hideaki-j / convsrch_literature_survey

Conversational search literature survey

0 stars 0 forks source link

🤔 Joint Detection and Coreference Resolution of Entities and Events with Document-level Context Aggregation (ACL21 SRW) #6

Open hideaki-j opened 2 years ago

hideaki-j commented 2 years ago

Joint Detection and Coreference Resolution of Entities and Events with Document-level Context Aggregation

Contribution summary

Kriman et al. proposed IE (including entity CR) model which handles full-document as a context,
to understand cross-sentence dependencies, and
achieved higher performance than baseline model, OneIE.

Authors

Samuel Kriman and Heng Ji (University of Illinois at Urbana-Champaign)

Motivation

Existing methods focus on extracting information from sentences or paragraphs, meaning cannot consider full-document context
To improve performance, Incorporating cross-sentence dependencies is needed

Method

Proposed IE model that can be used for IE subtasks, including entity CR (coreference resolution), at the document level
Input is a document and output are mention-entity and/or mention-events pairs
Steps
- Generate representation for the mentions, including pronouns
- ?

Results / Insight

Dataset: ACE05-E+, OIE dataset including pronoun mention annotations
Entity and event trigger extraction results
- Higher than baseline model of OneIE
Coref results
- F1 B³_sys: 84.83
- No baseline is set
  Questions
How do the representations are used to identify corresponding entities and events?

hideaki-j commented 2 years ago

Others

Code: https://github.com/sam1373/long_ie