September 20, 2016 / 16:00 ~ 17:15



Recently huge amounts of information from online documents are available. One of the main effort to organize this information for researchers and companies have been investigating the problem of automatic text analysis. The bulk of such work has focused on topic analysis, attempting to extract topics from many documents which is used for text categorization and opinion mining. However, on‐line discussion groups and review sites  (e.g.,, Amazon) are rapidly growing, where a crucial characteristics of the posted articles is their sentiment, or overall opinion towards the subject matter – for example, whether speeches of US presidential candidate are positive or negative, what is main topic and sentiment of terrorist speech and what people write in their SNS about these days political issues. These topics extracted form articles and authors also can be organized through social network analysis. The relations of each topics and authors can explain what the most important topic is and who the most influential person/user is in the target area. The result can be applied customer related management, marketing, web service design etc.
In this talk, I will describe these two analysis: sentiment analysis and social network analysis. Then, I’ll also briefly introduce my study on analyzing healthcare web forum through sentiment and social network analysis.



Ph.D. Majored in Industrial and Systems Engineering,  Department of Industrial and Systems Engineering,
KAIST (Korea Advanced Institute of Science and Technology), Korea,  March 2007 ~ August 2013
M.S.  Department of Industrial and Systems Engineering, KAIST, Korea, March 2005 ~ February 2007
B.S.  Department of Industrial and Systems Engineering, KAIST, Korea,  March 2000 ~ February 2005


July 2013 ~ Present
ETRI(Electronics and Telecommunications Research Institute) SW and contents technology future research team Senior Researcher,
March 2011 ~ June 2013
Department of Industrial Management, University of Koreatech Teaching Professor,