We as a team think that harassment is one of the biggest problems that the youth faces in the world. We all believe that it is important that as hackers, we should promote harassment awareness to combat and restrict this unfortunate act. This is why we built the Harassment FAQ Bot.
What it does
We have designed our Harassment FAQ Bot by compiling data from multiple trustworthy sources so that the product can convey the information in a succinct and user-friendly environment, the Microsoft Web Chat for Bot service. The user can access this FAQ Bot through a web app and also a text messaging interface. The QnA API also uses Natural Language Processing to identify the most relevant response to the question posed by the user.
How we built it
We built the web interface using Python Flask hosted on Microsoft Azure Cloud. We also used Microsoft Bot Framework and Cognitive Services APIs such as the QnA Maker. We also integrated Twilio API with the Bot Framework so that users can access the Bot from text messages. The data for the Bot FAQ was collected with the use of the QnA API by linking the URL for many trustworthy harassment awareness FAQ websites.
We manually pulled the data from following sources to power our bot's capability. https://www.stopbullying.gov/what-is-bullying/index.html http://www.cde.ca.gov/ls/ss/se/bullyfaq.asp
Challenges I ran into
While this project was incredibly satisfying, we encountered several challenges along the way. First, we had trouble setting up Python Flask on Azure adding custom domain.
Our original plan was to use speech-to-text API and language translation API to enable more users to comfortably communicate with the Bot. However, once we implemented the Bot Framework in accordance with QnA API, we realised that we could not tap into the input/output of the Web Chat, which simply meant that we could not host these features in this Bot. On the contrary, because we used Bot Framework, we were easily able to add Twilio API integration.
Accomplishments that I'm proud of
This was our first time using APIs and Cloud Hosting services like Microsoft Azure. We are extremely proud of the fact that we successfully used Bot Framework, QnA API, Microsoft Azure and much more. In fact, the most rewarding part of this project was the fact that we acquired much more knowledge that goes even beyond the span of our project.
What I learned
We learned how to use APIs efficiently and effectively to create powerful applications that can benefit society.
What's next for FAQ.ts
While Harassment FAQ Bot is great at providing information about harassment, we believe the next step should be an interface that lets user take action against harassment. In some form, users should be able to report harassment to officials so that harassment can be minimised in the larger picture.
We also believe that we need to add speech recognition and language translation. This would make this Bot much more practical.