Tuesday, May 7 4:30-5:55 PM
Room: Regency Ballroom D (West Tower, Ballroom/Gold Level)
In this session, we will uncover how machine learning algorithms can be used to transform the way information professionals approach large, complex searches. When faced with a research question that requires a comprehensive search likely to generate thousands or tens of thousands of results, machine learning can be used as a tool in place of limiters that are often artificially imposed simply to limit results (e.g., date limits, additional keyword sets). After providing an overview of basic concepts, we will demonstrate how we have used machine learning to refine search strategies, deduplicate search results, and reduce the volume of search results that must be screened manually. We will also facilitate small-group discussions of barriers to implementing AI-enabled searches at participants’ institutions and provide validation data that demonstrates the efficacy and efficiency of these approaches.
Station 1: Keyword Analysis Tool (KAT)
Station 2: AI-Enabled DeDuplication
Station 3: Feedback
Station 4: Clustering
Station 5: Supervised Clustering
Station 6: Machine Learning