I am a Ph.D student in School of Co-blmputer Science at University of Massachesetts Amherst (UMass). I am now working in Computer Networks Research Group (CNRG). My advisor is Professor Don Towsley.
My research interests includes Network Science, Machine Learning, Data Mining and Natural Language Processing. Brief introductions of my current and past projects are list below. Please refer to Project for more detail.
Projects in network science focus on solving problems in social networks. By studying the Spreading of Epidemic in Dynamic Network, we can understand how fast a rumor in a social network vanishes or we can make better strategies, given limited resources (time and funds), for social network advertising based on user-to-user recommendation.
We are good at collecting huge amount of data with today's technology. But sometimes, not every piece of data is neccessary in analysis. With Defussion Wavelet for Networks, we construct multiple scale of analysis on a network structure, where a coarse scale analysis can save the cost of computation but still lead to a satisfiable answer.
Machine Learning and Data Mining
Online Dating. From large amount of user data on a dating website, we discovered several interesing user behaviors with current data mining techniques. Our work has been reported in MIT Technology Review.
Meanwhile, companies try to learn user preferences based on their behaviors so as to make accurate recommendations. Recommender System based on LDA proposed an alternative method to learn user preference based on Latent Dirichelete Allocation (LDA) and makes recommendations for online dating users.
Natural Language Processing
A corpus of documents provide us information. More importantly, huge of amount of commonsense is hidden within the context. Memory Reconsoliation with Commonsense try to learn these commonsense and summarize a document based on knowledge bases such as ConceptNet.