Activity Inference for Constructing User Intention Model


Myunggwon Hwang, Do-Heon Jeong, Jinhyung Kim, Sa-kwang Song, Hanmin Jung




User intention modeling is a key component for providing appropriate services within ubiquitous and pervasive computing environments. Intention modeling should be concentrated on inferring user activities based on the objects a user approaches or touches. In order to support this kind of modeling, we propose the creation of object�activity pairs based on relatedness in a general domain. In this paper, we show our method for achieving this and evaluate its effectiveness.