Enhancing E-Commerce Retail Product Classification Based on Big Data Analysis and Natural Language Processing
Tsz Hon Shao, Maher Maalla, Xiaolan Lin, Xiaohong Ning, Allam Maalla · 2023
Currently, most E-commerce platforms employ conventional machine learning algorithms for product classification, particularly when dealing with Chinese product titles. This classification process primarily centers on techniques like word segmentation, stop word removal, and feature engineering, with limited exploration of advanced technologies in this domain. However, recent years have witnessed significant breakthroughs in natural language processing, particularly in classification tasks, addressing various challenges like text classification, named entity recognition, and summarization of key content. Extracting valuable insights from textual data and delivering value to relevant industries has garnered substantial interest from researchers and developers. To underscore the strengths of our algorithmic model, this study leverages the traditional machine learning approach of the naive Bayes algorithm, using experimental data as a foundation. Additionally, we harness the cutting-edge BERT model from the realm of natural language processing to serve as a classification model for handling concise Chinese product titles.