Towards intelligent and adaptive digital library services

In this paper, we introduced a 3-Layer digital library architecture that facilitates intelligent and adaptive digital library services. We aimed at integrating DL contents with domain-ontology, user-profile and usage-pattern by means of intelligent algorithms and techniques. On top of an open-source digital library system, we developed required modules to capture and manipulate necessary data with the help of efficient techniques such as ontology-driven topic inference, collaborative filtering, single exponential smoothing, etc. We verified that our approach is capable of enhancing and adapting user profile dynamically with the help of ontology-driven topic inference and usage-pattern analysis. Usage pattern and content -based collaborative-filtering techniques are used in developing adaptive recommendation service. We also proposed a User Interest-Drift algorithm based on single exponential smoothing techniques. Our preliminary experimental results and exploratory analyses show that our approach has created positive user experience in a small digital library environment. Large scale deployment of the proposed digital library system along with further refinement of algorithms is also planned.