An Emphatic Humanoid Robot with Emotional Latent Semantic Behavior

In this paper we propose an Entertainment Humanoid Robot model based on Latent Semantic Analysis, that tries to exhibit an emotional behavior in the interaction with human. Latent Semantic Analysis (LSA), based on vector space allows the coding of the words semantics by specific statistical computations applied to a large corpus of text. We illustrate how the creation and the use of this emotional conceptual space can provide a framework upon which to build "Latent Semantic Behavior" because it simulates the emotional-associative capabilities of human beings. This approach integrates traditional knowledge representation with intuitive capabilities provided by geometric and sub-symbolic information modeling. To validate the effectiveness of our approach we have simulated an Humanoid Robot Robovie-M on dInfoBots a linux based framework developed in Our Mobile Robot Lab.