Automatic Categorization of Japanese Patents based on Surrogate Texts
Yuen‐Hsien Tseng, Da-Wei Juang, Chi-Jen Lin · NTCIR · 2005
This paper describes our work at the fifth NTCIR workshop on the subtask of patent classification. We use KNN (K-Nearest Neighbors) as our classifier and the English PAJ (Patent Abstract Japan) as the patent surrogate for classification. Based on the knowledge and experience learned from our previous experiments with other document collections, we leverage on the parameters to achieve above-average performance in an efficient way.