Sequence Prediction for Journal Titles using Natural Language Processing

Abstract

Every day the number of journals published online has increased. The new text generation is also hugely increasing. The title of the journal describes the content of the text which it contains. Based on the journal title a human can analyze the text so that title for the journal is the heart of the corresponding text. In this research work we proposed a model that generates a title for any journal by using text analyzing models. In previous works so many models are used to generate a sequence of text. In this research work we are using deep learning neural networks to increase accuracy of predicting a sequence to eliminate the vanishing gradient problem. We proposed a Bidirectional recurrent neural network to increase the accuracy of predicting a sequence based on probability measure.