Collecting User Preferences from User Activities on E-Commerce Web-Site Focusing on Sending Targeted Advertisements

Abstract

Internet is growing as a global marketplace. For the foreseeable future, online advertising remains as the main source of revenue on the web. Taking advantage of Web mining techniques is useful in this way that it is possible to replace personalized ads by public advertisement by examining the user profile and extracting his interests. Since the data, associated with user performance, have a large volume and are generated in high speed and are very versatile, they can be called as big data. As it will be explained below, the processing of big data seems impossible using traditional technology. Using relational database in order to process huge data is very difficult; hence, the use of a non-relational database such as MongoDB can facilitate the processing of data in addition to resolve the mentioned issues. A major problem in the process of understanding the user's favorites and sending his favorite target advertising is to collect information that will help researchers in this field; In order to collect this information, the website designed for an electronics store database with MongoDB database has been redesigned in which the required factors, including the user login time to the site, number of visits per page, the number of rotating the mouse cursor over the product image etc. have been considered to understand the user interests, and thereby, their interest has been identified; then it has been used in a 6-month period by the customers of the store. After collecting and analyzing the data, it was observed that different users have different performance in site and suggestions have been provided to improve site performance, customer satisfaction and online store sales based on their tastes.