Fine-grained attribute analysis method in e-commerce comment analysis scene

Abstract

The invention belongs to the technical field of natural language processing, and discloses a fine-grained attribute analysis method in an e-commerce comment analysis scene, which comprises the following steps of: respectively extracting target attributes and emotions of e-commerce comments by using two isomorphic and independently trained sequence labeling models; using a relationship classification model to perform relationship classification on the target attributes and emotion extracted by the model in the step 1; wherein the training data of the model is carried out by using an efficient labeling strategy, and the data labeled by using the strategy is used for training the models in the step 1 and the step 2; the rule matching relation module is used for integrating target attributes and emotions output by all the models; and the post-processing module is used for integrating the output of the rule matching relationship module and the output of the relationship classification model. Through the solution provided by the invention, on the premise of ensuring the accuracy of the fine-grained emotion prediction result, the complexity of the fine-grained emotion annotation task is simplified, and the annotation efficiency is improved.

Description

technical field [0001] The invention belongs to the technical field of natural language processing, and in particular relates to a fine-grained attribute analysis method in an e-commerce comment analysis scenario. Background technique [0002] At present, the existing technologies commonly used in the industry are as follows: With the popularization of the Internet, the logistics industry is becoming more and more perfect, and the e-commerce industry is booming. More and more users choose to purchase various daily necessities, even food and electronic products through online shopping. . Compared with offline channels, e-commerce channels provide a fast and convenient product feedback mechanism - user reviews. E-commerce user reviews have built a high-speed channel connecting users and brands. There are hundreds of thousands or even millions of user reviews for a single product, which undoubtedly has extremely high value for brand and product improvement. Such a huge amount...

AI Technical Summary

Problems solved by technology

[0004] To sum up, the problem existing in the existing technology is: in the field of machine learning, the acquisition cost of labeled data is high

Images