This requires selecting a subset of the Digital Libraries which most probably contains relevant documents with respect to the user’s information need. The second task is that of distributed retrieval. The user's information need has to be processed against the resources selected in the first step. The third task is to fuse the results obtained from the selected resources. The retrieval results need to be merged into a single result and visualised in an effcetive way for the user.

Impact/deliverables
The key milestones of the project are months 6, 12, 16, 24 and 30 (end of project) and their associated deliverables. For the four phases of the project these are:

MS1. Phase 1: The specification of the architecture of the MIND prototype and the identification of target categories of users and Digital Library contents (3rd month); the specification of resource description formats for text, image, and speech resources (6th month).

MS2. Phase 2: The specification of models and methods for resource selection and data fusion for homogeneous resources (12th month).

MS3. Phase MS2: The implementation of tools for automatic extraction and verification of resource descriptions (16 month); the implementation of tools for resource selection and data fusion for homogeneous resources (16th month).

MS4. Phase 3: The definition of methods for automatic generation of surrogates and hierarchical overview of fused results, accompanied by prototype tools (21st month); The specification of methods for resource selection and data fusion for heterogeneous collections, accompanied by prototype tools (24th month).

MS5. Phase 4: A prototype system for resource selection and data fusion, which will have been subjected to an extensive evaluation by various user groups (30th month), the main findings of which will be publicised and discussed at an end of project workshop (30th month).