Wow. What an experience.
Obiegba Onoriode Jonathan
11

You probably wrote it as a comment to my other post on the nanodegree.

Udacity does not have any courses on network visualization. So , I’ll take most of the credit along with my teammates for this project. :D There’s a course on data visualization if you are interested though.

I’ve taken 4 network science based MOOCs in the year before I even got started in data science. The dataset comes from kolpokoushol.org who got it from some stats department in Dhaka. I wanted to work at the intersection of data and news in the hackathon.

I’ve worked on network visualization with Gephi in past. I ended up using NetworkX for creating the network. My teammates were also super helpful and worked with me on the presentation and demo.

Ananna Farzana was a journalism student who helped with the demo and Saad Akash from Dhaka university CSE department helped me with the management, logistic related issues while I coded. Ananna also gave me some lessons on public speaking.

You can also see that this project is not machine learning intensive at all. This project uses the idea of a ‘co-occurrence network’ only. The interesting part is that even using such a rudimentary idea can result in a pretty good interactive network that depicts current status of violence against Bangladeshi women. I could have refined the network, but 4 days was not sufficient.

Obviously, the nanodegree was super helpful too. I had more experience and confidence in data munging from the nanodegree projects.

Show your support

Clapping shows how much you appreciated Tahsin Mayeesha’s story.