Semantic Interpreter Pythia : GIS-based Expert Knowledge Algorithm for Automated Geotechnical Soil Profiling of Applicable Data

Abstract

Although the fields of geospatial data are growing rapidly, the result is still not satisfactory for the needs of engineers. No systematic information is available about geotechnical subsurface soil conditions and underground artificial infrastructures. This old-age problem is two-fold: (a) inadequate available digital geotechnical data, and (b) no concepts to improving the applicability and to updating data for engineering applications. On the second, the paper proposes the innovative GIS-based model-driven data processing methodology implemented into an expert knowledge algorithm named Semantic Interpreter Pythia (thereafter SI). From the point of view of geotechnical engineering, the subject of SI is the automated multi-thematic geotechnical soil profiling (GSP) by which it determines the geometry, the properties and the stratigraphy of the site-specific subsoil. From the point of view of geographic information science, the subject of this expert is to relate multi-thematic sets of data from databases, to interpret these data with a specialized data fusion model and, ultimately, to lead to unified information in a core relational database. The paper presents the innovative idea of this algorithm to propose the development of automated SI tools by the modern GIS and internet technology. These tools could help disseminate useful and up-to-date data for a wide range of uses. Based on the experiences distilled from an extensive geotechnical case study, the paper specifies what content is appropriate for engineering studies. The notions of data applicability and geotechnical semantic interpretation arise.