Scientific Goals
The IRF aims to promote and facilitate research in Information Retrieval:
- From large scale collections of structured, semi-structured and multimedia documents;
- By real users undertaking real search tasks interactively;
- Using novel and sophisticated document analysis and indexing techniques;
- By providing an easy to access research infrastructure including hardware platforms, software, and test collections.
Research projects
The IRF has been engaged in several research projects on a national and European level involving a wide range of technologies: Automated information extraction, automated image analysis, cross-language search, adaptive user interfaces, machine translation, etc.
Research Network
More than 250 scientists from around the world have joined the IRF as scientific members to access the large scale research infrastructure, conduct joint research projects and interact with information professionals from industry. Our research network spreads over 4 continents and includes most prestigious institutions.
Research Areas
- Automated document categorization
- Machine translation
- Image Mining in technical documents
- Information Visualization
- Interfaces and Workflows
- Professional search
Publications & Reports
The IRF is very active in publishing its research results. This page contains a list of all IRF publications. Furthermore, there are technical reports on research performed by the IRF and commissioned by the IRF available.
Evaluation Tracks
The IRF has launched two evaluation tracks dedicated to cross-lingual patent search and chemical text search within the two major international evaluation workshops in the USA and Europe: TREC-CHEM (Chemistry track) and CLEF-IP (Intellectual Property track).
Large Scale Infrastructure
The IRF maintains a semantic supercomputer infrastructure that facilitates large scale IR experiments and provides access to a large and high-quality corpus of scientific and technical data.