Information Recommendation by collaborative filtering incorporated with gaze detection
|脇山孝貴||Kouki Wakiyama||広島大学大学院 工学研究科||Graduate School of Engineering, Hiroshima University|
|吉高淳夫||Atsuo Yoshitaka||広島大学大学院 工学研究科||Graduate School of Engineering, Hiroshima University|
|平嶋宗||Tsukasa Hirashima||広島大学大学院 工学研究科||Graduate School of Engineering, Hiroshima University|
Collaborative filtering is one of the information filtering techniques, which recommends information based on the evaluation of others' feedbacks whose preferences are similar to a user. The evaluation on the user's preference is often given by explicit operations by the user, and it is considered to be a burden in applying this technique for recommendation. This issue affects the granularity, i.e., degree of detail of user's feedback, since it largely depends on the easiness of acquisition of the user's feedback. As an application of the framework we inplemented a system that recommends paintings to a person based on others' attention in watching paintings. The evaluation of preference is based not on individual objects(i.e. paintings) but on regions in an object (e.g., a person, a building, an animals), that is obtained by the duration of watching a region in the painting. The proposed framework enabled us to get better recommendation compared with object based recommendation.