[D] What analysis should I do to decide between a personalisation algorithm and a general algorithm?

I am currently working on a ranking problem. We want to incorporate dislike information into our platform, i.e. if a video will be disliked, we want to move it further down. Now there are two ways of doing this - one is to create a general algorithm that predicts whether a video will be highly disliked and use this information to rerank the videos that are recommended. Another is to create a personalised algorithm, i.e. downrank any video that we …

algorithm analysis general machinelearning personalisation