Research on Sentiment Analysis and Personalized Recommendation Based on Agricultural Product Reviews

This paper takes the online review of Shenyou Songhua Egg in the shopping platform Jingdong as an example, introduces text sentiment analysis into e-commerce online commentary, and more accurately judges the emotional tendency of online reviews of agricultural products e-commerce based on the BERT model. Word frequency analysis and word cloud drawing of accurately classified online reviews of agricultural products can summarize the crucial criteria for the product sales and capture the key points in agricultural products e-commerce online reviews in more detail. Empirical analysis shows that by introducing text sentiment analysis into the field of online reviews of agricultural products e-commerce and using the BERT emotion model for emotional judgment, it can more accurately judge users' emotional tendencies, and take corresponding measures according to the key points of comments extracted by word frequency analysis, so as to improve users' favorability and loyalty to the platform and increase user stickiness.