Teaching game programming and basic programming to a person having cerebral palsy and developmental delay.Learnt alphabetical sign language and took personal care of the student.

•Examined the impact of different feature extraction and segmentation window in block-based learning environments and estimates of goodness of fit metrics sensitive to misclassification was obtained using recall, F1 score and lift metrics.•Conducted resampling experiment that compared semantic and character encoding to extract features that characterized student produced programs.•Varied the size and step of segmentation window to segment the series of student edits to the program.•Key achievement: Increased the F1 and recall score by 18% and 13% and generated better fitting models for semantic encoding and longer edit sequences.