ad_nltk.py

# NLTK Project (2018) Natural Language Toolkit [online] available from 
# <https://www.nltk.org>[10 November 2018]
import nltk
# Text Mining Online (2014) Dive Into NLTK, Part II: Sentence Tokenize and Word Tokenize [online] available from
# <https://textminingonline.com/dive-into-nltk-part-ii-sentence-tokenize-and-word-tokenize>[10 November 2018]
from nltk.tokenize import word_tokenize
# GeekforGeeks (n. d.) How to get synonyms/antonyms from NLTK WordNet in Python? [online] available from
# <https://www.geeksforgeeks.org/get-synonymsantonyms-nltk-wordnet-python/>[12 November 2018]
from nltk.corpus import wordnet
# Text Mining Online (2014) Dive Into NLTK, Part IV: Stemming and Lemmatization [online] available from
# <https://textminingonline.com/dive-into-nltk-part-iv-stemming-and-lemmatization>[10 November 2018]
from nltk.stem import WordNetLemmatizer


def find_word_synonyms(word):
    """Insert a word to return a list of synonyms."""
    syns = wordnet.synsets
    word_synonyms = []    
    synsets = syns(word)
    for synset in synsets:
        synonym = synset.lemmas()[0]
        word_synonyms.append(synonym.name())
    return word_synonyms

########Functions that are not utilised in other codes.######## 
  
def lemmatize_texts(text):  #to lemmatize a word is to find its root words
    """Tokenize texts and lemmatize(find the root word of) the tokenized words. 
    Lemmatized words are returned as a list(lemmatized_tokens)."""
    lemmatizer = WordNetLemmatizer
    tokens = word_tokenize(text)
    lemmatized_tokens = []
    for token in tokens:
        #added additional argument "token", reasons for error still unknown
        token = lemmatizer.lemmatize(token, token, pos='v') 
        lemmatized_tokens.append(token)
    return lemmatized_tokens

  
def find_token_synonyms(tokens):
    """Utilise tokenized texts to return a list of synonyms."""
    #'tokens' need to be in list form
    syns = wordnet.synsets
    token_synonyms = []
    token_antonyms = []
    for token in tokens:     
        synsets = syns(token)
        for synset in synsets:
            synonym = synset.lemmas()[0]
            token_synonyms.append(synonym.name())
    return token_synonyms

  
def find_antonyms(tokens):
    """Utilise tokenized texts to return a list of antonyms."""
    #'tokens' need to be in list form
    syns = wordnet.synsets  
    token_synonyms = []
    token_antonyms = []
    for token in tokens:     
        synsets = syns(token)
        for synset in synsets:
            synonym = synset.lemmas()[0]
            if synonym.antonyms():  #ignoring synonyms that are found
                token_antonyms.append(synonym.antonyms()[0].name())
    return token_antonyms