Assimilation of Information in RDF-Based Knowledge Base

Abstract

The Resource Description Framework (RDF) defines an interesting way of representing data based on a simple principle of expressing any piece of information as a triple: subject-property-object. This notion gains a lot of interest as a promising form of representing any type of data and facts on the Web.

This paper is about automatic assimilation of information.

We propose two procedures: one for assessing relevance between two pieces of information, another for information integration. Both of them can be executed by an agent on the behalf of the user. In a nutshell, the idea is as follows: an agent browses the web and finds some facts, determines their relevance to the facts already known to the user, and integrates them. The proposed procedures are based on compatibility of user’s information with the web contents, and on the level of user’s confidence in the information already known to the user.