SubClause

Automatic subordinate clause extractor

This is a non-identical Python port of the Java library AutoSubclause.

The NLP library SpaCy is used for document annotation.

Extract subordinate clauses in English text and related information:

General

• Clause Text
• Clause Type
• Clause Finiteness
• Clause Subordinator
• Level of Embededness

Complement

• Type (Subject/Object)

Adverbial

• Semantic Function of the Adjunct
  • Time, Place, Condition, Reason, Concession, Purpose, Comparison, Manner, Results

Relative

• Restrictiveness
• Head Noun
• Head Noun Animacy
• Head Noun Role in Main Clause
• Head Noun Role in Subordinate Clause

Example Usage

import spacy
nlp = spacy.load('en_core_web_sm')

from subclause import SubClauseFinder
subclausefinder = SubClauseFinder()

text = "The door opened because the man pushed it. " \
       "I wondered whether the homework was necessary. " \
       "They will visit you before they go to the airport. " \
       "Before they go to the airport, they will visit you. " \
       "I went to the show that was very popular."

doc = nlp(text)

for sent in doc.sents:
    print('\nSentence:', sent)
    subclauses = subclausefinder.get_subclauses(sent)
    for sc in subclauses:
        print("\tclause text:", sc.get_clause_span())
        print("\tis finite:", sc.get_is_finite())
        print("\tsubordinator:", sc.get_subordinator())
        print("\tembeddedness:", sc.get_embeddedness())
        print("\tclause type:", sc.get_clause_type())
        # for complement clauses
        if sc.get_clause_type() == 'COMPLEMENT':
            print("\t\tcomplement type:", sc.get_complement_type())
        # for adverbial clauses
        elif sc.get_clause_type() == 'ADJUNCT':
            print("\t\tadjunct function:", sc.get_adjunct_function())
        # for relative clauses
        elif sc.get_clause_type() == 'RELATIVE':
            print("\t\tis restrictive:", sc.get_is_restrictive())
            print("\t\thead noun:", sc.get_head_noun())
            print("\t\tis head noun animate:", sc.get_head_noun_animacy())
            print("\t\thead noun role in main clause:", sc.get_head_noun_role_in_main_clause())
            print("\t\thead noun role in subordinate clause:", sc.get_head_noun_role_in_sub_clause())

See Also

Xiaobin Chen et al., Automatic extraction of subordinate clauses and its application in second language acquisition research [1]