Social opinion mining for supporting buyers’ complex decision making: exploratory user study and algorithm comparison

This article reports our study of the role of social content (i.e., user-generated content in social networking environment) in online consumers’ decision process when they search for an inexperienced product to buy. Through close observation of users’ objective behavior and interview of their reflective thoughts during an initial exploratory user study, we have first derived a set of system implications and integrated these implications into a three-stage system architecture. Furthermore, driven by the specific implication regarding the impact of user reviews in influencing users’ decision stages, we have presented a linear-chain conditional random-field-based social-opinion-mining algorithm, and have identified its higher effectiveness against related algorithms in an experiment. Finally, we present our system’s user interfaces and emphasize on how to display the opinion-mining results in the form of both quantitative presentation and qualitative visualization.