The same the other way. The day bin Laden died is notably "unhappy" because a lot of people said "dead" or "death" but I do not recall it being seen as a negative event at all.

I saw the point where Paul Manafort received a guilty verdict, and it is also one of the lower points. My first thought was that it's still a negative situation, so even if people are happy about the outcome, overall it's still a drag. It still brings back all the things that had to happen to get to that point.

Also you have to wonder about some of the associations like on the day of the shooting in Dayton, OH [0] one of the few positive words is "white".

[0] https://hedonometer.org/index.html?from=2018-02-15&to=2019-0...

I don't this is necessarily a good measure of how happy people actually are. However, it is a good measure of how people are going to perceive the mood of society.

In other words, if most of what an average twitter user sees on given day is people wishing each other merry Christmas/happy valentines day/etc. then they're going to perceive that things are mostly going OK for people.

I think "happy <holiday>" is usually used as a shorthand for "have a happy <holiday>"

I think the effect you're describing is real, but... I turned off all of the days except for Saturday and Monday and it was interesting to see that with few exceptions (and those exceptions laying mostly around holidays), Saturdays were rated as having a higher "happiness" than Mondays.

Can’t they just measure which emojis people use?

This actually makes a lot of sense to me. Nice idea.

Yeah. I think an NLP solution would be more accurate in actually analyzing the sentiment of tweets.

Even then sentiment analysis is comically broken. I was playing with a tool that analyzed Tweets and plugged in a variety of words and phrases. It thought the term “white male” generally had a pretty strong positive sentiment, but on drilling into the Tweets, every single one that it coded positive was actually very disparaging. However, those Tweets were often overtly sarcastic and/or used terms like “privilege” which S.A. models believe to be an indicator of positivity. Some S.A. models think that the term “white” is positive and “black” is negative (probably not a racist conspiracy, but rather because they are too simplistic to understand the racial connotations).

Yeah I've yet to see any NLP solution that comes close to being able to understand sarcasm. It's not an easy thing to do though it requires understanding the whole culture a statement comes from and the message as a whole because of how much earlier sentences change the actual meaning of words.

I doubt we'll have something that really grasps sarcasm until we get close to full human level AI given how much it can trip up even people when communicated by text.