The Improvement of a Prediction Study of Transactions Based on E-Commerce Site Search Data

Abstract

The search behavior of E-commerce consumers is conscious. And their shopping is another kind of consciousness behavior. In the shopping process of e-commerce website, these two consciousness behavior were joined together. So, the phenomenon reflects some trends and patterns, reflect the relationship between e-commerce site search volume and trading volume. This paper will attempt to establish the model of a theoretical framework, which explored the first - lag relationship between the amount of e-commerce site search and e-commerce transaction volume, by stepwise method which obtain comprehensive search index, and then which complete the empirical analysis, obtain the prediction result. The results showed that: there are a higher correlation between the search data and trading volume of e-commerce site; after adding a synthesis of the search index, the model fit increased to 0.901, which significantly improve the prediction result of the model. At the same time, the model has a stronger timeliness, which can more timely predict e-commerce transactions.