Patent Terminology Classification Using Relaxation Labeling

Abstract

In this paper, an approach to unknown term classification using relaxation labeling is proposed. The technique is used to classify the terminologies in patent documents. The terminologies on a CMP (chemical mechanical polishing) domain about the relations among components, the attributes and materials of components are be extracted by NLP technique and regular expressions. In the progress of relaxation labeling, the probability of each class for each unclassified term was updated. Based on the extracted semantic information, the experiment results show that relaxation labeling is sufficient for terminology classification and has achieved 72% of accuracy.