User interests analysis and its application on the linked brain data platform

Abstract

User-centric academic resource service platforms are essential for scientific researchers, since hundreds of new literatures and data sets are produced every single day, and no one can keep him/herself most up-to-date and handle these resources to find, read and understand the most relevant ones completely manually. In this paper, we introduce some user interests analysis methods and apply them to build personalized recommendation services as a user-centric sub system for the Linked Brain Data (LBD) platform, which is an integrated data and knowledge platform for users, especially Neuroscientists and Artificial Intelligence researchers, to explore and better understand the brain and support their research. For interests analysis, we obtain user related data from relatively static data sources (e.g. user profiles maintained by uses), and more dynamic resources (e.g. publications and online social network contents generated by users, which are with chronological information). For recommendation service, we automatically recommend extracted knowledge in the brain association graph and related articles based on the understanding of research interests of the LBD platform users. Through use case studies, we illustrate the importance and potential value of user-centric services for brain and neuroscience related research.