This is by way of a reblog from John Spiers, whose posts don’t have a ‘Reblog’ button.
As those who follow my poetic efforts will be aware, I am not at all confident with anything of my own that doesn’t rhyme. (See But is it Poetry?) I can usually recognise the poesy in others’ blank or free verse but, without rhyme, I am never sure that my own efforts aren’t better described as prose.
Can an algorithm evaluate poetry?
The two sites John has found give a score depending on how well your poem (or lyric) compares to language used in certain recognised poetic or lyrical successes.
Scores above zero (maximum = 5) indicate that the poem is linguistically more like poems by ‘established’ poets such as those published in the Poetry Foundation’s ‘Poetry Magazine’ and Scribner’s ‘Best American Poetry’ series.
Scores below zero (minimum = -5) indicate that the poem is linguistically more like poems by poets who have not been published by mainstream establishment poetry publishers.https://www.poetryassessor.com/hello
… which I thought was a tactful way of putting it. The lyric assessor gives a similar explanation once your song has been assessed.
The poem I’ve linked to in the second paragraph above (with a similar title to this post) achieved a negative assessment of -0.6, although others have scraped more positive scores.
My submission this month to the 12 poems in 12 months challenge at deadlinesforwriters.com ran to two versions of, essentially, the same poem. The rhyming version scored better on poetry assessor than the free verse. (Although feedback from real people on the 12 poems site disagrees. Are they just being kind?)
Both versions will appear on here in due course for your opinion. I have bookmarked both ‘assessor’ sites and will ‘assess’ my next monthly effort before I post it.
I shall, of course, also be assessing some of the other poems from the challenge – just out of interest…