EDAHT: An Expertise Degree Analysis Model for Mass Comments in the E-Commerce System

摘要

In order to help consumers to retrieve the most valuable information from amount of comments quickly, we present a method of evaluating the expertise degree of comments. Firstly, we propose an algorithm to construct automatically an attribute-word hierarchy tree from the massive comments data. Secondly, we develop an expertise degree analysis based on attribute-word hierarchy tree (EDAHT) to estimate the expertise degree of comments. The experiments results on 8,000 manual scoring data show that EDAHT model is high consistent with the manual scoring data, and this novel model is effective.