Idea Outline
It is common sense that teachers teach differently and students learn differently. The same teacher or student may vary a lot in their teaching or learning ways within given education contexts. Teaching materials should meet personal requirements.

The Challenge
The development may face challenges from 5 aspects: a. Course knowledge must be organised in ways comprehensible to the mechanisms of automatic content selection. b. Users for a course, i.e. teacher and students, must be characterised thoroughly and the importance of each characteristic must be balanced properly for generating different teaching materials. c. How the course model and user model could be leveraged to make feasible content selection and suitable content organisation for different teaching materials. d. Efficient realisation of the selected contents into language and hypermedia requires both theoretical and empirical knowledge. e. How teaching and learning experiences with the tailored teaching materials could be used to improve the tailoring performance.

The Solution

Objectives and solutions include:a. A knowledge engineering platform is designed to help the teachers collect the course knowledge from existing teaching materials and organize the knowledge in a hierarchy according to the granularity and hyponymy of the course knowledge.b. A survey platform is designed to capture user attributes and organize them with regard to their frequencies and importances.c. Sets of rules are built by analysing and aligning course and user models to guide automatic content selection and discourse for different users.d. Rules for language and hypermedia tailoring are collected from the course content features, user features, and the education context.e. Student attendance and examination results can be used to evaluate and improve the tailoring performance.

Target group and social impact
Teachers and students will benefit directly from the system. Although by using this software, modeling a course will take more time and efforts than otherwise, efficient and simple maintenance, extension, and much more time and efforts saved latterly may well pay off. The course model will contribute to teachers' better understanding of the knowledge structure, and the student models and tailored materials will help in organising teaching and learning activities. With personal requirements being considered, students will learn with more interest and enthusiasm, and in turn make greater achievements. Schools, colleges, and universities that successfully facilitate this software will produce more satisfactory education, and finally the society will benefit from the better education results.




Competences of the applicant
The applicant is a computing science post-doctoral research fellow with expertise in natural language understanding and natural language generation. His familiarity with statistical, machine learning, and data mining methods, skills in using object-oriented programming languages, and experiences of being PI and Co-I in quite a few projects, make him suitable for the implementation of this idea.