Context-based Music Recommendation for Desktop Users

• Main Authors: Hsieh, Chi An, 謝棋安
• Other Authors: Shan, Man Kwan
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/05252365813217140455

碩士 === 國立政治大學 === 資訊科學學系 === 97 === With the development of digital music technology, knowledge workers will be delighted if the music recommendation system is able to automatically recommend music based on the operating context in the desktop. The context model and context identification algorithm are proposed to define the operating context of users and to detect the transition of context based on the changes of focused windows. Two association discovery mechanisms, MMAL (Multi-attribute Multi-label) algorithm and PM (Probability Measure), are proposed to discover the relationships between context features and music features. Based on the discovered rules, the proposed music recommendation mechanism recommends music to the user from the music database according to the operating context of users. The context-based recommendation system is implemented using Windows Hook API. Experimental results show that near 70% accuracy can be achieved.