SARVE-2: Exploiting Social Venue Recommendation in the Context of Smart Conferences

Globally, the superfluity of scholarly research conferences in varying disciplines has introduced the issue of scholarly big data and information overload related to both research papers and conference proceedings/sessions. This evident scholarly expansion in different disciplines has increased the collaborative importance of conferences. Consequently, the problem regarding attendees selecting the right conference session(s) to attend in academic conferences requires further and urgent attention. Using a smart conference scenario, this paper aims to address the problem above by proposing an improved venue recommender algorithm called Socially-Aware Recommendation of Venues and Environments-2 (SARVE-2). Using a closeness centrality approach, SARVE-2 initially employs Breadth First Search (BFS) and Depth First Search (DFS) strategies to search for relevant presenters for a target attendee. Then, the tie strength of the (searched) presenter and target attendee is computed to generate reliable social (conference session) recommendations for the target attendee. Through the utilization of a relevant (real-world) dataset, our benchmark experiments reveal that, in comparison with other contemporary methods, SARVE-2 exhibits better performance in terms of effective social recommendation search, as well as social recommendation quality, coverage and accuracy.