Alright, I guess the word at stake here is “disruptive”. I’m not a native english speaker, and did not know it had a negative meaning.

For what I sensed, disruption is used a lot to describe technology that ultimately changed the whole paradigm. Honnest mistake here.

I’m not here to defend my skills, but I’ve studied mathematics, and this study never meant to compare with complex statistics about finance or anything. My point of view on that is that, it is (I stress that word) scientific. Even if simple. Because, look I’m sorry but vocabulary, word use, sentence length are, in a context like the one the speeches were used, a key to how we receive information. My study is not a mathematically complex one because this is not how one perceives a speech, or at least not in what I intended to show. I do think these features do say a lot about how the orators speak. I would agree the dataset is too short. I had many other speeches but I did not expect to make so much views so I did not bother.

Take for example sentiment analysis. When you integrate part of speech into it, you don’t even get one whole percent of better predictions. Most of the time, simple features account for much of the results, may it be in machine learning or other areas. If my plan was to analyse their program, I would not have made this. But I think I made a (maybe insignificant) point.

Thanks for your polite response. I’ll account what you say for a next article. Actually, this one is my first.