The LCS data is stored in file verbs-English.lcs.
----------------------------- REFERENCES --------------------------
Dorr, Bonnie J., "Large-Scale Dictionary Construction for Foreign Language Tutoring and Interlingual Machine Translation," Machine Translation, 12:4, pp. 271--322, 1997. http://www.umiacs.umd.edu/users/bonnie/Publications/icall-97.pdf
Dorr, Bonnie J. and Mari Broman Olsen, "Multilingual Generation: The Role of Telicity in Lexical Choice and Syntactic Realization", Machine Translation, 11:1-3, pages 37-74, 1996. URL = http://www.umiacs.umd.edu/users/bonnie/Publications/lex-choice-96.pdf
Dorr, Bonnie J. and Douglas Jones, "Acquisition of Semantic Lexicons: Using Word Sense Disambiguation to Improve Precision", in Breadth and Depth of Semantic Lexicons, Evelyne Viegas (ed.), Kluwer Academic Publishers, pages 79-98, Norwell, MA, 1999. URL = http://www.umiacs.umd.edu/users/bonnie/Publications/kluwer-00.pdf
Fellbaum, Christiane, WordNet: An Electronic Lexical Database, MIT Press, 1998. URL = http://mitpress.mit.edu/books/wordnet
Green, Rebecca, Lisa Pearl, Bonnie J. Dorr, and Philip Resnik, "Lexical Resource Integration across the Syntax-Semantics Interface", in Proceedings of the NAACL Workshop on WordNet and Other Lexical Resources: Applications, Customizations, Carnegie Mellon University, Pittsburg, PA, pages 71-76, 2001a. URL = http://www.umiacs.umd.edu/users/bonnie/Publications/naacl-01.pdf
Green, Rebecca, Lisa Pearl, Bonnie J. Dorr, and Philip Resnik, "Mapping WordNet Senses to a Lexical Database of Verbs", in Proceedings of the 39th Annual Meeting of the Association for Computational Linguistics, Toulouse, France, pages 244-251, 2001b. URL = http://www.umiacs.umd.edu/users/bonnie/Publications/acl-01.pdf
Levin, Beth, "English Verb Classes and Alternations: A Preliminary Investigation", University of Chicago Press, Chicago, IL, 1993.
Olsen, Mari Broman, Bonnie J. Dorr, and Scott C. Thomas, "Toward Compact Monotonically Compositional Interlingua Using Lexical Aspect," in Proceedings of the Workshop on Interlinguas in MT, MT Summit, New Mexico State University Technical Report MCCS-97-314, San Diego, CA, pages 33-44, 1997. URL = http://www.umiacs.umd.edu/users/bonnie/Publications/il-97.pdf
Palmer, Martha, Joseph Rosenzweig and Scott Cotton, "Automatic Predicate Argument Analysis of the Penn TreeBank," in Proceedings of HLT 2001, First International Conference on Human Language Technology Research, J. Allan, ed., Morgan Kaufmann, San Francisco, 2001. http://hlt2001.org/papers/hlt2001-10.pdf
NOTES ON THE ENTRIES
This is a lisp-readable file of property lists, each with slots such as:
:DEF_WORD The surface word in the language of interest :ROMAN A romanized form of that word :GLOSS An English gloss for the word :GLOSS_HEAD The head of the english gloss :CLASS The Levin class from Levin (1993) :WN_SENSE A hand-tagged WordNet sense by Dorr and Jones (1996) :PROPBANK 1 or more comparable PropBank-style role-list strings :THETA_ROLES A thematic grid produced by Dorr (1994-1996) with augmentations by Dorr, Olsen, Thomas (1997-1999) :LCS An LCS representation automatically produced by routines written by Dorr (1997) with augmentations by Thomas :VAR_SPEC Featural information about variable positions in the LCS by Dorr (1997) :COLLOCATIONS Collocations
Any of the slots may be missing, though meaningful entries will have :DEF_WORD, :THETA_ROLES and :LCS slots.
Slots such as :WN_SENSE, :VAR_SPEC and :COLLOCATIONS, if missing, are equivalent to NIL.
If :ROMAN or :GLOSS are missing, they are equivalent to :DEF_WORD. If :GLOSS_HEAD is missing, it is equivalent to :GLOSS.
The `Grid line' is a line of #-separated fields, used as input to the LCS-generation process. It is included in some LCS files mainly for debugging purposes. Its format is:
cl-no#g-no#grid#wd#rom#gloss#cnsts#wn-senses#field-etc#incorps#comments
where
cl-no = verb-class number g-no = grid number wd = the word, in some language-specific encoding rom = romanization of the word gloss = English gloss of the word cnsts = `constants' utilized in creating the LCS wn-senses = word-net offsets field-etc = other info for creating the LCS, such as semantic field incorps = other material to be incorporated into the LCS comments = arbitrary comments on this entry, as desired
Any of the fields may be blank. I.e.,
##########
is a valid grid line, though in practice, at least the word' and field-etc' fields must be present for a usable LCS entry to be created. (Older subroutines in the LCS-generation code require the class-number' and `word' fields.)
In some LCS files (i.e., lisp readable files), these grid lines are prepended with the string `;; Grid: '. A lisp-readable form with the corresponding LCS should follow, except in cases where the LCS-generation process failed on the given grid line, for whatever reason.
Release Data: LCS_0_17 (1.1) English_Lexicon_0_15 (1.1) AD-verbs-English.edited-grid,v 1.2 2001/10/24 19:28:40 (UTC) BJDMAP-verbs-English.edited-grid,v 1.2 2001/10/24 19:28:40 (UTC) EM-verbs-English.edited-grid,v 1.2 2001/10/24 19:28:40 (UTC) NR-verbs-English.edited-grid,v 1.2 2001/10/24 19:28:41 (UTC) SZ-verbs-English.edited-grid,v 1.2 2001/10/24 19:28:41 (UTC)