Extractive Models

Extractive text summarization techniques involve the selection of sentences from the source document to make up the final summary. Techniques involve ranking the relevance of sentences in order to retrieve the most important ones for the final summary.

Extractive technique is the more popular approach as it is an easier task compared to abstractive. In addition, since the extractive approach only selects, removes and rearranges text from the source article, the result is almost guaranteed to be grammatically correct, fairly readable and definitely related to the source

This chapter will detail the three different model used; pyTextRank, TextRank4ZH and LexRank.