Professor Stephen Robertson I'm now retired from paid work.
I spent 15 years, from 1998 to 2013, as a researcher at the Microsoft Research Laboratory in Cambridge (England, that is). I was a fellow of Girton College, Cambridge from 2003 to 2013, and remain a Life Fellow.
I spent much of my career at City University London, in what was the Department of Information Science. I worked there full-time from 1978 to 1998 and part-time from then till 2009, and was head of department from 1988 to 1996. I remain Professor Emeritus at City.
In 1987 I started the Centre for Interactive Systems Research in the Department. My co-director for a number of years was Micheline Beaulieu, subequently at the University of Sheffield. A subsequent director was Andy MacFarlane.
This is the poem one of my colleagues, Susan Jones, wrote for me when I moved to Microsoft in 1998.
I previously held a Royal Society Scientific Information Research Fellow at University College London, School of Library Archive and Information Studies, where I had done my PhD with B.C. Brookes. Before that I worked at Aslib. Before that, I took an MSc at City. My first degree was in Mathematics from Cambridge.
In 2013, I was interviewed (by Dr Lai Ma) as part of an oral history project, on behalf of the Association for Information Science and Technology, ASIS&T. The interview was recorded and transcribed and placed on a website for ASIS&T. That website no longer exists; ASIS&T has plans to restore it, but these seem to be taking a long time to come to fruition. In the meantime, I have made the recording and transcript available here.
My main research interests are in theories and models for information retrieval (specifically probabilistic models) and the design and evaluation of IR systems. I was the author (back in 1976, with Karen Sparck Jones) of a probabilistic theory of relevance weighting, which has become quite well established in the field. An extension of that model (work with Stephen Walker) led to the BM25 function for term weighting and document scoring, now used by many other research groups. A further extension (work with Hugo Zaragoza and Michael Taylor) led to the field-weighted version, BM25F.
Our main experimental vehicle in the Centre during my time at City was the Okapi system, also due to Stephen Walker. We used Okapi both for live experiments with real users in operational settings, and for laboratory experiments in the Cranfield tradition. In this latter category, we developed a strong involvement in the TREC programme.
Selected publications listed here, with a small number of web-accessible ones. See also the Sparck Jones / Robertson IDF page.
ISI Lazerow Lecture, November 1993
Transparencies from Mira meeting in Glasgow, May 1996
I have at various times taught the following:
and a few other things.
A family art gallery (have a look).
My father's poems (very good).
Another poetry website I maintain (contains a few of my own).
Last updated December 2021