I am a research scientist at Yahoo! Research. I graduated with a PhD from Carnegie Mellon University in 2008. I am interested in Information Mediation, with an emphasis on search engines. I work on several key tasks that search engines perform such as Web Crawling and Monitoring, Web Ranking and Search Advertising. Also, I study the exploration vs. exploitation tradeoff that arise in many of these tasks using the Multi-armed Bandit problem. Web Crawling and Monitoring: I work on different subtasks that crawling involves such as discovering new pages, acquiring newly discovered pages and recrawling known pages. I proposed a new query-driven crawling paradigm on this front. Search Advertising: Search engines advertise by showing ads on Web pages as well as by displaying ads next to search query results. I use the Multi-armed Bandit problem to study both these forms of advertising. Multi-armed Bandit problem: I develop and study new variants of the classical Bandit problem. These variants are motivated by real applications from databases and search engines. Publications (DBLP) Conference Publications Recrawl Scheduling Based on Information Longevity Chris Olston and Sandeep Pandey WWW 2008, 17th International World Wide Web Conference Crawl Ordering by Search Impact Sandeep Pandey and Chris Olston WSDM 2008, 1st International Conference on Web Search and Data Mining Multi-armed Bandit Problems with Dependent Arms (PPT) Sandeep Pandey, Deepayan Chakrabarti and Deepak Agarwal ICML 2007, 24th International Conference on Machine Learning The Discoverability of the Web Anirban Dasgupta, Arpita Ghosh, Ravi Kumar, Chris Olston, Sandeep Pandey and Andrew Tomkins WWW 2007, 16th International World Wide Web Conference Bandits for Taxonomies: A Model-based Approach (PPT) Sandeep Pandey, Deepak Agarwal, Deepayan Chakrabarti and Vanja Josifovski SDM 2007, SIAM International Conference on Data Mining Handling Advertisements of Unknown Quality in Search Advertising (PPT) Sandeep Pandey and Chris Olston NIPS 2006, 20th Annual Conference on Neural Information Processing Systems Shuffling a Stacked Deck: The Case for Partially Randomized Ranking of Search Engine Results (PPT) Sandeep Pandey, Sourashis Roy, Chris Olston, Junghoo Cho and Soumen Chakrabarti VLDB 2005, 31st International Conference on Very Large Data Bases User-Centric Web Crawling (PPT) Sandeep Pandey and Chris Olston WWW 2005, 14th International World Wide Web Conference