Education, Science, Technology, Innovation and Life
Open Access
Sign In

Research on online shopping behavior based on Long Short-Term Memory and Latent Dirichlet Allocation

Download as PDF

DOI: 10.23977/AICT2020045

Author(s)

Haochen He, Liang Xie, Lu Liu, Zhihan Xu

Corresponding Author

ABSTRACT

In this paper, we use natural language processing and machine learning methods to extract product and user characteristics from product descriptions and reviews to establish the sigmoid function dynamic weight factor evaluation model to evaluate the product's success in the market. When establishing the evaluation model, we fully consider the star rating, helpfulness vote, view submitted by the customer. When carrying out quantitative processing of the view, Nltk, Word2vec and Long Short-Term Memory (LSTM) are used to extract the emotional polarity of the text. We compare the results of the model with the actual situation to verify the accuracy of the model. In addition to the above work, we also use the Latent Dirichlet Allocation (LDA) model to investigate whether some words in the view submitted by customers are clearly associated with star rating.

KEYWORDS

Word2vec; Long Short-Term Memory; Sigmoid function dynamic weight factor evaluation model; Latent Dirichlet Allocation

All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.