Biomedical Literature Mining for Identifying Conclusive Finding Entities
Conclusive finding entities (CFEs) in a biomedical article a are specific entities on which conclusive findings in a are reported. Identification of the CFEs is essential for the analysis of highly related conclusive findings in biomedical literature. It can support various kinds of analysis tasks that biomedical scientists often routinely conduct. However, CFE identification is challenging as well, as it is difficult to identify the specific targets and then estimate how conclusive the findings on the targets are. In this project we plan to develop those techniques that, given candidate entities in the title and the abstract of a biomedical article a, identify CFEs in a. The techniques will be developed based on CFE-based association mining, with the following main idea: two candidate entities in an article may be more likely to be CFEs of the article, if a strong association between them is mined from a collection of articles. Based on the techniques developed, we also plan to implement a system that supports exploratory analysis of conclusive evidence by navigating through a network of CFEs. Results of the project are of technical and practical significance to the identification and exploration of highly related CFEs in biomedical literature.
Keywords: Conclusive finding entity, association mining, network navigation, biomedical literature.