Abstract:
In the text summaries, a description is the system of the scale of one or more documents
& the key detail of the document is provided by the curtail passage. In this present era
of information technology, we always rely on online information that is raised in need
of a summary of the original text. There is a lot of methods already implemented for
other languages like English text summarize but the effort is now underway for Bengali
text summarization. In this paper, we will propose RNN's ( Recurrent Neural Network)
deep learning model based extract text abbreviation strategy. So the categorization
strategy is useful for description or not. For the back extension process, we have used
Long Short-Term Memory(LSTM) and Gated Recurrent Units (GRU) in this article.
But we used Long Short-Term Memory (LSTM) so it's more positive than that, and in
terms of Rouge-I, Rouge-II and Rouge-III, we realize the typical F-1 scores being 0.65,
0.61, 0.58.
