d1="the best Italian restaurant enjoy the best pasta"

d2="american restaurant enjoy the best hamburger"

​

pip install textmining3

​

#termdocumentmatrix

import textmining

tdm=textmining.TermDocumentMatrix()

tdm.add_doc(d1)

tdm.add_doc(d2)

tdm=tdm.to_df(cutoff=0)

tdm

​

#inverted index

docs=[d1,d2]

inverted_index={}

for i,doc in enumerate(docs):

  for term in doc.split():

    if term in inverted_index:

      inverted_index[term].add(i)

    else:

      inverted_index[term]={i}

​

inverted_index

​

#tfidf

from sklearn.feature_extraction.text import TfidfVectorizer

tfidf=TfidfVectorizer()

result=tfidf.fit_transform(docs)

print(tfidf.vocabulary_)

print(result)

​

#cosine similarity

from sklearn.metrics.pairwise import cosine_similarity

print(cosine_similarity(tdm))