The QUO - Minerva uses data science to increase social impact
Minerva brings together volunteer data scientists and Australian not for profit organisations to work on problems using the power of data.
Minerva uses data science to increase social impact

Story

Minerva uses data science to increase social impact

01 March 2018
Play Video
Less than one third of young Australians experiencing mental distress get the help they need. ReachOut is combating this by providing online resources to young Australian who need them.

And they’re hoping to get even more targeted with their outreach and solutions with the help of some volunteer data scientists.

The data scientists are part of Minerva Collective, a group start in October 2016 by Michael Allwright, Anthony Tockar and Whitney Komor. Allwright, now the CEO, and Tockar, a Director, met while working at PwC. “We always thought of ourselves as a bit different,” says Allwright. “We spent a lot of time talking about data and asking, ‘how can we use data for good?’”

The idea for a collective blossomed from there. Minerva brings together volunteer data scientists and Australian not for profit organisations to work on problems using the power of data. Think: using supermarket purchase data to help stores help their customers make healthier choices.

Minerva uses a data exchange platform developed by Data Republic, which also partly seed funded Minerva for the first year of operation. It’s the first platform to allow for the exchange of customer-level data in a legal, secure way. It has the backing of the government and is funded by Westpac. When an organization contributes data (for example, a supermarket submitting purchase data), the data is stored securely on the platform. A token is applied to each instance, so while the attributes remain (items purchased, money spent), any personally identifiable information (name, credit card number) is removed. “This eliminates the risk of individuals being identified,” Allwright says.

In the case of ReachOut, the data being analysed are survey responses from ReachOut users over a period of several months. The data includes information on a range of topics including mental health symptoms, perceived need for care, mental health literacy, exposure to other people experiencing mental health problems and user impression of ReachOut.

At a November meeting, three volunteer data scientists were digging into this rich data set at a Minerva meetup. The core team working on the project over the past several months is Lance Abel, a former trader who dabbles in cryptocurrency; Dan Bridgman, who works at PwC and is currently pursuing a master’s in data science; and Allen Nugent, who studied physics and medical devices before finding data science. The goal is to develop a dashboard using Tableau, a data visualisation program, that will allow the organisation to dig further into their own data. “The dashboard gives them a much better platform to ask questions about their data,” says Bridgman. “The intention is to provoke curiosity.”

For ReachOut, the expertise provided by the volunteer data scientists allows them to harness skills they might not have in house, says Lorraine Ivancic, who leads data analytics efforts at ReachOut. “At ReachOut, we want our service to be evidenced-based, so we do a lot of research internally, but it’s great to have external resources to do research that we may not have the capacity to do.”

The team splits up responsibilities, with Abel working on a principal component analysis, which can reveal patterns in the data; Nugent working on a Sankey diagram, which will help ReachOut find patterns over time; and Bridgman working on the dashboard visualisation. Most of the intense data analysis happens outside the regular meetups, Bridgman says, and the team uses typical techie tools (such as Slack) for communication.

At Minerva’s first meet-up, there were 50 volunteer data scientists working on several different projects, from mental health to climate change to domestic violence. At the November meeting, there were about 30 working on four different projects, though Allwright says the total community is closer to 300. Minerva also holds longer hack-a-thons that give data scientists a longer time to brainstorm, analyse and create a product. In September, Minerva held a hack-a-thon focused on homelessness, and a hack-a-thon coming up in December will focus on domestic violence.

Minerva also offers a corporate engagement program that matches a company with a not for profit to work on a data project. “Often millennials really want to have purpose in the work that they’re doing,” Allwright says. “This is a way to create that employee engagement and corporate social responsibility, and creating social impact can be part of the journey.”

For Bridgman, volunteering allows him to use his data science skills in a meaningful way. “I wanted to give something back,” he says. “I thought the best way for me to do that was with the skills I use in my data job.”

On this current project, Bridgman is using his skills to fine tune the dashboard so that ReachOut can use it to analyse its own data, but the Minerva team also hopes to do some more in-depth analysis of their own on the survey data. Finding patterns in the data will help nonprofits get a better idea of the root causes of problems they’re trying to address. And that will ultimately lead to better solutions. “We really believe it’s a movement,” Allwright says.

Learn more about the Minerva Collective on their website.

Abby Callard

About Abby Callard

Abby Callard is an American writer based in Sydney. She has lived and written in Chicago, Washington, D.C., Spain, and India. She’s particularly interested in analysing the food system through a social justice lens and initiatives that make technology easier to access.

More from Abby Callard

Details

Technology for Purpose
Sydney NSW, Australia
1st March 2018

Share

Reactions

Find out more about us and become part of the QUOmmunity.

There are currently no ${type} listed. Start promoting your ${type} with us by signing up to one of our subscription plans!