Analyzing Trademark Similarity Decisions using Machine Learning

Abstract

Registered trademarks enjoy legal protection and can therefore take legal measures against (newer) trademarks that might be confused with the original trademark. Hence, the question whether two trademarks are similar is crucial for the trademark office’s decision. However, it is not yet entirely clear which criteria/features of two trademarks have an impact on the trademark office’s decision.

In collaboration with Prof. Dr. Tilmann Altwicker (Faculty of Law) and Prof. Dr. Florent Thouvenin (Faculty of Law), we aim at developing a machine learning algorithm that predicts decisions of trademark offices regarding the similarity of two trademarks. In a second step, we will analyse the algorithms behaviour in order to assess which features are informative with respect to the trademark office’s decision.

Principal investigators

Master thesis

We are currently looking for a Master thesis student to work on this project.