Cross-Lingual Information access, retrieval, extraction, ... needs, requirements and current state of the art

The main topics that were addressed were the following:\\ * The needs of large users of multilingual information archives and their requirements, expectations * The current offers for enterprise search and web * What is preventing the take off of cross-lingual search applications The theme of the panel discussion was why, despite 10 years of MLIA evaluations (within TREC, CLEF, NTCIR programmes), the adoption of such technologies is still a questionable issue. This session was organized around a series of questions which were circulated to the panelist beforehand. The list of questions was as follows:\\ * During the last ten years or so, the number of languages and domains, for which research and development on MLIA have been conducted, has increased massively; is there any killer application that emerged from such clusters? * Have we managed to assess performance and are these in line with application deployments; are these two concepts strongly related (performance, deployments)? * Did the evaluations consider the right needs and requirements, and by the way do we have a clear picture of these? * If no success story can be mentioned today, does that mean key players were left out of these evaluations? Why? * If players were among the participants, does that mean we do have applications/core-technology with performances that can not be offered to real users? * And what offers do MLIA players have that one can highlight today, in particular for use by professionals? * Can we imagine fulfilling the needs of all languages and all domains at a reasonable cost (money, time, ...)? * How can the panelists (and the participants) address these issues and establish pointers to high levels of cooperation between users, service providers and core-technology providers (including computational linguistics?