Air Travel Information Service Phase III (ATIS3) Speech and Natural Language Understanding Corpora December 1994 Multi-Site Evaluation Test Material NIST Speech Disc 17-5.1 March, 1995 This disc contains an evaluation test set drawn for the ATIS3 pool of multi-site data to be used in the December 1994 ARPA SLS benchmark tests. This test set is similar in scope and content to previous ATIS3 evaluation test sets. If you intend to implement the test protocols for the December 1994 ARPA SLS Benchmark Tests, do not examine the enclosed transcriptions and documentation until after you have run the tests. Index (*.ndx) files have been included which specify the exact test material to be used for each test under "atis3/sp_tst/dec94". Instructions for running the tests and preparing your results for scoring by NIST are included in the file, "readme.doc" under the "atis3/sp_tst/dec94" directory. PLEASE READ THIS FILE IN ITS ENTIRETY BEFORE EXAMINING ANY DATA OR DOCUMENTATION ON THIS DISC. A summary of the contents of this disc is as follows: atis3/ Waveforms, transcriptions, annotations, and documentation for the December 1994 ATIS3 Evaluation Test set. comp/ NIST comparator for scoring CAS-formatted answers output from ATIS NL/SLS systems against reference answers. rdb4_0/ ATIS 46-city/52-airport relational database. score/ NIST speech recognition scoring software. Includes dynamic programming string-alignment scoring code and statistical significance tests. sphere/ NIST SPeech HEader REsources toolkit. Provides command- line and programmer interface to NIST-headered speech waveform files. Also provides for automatic decompression of the Shorten-compressed waveform files on these discs. 17_5_1.dir Directory of files on this disc General information files named "readme.doc" have been included in the high-level directories on this disc and describe the contents of the directories. Note that the annotations on this disc conform to a new Principles of Interpretation (PofI) document which is a revision from previous ATIS3 PofI. This PofI pertains only to the 1994 Development Test Data (CD17-4.1) and Evaluation Test Data (this disc). The revised PofI is located under the "atis3/doc/pofi" directory. Note that the waveforms on this disc have been compressed using SPHERE-embedded Shorten. The collection of the ATIS3 corpus was sponsored by the Advanced Research Projects Agency Software and Intelligent Systems Technology Office (ARPA-SISTO). The corpus was annotated by SRI international and collated, documented and produced on CD-ROM by the National Institute of Standards and Technology under the sponsorship of the Linguistic Data Consortium.