Event Extraction in pieces: tackling the partial event identification problem on unseen corpora

Sophia Ananiadou, Cohen Kevin (Editor), Demner-Fushman Dina (Editor), Sophia Ananiadou (Editor), Tsujii Jun-ichi (Editor)

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


    Biomedical event extraction systems have the potential to provide a reliable means of enhancing knowledge resources and mining the scientific literature. However, toachieve this goal, it is necessary that current event extraction models are improved, such that they can be applied confidently to unseen data with a minimal rate of error. Motivated by this requirement, this work targets a particular type of error, namely partial events, where an event is missing one or more arguments. Specifically, we attempt to improve the performance of a state-of-the-art event extraction tool, EventMine, when applied to a new cancer pathway curation corpus. We propose a post-processing ranking approach based on relaxed constraints, in order to reconsider the candidate arguments for each event trigger, and suggest possible new arguments.The proposed methodology, applicable to the output of any event extractionsystem, achieves an improvement in argument recall of 2%-4% when appliedto EventMine output, and thus constitutes a promising direction for further developments.
    Original languageEnglish
    Title of host publicationBiomedical Natural Language Processing
    EditorsCohen Kevin, Demner-Fushman Dina, Ananiadou Sophia, Tsujii Jun-ichi
    PublisherAssociation for Computational Linguistics
    Number of pages11
    Publication statusPublished - 31 Jul 2015
    EventACL, BioNLP - Beijing, China
    Duration: 26 Jul 201531 Aug 2015


    ConferenceACL, BioNLP
    CityBeijing, China


    • Event extraction, biomedical text mining


    Dive into the research topics of 'Event Extraction in pieces: tackling the partial event identification problem on unseen corpora'. Together they form a unique fingerprint.

    Cite this