An Information Infrastructure for Comparing Accessibility Regulations and Related Information from Multiple Sources

This paper describes a research project that addresses the difficulties in dealing with regulatory documents such as national and regional codes. These documents tend to be voluminous, heavily cross-referenced, possibly ambiguous and even conflicting at times. There are often multiple documents that need to be consulted and satisfied; however it is a difficult task to locate all of the relevant provisions. In addition, sections dealing with the same or similar conceptual ideas sometimes lay down conflicting requirements. We propose a framework for regulation representation, analysis and comparison with emphasis on the extraction of similarities between provisions. We focus on accessibility regulations, whose intent is to provide the same or equivalent access to a building and its facilities for disabled persons. An XML regulatory repository is developed to extract structural as well as non-structural features from government regulations to help user understanding and computational analysis. A similarity analysis is performed between different sources of regulations. In order to achieve a better comparison between provisions, we employ a combination of feature matching and structural analysis. Results are shown on comparisons between American and European codes, as well as on the domain of electronic-rulemaking.