spaCy/website/docs/api/annotation.jade

145 lines
4.5 KiB
Plaintext
Raw Normal View History

2016-10-31 21:04:15 +03:00
//- 💫 DOCS > API > ANNOTATION SPECS
include ../../_includes/_mixins
p This document describes the target annotations spaCy is trained to predict.
+h(2, "tokenization") Tokenization
p
| Tokenization standards are based on the
| #[+a("https://catalog.ldc.upenn.edu/LDC2013T19") OntoNotes 5] corpus.
| The tokenizer differs from most by including tokens for significant
| whitespace. Any sequence of whitespace characters beyond a single space
| (#[code ' ']) is included as a token.
+aside-code("Example").
from spacy.en import English
nlp = English(parser=False)
tokens = nlp('Some\nspaces and\ttab characters')
print([t.orth_ for t in tokens])
# ['Some', '\n', 'spaces', ' ', 'and', '\t', 'tab', 'characters']
p
| The whitespace tokens are useful for much the same reason punctuation is
| it's often an important delimiter in the text. By preserving it in the
| token output, we are able to maintain a simple alignment between the
| tokens and the original string, and we ensure that no information is
| lost during processing.
+h(2, "sentence-boundary") Sentence boundary detection
p
| Sentence boundaries are calculated from the syntactic parse tree, so
| features such as punctuation and capitalisation play an important but
| non-decisive role in determining the sentence boundaries. Usually this
| means that the sentence boundaries will at least coincide with clause
| boundaries, even given poorly punctuated text.
+h(2, "pos-tagging") Part-of-speech Tagging
p
| The part-of-speech tagger uses the
| #[+a("https://catalog.ldc.upenn.edu/LDC2013T19") OntoNotes 5] version of
| the Penn Treebank tag set. We also map the tags to the simpler Google
| Universal POS Tag set. See
| #[+src(gh("spaCy", "spacy/tagger.pyx")) tagger.pyx] for details.
+h(2, "lemmatization") Lemmatization
p A "lemma" is the uninflected form of a word. In English, this means:
+list
+item #[strong Adjectives]: The form like "happy", not "happier" or "happiest"
+item #[strong Adverbs]: The form like "badly", not "worse" or "worst"
+item #[strong Nouns]: The form like "dog", not "dogs"; like "child", not "children"
+item #[strong Verbs]: The form like "write", not "writes", "writing", "wrote" or "written"
p
| The lemmatization data is taken from
| #[+a("https://wordnet.princeton.edu") WordNet]. However, we also add a
| special case for pronouns: all pronouns are lemmatized to the special
| token #[code -PRON-].
+h(2, "dependency-parsing") Syntactic Dependency Parsing
p
| The parser is trained on data produced by the
| #[+a("http://www.clearnlp.com") ClearNLP] converter. Details of the
| annotation scheme can be found
| #[+a("http://www.mathcs.emory.edu/~choi/doc/clear-dependency-2012.pdf") here].
+h(2, "named-entities") Named Entity Recognition
+table([ "Type", "Description" ])
2016-10-31 21:04:15 +03:00
+row
+cell #[code PERSON]
2016-11-25 17:53:40 +03:00
+cell People, including fictional.
2016-10-31 21:04:15 +03:00
+row
+cell #[code NORP]
2016-11-25 17:53:40 +03:00
+cell Nationalities or religious or political groups.
2016-10-31 21:04:15 +03:00
+row
+cell #[code FACILITY]
+cell Buildings, airports, highways, bridges, etc.
2016-10-31 21:04:15 +03:00
+row
+cell #[code ORG]
+cell Companies, agencies, institutions, etc.
+row
+cell #[code GPE]
2016-11-25 17:53:40 +03:00
+cell Countries, cities, states.
2016-10-31 21:04:15 +03:00
+row
+cell #[code LOC]
2016-11-25 17:53:40 +03:00
+cell Non-GPE locations, mountain ranges, bodies of water.
2016-10-31 21:04:15 +03:00
+row
+cell #[code PRODUCT]
2016-11-25 17:53:40 +03:00
+cell Objects, vehicles, foods, etc. (Not services.)
2016-10-31 21:04:15 +03:00
+row
+cell #[code EVENT]
+cell Named hurricanes, battles, wars, sports events, etc.
+row
+cell #[code WORK_OF_ART]
+cell Titles of books, songs, etc.
+row
+cell #[code LANGUAGE]
2016-11-25 17:53:40 +03:00
+cell Any named language.
2016-10-31 21:04:15 +03:00
p The following values are also annotated in a style similar to names:
+table([ "Type", "Description" ])
2016-10-31 21:04:15 +03:00
+row
+cell #[code DATE]
2016-11-25 17:53:40 +03:00
+cell Absolute or relative dates or periods.
2016-10-31 21:04:15 +03:00
+row
+cell #[code TIME]
2016-11-25 17:53:40 +03:00
+cell Times smaller than a day.
2016-10-31 21:04:15 +03:00
+row
+cell #[code PERCENT]
2016-11-25 17:53:40 +03:00
+cell Percentage, including "%".
2016-10-31 21:04:15 +03:00
+row
+cell #[code MONEY]
2016-11-25 17:53:40 +03:00
+cell Monetary values, including unit.
2016-10-31 21:04:15 +03:00
+row
+cell #[code QUANTITY]
2016-11-25 17:53:40 +03:00
+cell Measurements, as of weight or distance.
2016-10-31 21:04:15 +03:00
+row
+cell #[code ORDINAL]
+cell "first", "second", etc.
2016-10-31 21:04:15 +03:00
+row
+cell #[code CARDINAL]
2016-11-25 17:53:40 +03:00
+cell Numerals that do not fall under another type.