CONCEPTION OF A KNOWLEDGE MANAGEMENT SYSTEM FOR TECHNOLOGIES

摘要

For strategic product and technology planning of intelligent technical systems a holistic process and tool is needed to support a company in the innovation process. An important aspect within the innovation process, is technology planning. The Heinz Nixdorf Institute developed a technology-planning concept, which is supported by an Innovation-Database. The Innovation-Database synchronizes aspects from the market pull and the technology push. Furthermore, it is used as a knowledge management system for innovations and technologies. The challenge is filling the technology pool with a minimum set of technologies in order to use the whole functionality of the Innovation-Database. Many resources are needed in order to fill the technology pool or to update it. There is in increase in information in the World Wide Web and most of the information is fuzzy. Our approach is to identify them automatically by using crawler technologies. Therefore a software tool was developed, which automatically crawls for technologies in the database of the European Patent Office and in the journal database IEEE Xplore. In these data sources, the tool identifies technologies automatically in natural texts, prepares the data and visualize them. It is not suitable to scan a text for a defined set of technologies in a dictionary, because then the tool is not able to find new technologies or to detect weak signals. In this paper, the architecture and features of the software tool is described. It uses text mining-methods to detect technologies in unstructured, fuzzy data. As input data, the user has to provide a search field, which the search is performed for. Furthermore, the data sources have to be defined. The tool crawls these data sources in order to identify technologies, which can be relevant for a given search field. The tool suggests newly identified technologies for the search field, which are generally relevant. The user has to decide whether those technologies are important for him. This information is used as feedback for the text-mining model to improve the precision of future analyses. In order to support the user, a suggested technology is automatically provided with further information, like a description and an estimation of technological maturity. With this information a user can start an evaluation of the technology. For every technology a datasheet is created with detail information. It can easily be customized for specific demands. All technologies can be visualized on a radar. The radar shows all technologies for a search field according to its relevance. These information is used for strategic product and technology planning.