Na Li had an accepted paper in WWW2023 conference


A computational notebook search system called CNSVRE has been proposed to address shortcomings in existing solutions for Virtual Research Environments (VREs). Existing approaches lack focus on the specific information needs of scientific researchers and struggle with relevance evaluation due to the mixed nature of text and code in computational notebooks. CNSVRE incorporates scientific query reformulation and computational notebook summarization to better cater to researchers’ specialized and complex information needs. The system aims to improve effectiveness, efficiency, and user satisfaction in exploratory data analysis and algorithm prototyping tasks within VREs. A user study is conducted to validate the effectiveness of CNSVRE in meeting these objectives.

This work was partially funded by the European Union Horizon 2020 and Horizon Europe research and innovation programs through projects CLARIFY (860627), ENVRI-FAIR (824068), BlueCloud (862409), Blue-Cloud 2026 (101094227) and the LifeWatch ERIC.