The research project PH 3 deals with the fully automatic classification of patents into user-specific categories. The classification should allow typical intellectual property management processes such as competitor analysis, freedom-to-operate analysis, detection of market trends and landscaping analysis to be carried out significantly faster than before. These analyzes currently require the time-consuming and correspondingly expensive work of experts because, due to the complexity of the topics, they have to be carried out mostly manually. The aim of our project is the development of software based on big data technologies, which is able to combine semantic methods of text analysis and machine learning based classification algorithms so cleverly that IP analyzes can be automated as far as possible.
PH3 is a two-year research project as part of the “Central Innovation Program for Small and Medium-Sized Enterprises (ZIM)”, which Averbis GmbH is carrying out together with the Department for Knowledge Management in Bioinformatics at the Humboldt University in Berlin under the direction of Prof. Reader.