Personal History Based Recommendation Service System with Collaborative Filtering

Although the conventional ubiquitous home service provides services using the information of environments obtained by the analysis of sensors, it shows a lack of user information. If recommendation services are able to use the past item selection information related to the context information of users, individualized services would be achieved. Also, it is possible to solve a specialization tendency that makes not possible to avoid the taste of users themselves for recommended items when users use the item selection information of other users. This paper attempt to use Naive Bayesian for context model and propose recommendation service method based on personal history. And the recommendation service system (RSS) use a collaborative filtering (CF) to solve a specialization tendency on Open Service Gateway Initiative (OSGi).