People like to be experts, it is stated. Thus they want to share their expertice in restaurants, shops or different products among their peer group. This works conveniently for online services seeking relevant content.
It is also said that people like to utilize insight provided by those they trust as part of their purchase decision process, but there is a catch.
Challence 1: What is the motivation of the expert to create evaluations? Does volume go over quality? What is the relationship between expert and the target evaluated? These are not always clearly visible for those utilizing evaluations. Thus evaluator profiles become increasingly important for end user to be able to check relevance of evaluations.
Challence 2: If you want to use evaluations created someone you trust to have judgement, how do you find that among wast ammount of evaluations? Some services offer opportunity to prioritize those who belong to your own social media circles, but you do not neccessarily have an expert among people you know for all purposess. For these situations it would be neat to be able to select evaluations based on your own profile: people with similar values, preferences etc.
When tackling these challences in the back ground there is always the rude fact of small ammount of evaluations or limited capability to filter those. Usually inorder to get relevant and fresh evalauation, categorization of them is not really possible, because of limitted ammount or limitted information of evaluator to base categorization (language, date and stars is quite limited).