Rome, Latium, Italy
Fabrizio Silvestri is a Full Professor at Sapienza, University of Rome in Italy. Formerly a Research Scientist at Facebook AI in London. His interests are in AI applied to integrity-related problems and Natural Language Processing. In the past, he has worked on web search research, and in particular, his specialization is building systems to interpret better queries s from search users. Before Facebook, Fabrizio was a principal scientist at Yahoo, where he has worked on sponsored search and native ads within the Gemini project. Fabrizio holds a Ph.D. in Computer Science from the University of Pisa, Italy, where he studied problems related to Web Information Retrieval with a particular focus on Efficiency-related problems like Caching, Collection Partitioning, and Distributed IR in general.
I am leading the Ad Processing and Retrieval group at Yahoo Labs in London. The amazing group of scientists with which I'm working is responsible of devising, implementing, and testing algorithms for the evaluation of relevance of advertisements in the context of Sponsored Search. The main activities of my group are related to study how to interprete signals from user behavior in order to improve the list of advertisements shown to users. We have developed a solution based on information retrieval techniques to compute the relevance of an ad to a query in order to reduce the amount of ads considered not relevant; filtered out 80% of editorially annotated bad cases.
I have primarily worked on Yahoo Gemini platform. Contributed to the dwell time prediction component deployed to the Gemini Native ad serving platform with a solution based on random survival forest uplifting the measured dwell time by 20%. Developed a solution based on word embedding for query rewriting and query ad matching within Gemini Search resulting in more than 15% uplift in revenue.
I am currently working on problems related to computational advertising. My research is related to the Yahoo Gemini Marketplace (https://admanager.yahoo.com). In particular, I am studying how to interprete signals from user behavior in order to improve the list of advertisements shown to users. The projects I am involved in are mainly related to native advertising and sponsored search and have direct impact on products that are deployed online. In addition to computational advertising, I am involved in projects related to query log mining for query suggestion and ranking in general. In my research at Yahoo I employ a good deal of machine learning techniques applied to the problems mentioned above. Given my background on high performance computing I also put particular attention to scalability and efficiency in the solutions I propose.
I studied problems related to Search and Information Retrieval. In particular, I developed skills for the design, development, and evaluation of high performance computing solutions to the field of distributed information retrieval. In addition, I have done research in query log mining developing solutions for query suggestion, spell checking, and caching. Published a book on the subject: http://didawiki.cli.di.unipi.it/lib/exe/fetch.php/wma/paper.pdf