|
|
스마트폰
|
1. Motivation
Why intelligent interaction on mobile
device?
- Everyone has one or more mobile
device
- It contains many useful private
information
- Relatively small screen size,
difficult input interface, and limited computational power
- Increase of multimedia data (photo,
mp3, movie files) on the mobile device
The goals of this research are as
follows.
- Developing synthetic characters for
mobile devices
- Prediction of users' behavior based
on the information on the mobile device (call logs, location, short
message, etc)
- Context-awareness on mobile device
- Efficient summarization and
searching of multimedia data on the mobile device
2. Proposed Methods
- Developing software for gathering
information on the mobile device
- Estimation of high-level context
from the low-level sources using probabilistic models (Bayesian
networks)
- Detection of chances and keys from
huge information using key graph and Bayesian networks
- Reorganization of personal
experience and multimedia data for efficient search (semantic
search) using semantic network
- Prediction of user's future
location using recurrent Self-organizing map and Markov model
- Behavior generation of synthetic
characters on the phone using behavior network and rule-based action
selection
- Implementation on real PDA devices
Figure 1. Proposed architecture for synthetic characters on mobile
device
3. Experimental Results
- Demonstration of synthetic
characters [mpeg1]
- Demonstration of location-aware
service [mpeg2]
Figure 2. Bayesian networks for
inference of emotional status, business and closeness between two
persons
4. Publications
[1] S.-J. Han, and S.-B. Cho, "Putting
together AI techniques for personal assistants inside smartphone,"
Int. Journal of Hybrid Intelligent Systems, 2005.
[2] S.-J. Han, and S.-B. Cho, "Synthetic
character with Bayesian network and behavior network for intelligent
smartphone," Lecture Notes in Computer Science, 2005.
[3] S.-B. Cho, "Developing intelligent
assistants in digital convergence," 2005 US-Korea Conference on
Science, Technology and Entrepreneurship, 2005.
[4] S.-B. Cho, "Intelligent personal
assistant for natural interface to smartphone," Plenary Talk at Human
and Computers, 2004.
[5] S.-B. Cho, "Hybrid intelligent
techniques for intelligent personal assistant in digital convergence,"
Plenary Talk at Hybrid Intelligent Systems, 2004.
|
|
|
