A.I. Chatbot for Mental Health
Year
2021
Role
UX/ UI Designer
Team
Solo
Project Overview
A.I Chatbot ‘Lee luda’ just disappeared after they attracted more than 750,000 users in 20 days. It seemed that she was success by demonstrating natural interaction with humans but there were several issues with using chatbot such as ethical, unfairness, and even privacy problem.
My inspiration of this project is that I kept hearing about machine learning and chatbot and I was curious about it. So, I’ve learned from ‘Lee luda’ situation and I tried to put them into my learning.
Through this project, I would like to show you how I implement AI chatbot for mental health.
Aims and Goal
Build AI chatbot for mental health.
Target Audience
People who use Facebook and Instagram messenger.
People who has mental illness.
Requirements
Smart phone or Computer
Research
1. Market Research
What I found after competitive analysis was they are all pattern based chatbot, expensive, and also need to download it. After this analysis, I wanted to use Facebook and Instagram platform so people don't need to download the app with no cost.
Then, I wanted to know more about Facebook to see what's out there.
After Facebook research, they offer rule based chatbot and limited conversation.
Through those researches, I decide to use Natural Language Processing(NLP) so that they can talk like a real person.
User Empathy
To understand mentees and mentor, I explore users thoughts using personas, empathy maps,
user journeys and mental models.
Build A.I. Chatbot
After research, I start build A.I. Chatbot using Colaboratory.
Final Design
1. The A.I chatbot for people who has mental problem
Users can easily contact them without any hesitation at anytime and anywhere whenever they need.
2. Easy to interact using FB Messenger & Instagram
Anyone who has Facebook messenger or Instagram,
able to have conversation without paying
3. Have conversation with Natural Tone
Since the AI chatbot use Natural language processing (NLP), users can have natural conversation with them.
Reflection
Through this case study, I learned how rule-based and AI chatbot works and I created chatbots based on user research. In order to understand and write AI chatbot, I needed to acknowledge basic Python.
The reason I worked on this project was chatbots are everywhere and I heard about machine learning which is not familiar to me, but I wanted to get to know it and make my project work.
If I have any chance to improve this project, I would like to learn more about machine learning and deep learning, so that I can make a AI chatbot using real data in the future.