Agent Technology for Personalized Information Filtering: The PIA System

Abstract

As today the amount of accessible information is overwhelming, the intelligent and personalized filtering of available
information is a main challenge. Additionally, there is a growing need for the seamless mobile and multi-modal system
usage throughout the whole day to meet the requirements of the modern society (anytime, anywhere, anyhow). A personal
information agent that is delivering the right information at the right time by accessing, filtering and presenting
information in a situation-aware matter is needed. Applying Agent-technology is promising, because the inherent
capabilities of agents like autonomy, pro- and reactiveness offer an adequate approach. We developed an agent-based
personal information system called PIA for collecting, filtering, and integrating information at a common point,
offering access to the information by WWW, e-mail, SMS, MMS, and J2ME clients. Push and pull techniques are combined
allowing the user to search explicitly for specific information on the one hand and to be informed automatically about
relevant information divided in pre-, work and recreation slots on the other hand. In the core of the PIA system advanced
filtering techniques are deployed through multiple filtering agent communities for content-based and collaborative
filtering. Information-extracting agents are constantly gathering new relevant information from a variety of selected
sources (internet, files, databases, web-services etc.). A personal agent for each user is managing the individual
information provisioning, tailored to the needs of this specific user, knowing the profile, the current situation and
learning from feedback.