GLA - The University of Glasgow, UK (TERRIER Search Engine)

GLA - The University of Glasgow, UK (TERRIER Search Engine)

The School of Computing Science at the University of Glasgow is a world-class research centre, which has celebrated its 50th anniversary in 2007. In the recent UK RAE research assessment, the international stature of the Department has been further recognised with 80% of its research outputs classified as 4* or 3*. Moreover, the School was ranked as one of the top 5 UK computing departments in the Times Good University Guide 2010. The School is internationally recognised for its leading role in information retrieval (IR) research, setting trends in many aspects of IR research (e.g. theoretical models, high performance text and multimedia retrieval, evaluation, etc.). The IR group has active interests in the process of Information Retrieval as a whole, from the modelling of IR systems through large-scale experimentation to user-centred evaluation. The conducted research preserves a strong emphasis on theoretically-driven, still practical solutions for large-scale document collections. In particular, the School is home of the high performance Terrier IR search engine, the open source version of which has been widely used across the world by both academia and industry. The Terrier engine provides state-of-the-art MapReduce-based indexing, and cutting-edge retrieval models. The School brings strong expertise in a number of areas related to the proposal, including large-scale, cost-effective and high performance search engines building, semantic web information retrieval, large-scale data mining and knowledge discovery, as well as the deployment/development of MapReduce-based solutions. Indeed, several researchers within the School work on real-time large-scale search applications as well as semantic web and context-sensitive search and interaction. The School has a very strong tradition in evaluating search engines, whether using the TREC paradigm or in a user-centric fashion. In particular, members of the School have been coordinating the Blog track at the renowned TREC evaluation forum from 2006 until 2010, which will be morphed in 2011 into a Microblog track (Twitter), dedicated to real-time and social search. Researchers of the School are regular participants of TREC, CLEF, INEX and other major evaluation forums activities. The School has been at the foremost of many EU projects including K-space, SEMEDIA, SIMAP, SALERO, Mira, MIRO, FERMI, PROSPER, FIDE I & II, IDOMENEUS, COMMANDOS I & II, etc., and its researchers are involved in the organisation of premium Information Retrieval conferences (e.g. SIGIR, CIKM, ECIR, TREC, etc.). The School will host the 20th ACM International Conference on Information and Knowledge Management (CIKM) in Glasgow, 2011. Note that the University of Glasgow has the foremost Information Retrieval group in Europe, and has an unrivalled experience in building large-scale search engines, requiring advanced and efficient indexing and retrieval method. Its renowned state-of-the-art open source search engine Terrier is widely used across the world in both academia and industry. The Glasgow group has a very long and renowned expertise in developing Web search solutions, from theoretical models to combining many features using learning to rank techniques within a data-driven machine learning setting. The Glasgow group is also renowned for developing models for other search applications, including feed/stream/news search, Twitter, people search within Enterprises, and the selection of diverse search results. In addition, the group has a long history in tackling multimedia search and browsing applications, as well as a comprehensive experience in mining and searching ontologies within physical science domains (e.g. c.f. the EPSRC Explicator project). Finally, Glasgow leads international efforts in building corpora, testbeds and infrastructure for the evaluation of effective and efficient retrieval solutions for user generated content, through their leadership of the dedicated Blog & Microblog tracks of the TREC evaluation forum since 2006.

Role in the Project: Given its long experience in building large-scale search engines, GLA will lead the development of the SMART search engine core in WP5. As a consequence, GLA will also oversee the architecture requirements of the SMART search engine in WP2. GLA will be responsible of designing and implementing the social search functionalities of SMART, as well as its operational deployment in social search use cases (WP4/WP6). Finally, GLA will actively engage in the integration and evaluation activities within SMART, and will also lead its open sourcing tasks (WP6).

Dr Iadh Ounis is a Reader in the School of Computing Science at the University of Glasgow, which he joined as a Lecturer in 1999. He holds a PhD degree from the University Joseph Fourier, Grenoble, France. He has been an active researcher in information retrieval since 1994, is a member of the editorial board of the Information Processing & Management journal, a member of the advisory board of the Forum for Information Retrieval Evaluation, and has chaired a number of IR events and initiatives. Dr. Ounis has authored over 100 refereed articles and publications. Since joining the University of Glasgow, his research has concentrated on probabilistic IR Models, Enterprise/Intranet/Web/Blog search, large-scale IR and efficiency and effectiveness evaluation. Dr. Ounis is the principal investigator of the high-performance and scalable Terrier search engine, the open source version of which has been downloaded thousands of times since becoming available in November 2004. Since 2006, Dr. Ounis has been leading the international TREC Blog & Microblog track initiatives, which explore information seeking behaviour in the blogosphere and microblogging environments. Dr. Ounis is on various program committees of major information retrieval conferences and related events (e.g. SIGIR, CIKM, WWW, ECIR, WSDM, CORIA). Recently, he was Chair of the 6th ESSIR 2007 IR summer school and the 30th ECIR 2008 IR conference. He was the General Chair of the 20th ACM Conference on Information and Knowledge Management (CIKM), which was held in Glasgow in 2011.

Craig Macdonald is a postdoctoral researcher at the University of Glasgow. He received his PhD in Information Retrieval from the University of Glasgow in 2009. He is a principal developer of the Terrier search engine platform. His research interests include Information Retrieval in Enterprise, Web and Blog settings and has over 40 publications with research based on the Terrier platform. He has worked extensively on building and evaluating large-scale search engines, as well as blog search and expert finding systems. He has been a joint coordinator of the Blog track at TREC since 2006. Recently, he has been working extensively on distributed information retrieval systems, and efficient indexing (e.g. MapReduce indexing and mining) and retrieval strategies (e.g. dynamic pruning and caching) for very large-scale document collections. He was a Chair-at-Large and Workshops Chair of the 20th ACM Conference on Information and Knowledge Management (CIKM), which was held in Glasgow in 2011.

The Research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7/2007-2013)