[Solved] Automatic Summarization : Extraction Based [closed]

Question

There is not one single algorithm for extraction based summarization. There are several different algorithms to choose from. You should choose one that fits your specific needs.

There are two approaches to extraction based summarization:

Supervised learning – you give the program lots of examples of documents together with their keywords. The program learns what constitutes a keyword. Then you give it a new document, this time without any keywords, and the program extracts the keywords of this document based on what it learned during the training phase. There is a huge number of supervised learning techniques. To name a few, there are neural networks, decision trees, random forests and support vector machines.
Unsupervised learning – you simly give the program a document and it creates a list of keywords without relying on any past experience. A popular unsupervised algorithm for extraction based summarization is TextRank.

Accepted Answer

There is not one single algorithm for extraction based summarization. There are several different algorithms to choose from. You should choose one that fits your specific needs.

There are two approaches to extraction based summarization:

Supervised learning – you give the program lots of examples of documents together with their keywords. The program learns what constitutes a keyword. Then you give it a new document, this time without any keywords, and the program extracts the keywords of this document based on what it learned during the training phase. There is a huge number of supervised learning techniques. To name a few, there are neural networks, decision trees, random forests and support vector machines.
Unsupervised learning – you simly give the program a document and it creates a list of keywords without relying on any past experience. A popular unsupervised algorithm for extraction based summarization is TextRank.