An impact of online recommendation network on demand

作者：Leem Byunghak; Chun Heuiju^*

来源：Expert Systems with Applications, 2014, 41(4): 1723-1729.

DOI：10.1016/j.eswa.2013.08.071

• Social Network Analysis (SNA)
• Centrality
• Clustering coefficient
• PageRank

This study proposes roles of online recommendation network on online book marketplace with social network perspective. Also this paper is to provide a new perspective on how Social Network Analysis (SNA) can be used to study the influence on demand by the recommendation network. We first built the books' recommendation network based on the co-purchasing data of customers and then computed the five network centralities and clustering coefficients using NodeXL. We also analyzed our research model by correlation and multiple regression analysis. The results of correlation analysis show a significant correlation between the six SNA measures - degree centrality, closeness centrality, betweenness centrality, eigenvector centrality, PageRank centrality, and clustering coefficient-and demand, as well as the six SNA measures. The result of regression analysis demonstrates that five of six SNA measures in the recommendation network have a significant effect towards demand, and then the largest effect towards a book's demand is associated with degree centrality. According to our results, a book's position within this network affects its overall demand. Hence managers should build more effective and accurate recommendation network of books including data customers co-purchased, and in turn, position books with a high degree of centrality on the website.

• 出版日期2014-3