A Data Mining Technique for Real Time Process Monitoring with Mass Spectrometry : APC: Advanced Process Control

Abstract

As the semiconductor process becomes more complicated, process monitoring that reflects real time process conditions is important. The mass spectrometer is an effective tool to represent the process by monitoring process chemical reaction in real time. In order to apply the mass spectrometer data as the process-related data, it is necessary to use the data mining technique to process the large amount of collected data. In this study, we find out the correlation between the mass spectrometer data collected in real time and the process data describing the device performance with the data mining technique. We developed an automatic data analysis model to reduce the repetitive work of the analysts and improve the analysis efficiency about a large amount of the mass spectrometer data. We will contribute making a fault detection & classification system for fine control process by using advanced data analysis technology.