Alex kept complaining that I stole "his" pizza slice (I officially deny everything), but this alleged incident that did not occur gave us an idea: a mobile app to let people report crime swiftly, easily, comfortably. The problem is that many people who witness a crime are reluctant to come forward, and want an anonymous, non-face-to-face way to do so. No one wants to be the "snitch", and people are afraid of repercussions from the perpetrator, or being called forward in court, or being accused of doing the crime itself. Sherlock lets authorities launch/conduct the investigation without implicating the one reporting it.
What it does
Sherlock features an intelligent bot that aggregates information about the crime, through a casual, conversational approach.
How we built it
We built it by designing an intuitive user interface, developing our own natural-language processing system, and integrating it into the app.
Challenges we ran into
This whole fckin day was a challenge bruh. But more specifically, we faced difficulty coming up with an idea, creating a list of inquiries for Sherlock, developing a system to process user input, and generating responses.
Accomplishments that we're proud of
Our expectations were so goddamn low that we proclaimed victory after we got a simple while loop to work properly, lmao. After that though, we were proud of our nice-looking color palette (bright red and casual gray), our comprehensive list of responses, and our overall system for word analysis.
What we learned
We learned how to scan input for select keywords, randomize same-meaning responses to make statements less generic, and making scrollable text in Android. More significantly though, we learned a lot about crime and how it manifests, as well as the overall reporting process.
What's next for Sherlock
Integrating it with police and federal authorities' systems to alert them of the incidents that we pick up, like child abuse, murder, theft, etc.