@LANTIS: Language Technology for Information Services2022-08-24T19:09:57-04:00
Our lab is dedicated to research on NLP-driven techniques for science communication. We develop language technologies for convenient, accurate and equitable information services for scientists and the public.
PreCheck: Understanding Press Release Exaggeration of Scientific Research
The PreCheck project aims for detecting three common types of exaggerated claims in health-related press releases: causal claim from correlational findings, extrapolating animal study results to humans, and unwarranted health advice.
News2References: Link News to Research Literature
The News2Reference project aims to accurately link science news articles to the original research publications. Often missing in current news articles, the links can be used to measure the social impact of a research work or detect inaccuracies or distortions in science news.
Cora (Citation Opinion Retrieval and Analysis):
The CORA project aims to develop an automated tool that can plug into a full-text bibliographic database such as PubMed, extract the citation statements toward a cited article, separate substantial citations from perfunctory ones, and categorize substantial citation opinions by their purposes (e.g. comparison, critique, etc.), subjects (e.g. methods, results, etc.), and tones (e.g. positive, negative, and neutral). This tool will help find the most useful comments from large numbers of citations, and facilitate various downstream applications, such as literature review, citation bias detection, and research impact assessment.