We have a system that takes text from blogs and the like and converts them to video. Need help in the review stage.
1) First step is taking the paragraph and splitting it up into slides and bullets points or phrases that read well and can each stand on their own e.g.
"mary had a little lamb who's fleece was white as snow" would break up into
mary had a little lamb
who's fleece was white as snow
mary had a little lamb who's
fleece was white as
2) Each of these sentences will have one or more images needed to accompany it. e.g. a picture of a little girl with a pet lamb, or a picture of mary and a picture of a lamb.
Both steps 1 and 2 will have a first pass done by both computers and crowdsource workers, who may
1) do a great job and then all you have to do is accept (paying them),
2) Ok but require correction (accept, then correct, then submit)
3) not do the work at all (reject, then do the phrasing and image research, then submit)
4) do a horrible spammy job (reject +block this worker)
Each article is around 150 words and there are currently 150 with many more expected potentially enough work for several. On average it takes 5 minutes for me to QA/Proof read, and 10 minutes if there are many images needing to be found.