Towards an information filtering system in the Web integrating collaborative and content based techniques

Abstract

The amount of information currently available, the different media and presentation formats joined with the little time availability of the researchers and people in general, make necessary the implementation of automated tools selecting and evaluating information, aiming at not only optimizing resources but also obtaining useful and personalized results that optimize the daily work of its users. A technique known as Information Filtering could be seen as a solution to this problem. Within an information filtering system, a user introduces a profile in the System which represents his/her information needs; then, the system works to display the relevant information. This article contains a sample of the research carried out by us in this important area, focusing the work towards two of its most representative techniques: "Content Based filtering" and "Collaborative filtering." These techniques have been studied from different points of view, allowing to create a solid framework which involves the necessary criteria for designing and creating a tool using the most outstanding characteristics of each technique. They provide a view to facilitate the work of people devoted to the search, depuration and distribution of information.

Classification of Messages and Content based Filtering From OSN User Walls

“…This filtering is based on ML paradigm. ML techniques [3,9,10] and other strategies [8,7] have been used to exploit the Content-based filtering technique However the problem occurred is that content-based filtering on the contents of social network users has achieved few attention in the scientific community. For example Boykin and Roychowdhury [4] have proposed an automated system, exploiting the properties of social networks, able to find out the unsolicited commercial e-mail, spam and messages.…”