No abstract

This paper describes a methodology for navigation and exploration assistance intended to enhance user satisfaction when exploring three-dimensional virtual environments. The complexity of such environments often makes navigation and information retrieval difficult, making it necessary to add assistance components to the world in order to turn it easier to manipulate. This methodology uses three-dimensional "intelligent" avatars as interactive guides, along with information-based navigation strategies. The intelligence of the avatars is represented through physical features, behaviors, and knowledge about the user and the environment. These components establish the avatar's architecture. Content personalization according to the user's interest, navigation assistance according to the desired content, and avatar guides that make the virtual place more realistic and pleasant are proposed in order to involve users. A three-dimensional model of the Guggenheim Museum Bilbao, in Spain, is presented as a prototype for the validation of this methodology. In the museum, the guide is represented as a fish that, according to the user's preferences, assumes gender and age. The avatar swims through the museum, following navigation routes that lead through exhibitions previously chosen by the visitor.

Abstract. This paper describes the study on example selection in regression problems using µ-SVM (Support Vector Machine) linear as prediction algorithm. The motivation case is a study done on real data for a problem of bus trip time prediction. In this study we use three different training sets: all the examples, examples from past days similar to the day where prediction is needed, and examples selected by a CART regression tree. Then, we verify if the CART based example selection approach is appropriate on different regression data sets. The experimental results obtained are promising.

No abstract