A Research On Social E-commerce Community Detection Based On Interest Graphs

The emergence of the social media which is based on web2.0 technology,has changed the people’s way of lifetremendously, while the impact on the traditional e-commerce cannot be ignored. A large number of social media content generated by UGC, truly reflects the idea of a user in real life and interpersonal networks, so that people online and offline life living close together, it has become a global Internet business model of Third Wave. More and more e-commerce enterprises make social media as its main channel to achieve marketing goals. Social media will be the electricity supplier product promotion, maintaining customer loyalty and achieve an important channel of value added. This emerging e-commerce model is called social commerce, social e-commerce in practice to carry out marketing relationships fully considered this element to the social as a starting point, not only in the shopping closer to reality, but at the same time be able to buy enough psychological needs of consumers.On the other hand, the emergence of new information technologies and the application model, making the world an unprecedented amount of data showing explosive growth, the era of big data has arrived.In this background, the most of data are created by the social media, how to deal with the challenges of traditional electricity suppliers in these two areas will determine its future survival and development.The traditional electricity providers tend to ignore the needs of the users’ interests,just simply tap the potential needs of users from shopping basket history, and there is no emphasis on social relationships in social media for user shopping behavior. On the other hand, expanding e-commerce has led to the massive scale of the goods and the emergence of numerous electricity providers, and how they need to get from the mass merchandise goods, how to identify the electricity supplier’s reputation, which has become the attendant problems restrict the development of e-commerce model bottleneck. Interest-based online community atlas found that consumers have the same interests, give full play to the role of social media social networking will become a user views a friend to buy goods as a major decision factor, can solve these problems.Based on the background big data and referring to a large number of literatures, the paper carries out research from the perspective of common interest community found, taking social e-commerce as the research object and revolving around social media network theory and the theory of data mining research. Through the webcrawler technology to capture a little of data about social media which have many users. Through the analysis of these data and finishing, it sets up the users’ interest graph of social e-commerce by using Gephi. Based on the interest graph, the paper makes use of R data mining tool to implement the K-means, as well as discover the users’ interest communities in the sample. Research result shows that in the social e-commerce, users’ interest can be expressed accurately, at the same time, the boundaries between different network communities which are based on the different interests is obvious. Finally, according to the result of this study, it put forward some corresponding countermeasures and suggestions to merchants and platforms.Features of this article:First, to begin with community discovery, the paper researches social media impact of the e-commerce model, and use data mining algorithms to improve, to build maps based on similar interests, find community-based network of consumer interest in this map. So that electricity suppliers and platforms can be a good use of social relationships between consumers and added value to achieve the maintenance of customer relationships, but also help consumers to obtain more accurate information when making online shopping.Second, the choice of open-source R plus Gephi framework as a research tool using java crawler to grabs data from social media to achieve a mutual integration of computer application and management of knowledge.