This is a good start to measuring complexity. We do need to break down concepts that seem difficult to measure and come up with measures and test them on large amounts of data. In terms of images and text, I'm interested in the relationship between complexity and how much we like a piece of work. I bet the UK flag hits a sweet spot for a lot of people, as do other symmetrical forms that are not too simple. In text, we could compare different forms of poetry, limerick vs haiku, rhyming vs not, sensory vs abstract.
>"I'm interested in the relationship between complexity and how much we like a piece of work."
Yes, I'm wondering about that too. One thing I've been meaning to do is apply it to music, then see if there are trends over the years, and if number 1s might be higher or lower on average. Same for comparing classics in other domains. That would be interesting, if there is a relationship.
What would you predict as the results for the different types of poetry? I may give it a go, if I can get it to output the results in a batch. I have tried a few different types of poetry, and it was pretty consistently scoring higher than prose (of a similar length), which was cool to see.
This is a good start to measuring complexity. We do need to break down concepts that seem difficult to measure and come up with measures and test them on large amounts of data. In terms of images and text, I'm interested in the relationship between complexity and how much we like a piece of work. I bet the UK flag hits a sweet spot for a lot of people, as do other symmetrical forms that are not too simple. In text, we could compare different forms of poetry, limerick vs haiku, rhyming vs not, sensory vs abstract.
Thanks :)
>"I'm interested in the relationship between complexity and how much we like a piece of work."
Yes, I'm wondering about that too. One thing I've been meaning to do is apply it to music, then see if there are trends over the years, and if number 1s might be higher or lower on average. Same for comparing classics in other domains. That would be interesting, if there is a relationship.
What would you predict as the results for the different types of poetry? I may give it a go, if I can get it to output the results in a batch. I have tried a few different types of poetry, and it was pretty consistently scoring higher than prose (of a similar length), which was cool to see.
It's hard to predict for poetry. I have no intuition about it. But unrhyming poetry is like prose with words left out, maybe not as high as rhymes.