Using Large Language Models for Hypotheses and Claims Extraction from Scientific Literature
Year: 2025
Type: conference-paper
Abstract: We present a novel dataset for hypothesis and claim extraction from research papers in the interdisciplinary domain of social infertility. We use OpenAI's o1-mini reasoning LLM to create the dataset by extracting hypotheses and claims, which are then manually validated. Then we also use smaller closed and open-weight LLMs to create a benchmark for hypothesis and claim extraction.
Authors: Sagar Uprety, Boyana Buyuklieva, Prayag Tiwari
Institutions: University College London, Halmstad University
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4411549679