Big Mountain Data Aims to Change the World
Using Data to Determine and Deter Domestic Abuse Offenders
For Susan Scrupski, the possibilities of data are clear: with the founding of Big Mountain Data, her not-so-modest goal is to make all the information currently collected by towns, cities, and municipalities across the United States on domestic abuse offenders widely accessible. “I’m trying,” she says, to “solve a societal problem.”
In some ways, it’s no surprise that Scrupski would take on this mission. She is a believer in the power of the Internet to solve problems through collaboration and cooperation. Her background is in tech: she founded both the 2.0 Adoption Council and Change Agents Worldwide, and has a strong presence on social media. In 2010, Fast Company named her one of the most influential women in technology. Thirty years ago, Scrupski was a victim of abuse, which she’s written about movingly.
Before founding Big Mountain Data, she spent time researching domestic violence and forming connections with organizations and advocates. Scrupski is a relative newcomer to the field, though. “I don’t,” she says, “have any street cred.”
Perhaps it’s this outsider status that helped her to see another perspective for eradicating domestic violence. Because she was a new voice in the conversation, Scrupski comments, “I didn’t need to conform to a set of beliefs. I can look at this with fresh perspective and draw upon my knowledge from the tech community to look at the problem in a fresh way.”
Scrupski credits a conversation with a former coworker, Erik Huddleston, with focusing her on offenders, rather than victims. In talking to Huddleston, who is now an advisor to Big Mountain Data, they were able to “recognize that what you really have in domestic violence is big data.” The business plan for the company developed by “putting together the need and the opportunity.”
A Different Angle for Stopping Domestic Violence
Domestic violence is a massive problem; the numbers are both staggering and disheartening. The CDC reports that every minute, twenty people will be the victim of violence from their partner. And it’s an expensive problem: costs from domestic violence, including medical attention, lost time at work, and mental health support, amount to $5.8 billion each year. One in four women, according to Safe Horizons, will experience abuse in her lifetime.
As Scupski comments, “I don’t know a woman, who does not know another woman, who has been abused. I don’t know one.” It’s a tragic statement, but it reveals the volume of data available. Where there are many crimes, there is much information.
Big Mountain Data’s mission is to use all this available information on abusers to “datafy the repeat offender.” There is, it turns out, a lot of data on the perpetrators of abuse. The hitch? Too often, the details, charts, and information are available only on PDFs—useful to print, read, and circulate, but difficult to combine and analyze. Still in its infancy—the company launched in late 2014—Big Mountain Data’s goal is to make that information readily available. “We’d like to accumulate the data from offender-focused programs and start looking at it in aggregate and giving the benefit of the data science approach to develop specific applications and tools for law enforcement to be more insightful and informed about the prevention they’re doing,” says Scrupski.
From law enforcement to charitable organizations, there are many people—and millions of dollars—focused on victims and their families. But talk to anyone who works in a domestic violence organization, Scrupski says, and you’ll hear about the frustration of knowing that the abusers will likely reoffend—either with their current partner, or a new one. No amount of support for victims and their families can stop that cycle completely. Scrupski doesn’t want Big Mountain Data to displace any of these powerful agencies, but she does “see an opportunity to get at the root of the problem, which is the offender.”
With access to data from across the nation, would predictive modeling be able to reveal the best moment for deterrence? Could data scientists identify the external variables most likely to impact behavior? Ideally, datasets shared by Big Mountain Data will allow data scientists to draw conclusions, and make law enforcement more effective in intervention and stopping repeat offenders.
Charting the Conversation
From hashtag activism to front-page news, domestic violence is a big part of our national conversation lately. Scrupski approves—too often, abuse is a private matter, a shameful secret, known only to the women, children, and families mired in violence. After NFL player Ray Rice was seen on video abusing his fiancé, countless domestic abuse survivors shared their stories.
This outpouring of narratives spread across the social web led to one of Big Mountain Data’s first projects: a data visualization of tweets using the hashtags #WhyIStayed and #WhyILeft. “It blew my mind,” says Scrupski, “how many women were affected by this hidden in plain sight scourge upon our society.” Big Mountain Data’s goal was to provide a meaningful visual structure to the individual narratives.
What’s Next for Big Mountain Data
But exposing the stories shared on the Internet is just one facet of Big Mountain Data’s mission. Big Mountain Data’s biggest goal is to create a “massive national database of all this data, so that civic hackers, teams, universities, and anyone who wants to have access to the data can get in there and start developing programs for their own cities,” says Scrupski.
That’s where Socrata comes in. Big Mountain Data’s plan is to upload the data they have available to Socrata’s platform, and “give free, open, easy, elegant access to this data for any data scientist team that wants to create an application or visualization.” In their first hackathon, Big Mountain Data used a file-sharing service to distribute data, and while that worked, it wasn’t exactly an elegant or easy solution. The benefit of Socrata, Scrupski finds, is the ability to offer visualizations within the platform, and management techniques for data that isn’t clean or well organized.