mLife:

Indentifying Human Mobile Behaviors in Context

We study human behaviors from lifelogs. Using learning techniques we discover and recognize activity patterns of human mobile life to develop an intelligent agent which provides predictive services on mobile phones. In particular, we develop efficient algorithms for modeling and predicting sequential activities from a stream of multidimensional spatio-temporal mobile sensor data. This research is supported by the IT R&D Program of MKE/KEIT.

ΆΖ Outline ΆΖ

1. Abstract

  Development of an agent applied to an open mobile platform for learning automatically the model of mobile phone users

   1) Preferred learning technology by understanding the intention of user

  • Development of intention identification technology based on a variety of context inference of smart phone users, statistical data of users' behavior patterns and Web resource collection
  • Development of individual preferences extraction technology based on learned knowledge and statistical information

   2) Dynamic learning technology in mobile environments

  • Study to supplement the lack of data and information into transition using the ontology based knowledge
  • Research active incremental learning using a variety of data of series, order and rank

2. Primary research

  • Location- aware Learning
  •   Research inference model for main place of personalized mobile phone users
  • Social-aware Learning
  •   Research inference model for social relationship between user and associated people in mobile platform
  • Activity-aware Learning
  •   Research behavior recognition model based on activities of mobile phone users
  • Behavior Prediction Learning
  •   Research behavior prediction model based on multimodal data of mobile phone users

    Collaboration

 

ΆΖ Overall Concept  ΆΖ

 

structure


ΆΖ Publications ΆΖ


Project Title Development of a Cognitive Planning and Learning Model for Mobile Platforms
Duration March 2010 - Febrary 2015
Funding Korea Evaulation Institute of Industrial Technology
Principal Investigator Prof. Yung-Tak Park, Soongsil University
Researchers

Prof. Byong-Tak Zhang

Min-Oh Heo
Jin-Seok Nam
Myung-Goo Kang
Prof. Sung-Bae Cho Han-Saem Park
Bong-Whan Choe
Tae-Min Jung
Prof. Kyu-Baek Hwang In-Sik Suh
Sang-Hun Ju
Seong-Yong Bong
Prof. Soo-Won Lee Jong-Suck Song
Jong- Su Oh
Yung-Ho Kim

Collaborative
Labs

Soft Computing Laboratory, Yonsei University
Machine Learning Laboratory, Soongsil University
Mining Laboratory, Soongsil University


Contact Sang-Woo Lee
E-Mail slee at bi snu ac kr
Phone +82-2-880-1847
Fax +82-2-875-2240