Danielchris145/TruthCheck-AI

1# models/keyword_extractor.py2import spacy3from collections import Counter4 5class KeywordExtractor:6    def __init__(self): # Corrected __init__7        try:8            self.nlp = spacy.load("en_core_web_sm")9        except OSError:10            print("Please install spaCy English model: python -m spacy download en_core_web_sm")11            raise12    13    def extract_keywords(self, text):14        """Extract keywords and named entities from text"""15        doc = self.nlp(text)16        17        keywords = []18        19        # Extract named entities20        for ent in doc.ents:21            if ent.label_ in ['PERSON', 'ORG', 'GPE', 'PRODUCT', 'EVENT', 'DATE']:22                keywords.append(ent.text)23        24        # Extract noun phrases and important words25        for chunk in doc.noun_chunks:26            if len(chunk.text.split()) <= 3:  # Avoid very long phrases27                keywords.append(chunk.text)28        29        # Extract individual important words30        for token in doc:31            if (token.pos_ in ['NOUN', 'PROPN'] and 32                not token.is_stop and 33                not token.is_punct and 34                len(token.text) > 2):35                keywords.append(token.text)36        37        # Remove duplicates and return most common38        keyword_counts = Counter(keywords)39        return [word for word, count in keyword_counts.most_common(10)]40 41