Originally Posted by HomeInMyShoes
It generally depends on people who like this also like this data. As also has been said, it could also factor in people who look for this also look for this. There's a site that asks random questions like "do you start with your left or right foot when you tie your shoes" and uses similarities in answers to suggest things people would like. It uses non-domain knowledge to try to make suggestions within a domain.
It is fascinating stuff.
I like going to literature map to find random authors every now and then. I'll look at the close ones and then other outliers. I'm probably the one skewing the data.