Copyright 2002 IEEE. Published in the Proceedings of the Hawai’i International Conference on System Sciences, January 7–10, 2002, Big Island, Hawaii. Enabling Technology for Personalizing Mobile Services M.M. Lankhorst, H. van Kranenburg, A. Salden, and A.J.H. Peddemors, Telematica Instituut, P.O. Box 589, 7500 AN Enschede, the Netherlands {lankhorst,kranenburg,salden,peddemors}@telin.nl Abstract A major trend in the current information society is personalization, which is considered a key business enabler for mobile services because the mobile device is a personal assistant that offers unique opportunities such as real-time adaptation of services to a dynamic user environment. To offer personalized services in a consistent manner and to speed up the development of such services, we present a Personal Service Environment (PSE) that, besides profile management, provides generic service discovery, content adaptation and service adaptation functionality. The PSE implementation is based on a loosely coupled architecture based on the Web services paradigm. Specific PSE components will be generated on a development and deployment environment for negotiating brokerage agents with learning, anticipating, and adaptive capabilities. These brokerage agents are intended to handle matching and adaptation of service content and logic, ensuring optimal quality of service delivery. We illustrate the implementation of our PSE with a ‘Personalized Messaging’ prototype, which is based on the concepts of Unified Messaging and Instant Messaging, combined with new mobile network features such as location awareness. 1 Introduction Research and development for mobile services more and more focus on the personal aspect of the mobile phone. There is a growing awareness that the unique selling points of the mobile are not its anytime/anywhere capabilities, but the fact that it is a truly personal device, and enables context-aware and personalized data services. Examples are travel information while travelling, your local news, up-to-date schedules, your messages always on time, you local shopping guide, your nearest cinema, horse races betting at the latest possible moment, etcetera. In order to enable these unique selling points, it is necessary that a future mobile network infrastructure contain a personalization-enabling service infrastructure. In such a personalization infrastructure, metadata on users (often referred to as user profiles) and metadata on services are indispensable for adapting the content and service to the person and his location. But there is more. Besides user profiles and service characteristics, terminal and network capabilities are also necessary to adapt the quality of the content and service to the specific terminal and network capacities in use. In the near future, it will not only be GSM-evolved networks such as GPRS and UMTS, but also wireless LAN’s, Bluetooth and other wireless technologies that have to be integrated. Furthermore, mobility profiles (data on location, speed, direction, physical environment and ambient conditions) are of course an essential input to location-based services. Our paper is organized as follows. In section 2 we lay down our PSE specifications in terms of various standards. In section 3 we propose our PSE functional architecture, whereas in section we 4 we treat the agent technologies needed to make specific PSE components of the architecture operational. Finally, in section 5 we describe a prototype in which our PSE has been partially implemented. 2 Specification of a Personal Service Environment In 3G standardization bodies such as the 3GPP, the concept of the Virtual Home Environment (VHE) [1], [2] is an example of such a personalization-enabling infrastructure. However, the VHE is limited to the management of profiles. We propose a Personal Service Environment (PSE), which in addition provides generic discovery services (of potentially interested users in mobile services and, vice versa, potentially interesting services for users) as well as adaptation services to accommodate changing conditions and situations (e.g. in roaming, etc.). The PSE is a holistic approach to the personalization of mobile data services in a scalable and efficient way by making use of both centralized as well as distributed functionality and intelligence. A PSE, in our definition, assists a user in finding, adapting and using services that fulfil his needs given his personal profile, his mobility and his context [4]. Different stakeholders are involved in PSE provisioning and usage: