(Another telltale sign is the fact that the link refers to Rolf Heuer as the Director General, but in fact he is the former Director General, while the current one is Fabiola Gianotti.)
That's an excellent idea. I wrote the copy on that page trying to convey "can I get your help against these bad people?", rather than something blaming the user, or a scary looking warning.
But showing them the _reason_ why a certain website is blocked can become an opportunity to teach people critical thought, something that other comment threads point out.
I disagree on this slope being slippery (or actually, this being a slope at all).
Let me be more precise, and use tptacek's words in https://news.ycombinator.com/item?id=12999887: "fake news = spam sites built from scripts consisting largely of a backcatalog nonsense stories [...] with a one or two carefully produced fake stories as a "payload".
I think this definition is as objective as it gets, and clearly excludes things I might disagree with politically, but are not the target of this extension.
Given two clusters, I can predict the class of a new article by choosing the cluster whose centroid is closest to the article.
So, given some training data that produced two reasonable clusters with respect to the ground truth, I have a model that I can expect to generalize well on new data.
Now, this is not what that notebook shows, because it's missing the evaluating on testing data! The main point of the notebook is that the Jaccard Distance of the tokens of the HTML of the page, despite being very simple, appears to generate a reasonable model.
Consider that I built it this night for http://lauzhack.com/, so it's not intended to be a fully-featured solution to the problem.
The main issue I wanted to address is the fact that such a blocker must be widely installed to be useful. Therefore, the Facebook message is intended to prod current users of the extensions to ask that their friends install it as well.
Exactly. Those are the only websites I'm interested in targeting with this extension.
For example, I would never blacklist Breitbart, because I'm not interested in censoring political opinions I disagree with. I just want to free us from the burden of these websites that add nothing to the world and leech attention from everyone.
I had to use a similar approach when creating a cluster analysis of the amendments in the Italian Senate [0].
The Italian Senate offers a SPARQL endpoint [1], which unfortunately doesn't offer access to the texts of the amendments. So I had to roll my own and create a small spider for them using Scrapy [2].
It's interesting to note that High Energy Physicists all around the world will answer "yes", in fact they even created a C++ interpreter: https://root.cern.ch/drupal/content/cint.
I apologize for the self promotion, but some people told me they immediately grasped the Havel-Hakimi Algorithm by solving this puzzle I made: http://jacquerie.github.io/hh/