Policy Aware Geospatial Data

Abstract

Digital Rights Management (DRM) prevents end-users from using content in a manner inconsistent with its creator's wishes. The license describing these use-conditions typically accompanies the content as its metadata. A resulting problem is that the license and the content can get separated and lose track of each other. The best metadata have two distinct qualities--they are created automatically without user intervention, and they are embedded within the data that they describe. If licenses are also created and transported this way, data will always have licenses, and the licenses will be readily examinable. When two or more datasets are combined, a new dataset, and with it a new license, are created. This new license is a function of the licenses of the component datasets and any additional conditions that the person combining the datasets might want to impose. Following the notion of a data-purpose algebra, we model this phenomenon by interpreting the transfer and conjunction of data as inducing an algebraic operation on the corresponding licenses. When a dataset passes from one source to the next its license is transformed in a deterministic way, and similarly when datasets are combined the associated licenses are combined in a non-trivial algebraic manner. Modern, computer-savvy, licensing regimes such as Creative Commons allow writing the license in a special kind of language called Creative Commons Rights Expression Language (ccREL). ccREL allows creating and embedding the license using RDFa utilizing XHTML. This is preferred over DRM which includes the rights in a binary file completely opaque to nearly all users. The colocation of metadata with human-visible XHTML makes the license more transparent. In this paper we describe a methodology for creating and embedding licenses in geographic data utilizing ccREL, and programmatically examining embedded licenses in component data...