Order-preserving submatrix (OPSM) and frequent sequence mining based emotion classification method for e-commerce comments

Abstract

The invention discloses an order-preserving submatrix (OPSM) and frequent sequence mining based emotion classification method for e-commerce comments. The method comprises the following steps: (1) performing pretreatment and Chinese word segmentation on the e-commerce comments; calculating to obtain a TF-IDF weight vector of synonyms; and then mining a local mode in the weight vector through a biclustering algorithm based on OPSM; (3) mining classification frequent phrase characteristics through an improved PrefixSpan algorithm, and meanwhile, improving the capacity for distinguishing emotion tendency by the frequent phrases through limitation such as word intervals; and (4) converting the characteristics mined in steps (2) and (3) into a 0 / 1 vector to be used as an input of a classifier, and thus obtaining the emotion classification result of the e-commerce comments. With the adoption of the method, the emotion classification characteristics of the e-commerce comments can be accurately mined, so that potential customers can know the goods evaluation information before buying, and meanwhile, the businessman can fully know the suggestions of the customers and accordingly improve the service quality.

Description

technical field [0001] The invention belongs to the field of natural language processing technology and emotion computing, in particular to an e-commerce comment emotion classification method based on order-preserving sub-matrix and frequent sequence mining. Background technique [0002] With the development of e-commerce, it has become a part of daily life for users to evaluate online shopping products on e-commerce platforms. How to use machine learning and natural language processing technology to analyze comment texts and obtain opinion tendencies and emotional polarity has become an important research issue in the field of artificial intelligence. The techniques commonly used in text sentiment analysis are divided into rule-based methods and statistical-based methods. The rule-based method mainly starts from the perspective of linguistics, using artificial dictionaries and templates for sentiment analysis (Xu et al., 2008). Statistics-based methods start from the pers...

AI Technical Summary

Problems solved by technology

First of all, multiple words in online comments often express similar semantics, making the number of feature words in the corpus very large, so the feature vectors calculated by TF-IDF are very sparse, which affects the accuracy of sentiment analysis; secondly, TF-IDF is calculating The feature weight will be affected by the length of the sentence, and the network comments are long or short, resulting in emotional approximation and different lengths of sentences corresponding to the weights in the feature vector have different magnitudes; finally, the idea of ​​​​the TF-IDF algorithm is similar to the bag of words model ( Bag-of-Words), does not consider the word order information in the sentence, and the word order has an important impact on the semantics and emotional expression of the review text

Images