def get_keywords()

in scripts/interpret.py [0:0]

• 15 lines of code
• 6 McCabe index (conditional complexity)

def get_keywords(texts):
    """Returns candidate keywords based on POS as well as original form of keyword.
    """
    candidate_pos = ['ADJ', 'VERB', 'NOUN', 'PROPN']
    keywords = []
    tokens = [] # for referencing keywords in original text later on
    for text in texts:
        text_keywords = []
        text_tokens = []
        doc = nlp(text)
        for token in doc:
            if token.pos_ in candidate_pos and token.is_stop is False:
                text_tokens.append(str(token))
                text_keywords.append(token.lemma_.lower())

        keywords.extend(text_keywords)
        tokens.extend(text_tokens)
        
    return keywords, tokens