Open cheerfulstoic opened 9 years ago
I was just thinking of subjective ratings to start, but the rest is worth considering. And the cultural aspect is worth noting. Maybe ratings could be standardized somehow based on the user's overall ratings distribution. You'd probably need a fair number of ratings first...
Do you mean like (1) subjective ratings, (2) aggregated ratings or (3) something like "noderank", where an asset gets a rating based on how many people accessed it, and their respective ranking, etc. For (2), you're going to need (1), with relationships between users and assets to stop fraudulent ratings (with people voting more than once). Is a "noderank" algorithm generic enough for (3)? And could one just optionally layer the GraphAware noderank plug-in onto the Asset Portal and customise it.
PS: I find ratings very culturally specific. If I tell a European that a project's success was 7/10, that's really good. But to a Japanese, that could be very worrisome. Similarly, a 4/10 isn't a disaster for a European, it means there's room for improvement; but to a Japanese that's like a 2/10. Anyway, that has been my experience in the past. I can't speak for other cultures, or even for all Europeans/Japanese.