Identification of network behavioral characteristics of high-expertise users in interactive innovation: The case of forum autohome

Abstract

With the development of mobile internet, an increasing number of enterprises have made use of online forum, communities and other social media to closely cooperate with users for interactive innovation. Thereby, how to use users’ network behavioral data to identify those high-expertise users is an issue of great concern for enterprises’ identification of valuable lead users and implementation of products’ interactive innovation. In this paper, a questionnaire is conducted to divide automobile knowledge and ability into five grades. Following that, a 5-level automobile terminology word bank containing 560 terminologies is built based on statistical results of samples. According to the terminology word bank, the corpuses of 399 users are selected from the autohome.com to analyze and quantify their expertise. Finally, models based on the two-stage least square method is adopted for correlation analysis of users’ expertise and their network behavioral characteristics. Empirical results suggest that users with high expertise in the network generally demonstrate the following six characteristics: a high centrality in the social networks, a high frequency of replies, a high length of posts, a large number of post replies, a large number of fans, activeness of users in multiple subgroups. The main contribution of this study is twofold: propose a method to quantify the user’s expertise based on user’s corpus; provide a new perspective for finding high-value users in enterprise interaction innovation practice.