Lexicon Multiword Candidates

This page describes the Lexicon multiword candidate list generation and post-proessing processes:

• I. LMW Candidate Generation Process
  LSG developed serveral models to generate high precision multiword candidate list for the Lexicon buidling.
  • Expansions of abbreviations and acronyms in Lexicon Annual Releases
  • Expansions of Acronyms Pattern Matcher from the MNS
  • CUI and Endword Matcher from the DMNS
  • SpVar Pattern Matcher With Frequency from the DMNS

  • WordNet Data
  • Consumer Data

  • Other Patterns

• II. LMW Candidate Post-Process (after tagged)
  Computer programs are developed and conducted in the post-processes after a LMW candidate file is completed to find the performance for the specific candidate file and over-all candidate files.
  • LMW Candidate Post-Processes Overview
  • LMW Candidate Post-Processes Details

  • Results: performance of Previous LMW Candidate Lists

• III. MW Candidate Post-Process Associated files and Programs
  • lexAbbAcrExpCand

  • Not Base Forms and LMW (Lc) Files in the LexCheck
  • Total Data Set and Usage

  • validLmwFromLexicon
  • notBaseLmwFiles