Our group set out to write a predictive text model for recipes that would combine our knowledge of computer science principles into an application that would predict the instructions of a typed recipe.

The overarching topics that we focused on for our model were Machine Learning and Artificial Intelligence as well as some focus on Data Science. Focusing on ML/AI, we particularly honed into Natural Language Processing (NLP), where we looked into Recurrent Neural Networks (RNNs).

Through the exploration of different RNN architectures, we came across the Long Short-Term Memory (LSTM) architecture that performs the predictive text using a sequence of words as inputs and processing the output as a matrix of probability from each word. In this paper, we will discuss our research, website application, data configuration, and our two predictive text models- including the LSTM model and NodeJS model.