Interpretable commodity recommendation method based on fine-grained data

Abstract

The invention discloses an interpretable commodity recommendation method based on fine-grained data. The method aims to overcome the problems that in the prior art, the commodity comment content is insufficiently utilized, the interpretability is weak, and the effect of a user on commodity recommendation is not fully utilized due to the emotional tendency displayed in comments. The method comprises the steps that 1, fine-grained data of the user on the commodity comments are obtained through a data processing module; 2, a recommendation chain establishing module generates a recommendation chain of the to-be-recommended commodity for the target user according to the fine-grained data; and 3, the recommendation sequence is interpreted by the recommendation generation module according to therecommendation chain.

Description

technical field [0001] The present invention relates to an explainable recommendation method in the field of product recommendation, more precisely, the present invention relates to an explainable product recommendation method based on fine-grained data. Background technique [0002] The recommendation system has been widely used as a method of information filtering, which provides users with more humanized services and brings considerable economic benefits to businesses. The explainability of the recommendation system is defined as explaining the working principle of the recommendation system, making the system transparent, allowing users to understand when the system makes mistakes (testability), helping users make fast and high-quality decisions (effectiveness), affecting or Persuade users to choose products (persuasion) and improve user acceptance of recommended products (satisfaction). [0003] At present, there are mainly four types of methods for implementing interpr...

AI Technical Summary

Problems solved by technology

[0005] 1. Insufficient use of data, failure to fully mine and utilize various information reflected by users in product reviews

[0006] The user's comments on the product will reflect the product aspect that the user cares about, and the verbs and descriptors used on the aspect express the emotional tendency of the aspect. Most algorithms do not use this aspect when considering recommendations

[0007] 2. Lack of attention to the user's emotional factors of the product when generating explanations

[0008] The user's emotion on the product, the specific content of his likes or dissatisfaction is an important basis for generating personalized recommendations, but the current algorithm for using this content is relatively limited

Images