E-mail autocomplete function using RNN Encoder-decoder sequence-to-sequence model

Text Auto-complete feature suggests a stream of words which complete a user's text as the user types each character. Such a feature is used in search engines, email programs, source code editors, database query tools etc. Earlier people have used traditional language models for this problem. But for better performance, a Neural network-based language model is needed. Here, an encoder-decoder based sequence-to-sequence language model has been used for performing text generation and the empirical results show that the model effectively suggests the incomplete sentences.