We are pleased to announce that the paper “Technology Forecasting through Data-Driven Analysis of Patents” by Arman Arzani, Theodor Josef Vogl, Marcus Handte, and Pedro José Marrón has been accepted for publication at CIKM 2026.
The paper presents a domain-agnostic framework for forecasting technology lifecycle phases using patent data from the European Patent Office (EPO). By identifying a universal eight-phase technology lifecycle pattern and combining neural time series forecasting with lifecycle phase classification, the approach enables accurate prediction of technology evolution and upcoming lifecycle transitions without relying on domain-specific features or patent text.
Key results show that the proposed framework achieves strong forecasting and classification performance, with the best pipeline predicting lifecycle phase transitions with a median error of only 0.65 years and a mean error of 1.8 years.
Congratulations to all authors on this achievement!