Text Summarizer

Text Summarizer

1. Python

   • NLTK Library
     • Word/ Sentence Tokenizer

2. Extractive Summary

   • Sentence Score
     • Google page scoring algorithm
   • Technique 1
     • Tokenize Word
     • Remove Stop words/punctuations
     • Create frequency dictionary
     • Tokenize sentences
     • Sentence Score = sum of all the word score of the sentence
     • Total Sentence Score = Sentence1 + Sentence2 + ...
     • Average Value = Total Sentence Score / Total Number of Sentences
     • If Sentence Score > (1.3 * Averate Value) then add it to Summary
   • Technique 2
     • Tokenize Words
     • Create Word Frequency dictionary
     • Remove stop words/punctuations
     • Add these words to frequency dictionary
     • Count frequency of each words
     • New frequency = Previous frequency of word / maximum frequency
     • Remove words that have new frequency <= min_cut or >= max_cut
     • Sentence score = sum of new frequency of each words of that sentence
     • Extract sentence with highest sentence score
     • Technique 3
       • Combination of Technique 2 and Google Page Scoring Algorithm
         • Create GUI using Flask Framework
           • Web Based GUI
   • Remove Stopwords/Punctuations

Bilder und Dateien der Mindmap