Recognizing customers' mood in 3D shopping malls based on the trajectories of their avatars

This paper proposes a method to assess the cognitive state of a human embodied as an avatar inside a 3-dimensional virtual shop. In order to do so we analyze the trajectories of the avatar movements to classify them against the set of predefined prototypes. To perform the classification we use the trajectory comparison algorithm based on the combination of the Levenshtein Distance and the Euclidean Distance. The proposed method is applied in a distributed manner to solving the problem of making autonomous assistants in virtual stores recognize the intentions of the customers.