Design and Analysis of Web-Based Prediction System Using E-Commerce Data

Recommendation systems (RSs) are used by a number of e-commerce sites to help customer find the product to purchase based on their preferences. Developed RS is based on Web usage mining patterns. Web usage mining techniques aim to analyse and discover the user's navigational patterns and past buying behaviour patterns in e-commerce sites. RS can implement in different ways: Collaboration filtering and content-based filtering and other approaches are knowledge base, demographic. We can also combine the above techniques called hybrid recommendation for better recommendation to user. Here, we also compare different algorithms of RS with parameters. In this paper, we examine how recommender systems help e-commerce sites to increase sale and analyse the key challenges for RS.