Ontology-based decision support system for semiconductors EDS testing by wafer defect classification

Abstract

Wafer failures have be analyzed and managed to improve the performance of semiconductor EDS testing. Thereby, in this paper, we propose an ontology-based decision support system to predict which defects may happen in a given condition by classifying previous wafer failure patterns. Especially, ontologies are exploited to represent and maintain (i.e., classify) those wafer failures, so that heterogeneity problem among multiple end users (e.g., in the clean room) can be dealt with. More importantly, this ontology-based scheme can build a system entity structure (SES) that contains knowledge of decomposition, taxonomy, and coupling relationships of a system necessary to direct model synthesis. Also, given a certain wafer failure, the SES can be reduced into a substructure, called PES, by removing irrelevant entities. Consequently, this system can support the end users to efficiently evaluate and monitor semiconductor data by (i) analyzing failures to find out the corresponding causes and (ii) managing historical data resulting into such failures. Therefore, this study contributes to the increase in product quality and wafer yield.