Use of automatic speech recognition for invoice related activity recording processes in Swiss healthcare

Even the smallest enterprise has to manage so much information and documents, that a system for arranging these things is needed; even if it is a small binder. Now when we think about the amount of information which today exists in a company, we have surely to say, that information and knowledge management is not done by only one binder – the companies nowadays need something more sophisticated. What companies nowadays need is information about information – metadata. If metadata is available, then the finding and filing process can be dramatically improved. But if the metadata is not available, it needs to be created – and this has to be done in most of the cases by hand. Would it not be great to have an automatic approach? This thesis introduces an approach for creating metadata in an automatic way based on rules and a formal description of an enterprise. We often hear the statement that a company has the information available – ‚We have the information in our systems.‛ But it is the question how the information is available. The Linked Enterprise Models and Objects (LEMO) approach gives the possibility to formalise the information in an enterprise. And not only the infor-mation, LEMO tries to make the relationships / links between different enterprise objects, documents, people, customers, money, almost everything in an enterprise explicit and machine process able using an ontology called enterprise model ontology (EMO). This EMO can be seen as context description of an entire enterprise. And this context can be used to create metadata using rules....