A Quantitative Evaluation of Trademark Search Engines' Performances through Large-Scale Statistical Analysis

• DOI: 10.1145/3594536.3595149
• Publication Date: 2023-09-07T23:40:01Z
• Authors: Thomas Vandamme; Julien Cabay; Olivier Debeir
• Keywords: Openness to experience; Profiling (computer programming)

Intellectual Property Offices now offer their users trademark search engines to help them identify earlier trademarks in register. Such tools have proven be extremely useful given the growing number of registered but never been subjected thorough evaluation, despite necessity for openness and accountability justice systems. Additionally, performance is unknown, particular reliability results pertaining applicable legal rules. In fact, "black box nature" makes automatic at-scale evaluation hard perform directly, which why we propose a novel method evaluating using settled case-law ground truth, analysis. Based on this methodology, evidence two such systems, Benelux Office (BOIP) European Union (EUIPO), 8 126 opposition division decisions from EUIPO. We show important disparities between along with surprisingly good EUIPO's system.