Analysis of Online Shopping Time Series Model Based on Data Analysis

With the rapid development of information technology, the network has gradually penetrated into People’s Daily life. The rapid development of e-commerce, especially the network shopping, can let more and more netizens shop online conveniently. In this paper, our team is invited by Sunshine Company to analyze the product reviews to help it improve products and develop a reasonable sales strategy. Firstly, we preprocess the data. We delete irrelevant product reviews and make team frequency statistics on the text review. Secondly, in order to quantify the content of review, we use the statistical method TF-IDF to manually screen out 20 specific quality descriptors of text-based reviews and calculate the corresponding weights. In order to measure the emotional tendency of each review, we take a weighted average of the weight of adjectives in each review. At the same time, according to the length of the text, we calculate customer’s Attitude-Score. According to the results of model I, we evaluate Star-Score, Helpness-score, Review-score and Vine-Score linearly with different weights. The reliability of the model is tested by sensitivity analysis. Therefore, our model is applicable to most products reviews. Finally, in order to study the changes of product’s reputation over time, our team build Time Series Model.