Research On Multi Dimensional Comprehensive E-commerce Recommendation Method

With the rise of e-commerce,various e-commerce online shopping platforms have also emerged,making more and more consumers' shopping methods shift from offline to online.Online shopping methods are favored by consumers because of the convenience of the shopping process,the richness of e-commerce platforms and the richness of products.However,in the face of various e-commerce platforms and the massive amount of products on the platforms,consumers have difficulty making choices.In order to solve the above problems,this article first conducts a hierarchical sentiment analysis of the e-commerce review text.On this basis,combined with the analysis of the indicators of different dimensions of the e-commerce platform,a multi-dimensional comprehensive e-commerce recommendation system is realized to provide consumers with The e-commerce platform and the recommended solutions for the corresponding products in the platform,the following is the research content of this thesis:(1)Aiming at the problem that it is difficult to distinguish the multi-polar emotions of the product attributes in the e-commerce review content,this thesis constructs a hierarchical labeling model to perform label decomposition and sentiment analysis on the reviews.First,use the pruning method to extract aspect words from the review text,complete the construction of the secondary tags,and cluster the secondary tags,and then explain the weight calculation of the tags at all levels,and finally use the constructed hierarchical tag model to analyze the Extraction and sentiment analysis of each product attribute label included in the business review.The experimental results on the public data set show that the hierarchical labeling model constructed in this thesis can effectively improve the sentiment classification effect of comments.(2)Through the cluster analysis of the indicator data of each dimension of the e-commerce platform,the characteristics of the various e-commerce platforms are explained and explained,and the indicator data of the different dimensions of the e-commerce platform is weighted and calculated to obtain the ranking results of the e-commerce platform,and combined The hierarchical label model constructed in this thesis analyzes the sentiment of the commodity attributes contained in each e-commerce platform,calculates the corresponding sentiment recommendation index,and obtains the corresponding product ranking recommendation in the platform.Experiments show that the method of combining e-commerce platform indicators and the tag sentiment contained in comments to recommend platforms and products has a good effect,and it can also provide consumers with a multi-dimensional platform and product ranking scheme.(3)Designed and implemented a multi-dimensional e-commerce recommendation system based on comment tag sentiment and e-commerce platform indicator data.The application results show that the application of the recommendation method in this article can not only provide consumers with a smooth interactive experience,but also provide consumers with A flexible e-commerce platform and corresponding merchandise sorting scheme.