WNINPUT(5WN) WordNet™ File Formats WNINPUT(5WN)NAME
noun.suffix, verb.suffix, adj.suffix, adv.suffix - WordNet lexicogra‐
pher files that are input to grind(1WN)DESCRIPTION
WordNet's source files are written by lexicographers. They are the
product of a detailed relational analysis of lexical semantics: a vari‐
ety of lexical and semantic relations are used to represent the organi‐
zation of lexical knowledge. Two kinds of building blocks are distin‐
guished in the source files: word forms and word meanings. Word forms
are represented in their familiar orthography; word meanings are repre‐
sented by synonym sets (synsets) - lists of synonymous word forms that
are interchangeable in some context. Two kinds of relations are recog‐
nized: lexical and semantic. Lexical relations hold between word
forms; semantic relations hold between word meanings.
Lexicographer files correspond to the syntactic categories implemented
in WordNet - noun, verb, adjective and adverb. All of the synsets in a
lexicographer file are in the same syntactic category. Each synset
consists of a list of synonymous words or collocations (eg. "fountain
pen", "take in"), and pointers that describe the relations between this
synset and other synsets. These relations include (but are not limited
to) hypernymy/hyponymy, antonymy, entailment, and meronymy/holonymy. A
word or collocation may appear in more than one synset, and in more
than one part of speech. Each use of a word in a synset represents a
sense of that word in the part of speech corresponding to the synset.
Adjectives may be organized into clusters containing head synsets and
satellite synsets. Adverbs generally point to the adjectives from
which they are derived.
See wngloss(7WN) for a glossary of WordNet terminology and a discussion
of the database's content and logical organization.
Lexicographer File Names
The names of the lexicographer files are of the form:
pos.suffix
where pos is either noun, verb, adj or adv. suffix may be used to
organize groups of synsets into different files, for example noun.ani‐
mal and noun.plant. See lexnames(5WN) for a list of lexicographer file
names that are used in building WordNet.
Pointers
Pointers are used to represent the relations between the words in one
synset and another. Semantic pointers represent relations between word
meanings, and therefore pertain to all of the words in the source and
target synsets. Lexical pointers represent relations between word
forms, and pertain only to specific words in the source and target
synsets. The following pointer types are usually used to indicate lex‐
ical relations: Antonym, Pertainym, Participle, Also See, Derivation‐
ally Related. The remaining pointer types are generally used to repre‐
sent semantic relations.
A relation from a source to a target synset is formed by specifying a
word from the target synset in the source synset, followed by the
pointer_symbol indicating the pointer type. The location of a pointer
within a synset defines it as either lexical or semantic. The Lexicog‐
rapher File Format section describes the syntax for entering a semantic
pointer, and Word Syntax describes the syntax for entering a lexical
pointer.
Although there are many pointer types, only certain types of relations
are permitted between synsets of each syntactic category.
The pointer_symbols for nouns are:
! Antonym
@ Hypernym
@i Instance Hypernym
∼ Hyponym
∼i Instance Hyponym
#m Member holonym
#s Substance holonym
#p Part holonym
%m Member meronym
%s Substance meronym
%p Part meronym
= Attribute
+ Derivationally related form
;c Domain of synset - TOPIC
-c Member of this domain - TOPIC
;r Domain of synset - REGION
-r Member of this domain - REGION
;u Domain of synset - USAGE
-u Member of this domain - USAGE
The pointer_symbols for verbs are:
! Antonym
@ Hypernym
∼ Hyponym
* Entailment
> Cause
^ Also see
$ Verb Group
+ Derivationally related form
;c Domain of synset - TOPIC
;r Domain of synset - REGION
;u Domain of synset - USAGE
The pointer_symbols for adjectives are:
! Antonym
& Similar to
< Participle of verb
\ Pertainym (pertains to noun)
= Attribute
^ Also see
;c Domain of synset - TOPIC
;r Domain of synset - REGION
;u Domain of synset - USAGE
The pointer_symbols for adverbs are:
! Antonym
\ Derived from adjective
;c Domain of synset - TOPIC
;r Domain of synset - REGION
;u Domain of synset - USAGE
Many pointer types are reflexive, meaning that if a synset contains a
pointer to another synset, the other synset should contain a corre‐
sponding reflexive pointer. grind(1WN) automatically inserts missing
reflexive pointers for the following pointer types:
┌───────────────────────┬────────────────────────┐
│ Pointer │ Reflect │
├───────────────────────┼────────────────────────┤
│Antonym │ Antonym │
│Hyponym │ Hypernym │
│Hypernym │ Hyponym │
│Instance Hyponym │ Instance Hypernym │
│Instance Hypernym │ Instance Hyponym │
│Holonym │ Meronym │
│Meronym │ Holonym │
│Similar to │ Similar to │
│Attribute │ Attribute │
│Verb Group │ Verb Group │
│Derivationally Related │ Derivationally Related │
│Domain of synset │ Member of Doman │
└───────────────────────┴────────────────────────┘
Verb Frames
Each verb synset contains a list of generic sentence frames illustrat‐
ing the types of simple sentences in which the verbs in the synset can
be used. For some verb senses, example sentences illustrating actual
uses of the verb are provided. (See Verb Example Sentences in
wndb(5WN).) Whenever there is no example sentence, the generic sen‐
tence frames specified by the lexicographer are used. The generic sen‐
tence frames are entered in a synset as a comma-separated list of inte‐
ger frame numbers. The following list is the text of the generic
frames, preceded by their frame numbers:
1 Something ----s
2 Somebody ----s
3 It is ----ing
4 Something is ----ing PP
5 Something ----s something Adjective/Noun
6 Something ----s Adjective/Noun
7 Somebody ----s Adjective
8 Somebody ----s something
9 Somebody ----s somebody
10 Something ----s somebody
11 Something ----s something
12 Something ----s to somebody
13 Somebody ----s on something
14 Somebody ----s somebody something
15 Somebody ----s something to somebody
16 Somebody ----s something from somebody
17 Somebody ----s somebody with something
18 Somebody ----s somebody of something
19 Somebody ----s something on somebody
20 Somebody ----s somebody PP
21 Somebody ----s something PP
22 Somebody ----s PP
23 Somebody's (body part) ----s
24 Somebody ----s somebody to INFINITIVE
25 Somebody ----s somebody INFINITIVE
26 Somebody ----s that CLAUSE
27 Somebody ----s to somebody
28 Somebody ----s to INFINITIVE
29 Somebody ----s whether INFINITIVE
30 Somebody ----s somebody into V-ing something
31 Somebody ----s something with something
32 Somebody ----s INFINITIVE
33 Somebody ----s VERB-ing
34 It ----s that CLAUSE
35 Something ----s INFINITIVE
Lexicographer File Format
Synsets are entered one per line, and each line is terminated with a
newline character. A line containing a synset may be as long as neces‐
sary, but no newlines can be entered within a synset. Within a synset,
spaces or tabs may be used to separate entities. Items enclosed in
italicized square brackets may not be present.
The general synset syntax is:
{ words pointers ( gloss ) }
Synsets of this form are valid for all syntactic categories except
verb, and are referred to as basic synsets. At least one word and a
gloss are required to form a valid synset. Pointers entered following
all the words in a synset represent semantic relations between all the
words in the source and target synsets.
For verbs, the basic synset syntax is defined as follows:
{ words pointers frames ( gloss ) }
Adjective may be organized into clusters containing one or more head
synsets and optional satellite synsets. Adjective clusters are of the
form:
[
head synset
[satellite synsets]
[-]
[additional head/satellite synsets]
]
Each adjective cluster is enclosed in square brackets, and may have one
or more parts. Each part consists of a head synset and optional satel‐
lite synsets that are conceptually similar to the head synset's mean‐
ing. Parts of a cluster are separated by one or more hyphens (-) on a
line by themselves, with the terminating square bracket following the
last synset. Head and satellite synsets follow the syntax of basic
synsets, however a "Similar to" pointer must be specified in a head
synset for each of its satellite synsets. Most adjective clusters con‐
tain two antonymous parts. See wngloss(7WN) for a discussion of adjec‐
tive clusters, and Special Adjective Syntax for more information on
adjective cluster syntax.
Synsets for relational adjectives (pertainyms) and participial adjec‐
tives do not adhere to the cluster structure. They use the basic
synset syntax.
Comments can be entered in a lexicographer file by enclosing the text
of the comment in parentheses. Note that comments cannot appear within
a synset, as parentheses within a synset have an entirely different
meaning (see Gloss Syntax ). However, entire synsets (or adjective
clusters) can be "commented out" by enclosing them in parentheses.
This is often used by the lexicographers to verify the syntax of files
under development or to leave a note to oneself while working on
entries.
Word Syntax
A synset must have at least one word, and the words of a synset must
appear after the opening brace and before any other synset constructs.
A word may be entered in either the simple word or word/pointer syntax.
A simple word is of the form:
word[ ( marker ) ][lex_id] ,
word may be entered in any combination of upper and lower case unless
it is in an adjective cluster. A collocation is entered by joining the
individual words with an underscore character (_). Numbers (integer or
real) may be entered, either by themselves or as part of a word string,
by following the number with a double quote (").
See Special Adjective Syntax for a description of adjective clusters
and markers.
word may be followed by an integer lex_id from 1 to 15. The lex_id is
used to distinguish different senses of the same word within a lexicog‐
rapher file. The lexicographer assigns lex_id values, usually in
ascending order, although there is no requirement that the numbers be
consecutive. The default is 0, and does not have to be specified. A
lex_id must be used on pointers if the desired sense has a non-zero
lex_id in its synset specification.
Word/pointer syntax is of the form:
[ word[ ( marker ) ][lex_id] , pointers ]
This syntax is used when one or more pointers correspond only to the
specific word in the word/pointer set, rather than all the words in the
synset, and represents a lexical relation. Note that a word/pointer
set appears within a synset, therefore the square brackets used to
enclose it are treated differently from those used to define an adjec‐
tive cluster. Only one word can be specified in each word/pointer set,
and any number of pointers may be included. A synset can have any num‐
ber of word/pointer sets. Each is treated by grind(1WN) essentially as
a word, so they all must appear before any synset pointers representing
semantic relations.
For verbs, the word/pointer syntax is extended in the following manner
to allow the user to specify generic sentence frames that, like point‐
ers, correspond only to a specific word, rather than all the words in
the synset. In this case, pointers are optional.
[ word , [pointers] frames ]
Pointer Syntax
Pointers are optional in synsets. If a pointer is specified outside of
a word/pointer set, the relation is applied to all of the words in the
synset, including any words specified using the word/pointer syntax.
This indicates a semantic relation between the meanings of the words in
the synsets. If specified within a word/pointer set, the relation cor‐
responds only to the word in the set and represents a lexical relation.
A pointer is of the form:
[lex_filename: ]word[lex_id],pointer_symbol
or:
[lex_filename: ]word[lex_id]^word[lex_id],pointer_symbol
For pointers, word indicates a word in another synset. When the second
form of a pointer is used, the first word indicates a word in a head
synset, and the second is a word in a satellite of that cluster. word
may be followed by a lex_id that is used to match the pointer to the
correct target synset. The synset containing word may reside in
another lexicographer file. In this case, word is preceded by
lex_filename as shown.
See Pointers for a list of pointer_symbols and their meanings.
Verb Frame List Syntax
Frame numbers corresponding to generic sentence frames must be entered
in each verb synset. If a frame list is specified outside of a
word/pointer set, the verb frames in the list apply to all of the words
in the synset, including any words specified using the word/pointer
syntax. If specified within a word/pointer set, the verb frames in the
list correspond only to the word in the set.
A frame number list is entered as follows:
frames: f_num[,f_num...]
Where f_num specifies a generic frame number. See Verb Frames for a
list of generic sentences and their corresponding frame numbers.
Gloss Syntax
A gloss is included in all synsets. The lexicographer may enter a text
string of any length desired. A gloss is simply a string enclosed in
parentheses with no embedded carriage returns. It provides a defini‐
tion of what the synset represents and/or example sentences.
Special Adjective Syntax
The syntax for representing antonymous adjective synsets requires sev‐
eral additional conditions.
The first word of a head synset must be entered in upper case, and can
be thought of as the head word of the head synset. The word part of a
pointer from one head synset to another head synset within the same
cluster (usually an antonym) must also be entered in upper case. Usu‐
ally antonymous adjectives are entered using the word/pointer syntax
described in Word Syntax to indicate a lexical relation. There is no
restriction on the number of parts that a cluster may have, and some
clusters have three parts, representing antonymous triplets, such as
solid, liquid, and gas.
A cross-cluster pointer may be specified, allowing a head or satellite
synset to point to a head synset in a different cluster. A cross-clus‐
ter pointer is indicated by entering the word part of the pointer in
upper case.
An adjective may be annotated with a syntactic marker indicating a lim‐
itation on the syntactic position the adjective may have in relation to
noun that it modifies. If so marked, the marker appears between the
word and its following comma. If a lex_id is specified, the marker
immediately follows it. The syntactic markers are:
(p) predicate position
(a) prenominal (attributive) position
(ip) immediately postnominal position
EXAMPLES
(Note that these are hypothetical examples not found in the WordNet
lexicographer files.)
Sample noun synsets:
{ canine, [ dog1, cat,! ] pooch, canid,@ }
{ collie, dog1,@ (large multi-colored dog with pointy nose) }
{ hound, hunting_dog, pack,#m dog1,@ }
{ dog, }
Sample verb synsets:
{ [ confuse, clarify,! frames: 1 ] blur, obscure, frames: 8, 10 }
{ [ clarify, confuse,! ] make_clear, interpret,@ frames: 8 }
{ interpret, construe, understand,@ frames: 8 }
Sample adjective clusters:
[
{ [ HOT, COLD,! ] lukewarm(a), TEPID,^ (hot to the touch) }
{ warm, }
-
{ [ COLD, HOT,! ] frigid, (cold to the touch) }
{ freezing, }
]
Sample adverb synsets:
{ [ basically, adj.all:essential^basic,\ ] [ essentially, adj.all:basic^fundamental,\ ] ( by one's very nature )}
{ pointedly, adj.all:pungent^pointed,\ }
{ [ badly, adj.all:bad,\ well,! ] ill, ("He was badly prepared") }
SEE ALSOgrind(1WN), wnintro(5WN), lexnames(5WN), wndb(5WN), uniqbeg(7WN),
wngloss(7WN).
Fellbaum, C. (1998), ed. "WordNet: An Electronic Lexical Database".
MIT Press, Cambridge, MA.
WordNet 3.0 Dec 2006 WNINPUT(5WN)
