KL_Divergence.py

from build_index import Index
import numpy as np
def get_result(data):
    query_doc=data
    query_doc1=data[:-4]
    #print(query_doc1)
    inn=Index()
    inn.retrieve_file()
    inn.tok_lem_stem(type_op='lemmatize')
    inn.inverted_index_constr()
    inn.calculate_tf_idf(test_file=query_doc)
    inn.tfidf_of_query(query_doc1)
    #print(inn.tfidf_query_doc)
    final={}
    for i,key in enumerate(inn.all_files.keys()):
        wt=[]
        u=0
        for word in inn.doc_sim_score.keys():
            if(inn.doc_sim_score[word][i][1]==0 and inn.tfidf_query_doc[u]==0):
                p1=1
                p2=1
            else:
                p1=inn.doc_sim_score[word][i][1]/(inn.doc_sim_score[word][i][1]+inn.tfidf_query_doc[u])
                p2=inn.tfidf_query_doc[u]/(inn.doc_sim_score[word][i][1]+inn.tfidf_query_doc[u])
            wt.append(inn.doc_sim_score[word][i][1]*p1+inn.tfidf_query_doc[u]*p2)
            u=u+1
        v=0
        d1=0
        d2=0
        sim=0
        for word in inn.doc_sim_score.keys(): 
            d1+=inn.tfidf_query_doc[v]*((inn.tfidf_query_doc[v]+1)/(wt[v]+1))
            d2+=wt[v]*np.log((inn.tfidf_query_doc[v]+1)/(wt[v]+1))
            v=v+1
        #final[key]=d1
        u=0
        for word in inn.doc_sim_score.keys():
            if(inn.doc_sim_score[word][i][1]==0 and inn.tfidf_query_doc[u]==0):
                p1=1
                p2=1
            else:
                p1=inn.doc_sim_score[word][i][1]/(inn.doc_sim_score[word][i][1]+inn.tfidf_query_doc[u])
                p2=inn.tfidf_query_doc[u]/(inn.doc_sim_score[word][i][1]+inn.tfidf_query_doc[u])
            sim=sim+(p1*d1)+(p2*d2)
            u=u+1
        final[key]=sim
    return final