Discovering target groups in social networking sites: An effective method for maximizing joint influential power
Highlights
► Effectively discover the most influential users from social networking sites (SNS). ► Computational models for discovering user groups with max joint influential power. ► The empirical evaluation shows the validity of the method.
Introduction
In this Web 2.0 era, users are able to express their opinions on products through many channels such as online forums, shopping websites, blogs, and wikis. These opinions can influence other users’ buying decisions and their views on companies (Cheung et al., 2009, Chevalier and Mayzlin, 2006, Dellarocas, 2003, Hennig-Thurau and Walsh, 2003, Koh et al., 2010, Mayzlin, 2006, Park et al., 2007). Recently, the emerging channel of social networking sites (SNS), such as Facebook, Twitter, and Epinions, has attracted the attention of marketing practitioners and researchers. These sites not only permit users to express comments and opinions on products, people, organizations, and many other entities, but also enables users to build various social relationships. For example, on the Epinions site (Epinions 2010), a user can build a trust relationship with another by adding him or her to a trust list, or the user can block him or her with a block list. The site then shows the trusted users’ opinions at the top of the list. With these social relationships, opinions will have greater impact on users than those expressed on the other channels (such as shopping websites) because people always believe or accept more easily the opinions of those with whom they have social relationships (Golbeck, 2005, Lu et al., 2010, Massa and Avesani, 2007). In addition, the influence of opinions on SNS can be disseminated more widely and quickly than that of other channels. Thus, some users’ opinions captured on SNS can greatly influence other users’ buying decisions or their views on certain companies.
Many business entities have recently come to recognize this phenomenon, and some companies have begun to identify certain users of SNS to conduct online marketing and reputation management (Conlin and MacMillan 2010, Marks 2010, Miller and Dickson 2001) in e-commerce and e-business. For companies to better utilize SNS for cost-effective, targeted marketing and reputation management, they must address an important question, given the huge number of social network users and companies’ limited budgets. That brings us to the question of which users’ opinions will most influence others’ actions. If the most influential group of users could be identified, companies could consume minimal resources to improve product sales and enhance their reputations.
Although there are many existing studies on measuring node importance in social network analysis (Wasserman and Faust 1994), as well as studies that explore the spread of influence in social networks (Kempe et al., 2003, Kempe et al., 2005), these works emphasize the importance of each node, without considering the joint influential power of a group of nodes. According to the latest findings from marketing research (Katona et al., 2007, Katona et al., 2011), if the customers are provided with positive information on products or enterprises by all related users in online communities, there may be a higher probability of customers purchasing such products or having positive perceptions of these enterprises. This is known as the joint influential power of a group of users. Previous research also indicates that the joint influential power of a small group of users could have considerable impact on a large number of consumers (Domingos and Richardson, 2001, Richardson and Domingos, 2002). Therefore, marketing personnel should identify the users who have great joint influential power on SNS, and find ways to encourage these users to express positive opinions about companies and the companies’ products through the strategy of targeted marketing. As a result, companies could maximally promote product sales and improve enterprises’ reputations through the joint influential power of the specific group of users.
Effectively discovering the group of users with maximal joint influential power from the huge number of users on SNS has become a key issue for companies to conduct targeted marketing and reputation management. Although previous research has examined the problem of discovering a group of influential users, the heuristic method used does not address the issue of identifying a group with maximal joint influential power (Zhang et al. 2008). One of the weaknesses of the previous research is that the users are added into the target group one by one, according to their attributes and parts of trust relationships between them, without considering the influential power of the target group as a whole. Thus, these method usually do not discover target groups having the most joint influential power (Kempe et al. 2003). In contrast, our proposed method represents global influential relationships among all users as a directed graph, and uses mathematical programming as the computational apparatus to discover the group of users with maximal joint influential power. Considering the cost of marketing, the proposed models can discover the target group with flexible costs. The empirical evaluation with real data from Epinions and Twitter websites shows that the proposed method shows much better performances, compared to the benchmark methods. In summary, the main contribution of our research is the development of a novel method that discovers the user group with maximal joint influential power in a cost-effective way; it also overcomes the disadvantage of existing methods (Zhang et al. 2008) which only consider the attributes of users and parts of influence relationship. Our research opens the door to apply widely available data captured on SNS to conduct targeted marketing and enterprise reputation management in e-commerce and e-business.
The rest of the paper is organized as follows. Section 2 discusses related research. Section 3 provides an overview of the proposed method for cost-effective targeted marketing and enterprise reputation management. Section 4 presents the computational models for discovering the most influential groups of users. Section 5 reports the experimental evaluation of the proposed method, and Section 6 summarizes our research work and discusses directions for future work.
Section snippets
Target group
A few studies exist that involve discovering a target group, and these bear some similarity to this study. In (Zhang et al. 2008), an algorithm based on the trust relationships between users is proposed to detect the influential target groups. The users are ranked according to the number of their written reviews, and then the users and the trust relationships are sequentially added into the target group until the clustering coefficient of the target group is less than a threshold. However, the
Influence network
On SNS, the influence relationships exist only among some users, and the influence strengths vary between different pairs of users. In addition, opposed to the friendship relationship, the influence relationship is asymmetric. Here, the influence network of users is proposed to describe the influence relationships among users on SNS represented as a directed graph. A simple influence network is shown in Fig. 1.
The influence network can be represented as G = (V, E, W), and includes the following
Computational models for discovering the most influential groups
Here, we assume that the users of the target group have bought products or hold positive opinions on products through some marketing strategies, and that these users can influence others’ buying decisions. Hence, the joint influential power of one group is defined as the sum of the influence strengths of all users in this group on other users, without considering the inter-influence between them (also, the target group size is much smaller than the number of other users, and thus the
Experimental evaluation
In order to evaluate the effectiveness of the proposed methods, two experimental evaluations using two real datasets were conducted: The first evaluation adopts the dataset from the Epinions website, and the proposed Basic-MIG, CS-MIG, and CC-MIG models are compared with two benchmark methods; the second evaluation adopts the dataset from the Twitter website, and the performance of the proposed models is evaluated based on the large-scale dataset with sparse influence relationships.
Conclusions and future work
The latest marketing research shows that the joint influential power of a group of users is a very important factor influencing other users’ buying decisions and their views on companies. A novel method is proposed in this study for helping companies to discover the target group with the greatest joint influential power. First, the influence relationships are mined from SNS, and represented as a directed graph. Next, the proposed mathematics programming models, which can be transformed to SDP
Acknowledgments
This study was partially funded by Research Fund for Young Teachers provided by School of Business, Nanjing University, the National Science Foundation of China Grant (71101037, 70890082, 90924015) and China Postdoctoral Science Foundation Grant (2011M500679). The authors highly appreciate the time and effort of the volunteer respondents. We would like to thank the Editor-in-Chief, Prof. Robert Kauffman, the special issue Editors and three anonymous reviewers for their insightful comments and
References (57)
- et al.
Do online reviews reflect a product’s true perceived quality? An investigation of online movie reviews across cultures
Electronic Commerce Research and Applications
(2010) - et al.
Identifying influential reviewers for word-of-mouth marketing
Electronic Commerce Research and Applications
(2010) - et al.
From virtual community members to C2C e-commerce buyers: trust in virtual communities and its effect on consumers’ purchase intention
Electronic Commerce Research and Applications
(2010) - et al.
The architecture of complex weighted networks
Proceedings of the National Academy of Sciences
(2004) - Ben-Zwi, O., Hermelin, D., Lokshtanov, D., and Newman, I. An exact almost optimal algorithm for target set selection in...
- et al.
Opinion mining and sentiment analysis
Foundations and Trends in Information Retrieval
(2008) - Carnes, T., Nagarajan, G., Wild, S., and van Zuylen, A. Maximizing Influence in a competitive social network: A...
- Chen, Y., Ni, C., Ma, R., and Xiong, Z. Mining core group from network community. In Proceedings of 4th International...
- Chen, W., Wang, Y., and Yang, S. Efficient influence maximization in social networks. In Proceedings of the 15th ACM...
- et al.
Credibility of electronic word-of-mouth: informational and normative determinants of on-line consumer recommendations
International Journal of Electronic Commerce
(2009)
The effect of word of mouth on sales: online book reviews
Journal of Marketing Research
The digitization of word of mouth: promise and challenges of online feedback mechanisms
Management Science
Indirect social influence
Science
Self-organization and identification of web communities
Computer
Graph clustering and minimum cut trees
Internet Mathematics
Maximizing influence propagation in networks with community structure
Physical Review E
Community structure in social and biological networks
Proceedings of the National Academy of Sciences of the United States of America
Semidefinite programming in combinatorial optimization
Mathematical Programming
Improved approximation algorithms for maximum cut and satisfiability problems using semidefinite programming
Journal of the ACM
Electronic word-of-mouth: motives for and consequences of reading customer articulations on the internet
International Journal of Electronic Commerce
Cited by (40)
How to repair public trust effectively: Research on enterprise online public opinion crisis response
2021, Electronic Commerce Research and ApplicationsThe dynamic effect of interactivity on customer engagement behavior through tie strength: Evidence from live streaming commerce platforms
2021, International Journal of Information ManagementCitation Excerpt :Subsequently, Levin and Cross (2004) showed that knowledge received from strong ties is perceived as more useful than that from weak ties. Strong ties also lead to frequent WOM recommendations that are consistent with the needs of the recipient (Dong & Wang, 2018; Koo, 2016; Xu, Guo, Li, Lau, & Liao, 2012). All these non-transactional behaviors are the key to the formation of customer engagement.
Structural correlation between communities and core-periphery structures in social networks: Evidence from Twitter data
2018, Expert Systems with ApplicationsCitation Excerpt :It has been a major channel for media, brands, individuals and governments for communication and marketing. Since people tend to give more credit to the information from their social connections (Xu, Guo, Li, Lau, & Liao, 2012), identifying influential users in a community is critical to understand how messages are spread across social media(Yang, Qiao, Liu, Ma, & Li, 2016). The rest of this paper is organized as follows.
Understanding the determinants of social network sites adoption at firm level: A mixed methodology approach
2016, Electronic Commerce Research and ApplicationsCitation Excerpt :During the last decade not only has the number of SNS increased, but also their complexity and number of registered users, which currently is in the hundreds of millions. This massive adoption has been encouraging several types of actions from entities belonging to almost every sector of society and transforming SNS into a hot topic and a venture for new business opportunities (Fang et al., 2013; Kane et al., 2012; Xu et al., 2012). By analysing today’s business environment, one can easily perceive that it is becoming increasingly complex and more technological.
LCP segmentation: A framework for evaluation of user engagement in online social networks
2015, Computers in Human BehaviorCitation Excerpt :However, there are scant researches developing this method in scope of brand fan pages of online social networks specially when we look at customers as users of a fan page (Jahn & Kunz, 2012). Despite several researches proposing methods to identify key customers or user groups in social networking sites based on communication network of users by using social network analysis (Wang, Ting, & Wu, 2013; Xu, Guo, Li, Lau, & Liao, 2012; Xu, Li, & Song, 2012), there is still lack of a proper segmentation method for identifying user groups in fan pages. Social media are promising for targeting influencers, enabling the cost-effective delivery of personalized messages and engaging with many customer segments in a distinguished way.