(+) Web Data Analysis: user-behavior, (user- and content-) centric understanding & modeling
               Data Mining and IR of Web Document: recommendation and modelization
               Computer Science + Economics + Socialogy + Web Engineering + ...

         (-) For this, I have been doing my research in two axis: online users and contents of the web and I believe that
              it is important to investigate both sides when we try to understand the behaviors and characteristics of web.

              For user-level analysis, I study users' favoring behaviors in different web services and try to model them.
              We call 'reciprocity' as in-kind behaviors for others and it can be a tool to understand human behaviors.
              Now I am studying favoring reciprocity - when I favor a content of my friend, does she favor one of my contents?
              This is an interesting question because in social networking services users influence and affect other users.

              For content-level analysis, I am interested in seeing the characteristics of the popularity of online contents
              and modeling them with statistical approaches. The models can allow useful applicatoins for both service providers and users.
              Additionally, I like to use topic models which enable to classify or cluster documents according to their topics.
              This task is very exciting for me because I could investigate to underlying users' intention in various online contents.

        (+) Traffic Monitoring: monitoring infrastructure, machine
               Traffic Analysis and Modeling: descriptive and statistical
               Traffic Classification and Identification: automatic and machine learning

         (-) I have much experience about network monitoring and traffic analysis and modeling.
              I built a monitoring infrastructure in a campus network and with the infra I collected and analyzed large size of network traces.
              My research interst is to classify the traces according to applications and services, such as web, games, and streaming
              especially with un unsupervised fashion. This approach can be applied to classify web traffic according to a set of services.

        (+) Massive Data Analysis with Various Application Domain
               Batch Processing with Cloud Computing

         (-) I have used various kinds of datasets, such as network traffic, crawled datasets from web, and NLP datasets,
              very frequently needed much system resources for collecting or analyzing them.
              Now we have a very convenient enviroment, "a clould computing service".
              Naturally, as a researcher, I investigate in how to use the cloud computing in analyzing datasets efficiently.