Start Date

12-16-2013

Description

With the rapid proliferation of information technology, the increasing amount of information available has posted significant challenges on relevant information discovery for the users. An alternative way is to find an expert with specific expertise. Expert recommendation is important in variety of contexts ranging from industry to academia. Information retrieval methods or graph-based methods have been proposed to approach this problem in previous research while some important contextual factors are ignored. In this paper, considering the factors of topic relevance, expert quality, and researcher connectivity, we propose a novel researcher modeling approach to recommend experts in scientific communities. The proposed method has been implemented in ScholarMate (www.scholarmate.com), an online research social network platform. The preliminary experiment is conducted and the evaluation results exhibit that the proposed method is more effective than baseline methods.

Share

COinS
 
Dec 16th, 12:00 AM

Finding an expert: a model recommendation system

With the rapid proliferation of information technology, the increasing amount of information available has posted significant challenges on relevant information discovery for the users. An alternative way is to find an expert with specific expertise. Expert recommendation is important in variety of contexts ranging from industry to academia. Information retrieval methods or graph-based methods have been proposed to approach this problem in previous research while some important contextual factors are ignored. In this paper, considering the factors of topic relevance, expert quality, and researcher connectivity, we propose a novel researcher modeling approach to recommend experts in scientific communities. The proposed method has been implemented in ScholarMate (www.scholarmate.com), an online research social network platform. The preliminary experiment is conducted and the evaluation results exhibit that the proposed method is more effective than baseline methods.