In the E-Commerce system, sellers disseminate commodity information, as well buyers look for befitting commodities via the sellers' websites. Yet, the scale and source of commodity information have been increasing fleetly along with the quick development of E-Commerce. So sellers are insensible of how to find target market efficiently, and buyers could not find their needed commodities puzzled by the ocean of information. The intelligent agent technology (IAT) can provide good means to resolve the information matching between the seller and buyer in E-Commerce. The IAT can collect, index and filter the domain model and user model knowledge that acquired automatically, and then submit the interested and useful information to the users automatically. This paper presents the holistic framework of personalized service model, and then implements the model by clustering analysis and collaborative filtering recommendation.