The broad topic for my PhD study is social media and network analysis, involving complex networks, data mining, machine learning, regression analysis, etc.
My current project is about modeling user lifecycles in online communities (e.g., Wikipedia, Q&A sites) using dynamic topic models. By considering multiple dimensions of features (e.g., the types and quality of contribution, social interaction, etc) for online user behaviors, we will be able to identify and distinguish different user groups and different life stages. Understanding how individual users develop over time in online communities makes it possible to develop techniques for important applications such as churn prediction, expert finding, personalized recommendation, malicious user and content detection.
Specialties: Data mining, machine learning, social network analytics, user behavior analytics, regression analysis.
Application areas: Churn prediction, Expert finding, Malicious user and content detection, Personalized recommendation.
Development skills: Java, C, Matlab, Python, Latex