Application of Extracting Personal Recommendation Knowledge Based on Ant Colony Optimization

• Main Authors: Tzu-Ying Wu, 吳姿瑩
• Other Authors: Feng-Hsu Wang
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/52302906943498553259

碩士 === 銘傳大學 === 資訊工程學系碩士班 === 102 === Ubiquitous e-learning removes the restriction of time and place in learning, as long as the learner can have an easy access to the internet to achieve the learning objectives. The most important thing is personalization for each user based on the behavior characteristics of learners, individual needs and delivery of tailored materials. In the present thesis, an off-line extraction system for personalized recommendation knowledge is developed. Using the ant colony algorithm, knowledge extraction of recommended learning path can be conducted according to different learning styles and knowledge level of learners. This study validates the proposed method on a practical course to explore the effects of different learner abilities and learning styles on its effectiveness. Finally, the knowledge extraction model is evaluated based on correlation between learner final grade and the similarities of knowledge structure and the past path of student using the Pearson correlation statistics. It is found that they are positively correlated. It is expected the method will extract good learning experiences for reference by future learners to enhance their learning performance.