Using Luong's attention mechanism and simple classifiers to make people overcome psychological illnesses

Abstract

Conversational AI is the set of technologies behind automated messaging and speech-enabled applications that offer human-like interactions between computers and humans. It can communicate like a human by recognizing speech and text, understanding intent, deciphering different languages, and responding in a way that mimics human conversation. The objectives of this research are to explore the applicability of conversational AI technology in creating a chatbot for assisting people struggling with psychological illnesses and mental dysfunctions. The main hypothesis is that having an NLP system containing an NLG submodule (module for generation of the Natural text) and an NLU submodule (module for recognizing the emotional state of the person using this chatbot. We use an NLU submodule because we can’t rely only on the artificially generated text as a response for a person in an awful emotional state. Even more, we can use the information from the NLU submodule for stronger strategies generation to ensure emotional support. The system represents a chatbot with two NLP modules, Natural Language Generation, being represented by a Seq2Seq Neural Network with the Loung’s attention mechanism, and a Natural Language Understanding module represented by a classical classification NLP Pipeline that classifies the text in multiple emotional state classes. To interact with the user it uses the Telegram API and is able to save the user messages and the chatbot answers into a simple SQLite Data Base. Even if this implementation wouldn’t replace the real psychologists, with accurate management and maybe with additional inputs for professionals in psychology it may become a tool for detecting people with possible psychological and mental illnesses which can become the first step in further therapy with a real psychologist.