Net Haven

Inspiration

We originally wanted to win a prize so we looked at different categories. The idea of Safe Haven was born when one of our group members thought it would be a cool idea if we could create a web screening for people who may be young and going through difficult phases in life and are affected by bullying or harassment, or who simply struggle with serious anxiety. The idea was to be able to be warned about possible the hostility that can appear on the web. This is just a demo but the goal is to create a filter for those who want them.

What it does

Net Haven gets a URL from the user then checks all the text on that web page via a 2 step analysis. The first analysis step is IBM Watson's Tone Analyzer to determine the levels of Anger, Disgust, Fear, and Sadness in the content of the given page. Then we used bokeh, a python graphing library, to show these levels in a graphical model. The second analysis step grabs the text from the given site and compares it to a dictionary of racist, sexist, and overall offensive content. Then we get an average of how much vulgar content is found and display it graphically through bokah. Lastly we compare the 2 analysis levels to get a range of the harassing content on that web page and determine if it’s safe, partially dangerous or dangerous.

How we built it

We used Python Jquery and Javascript with some JSON files, to create the string parser that separates the text from the HTML on the webpage. Then we multi-threaded the analysis process to run both levels of analysis simultaneously. Later we created the Web version for the Chrome extension using more Javascript and ngrok to allow the software to be used as a Chrome extension, and we used sockets to transfer data from python to Javascript. .

Challenges we ran into

Connecting the Python to the Javascript in a fast and effective manner, displaying the software as a non-invasive Chrome extension, and rendering the result of the analysis in a timely manner, using IBM Bluemix API's without accessing the credentials.

Accomplishments that we're proud of

Everything! We were happy to use python and its multiple libraries to run the analysis as fast as possible, being able to use Watson API's in a project, using socket programming to connect Python to Javascript and creating an app that can actually help and make an impact in the global community.

What we learned

Lots of python code, a better implementation of socket programming with Javascript and Python, focusing on creating apps that run on a timely manner for websites.

What's next for Net Haven

We would like to add support for different web browsers and hopefully create a system for comparing images and videos to see every possible form of harassment.

Built With

Submitted to

MangoHacks '17

Created by

I worked mainly on the server side of the application. I connected the python scripts to the IBM Watson api and parse the data into a graph using bokeh. All of these scripts were feed to main script that created threads to help keep order.

Private user
I worked predominantly on the back end of the software. I worked with the API calls, the threading, the creating on the graphs using bokeh and the data comparison analysis.

Jim Louro
I worked on getting the communication between the JavaScript from the chrome extension and our python code that handled the logic of our application up and running. This was difficult for me because I had not done much developing with python before this hackathon but I ended up learning a lot. I also worked on the data parsing element of our logic so that the data sent to our tone analyzing function was in a valid format. I also contributed to the debugging phase of development which took longer than expected, but worked out in the end.

Kyle Roque

Updates

Private user started this project — Feb 26, 2017 08:23 AM EST

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.