SP500215 NIST Special Publication 500-215: The Second Text REtrieval Conference (TREC-2) Appendix B: System Features Appendix National Institute of Standards and Technology D. K. Harman HA. QUERY CONSThUCIION -- AUTOMAIICALLY BUILT QUERIES (ADHOC SYSThM NAME [OCRerr] ADS f UIc [OCRerr] DAUIOUSIE__L__MEAl) UIF J IDSR Topic Fields Concept fields: each unique word stem Total Computer Time to Build (cpu sec) 1 second Methods Used to Construct None - Term Weighting (1)ased on topic terms) - Phrase extraction - Syntactic Parsing ___________________ _________________ ________________ ___________ - Word Sense Disambiguation __________________ ________________ ________________ __________ Proper Noun Identification - Tokenizer -- Which Patterns - Heuristic Associations - Expansion of Queries Using Previous Constructed Data Structure -- Which Structure - Automatic addition of Boolean Connectors/Proximity Operators __________________ ________________ _______________ __________ - Other i