ITL published the proceedings of the seventh annual Text REtrieval Conference in August 1999 as NIST Special Publication 500-242, The Seventh Text REtrieval Conference (TREC-7). The TREC workshop series is sponsored by NIST and the Department of Defense to support the text retrieval industry by providing the infrastructure necessary for large-scale evaluation of text retrieval methodologies. TREC-7, the most recent workshop in the series, was held at NIST on November 9-11, 1998, with 150 participants. Fifty-six groups from 13 different countries and 19 companies attended.

The proceedings contain an overview summarizing the retrieval tasks and main results of the conference, papers that were presented at the conference, and complete evaluations of individual group results. The proceedings also contain "track" reports, where a track is focused work on a particular retrieval subproblem. This year's reports include a summary of the "Very Large Corpus" (VLC) track in which participants searched 100 GB of Web pages. VLC systems were not only forced to handle much greater amounts of data (standard TREC collections are approximately 2 GB), but also to accommodate the different genre of data represented by Web documents. An electronic version of the proceedings is available on the TREC Web page, http://trec.nist.gov.

In other TREC news, the Japanese National Center for Science Information Systems (NACSIS) held the first NTCIR Workshop on Research in Japanese Text Retrieval the end of August in Tokyo. The evaluation leading to this workshop was heavily based on the NIST TREC model of evaluation; ITL supported this effort with advice and tools during the year. Donna Harman presented the keynote for this workshop, which was attended by 28 participating groups including the major Japanese companies involved in text retrieval and 3 U.S. groups.

CONTACT: Ellen Voorhees, ext. 3761

Donna Harman, ext. 3569