Protocol

Introduction

We analysed the controversy on the most visited video-sharing social network in the world.

Knowing that the user base would lean towards pro-net neutrality, we wanted to dive into the debate that was happening inside and outside the videos: we scraped speeches and dialogues recorded on video, and the relative comments and reactions of those who watched and shared them.

We started from a sample of 20 videos, obtained filtering the whole database by view count during 2014 under the "net neutrality" keyword. We watched them and removed 3 misleading results, due to inflated view count or deceptive content, replacing them with subsequents videos on the ranking. Videos were manually classified into 4 macro-areas based on their production scale:

video blogs

web shows

talk shows

politics

Youtube thus allowed us to examine the controversy on two layers: on one side the voice of authors who strive to spread knowledge, on the other side thoughts and concerns of users looking for enlightenment or willing to share their point of view.

Step #1 — comments

For each video we loaded comments, ordered them by popularity and scraped 800 to 2000 of them along with the username, depending on the whole number available, with Kimono. We cleaned the formatting, removing new lines inside comments and misleading symbols, with Sublime Text in order to produce a suitable CSV table.

With Sven's TF analysis we obtained the frequency of the most recurring words and, after grouping similar concepts, we defined 10 keywords. Finally, we went through the keywords and classified them as ethical or technical.