Research On Information Resource Management On Negative Word-of-Mouth Processing In Social Media

With the fast development of social media, the huge amount of user generated content (UGC) and complex social networks relationships have constituted huge amount of information resources. Consumers frequently post negative word of mouth (NWOM) with unpleasant experience towards sellers’products in social media, such that virtual economy was strongly shocked. Meanwhile, the huge amount of information resources leads to "information overload" phenomenon. Consumers, who complain products, cannot timely discover effective information resource to fix the problems that they encounter when enjoying products. In real world, consumers lose their trickiness day by day; such that social media face a strong challenge. Social media, as the owner and manager of huge amounts of information resources, can provide solutions by compensating complaining consumers with helpful resources, and realize the personalized recommendation processes. However, disordering, value-sparse and huge amount of UGC builds a strong difficulty to the information resources management.Extant research on consumer complaining behavior suggested that sellers could compensate consumers from three dimensions: emotion, money and information. However, little research on social media explores the information processes model to realize information compensation from the perspective of information resources management. Value co-creation theory is the new theory in the field of marketing and consumers relationship management. It emphasizes the processes that both sellers and consumers join in the product design and manufacturing together and achieve their own values simultaneously. The key is sharing resources and obtains values in co-creation processes through information exchanging to satisfy their value demands separately. However, previous research almost focuses on the co-creation actions in traditional context. And few academics explore value co-creation by concentrating on information resources in social media.This study has summarized existing research in mainland and abroad on value co-creation, UGC processes, helpfulness, personalized recommendation and other fields. On the basis of value co-creation, I focus on multi-sided markets characters of information resource management, explore the processes model on disordering, value-sparse and huge amounts of UGC and complex social networks relationships, and further build the effective personalized recommendation processes. In particular, this paper explore the following aspects:the effect of co-creation, UGC ordering model, helpful UGC identification and personalized recommendation.1) On the basis of the multi-sides market characters of information resource management of the explored issues, this study defines the social media, value co-creation demand of multi players, characters of NWOM and characters of information resources, proposes the information resource management processes by focusing on physical mapping and logistic mapping, realizes the value co-creation processes from extant resources and latent resources. At last, this paper constructs a NWOM-centric information management framework against disordering, value-sparse and huge amounts of UGC.2) On the basis of value co-creation strategies in social media, this study uses linear threshold model to capture the interactions among consumers, use susceptible-infected-recovered(SIR) model to capture consumers’perception status during NWOM diffusion processes, designs value co-creation model of diffusion processes based on the assumption to consumer behavior from value co-creation theory. It further uses the real-world dataset, explores the NWOM diffusion processes in social media, validates the effect of the heterogeneity of consumers, and checks the effect of value co-creation strategies.3) In allusion to the physical mapping of information resources in social media, this paper describes the disordering characters, which is manifested in unstructured, fragmented and decentralized feature of UGC. Based on the proposed information resource management framework, this study explores the mapping processes from UGC to extant processes, designs structuralization model, converging model and centralization model to fix this problem, and defines the output formalization of these three processing model.4) In allusion to the logistic mapping of information resources in social media, this study explores the method to fix the problem of value sparsity, and defines the logistic mapping processes as the value identification processes. The value of information resources in social media is defined as the helpfulness. The author collects comments as datasets, preprocesses comments as suggested by physical mapping model, builds a brand-centric product knowledge base, extracts review length, usage and comparison strength from comments, empirically investigates the critical factor to review helpfulness by constructing a logistic regression model from the perspective of econometric analysis.5) Considering the features of NWOM, this study designs a personalized recommendation system, which is keyword centric. Our proposed framework can recommend the demanded information to someone needing it by integrating information retrieval technology. The recommendation system in social media chooses collaborative filtering recommendation model as benchmark model, integrates social networks together, measures consumes’ preference to keywords better based on information diffusion processes. Social networks structure is used to adjust consumers’ preference. In this research, keywords, which are extracted from physical mapping processes, are treated as the recommended objects. It helps to fix the data sparsity and the problem of cold-start. The author collects datasets from Sina Weibo, uses mean absolute error, precision and recall as evaluation index and validates the effectiveness of the proposes personalized recommendation model.At last, the author summarizes the research in this dessert, analyze the limitation and points out the possible direction of future research.