Item-Based Filtering and Semantic Networks for Personalized Web Content Adaptation in E-Commerce

摘要

Personalised web content adaptation systems are critical constituents of successful e-commerce applications. These systems aim at the automatic identification, composition and presentation of content to users based on a model about their preferences and the context of interaction. The paper critically reviews related work in the field and presents an integrated approach for the design of personalization and web content adaptation in e-commerce that places emphasis on item-based collaborative filtering and on short-term, dynamic user models represented as semantic networks. The proposed approach for personalised web content adaptation can provide different types of interesting recommendations taking into account the current user interaction context in a computationally inexpensive way. It is also respectful of user personal information and unobtrusive with respect to user feedback.