-
Bark
During the opening ceremony, we saw that there was potential in IBM Watson's available APIs, and when Hack Harassment reminded us of the threat of cyber-bullying, we had our goal and hit the ground running as hard as we could, especially due to the the suicide at one of our team mates' school due to the unfortunate cyber bullying and online harassment.
What does DoggoBot do?
It scours a given Twitter handle's timeline, using IBM Watson's Blue Mix to input aggregated tweets into the Natural Language Processing API and Tone Analyzer to determine both the emotions and level of negativity of the tweet. These emotions then became statistics, which we could compare to a level of significance. From there, we recommended the user to determine their own course of action, helpfully leaving a report button to their bidding.
Parallel to that, we also allowed the user to enter their own Twitter ID Key information, which we would eventually use to watch the user's timeline if another handle were to harass the user. The tweets were graded equitably, but this time we allowed the user to report, mute or even block the handle.
We came up with the concept of DoggoBot over two things: 1) Dogs are insanely adorable, 2) Dogs like to dig--for bones, for people, for information, and 3) Dogs bark, warning and informing others about the threat and existence of cyber-bullying and From IBM's first workshop, we learned that Twitter was the perfect social media platform to delve into. It was public, it was uncensored at times, and it was volatile. There were perfect handles to use as data.
Initially, Kevin had to first retrieve the data, so once we had agreed on using Java in Eclipse, he got to work tinkering with the installation process for the Twitter Developer API. William worked on the front end of the program, meagerly only setting up the UI. Kamyar, however, was the noble worker of our group. He stepped forward into unknown territories: IBM BlueMix. With him, DoggoBot was able to detect human tone with Watson's technology.
We faced many hardships, but primarily Kamyar. Numerous bug appeared in our code, and halfway through our project, we learned that Twitter forbid third-party softwares from collecting user's login information. Having spent several hours trying to find a work-around, we eventually submit to their will. Our biggest gripe, however, was that we did not have access to a credit card in order to purchase a domain and build our own website.
All of us our new to the hackathon environment and we are incredibly proud of what we have achieved over this time period, as it was our first immersion into this sort of experience. The intense time limits were restricting, yet motivating. Our ability to rapidly adapt to the new libraries and APIs was particularly empowering.
Had we been given more time to work on our project, we would've included some of our ideas to provide cross-platform support for our project, particularly for Android, IOS, and Pebble Time. We were also considering creating a web app, which is likely the best solution for providing DoggoBot's services to others.
Log in or sign up for Devpost to join the conversation.