Issue No. 03 - July-September (2007 vol. 6)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MPRV.2007.49
Sung-Bae Cho , Yonsei University
Kyung-Joong Kim , Yonsei University
Keum Sung Hwang , Yonsei University
In-Ji Song , Yonsei University
Owing to the advance of ubiquitous computing technology, various information sources make the events of a given user's daily life readily available. Mobile devices such as cell phones, smart phones, and PDAs let users capture rich media with ease and store personal information such as context, schedules, short messages, photos, and videos. We can regard usage information and stored multimedia content as relevant evidence for inferring a given user's daily actions. The AniDiary (Anywhere Diary) system summarizes a given user's daily life with as a cartoon-style diary. Because showing all the events for a day isn't appropriate, landmark (memorable) events are automatically detected by modeling the cause-effect relationships of various events with Bayesian networks. The system then converts the events into cartoon images for a diary. Experimental results on synthetic and real data show that AniDiary provides an efficient, user-friendly way to summarize the user's daily life.
life logging, Bayesian networks, landmark detection, comic generation, ubiquitous computing
S. Cho, I. Song, K. Kim and K. S. Hwang, "AniDiary: Daily Cartoon-Style Diary Exploits Bayesian Networks," in IEEE Pervasive Computing, vol. 6, no. , pp. 66-75, 2007.