Text-Similarity

Package for texts similarity, based on textdistance

Usage

1. installation

pip install textsimi

2. example usage:

from testsimi import textsimilarity as ts
ts.print_all_algorithms()

print(ts.compute_similarity('hello','hi'))

print(ts.top_K_similarity_between_one_mention_and_many_candidates(5,'hi',['hi','hihi','hello',"what's up", 'greetings', 'how are you', 'hallo', 'wie gehts']))


'''
['hamming', 'mlipns', 'levenshtein', 'damerau_levenshtein', 'jaro_winkler', 'strcmp95', 'needleman_wunsch', 'gotoh', 'smith_waterman', 'jaccard', 'sorensen', 'tversky', 'overlap', 'tanimoto', 'cosine', 'monge_elkan', 'bag', 'ratcliff_obershelp', 'arith_ncd', 'rle_ncd', 'bwtrle_ncd', 'sqrt_ncd', 'entropy_ncd', 'bz2_ncd',   'zlib_ncd', 'editex', 'prefix', 'postfix', 'length', 'identity', 'matrix']
Current algorithm is Jaccard({'qval': 1, 'as_set': False, 'external': True})
0.16666666666666666
['hi', 'hihi', 'wie gehts', 'hallo', 'hello']
'''

3. colab notebook for tf-idf text similarity