Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- import os
- from numpy import vectorize
- from sklearn.feature_extraction.text import TfidfVectorizer
- from sklearn.metrics.pairwise import cosine_similarity
- sample_files = [doc for doc in os.listdir() if doc.endswith('.txt')]
- sample_contents = [open(File).read() for File in sample_files]
- vectorize = lambda Text: TfidfVectorizer().fit_transform(Text).toarray()
- similarity = lambda doc1, doc2: cosine_similarity([doc1, doc2])
- vectors = vectorize(sample_contents)
- s_vectors = list(zip(sample_files, vectors))
- def check_plagiarism():
- results = set()
- global s_vectors
- for sample_a, text_vector_a in s_vectors:
- new_vectors = s_vectors.copy()
- current_index = new_vectors.index((sample_a, text_vector_a))
- del new_vectors[current_index]
- for sample_b, text_vector_b in new_vectors:
- sim_score = similarity(text_vector_a, text_vector_b)[0][1]
- sample_pair = sorted((sample_a, sample_b))
- score = sample_pair[0], sample_pair[1], sim_score
- results.add(score)
- return results
- for data in check_plagiarism():
- print(data)
Advertisement
Add Comment
Please, Sign In to add comment