14.11.2024New Paper in ISWC 2024!

Last week, Rosni Vasu presented the paper “SciHyp: A Fine-grained Dataset Describing Hypotheses and Their Components from Scientific Articles” at the 23rd International Semantic Web Conference, held in Maryland from November 11–15. Together with Cristina Sarasua and Abraham Bernstein, the work introduces SciHyp, a novel dataset containing RDF descriptions of 689 hypothesis from 479 computer science articles. Using a multi-step annotation pipeline with expert annotation, Language Models like BERT and Sci-BERT, and GPT-4 for hypothesis component extraction, SciHyp aims to benefit the scientific community by providing a structured scientific hypotheses dataset for model training and evaluation. The preprint is available here.

Excitingly, the paper received a nomination for the Best Resource Paper Award, and Rosni received a Student Travel Grant.

Congratulations to the authors!