A Bengali Text Summarization Using Encoder-decoder Based on Social Media Dataset

Abstract

Text summarization defines artifices of reducing a long document to create a tale of the main aims of the original text. Due to the huge number of long posts nowadays, the value of summarization is produced. Reading the main document and getting a desirable summary, time and stress are worth it. Using Machine learning & natural language processing built an automated text summarization system can solve this problem. So, our proposed system will distribute an abstractive summary of a long text automatically in a period of some time. We have done the full analysis with the Bengali text. In our planned model we used a chain-to-chain models of RNN with LSTM in the encrypting layer. The structure of our model works applying an RNN decoder and encoder where the encoder inputs text documents and creative output as a short summary at the decoder. This system improves two things namely, summarization & establishing great performance with ignoble train loss. To train our model we use our dataset that was created from various online media, articles, Facebook, and some people's personal posts. The difficulties we face most here are Bengali text processing, limited text length, enough resources for collecting text.