I have worked on algorithmic and statistical challenges arising in a broad range of problems primarily derived from Web Search, Computational Advertising, and Graph Mining. Web Search: The Web is characterized by multiple sources of information: searches, anchort- exts, user tags, etc. Combining these to extract the “true signal” has been the underlying theme of my work, which includes webpage segmentation, template detection, and title generation. Computational Advertising: In the same spirit, we combined IR scores with click feedback, & merged data from coarser to finer scales, for CTR estimation as well as explore-exploit strategies. Graph Mining: My thesis, titled Tools for Large Graph Mining, under Dr. Christos Falout- sos, looks at realistic graph generators, analysis of network epidemics, and community detection.