text_sum.py

import nltk
from nltk.corpus import stopwords
from nltk.tokenize import sent_tokenize,word_tokenize
import heapq  

def text_sumarize(row_text):
    
    stopWords=set(stopwords.words("english"))
    word_fequence={}

    for word in nltk.word_tokenize(row_text):
        if word  not in stopWords:
            if word not in word_fequence.keys():
                word_fequence[word]=1
            else:
                word_fequence[word] +=1
    maximum_freq=max(word_fequence.values())     
    
    for word in word_fequence.keys():
        word_fequence[word]=(word_fequence[word]/maximum_freq)
        
    sentence_list=nltk.sent_tokenize(row_text)
    
    sentence={}
    for i in sentence_list:
        for word in nltk.word_tokenize(i.lower()):
            if word in word_fequence.keys():
                if len(i.split(" "))<30:
                    if i not in sentence.keys():
                        sentence[word]=word_fequence[word]
                        
                    else:
                        sentence[word]+=word_fequence[word]
                        
    summary_test=heapq.nlargest(7,sentence,key=sentence.get)    
    summary=" ".join(summary_test)
    
    return summary


# def nltk_summarizer(raw_text):
# 	stopWords = set(stopwords.words("english"))
# 	word_frequencies = {}  
# 	for word in nltk.word_tokenize(raw_text):  
# 	    if word not in stopWords:
# 	        if word not in word_frequencies.keys():
# 	            word_frequencies[word] = 1
# 	        else:
# 	            word_frequencies[word] += 1

# 	maximum_frequncy = max(word_frequencies.values())

# 	for word in word_frequencies.keys():  
# 	    word_frequencies[word] = (word_frequencies[word]/maximum_frequncy)

# 	sentence_list = nltk.sent_tokenize(raw_text)
# 	sentence_scores = {}  
# 	for sent in sentence_list:  
# 	    for word in nltk.word_tokenize(sent.lower()):
# 	        if word in word_frequencies.keys():
# 	            if len(sent.split(' ')) < 30:
# 	                if sent not in sentence_scores.keys():
# 	                    sentence_scores[sent] = word_frequencies[word]
# 	                else:
# 	                    sentence_scores[sent] += word_frequencies[word]



# 	summary_sentences = heapq.nlargest(7, sentence_scores, key=sentence_scores.get)

# 	summary = ' '.join(summary_sentences)  
# 	return summary