Scalable Visual Patent Analysis


Professor Dr. Thomas Ertl, Stuttgart
Universität Stuttgart
Institut für Visualisierung und Interaktive Systeme (VIS)

Professor Dr. Hinrich Schütze
Universität Stuttgart
Institut für Maschinelle Sprachverarbeitung


Scalable Visual Patent Analysis


Patent analysis has become an important economic task in recent years. Due to the multidimensional nature of patent data, their heterogeneity, and often poor quality, exhaustive analysis of this data is time-consuming and error-prone. Exhaustiveness in terms of good data coverage (recall), however, is a key-factor for high-quality patent analysis, because missing important patent information can result in severe economic consequences. The main idea of this proposal is therefore to research and develop new approaches for the integration of usertailored interactive visual exploration methods and advanced text analytics methods in the field of intellectual property analysis. The power of interactive visualization can be enhanced greatly if the underlying text data are analyzed, filtered and classified using advanced text analytics methods. Classical information retrieval is not sufficient with respect to recall in the patent domain. By integrating the comprehensive experiences of the applicants within the fields of visualization, human computer interaction, natural language processing, and information retrieval, this project aims to develop new methods for patent analysis that greatly improve the state of the art in terms of effectiveness and reliability, thus establishing a new form of scalable visual patent analysis.