Community Clustering: Leveraging an Academic Crowd to Form Coherent Conference Sessions.

摘要: Creating sessions of related papers for a large conference is complex and time-consuming task. Traditionally, few organizers group into manually. Organizers often fail to capture the affinities between beyond created sessions, making incoherent difficult fix alternative groupings hard discover. This paper proposes committeesourcing authorsourcing approaches session creation (a specific instance clustering constraint satisfaction) that tap expertise interest committee members authors identifying affinities. During planning ACM CHI'13, on human-computer interaction, we recruited using two online distributed methods. To refine these — evaluate methods against existing manual automated identify fit well in with their own. Results show found grouped by be as relevant as, or more than, suggested through in-person meeting. also demonstrate communitysourced results provide flexibility during scheduling.