Future-oriented technology analysis of technology roadmap based on text mining

Future-oriented technology analysis (FTA) is a term derived from a collective description given to the range of technology-oriented forecasting methods and practices by a group of futures researchers and practitioners^[1]. Future-oriented Technology analysis are concerned with complex dynamic systems and processes and engage multiple stakeholders in participatory and interdisciplinary processes to assure distributed understanding and sustainable development. Thus, there are variety of methods, and variety of possible classification criteria and the combined use of this extensive variety of methods in this research processes, and in which there are some important foresight methods such as `Scenario building´, `Delphi´ and `Roadmapping´. As one of important FTA methods, the practice of technology roadmapping (TRM) has received much attention from researchers and practitioners, to support planning and forecasting in companies and sectors. However, little research has focused on the intelligence analysis of the text content of TRM but a lot of research has focused on the optimization of TRM methods. The paper proposes a triple co-occurrence algorithm to build the future-oriented technology analysis thesaurus of technology roadmap based on text mining combining the method of scientometrics and natural language processing, which reveals the future-oriented technology development direction and level characteristics of special technical field and achieves preliminarily the target of future-oriented technology analysis of technology roadmap. The experiment shows that this method can support the FTA of technology roadmap to some extent.