pythainlp.parse

The pythainlp.parse is dependency parsing for Thai.

Modules

pythainlp.parse.dependency_parsing(text: str, model: str | None = None, tag: str = 'str', engine: str = 'esupar') → List[List[str]] | str[source]

Dependency Parsing

Parameters:

  • text (str) – text to do dependency parsing

  • model (str) – model for using with engine (for esupar and transformers_ud)

  • tag (str) – output type (str or list)

  • engine (str) – the name dependency parser

Returns:

str (conllu) or List

Return type:

Union[List[List[str]], str]

Options for engine

  • esupar (default) - Tokenizer POS-tagger and Dependency-parser with BERT/RoBERTa/DeBERTa model. GitHub

  • spacy_thai - Tokenizer, POS-tagger, and dependency-parser for Thai language, working on Universal Dependencies. GitHub

  • transformers_ud - TransformersUD GitHub

  • ud_goeswith - POS-tagging and dependency-parsing with using goeswith for subwords

Options for model (esupar engine)

  • th (default) - KoichiYasuoka/roberta-base-thai-spm-upos model Huggingface

  • KoichiYasuoka/deberta-base-thai-upos - DeBERTa(V2) model pre-trained on Thai Wikipedia texts for POS-tagging and dependency-parsing Huggingface

  • KoichiYasuoka/roberta-base-thai-syllable-upos - RoBERTa model pre-trained on Thai Wikipedia texts for POS-tagging and dependency-parsing. (syllable level) Huggingface

  • KoichiYasuoka/roberta-base-thai-char-upos - RoBERTa model pre-trained on Thai Wikipedia texts for POS-tagging and dependency-parsing. (char level) Huggingface

If you want to train model for esupar, you can read Huggingface

Options for model (transformers_ud engine)

  • KoichiYasuoka/deberta-base-thai-ud-head (default) - DeBERTa(V2) model pretrained on Thai Wikipedia texts for dependency-parsing (head-detection on Universal Dependencies) as question-answering, derived from deberta-base-thai. trained by th_blackboard.conll. Huggingface

  • KoichiYasuoka/roberta-base-thai-spm-ud-head - roberta model pretrained on Thai Wikipedia texts for dependency-parsing. Huggingface

Options for model (ud_goeswith engine)

  • KoichiYasuoka/deberta-base-thai-ud-goeswith (default) - This is a DeBERTa(V2) model pre-trained on Thai Wikipedia texts for POS-tagging and dependency-parsing (using goeswith for subwords) Huggingface

Example:

from pythainlp.parse import dependency_parsing

print(dependency_parsing("ผมเป็นคนดี", engine="esupar"))
# output:
# 1       ผม      _       PRON    _       _       3       nsubj   _       SpaceAfter=No
# 2       เป็น     _       VERB    _       _       3       cop     _       SpaceAfter=No
# 3       คน      _       NOUN    _       _       0       root    _       SpaceAfter=No
# 4       ดี       _       VERB    _       _       3       acl     _       SpaceAfter=No

print(dependency_parsing("ผมเป็นคนดี", engine="spacy_thai"))
# output:
# 1       ผม              PRON    PPRS    _       2       nsubj   _       SpaceAfter=No
# 2       เป็น             VERB    VSTA    _       0       ROOT    _       SpaceAfter=No
# 3       คนดี             NOUN    NCMN    _       2       obj     _       SpaceAfter=No