Internet Archive Scholar

Event Prominence Extraction Combining a Knowledge-Based Syntactic Parser and a BERT Classifier for Dutch

Thierry Desot, LT3, Language and Translation Technology Team Ghent University, 9000 Gent, Belgium, Orphée De Clercq, Veronique Hoste, LT3, Language and Translation Technology Team Ghent University, 9000 Gent, Belgium, LT3, Language and Translation Technology Team Ghent University, 9000 Gent, Belgium
2021 Proceedings of the Conference Recent Advances in Natural Language Processing - Deep Learning for Natural Language Processing Methods and Applications   unpublished

Preserved Fulltext

fulltext thumbnail
Web Archive Capture PDF (535.3 kB)
https://web.archive.org/web/20211114084005/https://acl-bg.org/proceedings/2021/RANLP%202021/pdf/2021.ranlp-1.40.pdf

A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL. The file type is application/pdf.

A core task in information extraction is event detection that identifies event triggers in sentences that are typically classified into event types. In this study an event is considered as the unit to measure diversity and similarity in news articles in the framework of a news recommendation system. Current typology-based event detection approaches fail to handle the variety of events expressed in real-world situations. To overcome this, we aim to perform event salience classification and
more » ... e whether a transformer model is capable of classifying new information into less and more general prominence classes. After comparing a Support Vector Machine (SVM) baseline and our transformer-based classifier performances on several event span formats, we conceived multi-word event spans as syntactic clauses. Those are fed into our prominence classifier which is fine-tuned on pretrained Dutch BERT word embeddings. On top of that we outperform a pipeline of a Conditional Random Field (CRF) approach to event-trigger word detection and the BERTbased classifier. To the best of our knowledge we present the first event extraction approach that combines an expert-based syntactic parser with a transformer-based classifier for Dutch.
doi:10.26615/978-954-452-072-4_040 fatcat:d6qcknoc5fe4bd5amleredp5yy
Citation

Thierry Desot, LT3, Language and Translation Technology Team Ghent University, 9000 Gent, Belgium, Orphée De Clercq, Veronique Hoste, LT3, Language and Translation Technology Team Ghent University, 9000 Gent, Belgium, LT3, Language and Translation Technology Team Ghent University, 9000 Gent, Belgium. "Event Prominence Extraction Combining a Knowledge-Based Syntactic Parser and a BERT Classifier for Dutch." Proceedings of the Conference Recent Advances in Natural Language Processing - Deep Learning for Natural Language Processing Methods and Applications (2021)