FOR THE DEGREE OF

First, I would like to thank my advisor Prof. Jaideep Srivastava for his continued support and encouragement during my graduate stay at UMN. I thank him for the freedom and belief that he had in me during these years, in allowing me to pursue interesting research questions. I learnt several key qualities from him, that I personally consider very important for a researcher, such as challenging fundamental ideas, asking visionary questions, and being able to relate and apply new scientific results to ongoing research. With no doubt, his technical prowess on data mining and social network analysis has been a strong support for my thesis research. On a personal level, Prof. Srivastava was always supportive and backed me up whenever I needed his time for a friendly discussion. I consider myself gifted to have a co-advisor like Prof. Arindam Banerjee. His enthusiasm for research and drive for excellence in infectious. Every discussion with Prof. Banerjee has been invaluable, where I learned something, not just machine learning, but the ethics of doing good research and teaching in general. Some of the papers I co-authored with him made me realize how concise and simple communication, can be precise and effective at the same time. To summarize, I would give Prof. Banerjee most of the credit for becoming the kind of scientist I am today. I am grateful to my mentor Dr. Charu Aggarwal at IBM Research, for his valuable inputs and critical feedback in shaping my thesis research. I am indebted to him forever, for the guidance and mentoring he has offered me throughout my research career. I want to thank Prof. Vipin Kumar and Prof. Ravi Bapna for being a part of my committee and providing me valuable feedback and suggestions at various levels of my PhD, which has helped to shape my work. I thank IBM for supporting my thesis research, during the academic year 2014-15, through a Ph.D. fellowship. I also owe my sincere thanks to Prof. Abhishek Chandra, Prof. Dan Boley, and Prof. Yousef Saad for several hours of valuable discussions. I would also like t