> I'd say it's more likely that this is a complex problem, and will take a bit of time and computing power to work through...
It is a complex problem - but amazon has a serious advantage over other sites that have to deal with such issues (i.e. Twitter) - in that they have significantly more information on each user. I don't think Amazon is short on computing power either.
Taking into account order and browsing history, product review trends, linguistic similarities in review posts etc. They should be able to get very low error rates in identification.
Further unlike something like twitter feeds, it's quite possibly to silently de-prioritize abusive reviewers and associated products. Really, I'm quite surprised at how bad of a job they are doing - most of these cases are so blatant and obvious they should not require an author and a live representative to resolve.
It is a complex problem - but amazon has a serious advantage over other sites that have to deal with such issues (i.e. Twitter) - in that they have significantly more information on each user. I don't think Amazon is short on computing power either.
Taking into account order and browsing history, product review trends, linguistic similarities in review posts etc. They should be able to get very low error rates in identification.
Further unlike something like twitter feeds, it's quite possibly to silently de-prioritize abusive reviewers and associated products. Really, I'm quite surprised at how bad of a job they are doing - most of these cases are so blatant and obvious they should not require an author and a live representative to resolve.