---
title: "Common-Sense Knowledge of Space: Learning from Experience"
authors:
  - "Benjamin Kuipers"
doi: ""
url: "https://ijcai.org/proceedings/1979/111"
pdf: "https://ijcai.org/proceedings/1979/0111.pdf"
venue: "Proceedings of the Sixth International Joint Conference on Artificial Intelligence"
year: 1979
pages: ""
track: ""
keywords:
---

# Common-Sense Knowledge of Space: Learning from Experience

Benjamin Kuipers

*Proceedings of the Sixth International Joint Conference on Artificial Intelligence*.

## Abstract

The TOUR model is a computational model of human commonsense knowledge of large-scale space. It shows how ob sN \' alions arc assimilated into a description, from multiple perspectives, of the spatial environment. In this paper we propose a representation for sensory events at a level suited to this investigation, and an inference s trategy by which the se ~en s ory events are assimilated into descriptions of spatial structure. We al so di sc u~. s the stales of partial knowledge that occur during the learning process, and show that the represenl<~llon exh t h•l ~ graceful degradatton of performance under resource limitations.