Objective
This research addresses problems associated with the emergence of thousands of heterogeneous multimedia digital libraries distributed internationally on multiple platforms. Users have problems with resource selection as they are unaware of the contents of each individual library in terms of quantity, quality, information type, provenance, and likely relevance. Once a set of relevant libraries has been selected the user must organise and interpret the information in a common format and environment. Typically this is performed through visual evaluation and ad hoc integration, which forces users to restrict their attention to a small subset of the information retrieved. MIND will assist users to know where to search, how to query different media, and how to combine information from diverse sources.
Objectives:
The key objective of the MIND project is to address the problems faced by users in terms of their ability to access and exploit the increasing number of digital libraries available internationally through networks, like the Internet and the World Wide Web (WWW). More specifically MIND aims at designing models and building set of tools and associated test-beds to improve the effectiveness of resource selection, multimedia information access, retrieval and fusion of the retrieval data.
The achievement of these objectives will involve:
- the development of a variety of metadata generation methods for different media;
- the design of algorithms for improving the selection of the most relevant collections of information;
- the development of data fusion techniques for merging ranked lists of items retrieved from the selected collections;
- the evaluation of these methods on different sources of data and cohorts of users.
Work description:
MIND addresses issues that arise when people have routine access to thousands of heterogeneous and distributed multimedia Digital Libraries. When so many Digital Libraries are available, the first information access task is resource selection (that is the selection of the Digital Library/ies that is most likely to contain the information sought). This is predominantly an ineffective manual task as users are unaware of the contents of each individual library in terms of quantity, quality, information type, provenance and likely relevance. People need accurate automatic resource selection tools to assist them. Once a set of libraries is selected and searched, a person must organise and interpret the (often multimedia) information supplied by the different Digital Libraries. This involves data fusion (i.e. the merging of different data retrieved from different Digital Libraries) and information visualisation (in particular visualisation techniques aimed at displaying a summary of the retrieved data). Typically this is performed through visual evaluation and ad hoc integration, which forces users to restrict their attention to a small subset of the information retrieved. As the number of Digital Libraries increases the problem is exacerbated and the aspirations of information providers to increase access to scarce and/or unique resources is hindered. People need user interfaces that enable them to more fully exploit the multimedia information they find. MIND will assist users to know where to search, how to query different media, and how to combine information from diverse sources.
The information access environment that we envision requires advances in several different areas of research of which MIND partners are world known experts. MIND is an end-to-end solution, covering how Digital Libraries are described to external parties, how appropriate resources are selected automatically, how text, image and audio (recorded speech) databases are searched, and how multimedia search results are displayed. Solutions deployed on a worldwide scale require a solid theoretical foundation capable of coping with significant heterogeneity, which we will develop. Every aspect of the research will be tested, at the component level, and through a set of test-beds and user studies covering research and resources produced by several sites around the world.
Milestones:
The key milestones of the project are months 4, 16, 24 and 30 (end of project) and their associated reports, prototypes and workshop.
These are:
1. A report on the test-bed architecture and on the identification of target categories if users and Digital Libraries informative contents;
2. A report on theoretical work on content metadata extraction and verification and a report on methods for resource selection and data fusion for homogeneous collections (accompanied by prototype tools);
3. A report on methods for data fusion, automatic generation of surrogates and hierarchical overview of fused results and a report on methods for resource selection and data fusion for heterogeneous collections (accompanied by prototype tools);
4. A demonstrator of a system for resource selection and data fusion, which will have been subjected to an extensive evaluation by various user groups.
MIND has delivered the expected outcomes. More specifically it has produced: a theoretically founded resource selection technique for different data and media types; a schema mapping technique using DAML+OIL, XSLT and probabilistic logics; a specification of the MIND system which is highly distributed and which utilises separate components for each task which run on different machines; a new method for resource description of image libraries; a new method for data fusion of image libraries; wrappers for various digital libraries which receive a query, transform it, send the transformation to a third-party digital library, receive documents, transform them into a standard format, and present them to the user; and a multi-objective resource selection method.
Fields of science (EuroSciVoc)
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
- natural sciencescomputer and information sciencesdatabases
- natural sciencescomputer and information sciencesinternetworld wide web
You need to log in or register to use this function
Call for proposal
Data not availableFunding Scheme
CSC - Cost-sharing contractsCoordinator
G1 1XQ GLASGOW
United Kingdom