Skip to content

att.linguistic

provides a set of attributes concerning linguistic features of tokens, for usage within token-level elements, specifically w and pc in the analysis module.

Module analysis

Attributes

@lemmaprovides a lemma (base form) for the word, typically uninflected and serving both as an identifier (e.g. in dictionary contexts, as a headword), and as a basis for potential inflections.
Status
Optional
Datatype
teidata.text
<w lemma="wife">wives</w>
@lemmaRefprovides a pointer to a definition of the lemma for the word, for example in an online lexicon.
Status
Datatype
teidata.pointer
<w type="verb" lemma="hit" lemmaRef="http://www.example.com/lexicon/hitvb.xml">hitt<m type="suffix">ing</m>
</w>
@pos(part of speech) indicates the part of speech assigned to a token (i.e. information on whether it is a noun, adjective, or verb), usually according to some official reference vocabulary (e.g. for German: STTS, for English: CLAWS, for Polish: NKJP, etc.).
Status
Optional
Datatype
teidata.text

The German sentence Wir fahren in den Urlaub. tagged with the Stuttgart-Tuebingen-Tagset (STTS).

<s>
  <w pos="PPER">Wir</w>
  <w pos="VVFIN">fahren</w>
  <w pos="APPR">in</w>
  <w pos="ART">den</w>
  <w pos="NN">Urlaub</w>
  <w pos="$.">.</w>
</s>
@msd(morphosyntactic description) supplies morphosyntactic information for a token, usually according to some official reference vocabulary (e.g. for German: STTS-large tagset; for a feature description system designed as (pragmatically) universal, see Universal Features).
Status
Optional
Datatype
teidata.text
<ab>
  <w pos="PPER" msd="1.Pl.*.Nom">Wir</w>
  <w pos="VVFIN" msd="1.Pl.Pres.Ind">fahren</w>
  <w pos="APPR" msd="--">in</w>
  <w pos="ART" msd="Def.Masc.Akk.Sg">den</w>
  <w pos="NN" msd="Masc.Akk.Sg">Urlaub</w>
  <pc pos="$." msd="--">.</pc>
</ab>
@joinwhen present, provides information on whether the token in question is adjacent to another, and if so, on which side.
Status
Optional
Datatype
teidata.text
Legal values
no
the token is not adjacent to another
left
there is no whitespace on the left side of the token
right
there is no whitespace on the right side of the token
both
there is no whitespace on either side of the token
overlap
the token overlaps with another; other devices (specifying the extent and the area of overlap) are needed to more precisely locate this token in the character stream

The example below assumes that the lack of whitespace is marked redundantly, by using the appropriate values of @join.

<s>
  <pc join="right">"</pc>
  <w join="left">Friends</w>
  <w>will</w>
  <w>be</w>
  <w join="right">friends</w>
  <pc join="both">.</pc>
  <pc join="left">"</pc>
</s>

Note that a project may make a decision to only indicate lack of whitespace in one direction, or do that non-redundantly. The existing proposal is the broadest possible, on the assumption that we adopt the "streamable view", where all the information on the current element needs to be represented locally.

Classes and content

Used by
Members

Note

These attributes make it possible to encode simple language corpora and to add a layer of linguistic information to any tokenized resource. See section Lightweight Linguistic Annotation for discussion.

These guidelines provide no semantic basis or suggested precedence when both @lemma and @lemmaRef are provided. For this reason simultaneous use of both is not recommended for interchange unless documentation explaining the use is provided, probably in an ODD customization.