Fernando Carlos das Neves Pereira was born in Queluz in May 12^th, 1952.  His main research areas are computational linguistics, machine learning and computational biology.

Currently he is an endowed professor (Andrew and Debra Rachleff chair) and department chairman at the Department of Computer and Information Science of the University of Pennsylvania. He graduated in 1975 in Mathematics at Faculdade de Ciências de Lisboa after an initial passage through Electrical Engineering. While doing his undergraduate degree and for two years after graduating, he worked at LNEC (Laboratório Nacional de Engenharia Civil – Lisboa), as a systems programmer, and also did research on architectural CAD, urban traffic modeling, artificial intelligence and logic programming. In 1977 he took a British Council’s scholarship for graduate study at the University of Edinburgh, completing there his PhD in Artificial Intelligence in 1982. His main research areas in Edinburgh were natural language processing and logic programming, but he also did some further work on architectural CAD.

In 1982, his research in logic programming and natural language processing brought him an attractive offer to join SRI International, a research institute in Menlo Park, California. At SRI, he worked in logic programming, natural language processing, and speech recognition, and he directed for one year SRI’s research center in Cambridge, England. In 1989 he moved to Bell Labs, where he continued work on speech recognition, and developed new work in probabilistic language models, finite-state methods for speech and text processing, speech retrieval, and several other topics.  In 1996, he started of the machine learning and information retrieval department at AT&T Labs, which became the internationally most recognized group in those fields in the late 90s. During the 2000-01 academic year he held a position as a research scientist at Whizbang! Labs, a startup company where he developed finite-state models and algorithms for information extraction from the Web. Finally, in 2001 he was recruited for his current position at the University of Pennsylvania.

The reasons for his moves were opportunities to do research in new areas such as speech recognition, machine learning, web information extraction, and computational biology, and offers to lead growing organizations, which included his research department at AT&T Labs and the Department of Computer and Information Science at Penn, where he has led the recruitment of eleven new faculty members in areas ranging from artificial intelligence to networking and security.

At the University of Pennsylvania he is a PI or co-PI on three NSF grants and one DARPA contract, and he is a senior personnel on a large NIH grant. His group consists of one postdoctoral fellow, eight PhD students, and one programmer, carrying research on machine learning, biomedical text information extraction, natural language parsing, and computational gene prediction.

He has authored or co-authored 92 refereed publications, one textbook and five patents.

Fernando Pereira says: “My main research goal is to create better learning algorithms for structured natural data, in particular natural language and biological sequences. My dream is that machines can learn to grasp meaning in language and, more generally, in biological communication, through observation and experience”.

In the machine learning and language processing communities, he is best known as one of the creators of conditional random fields and related discriminative learning methods for structured data, and their applications to natural-language processing; of weighted finite-state transducer methods for text and speech processing; and of novel models of distributional similarity in language. He is also known in computational linguistics and semantics research for advances in logic-based methods for parsing and semantic interpretation, and in the logic programming community for contributing to the development of Prolog and logic grammars.

He received a few informal inquiries about returning to Portugal early in his career, but none were serious or attractive enough to be pursued further.

He is a Fellow of the American Association for Artificial Intelligence.

pereira at cis.upenn.edu