Thomas Roelleke EECS QMUL Homepage

Dr Thomas Roelleke

London E1 4NS

Phone +44 20 7882 7988

Fax +44 20 8980 6533

The Probabilitic DB+IR Team at Work

Thomas Roelleke: Abstraction	Hany Azzam: Semantic Data Retrieval	Miguel Martinez-Alvarez: Classification Logic	Marco Bonzanini: Knowledge-based Summarisation	Sirvan Yahyaei: Machine Translation

Alumni

Fred Forst: Summarisation Logic, PhD 2008	Hengzhi Wu: DB+IR Efficiency, PhD 2010	Jun Wang: Retrieval Models and Theory, PhD 2010

Research profile

Information retrieval models: theories, foundations & relationships
DB+IR: Seamless integration of database (DB) and information retrieval (IR) technologies
Descriptive and logical modelling of retrieval models and tasks
High-level abstraction and coding in probabilistic Datalog (PDatalog)
Probabilistic and object-relational knowledge representations
Knowledge-oriented (semantic) retrieval, classification and summarisation
Large-scale probabilistic reasoning

Harmony Assumptions: Extending Probability Theory for Information Retrieval LWDA 2016 (Lernen, Wissen, Daten, Analysen): Slides
Harmony Assumptions in Information Retrieval and Social Networks Full Text, Computer Journal
Many years of research on IR models in a small book: IR Models: Foundations & Relationships: Morgan & Claypool List of Corrections
Tutorial at ICTIR 2013 IR Models: Foundations & Relationships
Musing on the theory of it all, By Ingo Frommholz

Some pointers to slides and papers in the period 2008 - 2012:

TF-IDF Uncovered (IR Theory), Yahoo Labs Barcelona, January 2012 slides (pdf)
IR Models, IR Herbstschule, Dagstuhl, September 30th slides (pdf)
IPI ConfEx 2009 IR4IP (Information Retrieval for Intellectual Property) Tutorial. IPI ConfEx home; IPI ConfEx workshops; handout.
Database Spektrum 2009: Information Retrieval: Concepts and Practical Considerations for Teaching a Rising Topic: Daniel Blank, Norbert Fuhr, Andreas Henrich, Thomas Mandl, Thomas Roelleke, Hinrich Schuetze, Benno Stein. article (pdf). Springer link.
VLDB Journal 2008 paper: Roelleke/Wu/Wang/Azzam: Modelling Retrieval in a Probabilistic Relational Algebra with a new Operator: The relational Bayes. Springerlink portal, local download.
Talks about the relational Bayes:

SIGIR 2008 paper: Roelleke/Wang: TF-IDF Uncovered: A Study of Theories and Probabilities. ACM portal, DBLP entry. Slides
IR Herbstschule Schloss Dagstuhl 2008: Retrieval Models. Slides.

Some pointers to slides and papers before 2008:

CWI Amsterdam Dec 2007 seminar talk: Modelling Retrieval Models with the relational Bayes. slides (pdf).
Max Planck Institute Saarbruecken Oct 2006 seminar talk: On Information Retrieval Models and DB+IR. slides.
SIGIR 2006 paper: Roelleke/Wang: A Parallel Derivation of Probabilistic Retrieval Models. ACM portal, DBLP entry.
Barcelona Universitat Pompeu Fabrai Mar 2006 seminar talk: On the Generalisation and Logical Implementation of Retrieval Models. slides.
IPM Journal 2006 paper: Roelleke/Tsikrika/Kazai: A General Matrix Framework for Modelling Information Retrieval. IPM download, Science Direct IPM Volume 42(1), DBLP entry.
Glasgow IR Festival 2005 talk: On Occurrence and Informativeness Probabilities. slides.
SIGMOD 2005 panel: DB and IR - Rethinking the great divide. slides. SIGMOD record report, Dec 2005 Volume 34(2) by Sihem Amer-Yahia.

Teaching

HySpirit

Other