What you get is what you understand: Assessment in SACHS

Andrea Kohlhase, Michael Kohlhase

In: Christoph Lange , Jochen Reutelshöfer (Hrsg.). Wissens- und Erfahrungsmanagement: LWA (Lernen, Wissensentdeckung und Adaptivität) Conference Proceedings. GI-Workshop "Wissens- und Erfahrungsmanagement" (FGWM-09) befindet sich LWA (Lernen, Wissensentdeckung und Adaptivität) September 21-23 Darmstadt Germany Seiten 22-29 TU Darmstadt 2009.


Spreadsheets are heavily employed in administration, financial forecasting, education, and science because of their intuitive, flexible, and direct approach to computation. In previous work we have studied how an explicit representation of the background knowledge associated with the spreadsheet can be exploited to alleviate usability problems with spreadsheet-based applications. The SACHS system implements this approach to provide a semantic help system for DCS, an Excel-based financial controlling system. In this paper, we evaluate the coverage of the SACHS system with a "Wizard of Oz" experiment and see that while SACHS fares much better than DCS alone, it systematically misses important classes of explanations. We provide a first approach for an "assessment module" in SACHS, that assists the user in judging the situation modeled by the data in the spreadsheets and possibly remedying shortcomings.


SACHS_Assessment.pdf (pdf, 651 KB )

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence