Study On Exploration Of Reputation Dimensions Based On Text-Clustering And Corpus

Most existing online retail sites commonly used some simple online reputation systems to try to solve the lack of trust in online transactions. Though played a role, these online reputation systems still have some problems, for example, mostly there are different products share the same evaluation dimensions; evaluation dimensions cannot discriminate clearly; reputation systems cannot accurately express the true customer choices. To solve these problems, we shall study the text comments from the users themselves, then use an objective clustering method to explore the true customer choices, and analyze online reputation systems qualitatively. Finally we want to get a reputation dimension system that can shows accurately what users concerned about.In this paper, after a review of previous studies, we analyze text comments with the method of clustering analysis. We code a asp.net program to grab the original data, then analyze and cluster the data with Visual C++, Java and Matlab. After the process of grabbing text reviews data, words segmentation, the generation of characteristics collection, code marking to determine the semantic similarity calculation, machine clustering and cluster analysis, we get a more reasonable reputation dimension. Based on the specific circumstances, this paper improved the specific steps to caculate the TFIDF, applied semantic similarity to clustering analysis, and improved the DBSCAN method with a process of filtering with TFIDF. After a comparision of the original reputation dimensions and the new dimensions from clustering analysis, we find that the old dimensions missed some important dimensions and had some dimensions incomplete and unclearly, but the dimensions from clustering analysis show us a better result.As a new method to establish the reputation dimensions system, clustering analysis method can minimize human intervention and the interference of the subjective impact on the study, and ensure that we can get an objective reputation dimensions system from the user comments scientificly and rationally.