| Hi-index | 0.00 |
Current BtoC recommendation services utilize consumers'purchased log as criteria for selecting information,yet it includes little information of the reason why hebought the items. Thus it is difficult to recommend thesuitable information for each consumer.We have observed how each consumer judges his likesand dislikes on objects viewing them. We have modeledeach consumer's evaluation process by relationshipsamong physical features of objects, each consumer's subjectiveinterpretations and preferences. Thus, based onthe models, our system can estimate users' subjectiveevaluations and preferences from physical features ofobjects to perform a suitable recommendation.We have also build situated preference models accordingto the usage of the items. The recommendation systemrefers such situated preference models together with thesubjective evaluation model as the consumer's decision-makingmodel for selecting objects.