Research On Object Extraction Method Of Commodity Evaluation

Evaluation objects extraction as digging,according to the opinion of an important task,is the foundation of other steps,and also is the whole opinion dig according to priority.This article is based on association rules method to implement the evaluation object extraction,selected the taobao,amazon website commodity comments as a data source,in view of the comments made by real consumers,using Fp-Tree algorithm to extract goods evaluation objects.First of all,using the crawler software in taobao,amazon collected experimental data.Second,the collected data for word segmentation and part-of-speech tagging,syntactic analysis.And then by extracting the nouns and noun phrases as candidate evaluation objects.Finally,through the Fp-Tree algorithm to achieve the extraction of the final evaluation objects.After compares the results of the evaluation objects and artificial extraction,calculation accuracy and recall rate and F values,can come to the conclusion that to some extent can be extracted the commodity evaluation objects.