Mind Mate

Project Info

Project Description


Artificial chatbot buddy to support mental health well-being.

Using various sets of statistics surrounding suicide in New Zealand surrounding different demographics / categories (age, gender, ethnicity, employment) we intend to create a very simple scoring system which will be used to give the Mind Mate user an initial 'score' / risk factor. This score can change and evolve over time and as the user continues to use the app.

The reason behind the score is that it will be used to slightly tailor the responses from the chatbot dependant on the user. Ultimately the idea would be for the app to have other features (meditation, local events, notifications, etc) which will also take in to account the user's score and be specified accordingly.

We have decided not to try build an app at this stage, but to focus on creating a prototype of the bot using Google's DialogFlow software and also build a simple application wireframe using Adobe XD. Ultimately, the goal would be to integrate the bot in to a fully fledged mobile app that would be publicly available.

We are not trying to create an artificially intelligent mental health counsellor, as that is outside our area of expertise (however this is a potential area for future development by harnessing the power of machine learning), we are trying to create a simple, friendly bot which promotes positive mental health to people across the spectrum of current mental wellbeing.

The bot will be able to recognise if the user may be a serious risk to themselves if they have a low score and are asking certain trigger questions, in which case we intend to provide the user with links to services they can reach out to.


Data Story


We are first using historical data surrounding suicides to address the issue in New Zealand and how it has worsened over the years.

We then utilise data on suicides which are categorised by various demographic type data to help come up with a 'risk factor' for individuals dependant on their background. Specifically, we look at suicide rates across different ethnicities, genders, locations, age and employment status.

The risk factor for each category is taken in to account when the user signs up to the app and provides basic information about themselves. They are given a 'score' which will slightly alter how the chatbot responds and reacts to the user's input.


Evidence of Work

Video

Team DataSets

2015 Suicide Facts from Health.Govt

Description of Use: Pre-created graphs we can use in our presentation / video.

Data Set

Various graphs on suicide from Stats

Data Set

Various graphs on suicide from Stuff

Data Set

Serious Injury Outcome Indicators 2000-2017

Description of Use: Use in a dashboard for the pitch

Data Set

Challenges

Best Way to Promote Positive Mental Health

Region: New Zealand

Challenge

Innovate New Zealand

Region: New Zealand

Challenge

Best Creative Use Of Technology

Region: New Zealand

Challenge

Back to Projects