1WNGLOSS(7) WordNet™ WNGLOSS(7)
2
6 wngloss - glossary of terms used in WordNet system
7
9 The WordNet Reference Manual consists of Unix-style manual pages
10 divided into sections as follows:
11 ┌────────┬─────────────────────────────────────────┐
14 │Section │ Description │
15 ├────────┼─────────────────────────────────────────┤
16 │ 1 │ WordNet User Commands │
17 │ 3 │ WordNet Library Functions │
18 │ 5 │ WordNet File Formats │
19 │ 7 │ Miscellaneous Information about WordNet │
20 └────────┴─────────────────────────────────────────┘
21 System Description
23 The WordNet system consists of lexicographer files, code to convert
24 these files into a database, and search routines and interfaces that
25 display information from the database. The lexicographer files orga‐
26 nize nouns, verbs, adjectives and adverbs into groups of synonyms, and
27 describe relations between synonym groups. grind(1) converts the lexi‐
28 cographer files into a database that encodes the relations between the
29 synonym groups. The different interfaces to the WordNet database uti‐
30 lize a common library of search routines to display these relations.
31 Note that the lexicographer files and grind(1) program are not gener‐
32 ally distributed.
33 Database Organization
36 Information in WordNet is organized around logical groupings called
37 synsets. Each synset consists of a list of synonymous words or collo‐
38 cations (eg. "fountain pen", "take in"), and pointers that describe the
39 relations between this synset and other synsets. A word or collocation
40 may appear in more than one synset, and in more than one part of
41 speech. The words in a synset are grouped such that they are inter‐
42 changeable in some context.
43 Two kinds of relations are represented by pointers: lexical and seman‐
45 tic. Lexical relations hold between semantically related word forms;
46 semantic relations hold between word meanings. These relations include
47 (but are not limited to) hypernymy/hyponymy (superordinate/subordi‐
48 nate), antonymy, entailment, and meronymy/holonymy.
49 Nouns and verbs are organized into hierarchies based on the hyper‐
51 nymy/hyponymy relation between synsets. Additional pointers are be
52 used to indicate other relations.
53 Adjectives are arranged in clusters containing head synsets and satel‐
55 lite synsets. Each cluster is organized around antonymous pairs (and
56 occasionally antonymous triplets). The antonymous pairs (or triplets)
57 are indicated in the head synsets of a cluster. Most head synsets have
58 one or more satellite synsets, each of which represents a concept that
59 is similar in meaning to the concept represented by the head synset.
60 One way to think of the adjective cluster organization is to visualize
61 a wheel, with a head synset as the hub and satellite synsets as the
62 spokes. Two or more wheels are logically connected via antonymy, which
63 can be thought of as an axle between the wheels.
64 Pertainyms are relational adjectives and do not follow the structure
66 just described. Pertainyms do not have antonyms; the synset for a per‐
67 tainym most often contains only one word or collocation and a lexical
68 pointer to the noun that the adjective is "pertaining to". Participial
69 adjectives have lexical pointers to the verbs that they are derived
70 from.
71 Adverbs are often derived from adjectives, and sometimes have antonyms;
73 therefore the synset for an adverb usually contains a lexical pointer
74 to the adjective from which it is derived.
75 See wndb(5) for a detailed description of the database files and how
77 the data are represented.
78
80 Many terms used in the WordNet Reference Manual are unique to the Word‐
81 Net system. Other general terms have specific meanings when used in
82 the WordNet documentation. Definitions for many of these terms are
83 given to help with the interpretation and understanding of the refer‐
84 ence manual, and in the use of the WordNet system.
85 In following definitions word is used in place of word or collocation.
87 adjective cluster A group of adjective synsets that are orga‐
89 nized around antonymous pairs or triplets. An
90 adjective cluster contains two or more head
91 synsets which represent antonymous concepts.
92 Each head synset has one or more satellite
93 synsets.
94 attribute A noun for which adjectives express values.
96 The noun weight is an attribute, for which the
97 adjectives light and heavy express values.
98 base form The base form of a word or collocation is the
100 form to which inflections are added.
101 basic synset Syntactically, same as synset. Term is used
103 in wninput(5) to help explain differences in
104 entering synsets in lexicographer files.
105 collocation A collocation in WordNet is a string of two or
107 more words, connected by spaces or hyphens.
108 Examples are: man-eating shark, blue-collar,
109 depend on, line of products. In the database
110 files spaces are represented as underscore (_)
111 characters.
112 coordinate Coordinate terms are nouns or verbs that have
114 the same hypernym.
115 cross-cluster pointer A semantic pointer from one adjective cluster
117 to another.
118 derivationally related forms
120 Terms in different syntactic categories that
121 have the same root form and are semantically
122 related.
123 direct antonyms A pair of words between which there is an
125 associative bond resulting from their frequent
126 co-occurrence. In adjective clusters, direct
127 antonyms appears only in head synsets.
128 domain A topical classification to which a synset has
130 been linked with a CATEGORY, REGION or USAGE
131 pointer.
132 domain term A synset belonging to a topical class. A
134 domain term is further identified as being a
135 CATEGORY_TERM, REGION_TERM or USAGE_TERM.
136 entailment A verb X entails Y if X cannot be done unless
138 Y is, or has been, done.
139 exception list Morphological transformations for words that
141 are not regular and therefore cannot be pro‐
142 cessed in an algorithmic manner.
143 group Verb senses that similar in meaning and have
145 been manually grouped together.
146 gloss Each synset contains gloss consisting of a
148 definition and optionally example sentences.
149 head synset Synset in an adjective cluster containing at
151 least one word that has a direct antonym.
152 holonym The name of the whole of which the meronym
154 names a part. Y is a holonym of X if X is a
155 part of Y.
156 hypernym The generic term used to designate a whole
158 class of specific instances. Y is a hypernym
159 of X if X is a (kind of) Y.
160 hyponym The specific term used to designate a member
162 of a class. X is a hyponym of Y if X is a
163 (kind of) Y.
164 indirect antonym An adjective in a satellite synset that does
166 not have a direct antonym has an indirect
167 antonyms via the direct antonym of the head
168 synset.
169 instance A proper noun that refers to a particular,
171 unique referent (as distinguished from nouns
172 that refer to classes). This is a specific
173 form of hyponym.
174 lemma Lower case ASCII text of word as found in the
176 WordNet database index files. Usually the
177 base form for a word or collocation.
178 lexical pointer A lexical pointer indicates a relation between
180 words in synsets (word forms).
181 lexicographer file Files containing the raw data for WordNet
183 synsets, edited by lexicographers, that are
184 input to the grind program to generate a Word‐
185 Net database.
186 lexicographer id (lex id)
188 A decimal integer that, when appended onto
189 lemma, uniquely identifies a sense within a
190 lexicographer file.
191 monosemous Having only one sense in a syntactic category.
193 meronym The name of a constituent part of, the sub‐
195 stance of, or a member of something. X is a
196 meronym of Y if X is a part of Y.
197 part of speech WordNet defines "part of speech" as either
199 noun, verb, adjective, or adverb. Same as
200 syntactic category.
201 participial adjective An adjective that is derived from a verb.
203 pertainym A relational adjective. Adjectives that are
205 pertainyms are usually defined by such phrases
206 as "of or pertaining to" and do not have
207 antonyms. A pertainym can point to a noun or
208 another pertainym.
209 polysemous Having more than one sense in a syntactic cat‐
211 egory.
212 polysemy count Number of senses of a word in a syntactic cat‐
214 egory, in WordNet.
215 postnominal A postnominal adjective occurs only immedi‐
217 ately following the noun that it modifies.
218 predicative An adjective that can be used only in predi‐
220 cate positions. If X is a predicate adjec‐
221 tive, it can only be used in such phrases as
222 "it is X" and never prenominally.
223 prenominal An adjective that can occur only before the
225 noun that it modifies: it cannot be used pred‐
226 icatively.
227 satellite synset Synset in an adjective cluster representing a
229 concept that is similar in meaning to the con‐
230 cept represented by its head synset.
231 semantic concordance A textual corpus (e.g. the Brown Corpus) and a
233 lexicon (e.g. WordNet) so combined that every
234 substantive word in the text is linked to its
235 appropriate sense in the lexicon via a seman‐
236 tic tag.
237 semantic tag A pointer from a word in a text file to a spe‐
239 cific sense of that word in the WordNet data‐
240 base. A semantic tag in a semantic concor‐
241 dance is represented by a sense key.
242 semantic pointer A semantic pointer indicates a relation
244 between synsets (concepts).
245 sense A meaning of a word in WordNet. Each sense of
247 a word is in a different synset.
248 sense key Information necessary to find a sense in the
250 WordNet database. A sense key combines a
251 lemma field and codes for the synset type,
252 lexicographer id, lexicographer file number,
253 and information about a satellite's head
254 synset, if required. See senseidx(5) for a
255 description of the format of a sense key.
256 subordinate Same as hyponym.
258 superordinate Same as hypernym.
260 synset A synonym set; a set of words that are inter‐
262 changeable in some context without changing
263 the truth value of the preposition in which
264 they are embedded.
265 troponym A verb expressing a specific manner elabora‐
267 tion of another verb. X is a troponym of Y if
268 to X is to Y in some manner.
269 unique beginner A noun synset with no superordinate.
271WordNet 3.0 Dec 2006 WNGLOSS(7)