Kune

Welcome to our Project Kune - Safer in Numbers, together.

Inspiration

When it comes to keeping the APIDA and even the elderly community safe, we thought about what would make the most impact. While we can choose to virtually expunge hurtful words, block hateful people, and leave unwelcoming online spaces, we can't do the same physically in public. Even a Latina grandmother received undue violence because she was mistaken for being Asian. (https://www.independent.co.uk/news/world/americas/latina-grandmother-bus-antiasian-attack-b1832905.html)

So when it comes to leveraging artificial intelligence and machine learning to detect, stop, and mitigate violence directed towards Asians and the elderly, it starts with having a good dataset.

While the FBI has the most comprehensive set of hate crime data, it's far from being as comprehensive as we'd like. All data is voluntarily reported to the Bureau - and people like you and me generally don't have the FBI on our speed dial list. As such, most hate crimes end up going unreported. (https://www.apa.org/advocacy/interpersonal-violence/hate-crimes)

The numbers are dismal:

• only 14 percent of the over 15,000 agencies that participated in the FBI’s hate crimes program actively reported incidents to the FBI in 2019
• 40 percent of states don’t require hate crime data collection or don’t have hate crime laws in place; and

The most thorough data collection we found was in New York City, https://app.powerbigov.us/view?r=eyJrIjoiYjg1NWI3YjgtYzkzOS00Nzc0LTkwMDAtNTgzM2I2M2JmYWE1IiwidCI6IjJiOWY1N2ViLTc4ZDEtNDZmYi1iZTgzLWEyYWZkZDdjNjA0MyJ9

Additionally, there is no national, standardized definition of what a hate crime is

As such, we designed Kune, a mobile app that crowdsources data from the ground up. It's designed so that you use it because you care about not only your own but also the safety of the people you care about.

What it does

Kune means "Together" in Esperanto because ultimately we are stronger and safer together. We mean this in several different ways:

First, Kune is a social app where you can connect with and chat with friends to walk over together to the same destinations. Individual offenders tend to think twice when attacking people in groups because of sheer numbers. As an ally, you can also use this app to walk over with others and be there if one of your friends or someone nearby needs help.

That is, if you're walking down a street and you see someone giving you unwanted gestures, you could signal on the app that you'd like help. Tapping once sends a ping to your friends to let them know your location and that you may be in a vulnerable situation, but it's not confirmed. Depending on your settings, tapping twice could mean sending a ping to any friends who are nearby and the option to notify nearby strangers in the area to start heading your way as potential helpers or witnesses. Tapping three times rapidly could send a ping to your local public safety officers, ping your friends that you've confirmed this is not a safe space, and activate your camera to take pictures or videos, if possible. Those pictures and videos where you felt unsafe to go into your profile where your friends can freely view and offer their support and empathy and maybe reach out to say, hey, next time you go to this cafe, feel free to bring me with you.

The second purpose of this app is to crowdsource data for those moments when people are feeling unsafe or have already experienced that brief moment of hate. The data signaling when you felt unsafe would be recorded in the backend anonymously with only location coordinates and a date timestamp, and the data would be made available to public safety authorities and the entities they cooperate with local college campus public safety officials or community health organizations for further AI and machine learning or manual analysis. That way, they can distribute resources accordingly to see why people might be feeling a certain way at certain locations on certain days/times.

Based on routine activity theory, which suggests that offenders find targets within the context of their routine activities like hanging out at an area during lunch (meaning they don't tend to go too far out of their way to commit crimes), AI and machine learning can detect, stop, and mitigate these types of occurrences. It can also reveal higher-level insights like how most crimes happen during the day in big cities, in part because there are so many more people out and about during the day. (https://www.usnews.com/news/cities/articles/2019-06-12/study-finds-crime-in-big-cities-is-more-likely-during-the-day)

Third, this app allows a way for users to explore that data and be empowered to come to their own conclusions or even mention patterns to authorities who may be able to do something about crime patterns. By integrating data from Kune users and publicly available crime data sources, users can have the bigger picture when they make decisions when traveling, especially when traveling alone.

To be clear, rather than being the official source of hate crime data, our goal is to provide additional context not only for algorithms to analyze but also for regular people to make decisions concerning their own safety.

A few data points from one person is a start. But when we collect data points from many people, we can see the bigger picture and do better to protect not only ourselves but also our community. That's why with Kune, we can be safer, in numbers, together.

We started building a data dashboard for the exploration of Asian hate crime in New York City. https://datastudio.google.com/reporting/a95d1075-641f-45fc-8040-319a424d37b9

How we built it

Figma, HTML, and Google Data Studio.

Challenges we ran into

The major challenge we ran into was finding hate crime data. There isn't a lot of data available. We faced some issues when creating the bubble maps in Google Studio.

Accomplishments that we're proud of

We are really proud to have started working on an app that can keep people safe.

What we learned

We learned how to use Figma, we learned that hate crime data is not widely reported - it is voluntary. There are no data formats to submit this data. The most important thing is that we learned more about the APIDA community.

What's next for Kune

If we had more time, we would've built a mobile app in React Native with a cloud database like MongoDB. One team member had trouble setting up a mobile dev environment, as the emulator would not connect with the IDE, and this member has never created a mobile app before.

Built With

• figma
• google-data-studio
• html5