This document discusses using data mining techniques to analyze process data from a re-rolling mill that manufactures TMT steel bars. It aims to discover relationships between process parameters and the grade of steel bar produced to enable quality prediction and process optimization. Specifically, it applies Naive Bayes, association rule mining, neural network, and logistic regression algorithms to process data involving variables like material composition and temperatures at different stages. The results identify key parameters like billet and bar temperatures that influence the bar grade. The extracted knowledge could be used to set process parameters to achieve the desired quality grade.