[toxicFork]

Let's develop a machine learning algorithm which can decide the likelihood of misleading title by checking what contexts are involved

[misterman0]

Before we have developed a machine learning algorithm which can decide what contexts are involved in a corpus by checking which words are included, and in what order and within what subcontext?

[toxicFork]

Maybe.

[misterman0]

I like your reasoning, it is worthy of both karma and LOL's.

[snaky]

Good old list of hype-words is enough.