An Overview of Personalized Recommendation System to Improve Web Navigation

Abstract

We present a new personalized recommendation system, which means the searches of each user is done according to their interest which is based on ranking or preference method. It also maintains the logs which records the sessions of each user and brings out the exact data required by the user. This is done by fetching the data that is already stored in the database. Web server logs maintains history of page results and consists of a log file which automatically creates and maintains the list of activities performed by the users. For extracting the data according to the user’s previous searches, we are using Stemming Algorithm. The Stemming Algorithm is a process where the exact, meaningful words are extracted from the URL. Because of this process the user’s search time will be reduced. It also improves the quality of web navigation and overcomes the limitation of existing system. In the proposed system we extract user’s behaviour from web server logs in the actual process whereas, in the anticipated system, the user’s behaviour is done with the help of cognitive user model and we perform the comparison between the two usage processes. The data produced from this comparison can help the users to discover usability issues and take actions to improve usability. In the anticipated usage the cognitive user model is done that can be used to simulate or predict human behaviour or by performance and task. Finally, the system is executed by using the top-k ranking algorithm. The advantage of this system are accuracy and better processing speed. The user’s convenience deals with the ease of navigation which helps the users to interact with their interface.