Copy-cat Bot for Narendra Modi which generates plausible new speeches in Modhi’s style using machine learning approaches

Abstract

Many consequences in the human past can be traced back to that one well-written, well-presentedspeech. Speeches grasp the power to move nations or touch hearts as long as they are well-thought-out.This is why gaining the expertise of speech giving and speech writing is something we should all intent togain. A copy-cat bot is a model that can learn the writing and talking style of a certain person and replicateit. The main objective of this research study is to apply simple Recurrent Neural Network (RNN), LongShort-Term Memory (LSTM) Recurrent Neural Networks and Gated Recurrent Unit (GRU) in developinga speech generation system that deep learns one text and then generates new text. This research looks intothe generation of English transcripts of Narendra Modi’s speeches. The generated text using LSTM andGRU models has great potential. The output resulted by RNN is less realistic and pragmatic, but itsvariants LSTM and GRU performed better. Though the grammatical correctness and the sentencetransitions were absent in generated text of LSTM and GRU, but their output is somewhat logical ascompared to RNN. LSTM and GRU performed better as it generated more realistic text and training lossis small, perplexity is small and mean probability is high compared to RNN.