论文部分内容阅读
Reviews of electronic products are important resources for manufacturers to gather feedbacks from customers.Meanwhile,theyre acted as an crucial reference for customers to make quality assessments.How to effectively classify the reviews of electronic products is an open issue.Based on the two-value feature function,a maximum entropy classification of electronic products reviews is proposed to avoid the noise caused by the term frequency based feature function.For the singleness of the topic in reviews,a text classifier integrated with the features of descriptive anaphor is provided to enhance the models fitting degree.Furthermore,how to identify the descriptive anaphor is addressed.A library for feature resources is built by the if-idf algorithm to implement the location of the description,and the descriptive anaphors are identified by means of the minimum IP tree of the parsing tree.The experiment shows the proposed classifier has the relatively high stability and accuracy,which makes the reviews much more usable for the product analysis.