Spaces:
Runtime error
Runtime error
| import pandas as pd | |
| from sklearn.feature_extraction.text import TfidfVectorizer | |
| from sklearn.metrics.pairwise import cosine_similarity | |
| import numpy as np | |
| df = pd.read_csv('qna.csv',encoding = 'utf-8',delimiter=';') | |
| print(df) | |
| questions = df['question'].tolist() | |
| print(questions) | |
| answers = df['answer'].tolist() | |
| vectorizer = TfidfVectorizer() | |
| tfidf_matrix = vectorizer.fit_transform(questions) | |
| def get_most_similar_question(new_sentence): | |
| new_tfidf = vectorizer.transform([new_sentence]) | |
| similarities = cosine_similarity(new_tfidf,tfidf_matrix) | |
| most_similar_index = np.argmax(similarities) | |
| similarity_percentage = similarities[0, most_similar_index]*100 | |
| return answers[most_similar_index], similarity_percentage | |
| def AnswertheQuestion(new_sentence): | |
| most_similar_answer, similarity_percentage = get_most_similar_question(new_sentence) | |
| if similarity_percentage > 70: | |
| response = { | |
| 'answer': most_similar_answer | |
| } | |
| else: | |
| response = { | |
| 'answer': 'Sorry, I am not aware of this information :(' | |
| } | |
| return response | |
| print(AnswertheQuestion('Who is the Ninad')) | |