Abstract
In this paper, we propose a new unsupervised method of extracting events from biomedical literature, which uses the score measures of events and patterns having reciprocal effects on each other. We, first, generate candidate events by performing linguistic preprocessing and utilizing basic event pattern information, and then extract reliable events based on the event score which is estimated by using co-occurrence information of candidate event's arguments and pattern score. Unlike the previous approaches, the proposed approach does not require a huge number of rules and manually constructed training corpora. Experimental results on GENIA corpora show that the proposed method can achieve high recall (69.7%) as well as high precision (90.3%).
Original language | English |
---|---|
Pages (from-to) | 777-786 |
Number of pages | 10 |
Journal | Lecture Notes in Artificial Intelligence (Subseries of Lecture Notes in Computer Science) |
Volume | 3248 |
DOIs | |
Publication status | Published - 2005 |
Event | First International Joint Conference on Natural Language Processing - IJCNLP 2004 - Hainan Island, China Duration: 2004 Mar 22 → 2004 Mar 24 |
ASJC Scopus subject areas
- Theoretical Computer Science
- Computer Science(all)