Artificial Tutor for Handwriting Teaching

Many students struggle to produce neat, expressive written work which may be due to physical or cognitive difficulties. The traditional solution to this problem is to have a dedicated teacher for long hours of handwriting practice. Unfortunately, this is not feasible in many cases. As an example, in Egypt the number of the admitted male and female students in the all pre college stages hit 15511818 (8042701 male, 7469117 female) in 2004/05. The goal of this project is to develop an automated tool for teaching Arabic handwriting for children aged from seven to ten using tablet PCs and on-line handwriting recognition.

This tool should simulate the tasks performed by the human teacher:

• Detect the segments of hypothesized writing problems through targeting specific and narrow objectives
• Produce instructive real time feedback to help the student to improve his handwriting quality
• Praise efforts that are well-done

In order to build this handwriting tutoring tool, we have to build the necessary understanding of the domain of human writing behavior and how this behavior evolves with age from youth to adulthood. With this understanding, we will be building computational models capable of analyzing, spotting human writing errors and providing feedback to the young learners. Through this project we will help students to become good writers with fluent movements and good quality of writing in shorter time frame compared to traditional human teacher-based methods, also help the students learn to; analyze their own writing and understand how they can improve it.

Key Investigators
Prof Dr. Aly Fahmy, Dean, Faculty of Engineering
Dr. Sherif Mahdy Abdou, Assistant Professor, Information Technology Department

Back