ڹֱ鳵ס

Web-Based Five Senses Input Simulation --- Ten Years Later

Rafal Rzepka (̳ƻ ر ʳظ)



A decade has passed from our paper in which we question the sensors technology not being ready to give a machine the five senses with the recognition level close to human's and, as we expected, sensing devices are still not sufficient to become a base for the world knowledge based reasoning. Although we did not developed the proposed system any further because of limited access to Internet data, the need for machine experience is even more urgent than before. Existing world knowledge retrieval systems tend more and more to concentrate on name-entity information ignoring commonsensical knowledge which is still one of the most difficult obstacles on the path for reasoning about everyday situations. In our paper we will introduce results of our latest trials to implement universal theories to enrich machine's capability to simulate environmental perception and human behavior understanding.