Why does the first column mean in the evaluation confusion matrix?

For example, for this 2-class matrix, the first column ‘9’ for chitchat, does that mean there are 9 instances of chitchat mistakenly classified as unknown category?

It looks like your labels or data(for the confusion matrix) was not setup correctly. Every column should have a label. somehow it looks like you have 3 categories in labels but 3 categories in your confusion matrix.