Next Word Prediction in Telugu using RNN Mechanism
Vijaya Saraswathi R, Ratna Keerthana Nanduri, Bhavana Pinni, Marthala Manasa, Sai Siddhartha M · 2022 International Conference on Augmented Intelligence and Sustainable Systems (ICAISS) · 2022
Exchanging textual content by entering information and sending it to others has become one of the most popular ways of information exchange these days. In this advanced age, we spend huge amount of time sending and receiving emails, additional facts and data to others through social networking websites. To make the exchange of textual information faster and better, word prediction systems are introduced that can anticipate the most likely word to come next, so that people do not have to type the following word, but rather can select a word from the suggested ones. In this work, we built a model to predict the following word for Telugu. Predicting the next word in a sequence can reduce the count of number of letters typed made by the user. Next word prediction is an application of NLP, also known as language modelling. Many applications use these systems such as, Autocorrect and is primarily used in emails/messages. It is also used by MS Word, Google search that predicts the next word based on your search history. In this work, we studied NLP and applied LSTM, a deep learning technique to carry out the next word prediction process.