Authors
Albert Louca , Omar Farouk, Youssef Mohamed , Michel Ashraf
Supervised by: Dr Eslam Amer , TA Ahmed Hazem
Publishing Date
27-October-2020
Abstract
To anticipate our needs, machines or computers need to be able to understand human behavior deeply.This document proposes a system that detect and then diagnose the emotional responses made by humans in a conversation . Hence, The main idea behind this project revolves around the fact that machines must understand emotions in social dialogue to provide emotionally aware responses to users. In particular, understanding and expressing emotion is a crucial element of human behavior. Emotions must be deeply understood by machines and computers to be able to anticipate human needs, this can be achieved by using deep learning models specifically generative ones and facial expressions (detection). It will be followed by second response generation (improve chat) so as to lighten the burden on the user to keep the conversation going.
1.1 Background
A Chatbot is a conversational agent that interacts with users in a certain domain or on a certain topic with natural language sentences.[2] Artificial Intelligence is a big tool that help people combine data and analyze the information to get the most of it while also self learning. AI simulate how the human think and how he process the data to help in its usage.AI based projects are designed to make decisions in a non passive way which can change according the given data or the situation. It saves time as it does not need to be hard coded .Natural Language Processing (NLP) is a branch of AI that helps in the recognition of human language whether it was written or spoken. It’s main purpose is to read, decipher and understand the human user to be able to collect as much data as possible and it is an important part in the chatbot.
1.2 Motivation
- Despite the early start of chatbots in the 1960s, although it was always missing the factor of human like interacting with the system. This problem is interesting as when you are engaging with a chatbot it feels like there is something missing or that you are talking to a wall. This occurs as chatbots were traditionally made to just answer the question and/or query without putting into consideration the other person on the other side of that conversation. There is few companies that accomplished a great progress like NURDANC although they are trying to implement more on the hardware side.
Speed and reliability are both some great characteristics for a business and our Chatbot can provide both. Time is money and the quicker we are the better we get. Speed is a crucial element in some cases by looking at our current state and due to the COVID-19 status, health organizations are receiving so much calls that it became a great difficulty responding to all of them in a quick and authentic way. According to the state information service : Ministry of health receives over 344,000 calls via COVID-19 hot-lines in March. Coronavirus hot-line receives 3,650 calls per day.
1.3 Problem Statement
Businesses that use chatbots are having two big challenges: getting the right information from the client/patient and having enough people to respond and answer to the received calls. Our challenge is to gather the right information from the user using voice analysis and behavior scanning and with our chatbot responding to the user it can’t be any easier