Searching for experts in a context-aware recommendation network. (October 2015)

• Record Type: Journal Article
• Authors: Carchiolo, Vincenza; Longheu, Alessandro; Malgeri, Michele; Mangioni, Giuseppe
• Is Part Of: Computers in human behavior. Volume 51:Part B(2015)
• Page Start: 1086
• Page End: 1091
• Publication Date: 2015-10
• Subjects: Expertise -- Recommendation network -- Trust -- Context-aware; Interactive computer systems -- Periodicals; Man-machine systems -- Periodicals; 004.019
• DOI: 10.1016/j.chb.2015.03.028
• Languages: English
• ISSNs: 0747-5632

Abstract

Highlights: We apply expertise to improve the quality of recommendation systems. We consider how context-aware recommender systems can be used in learning context. Our approach save resources since it reduces the number of nodes to query. Abstract: The huge amount of data in the web made and is still making harder the issue of finding the right information. To help users in their choices, recommender systems are used as a valuable tool when dealing with innumerable choices of data, products and services. In this work, expertise is used to improve the quality of recommendations by selecting those provided by users that are considered expert in the same context their recommendations are about, since we believe they are more relevant with respect to recommendation coming from non-expert users. We present an approach of searching for a "guru" user (expert in a specific context) using context-dependent expertise information within the Epinions.com recommendation network, also considering how this can be exploited within technology enhanced learning context. Results show that context-based search can be used to significantly reduce the number of nodes (users) to query with a limited loss of expert nodes.