In many areas we need to model personal perceptions of people relating instances of concepts. Take the famous pizza topping ontology and extend it to perceptions of guests in a restaurant. Those guests of course need an interface to communicate (or vote) for their perception (say, salty or spicy), which is then incorporated into the ontology as a property value (say all_votes_for_spicy/all_votes_for_perceptions, and so on). If the value is too low, the corresponding topping wouldn't appear on the electronic menu. What's your view on this?