarity between event pairs is calculated using a pair scorer, allowing for the recognition of co-referenced events through clustering. To validate this approach, a large-scale Chinese cross-document event coreference dataset is developed, complementing existing English datasets. Experimental results demonstrate significant performance improvements over baselines on both English and Chinese datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper solves a problem in computer science by helping machines understand what’s happening in different documents about the same topic. Right now, computers are bad at recognizing when events in one document refer to the same event mentioned elsewhere. The new approach builds a special kind of tree that shows how ideas are connected within each document and then connects these trees across multiple documents. This allows computers to better understand long-distance relationships between events. To test this idea, researchers created a large dataset of Chinese texts about different topics, which is important because most existing datasets are only in English. |
Keywords
» Artificial intelligence » Attention » Clustering » Coreference
