Terrific feedback and thanks very much. This is exactly what we hope will come out of testbed uses.

FYI. The AMICO working meeting in Los Angeles in September formed an "Editorial Committee" which we hope will deal with some of the inconsistency problems systematically. Meanwhile we've just introduced an online web based editing tool for AMICO members that gives them the ability to report term occurences to help identify less systemmatic problems. We're hoping that these two approaches contribute to improved data both in the short term (this year) and the future.

Also, we're also exploring AAT, ULAN and TGN term matching. We've run all the AMICO terminology against these vocabularies and diagnosed the nature of the inconsistencies in the data and the lacunae in the vocabularies. We're currently studying ways to augment each and RLG is exploring a query assistant tool that will exploit the vocabularies between users and the data next year. We'll report more on these as they go forward.