We present a framework for adaptive news access, based on machine learning
techniques specifically designed for this task. First, we focus on the syst
em's general functionality and system architecture. We then describe the in
terface and design of two deployed news agents that are part of the describ
ed architecture. While the first agent provides personalized news through a
web-based interface, the second system is geared towards wireless informat
ion devices such as PDAs (personal digital assistants) and cell phones. Bas
ed on implicit and explicit user feedback, our agents use a machine learnin
g algorithm to induce individual user models. Motivated by general shortcom
ings of other user modeling systems for Information Retrieval applications,
as well as the specific requirements of news classification, we propose th
e induction of hybrid user models that consist of separate models for short
-term and long-term interests. Furthermore, we illustrate how the described
algorithm can be used to address an important issue that has thus far rece
ived little attention in the Information Retrieval community: a user's info
rmation need changes as a direct result of interaction with information. We
empirically evaluate the system's performance based on data collected from
regular system users. The goal of the evaluation is not only to understand
the performance contributions of the algorithm's individual components, bu
t also to assess the overall utility of the proposed user modeling techniqu
es from a user perspective. Our results provide empirical evidence for the
utility of the hybrid user model, and suggest that effective personalizatio
n can be achieved without requiring any extra effort from the user.