In Online Inquiry-Based Learning (OIBL)learners search for information to answerdriving questions. While learners conductsequential related searches, the search enginesinterpret each query in isolation,and thus are unable to utilize task context.Consequently, learners usually get lessrelevant search results. We are developinga NLP-based search agent to bridge thegap between learners and search engines.Our algorithms utilize contextual featuresto provide user with search term suggestionsand results re-ranking. Our pilotstudy indicates that our method can effectivelyenhance the quality of OIBL.
展开▼