MIRO: Multimedia Information Retrieval
Programme: THIRD FRAMEWORK PROGRAMME Programme Acronym: ESPRIT 3 Contract Type: NO CONTRACT TYPE Start Date: 1993-01-01 End Date: 1995-12-31 Contract No: N/A Role for TUC/MUSIC: Contractor Funding for TUC/MUSIC: 20000.0 Euros Principal Investigator for TUC/MUSIC: Stavros Christodoulakis
The approach is multi-disciplinary, bringing together ideas from IR, DBMS, AI and HCI. A major aim is to develop example logics in the context of object-oriented approaches to data models, and storage structures within which new retrieval models can be described, implemented, and tested. The investigations of new retrieval models, based on formal logic and a theory of uncertainty, will address the claim that retrieval can be viewed as a form of inference. To produce convincing evidence of the practical viability of the work, the Group will study the scaleability and performance characteristics of models and designs for very large multimedia information bases. A major feature of the approach is that IR systems must be constructed to be adaptive to user interaction from the beginning. Such a process of adaptation will be useable only if thought is given to the design of appropriate user interfaces supporting the multimedia information-seeking and browsing activities of users. Hence the partners in their research will emphasise the relationship between user interfaces and the formal models.
Activities will be focused on four main themes: the modelling of retrieval, run-time support for information retrieval, human computer interaction and evaluation techniques. Meetings will be held on specific topics and a plenary workshop convened on the research activity of all the participants. Representatives from industry will be invited to attend the workshop and to evaluate the lines of research and the results produced. If specific topics are found to be of special interest or relevance, the Group will organise appropriate workshops. An e-m distribution list has been already set up for the fast dissemination of information among Group partners.
The cooperation between the research groups involved in this project will promote or reinforce new research areas related to multimedia information retrieval. At first it will greatly facilitate the integration of the multiple aspects of the problem and the corresponding approaches within a common theoretical framework, so that results will initially have their main impact on existing research projects and prototypes on one hand, and improvement scientific collaboration IR community on the other hand. In the longer term, the potential for exploitation of these results by the IT industry is enormous, as the increasing use of wide-area networks and multimedia technologies is dramatically increasing the need for the new type of IR system that the Group is investigating.
Swiss Federal Institute of Technology Zurich
University of Glasgow
Consiglio Nationale delle Ricerche - Instituto di Elaborazione della Informazione
Gesellschaft fur Mathematik and Datenverarbeitung - IPSI
Université Joseph Fourier du Grenoble