Due to the complex shapes and various writing styles of Chinese characters, it is a challenge to write a Chinese character correctly, especially for children and foreigners who do not have much prior knowledge in Chinese handwriting. There are many potential errors in writing a Chinese character which include stroke production, sequence and relationship errors. Traditionally, a teacher is required to indicate those handwriting errors in the student’s handwriting during the class. The students can then correct their handwriting according to the teacher’s advice. After the students have practiced the handwriting many times, they will know how to write a Chinese character correctly. In order to reduce the teachers’ workload, intelligent computer- aided tutoring tools can be used to assist teachers in Chinese handwriting education. In this thesis, we focus on delivering an intelligent Chinese handwriting education system which can automatically check and notify the students about their input handwriting errors. With our proposed system, students can practice their Chinese handwriting whenever and wherever they want. The main research works and contributions are described as follows. In the baseline approach of our Chinese handwriting education system, several pattern matching techniques have been applied to identify students’ handwriting errors. The student is first shown a template character written by the teacher and then inputs his own handwriting (sample character) with a pen-based device. The attributed relational graph (ARG) is used to represent a Chinese character because it can characterize both the appearance and the structure of the template and sample characters. The nodes in the ARG are used to describe the strokes of the character and the edges in the ARG denote the spatial relations between any two strokes. Due to the noise and distortion of the handwritings, the error-tolerant graph matching (etgm) is applied to find the optimal matching between two ARGs. The optimal matching between two ARGs denotes the difference between the template and sample characters. It can be further analyzed to identify the possible stroke production, order and relationship errors of the students’ handwriting. The original definition of relationship is not sufficient to describe the various relations between the strokes in a Chinese character. We are thus motivated to refine the spatial relationship to increase its granularity by considering the relative distance so that it contains more discriminative information about the spatial relationships. K-means clustering is applied to derive our refined spatial relationship. The difference between two spatial relationships can be evaluated by the interval neighborhood graph. With the refined spatial relationship，the original interval neighborhood graph also needs to be extended. We generate a refined interval neighborhood graph by considering the deformations among our refined spatial relationships. In the original interval neighborhood graph, the edges that describe changes in spatial relationship are associated with uniform weights. This implies that the changes in all interval relationships are equally significant. To account for the fact that changes in various interval relationships may have different significance, we propose two ways to adapt the weights of the interval neighborhood graph according to the significance of change in spatial relationships observed in the training data. The first method is based on statistical analysis of relationship change and the second method relies on back-propagation neural network to generate the weight-adaptive interval neighborhood graph. With the proposed ARG matching, we can locate the stroke production and stroke order errors with high accuracy compared with the previous work. With the proposed refined spatial relationship, we can detect the stroke relationship error which has seldom been mentioned in the previous work. With the proposed weight-adaptive interval neighborhood graph, the results on finding the stroke production, order and relationship errors have been improved. The above contributions facilitate an intelligent handwriting education tool that can provide a useful feedback to both students and teachers. With the analysis result of the students’ handwriting, teachers can easily identify the handwriting problems of students and monitor their progress. On the other hand, with the automated error detection, students can find the problems in their handwriting and improve their handwriting skill. The outcome of this research helps delivering an effective teaching and learning platform for Chinese handwriting.