Acquisition and Sharing of Multimedia Data in Mobile Environment
| Project/Area Number |
17500070
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Media informatics/Database
|
|---|
| Research Institution | HIROSHIMA UNIVERSITY |
|---|
Principal Investigator |
ATSUO Yoshitaka Hiroshima University, Graduate School of Engineering, Assistant, 大学院工学研究科, 助手 (60263729)
|
|---|
| Project Period (FY) |
2005 – 2006
|
| Project Status |
Completed (Fiscal Year 2006)
|
| Budget Amount *help |
¥2,200,000 (Direct Cost: ¥2,200,000)
Fiscal Year 2006: ¥900,000 (Direct Cost: ¥900,000)
Fiscal Year 2005: ¥1,300,000 (Direct Cost: ¥1,300,000)
|
|---|
| Keywords | Mobile Environment / Information Sharing / Eye movement / GPS / 情報フィルタリング / アノテーション / モバイルコンピューティング / 視線 / 位置合わせ / 垂直エッジ |
|---|
| Research Abstract |
The purpose of this study is to establish the framework of multimedia information sharing in mobile environment base on implicit acquisition of user' s preference or activity context. In the proposed method, the user' s state of concentration to an object is detected based on eye movement. In that process, the target of his/her concentration is recognized not by the unit of object, but by the unit of component. The user interest model for objects in the real world is constructed based on the detected target objects in concentration. Then, we can provide the users with the mechanism for recommending multimedia information that he/she is not accessed yet but is inferred to be in his/her interest, by sharing the user interest model that is organized by implicitly accumulated preference information.
According to the experimental result, our method is advantageous since it does not require explicit specification on user preference. Moreover, component object-based preference acquisition offers better quality of recommended information. Another remarkable result is that the proposed method is compatible with the information recommendation based on explicitly specified user interest model in the sense of the quality of recommended information.
When we focus on occasions of retrieving information that helps his/her next action in the real world, the importance of information related to the object, place, or facility, needs to be evaluated based on the human activity in the real world. In order to evaluate the importance of the information, we studied a framework of modeling users' activity with so-called 'travel graph'. It represents the users' traffic between nodes that corresponds to places or facilities. The value of information is re-evaluated based on the travel graph, and it enables to evaluate the importance of from the point of view of human activity, and thus it offers new framework of information retrieval.
|
Report
(3 results)
Research Products
(13 results)
