I graduated from the University of Cambridge Computer Laboratory with a Ph.D. in computational linguistics. During my tenure as an Advanced Research Fellow in the University, the overall thrust of my work was in developing infrastructural support for natural language applications. Pioneering projects on large scale grammar and lexicon development resonated with IBM's early and ground-breaking work in facilitating a resource-rich text analysis framework, and I joined IBM Watson's Lexical Systems Group. At the time referred to as computational lexicology, my work there was a precursor to the empirical, corpus-driven methods for language analysis, underlying the applicability of background knowledge for text understanding.
As the Internet was taking off, and opportunities for new text processing applications — beyond interfaces and dialog systems — were emerging, I was managing the natural language program at Apple's Advanced Technologies Group. In the context of bringing intelligence to the desktop, the focus of that program was on developing embedded linguistic capabilities for discovering, skimming, summarising, and presenting topical news; we leveraged novel interface paradigms for dynamic news content delivery with base language functions including feature-rich parsing, coreference resolution, salience determination, topic detection and tracking, and summarisation.
I re-joined IBM Research in 1998, continuing work primarily on language engineering for content analysis; scalable technologies and infrastructure for text analytics, and language processing for business applications. In the Talent project, we developed one of the first industrial strength pipelined NLP architectures, and deployed it in a number of large-scale text management scenarios.
Most recently, I have been working in the Semantic Analysis and Integration department on the Watson project, otherwise known as the IBM Jeopardy! Grand Challenge. My work entails developing algorithms in the area of question analysis, with an emphasis on identifying prominent content from questions and available sources used to generate and assess possible answers. In December 2010, ibm.com showcased our work on its Watson page.
Until recently, I was the Executive Editor of the Cambridge University Press Studies in Natural Language Processing. I have been a member of the editorial boards of Computational Linguistics and Journal of Semantics; and continue to serve as one of the founding editors of Natural language Engineering. I am involved in standards setting activities, for language resources management (specifically the definition of ISO standard for representing and annotating time and events), and I serve on the advisory boards of NSF and international linguistic infrastructure projects.