Method of learning a knowledge-based database used in automatic defect classification

A computer-based method for teaching a knowledge-based database for automatic defect classification, which comprises:

• (a) accepting, using a specially programmed computer, selecting a user selection of a review data file;
• (b) accepting, using the specially programmed computer, inputting parameters and data by a user on one page of a learning mode whereby the parameters and the data are known to the user;
• (c) starting, using the specially programmed computer, an alignment procedure and a procedure for adjusting light intensity, the alignment procedure with respect to at least one point on a wafer;
• (d) automatically adjusting, using the specially programmed computer, the optimal intensity of the light intensity accepting a selection of a first specific number of defects to approach on a first wafer and taking respective pictures of the first specific number of defects on the first wafer and if necessary regulating to the optimal illumination using the respective pictures;
• (e) checking, using the specially programmed computer, a detection, whereby an optimization of the detection parameters is carried out by accepting a selection of a second specific number of defects to approach on a second wafer, taking pictures of the second specific number of defects on the second wafer, displaying the pictures, and using the pictures to adjust a detection threshold using pictures;
• (f) automatically approaching all defects of a wafer or wafers, whereby the respective defect is detected and a descriptor is assigned, by the specially programmed computer, to the respective defect;
• and,
• (g) analyzing and automatically grouping, using the specially programmed computer, the descriptors of the defect.