Sentic API for mental health detection

Abstract

Sentiment text analysis, which is a pivotal aspect of Natural Language Processing (NLP), involves reading different texts and identifying their labels (positive, negative, neutral). This report will dive into developing a Sentic API with the testing of different models and techniques and comparing the result of the different methods used. In this project, we have explored the different techniques namely Sentic API, TextBlob and Valer Aware Dictionary and sEntiment Reasoner (VADER). In addition, after we have done the sentiment text analysis, we will be feeding this data into models for training. This is a form of supervised training and the models that we have explored into are Recurrent Neural Networks and Long Short-Term Memory (LSTM) networks. Within this model, we will also be training the models with different hyperparameters to compare and find the best parameters for the model that we have come up with.