CAJ | 학술논문

协同过滤算法是目前最成功的个性化推荐技术之一，但传统的协同过滤算法没有考虑用户兴趣随时间而产生的变化，影响了推荐质量。本文针对这个问题，提出一种改进的指数遗忘函数对用户-资源评分矩阵进行修正，并将修正的评分矩阵用于协同过滤算法，从而得到一种改进的协同过滤算法。实验表明，与传统协同过滤算法相比，改进的算法在推荐准确度上有显著提高。
협동과려산법시목전최성공적개성화추천기술지일，단전통적협동과려산법몰유고필용호흥취수시간이산생적변화，영향료추천질량。본문침대저개문제，제출일충개진적지수유망함수대용호-자원평분구진진행수정，병장수정적평분구진용우협동과려산법，종이득도일충개진적협동과려산법。실험표명，여전통협동과려산법상비，개진적산법재추천준학도상유현저제고。
Collaborative filtering algorithm is one of the most successful technologies applied in recommendation system, but traditional collaborative filtering algorithm does not take into consideration of user's interest drifting over time, and this makes the quality of recommendation system decreased. In order to solve this problem, an improved exponential forgetting function which is used to correct the user-resource scoring matrix is proposed and the revised scoring matrix is applied to collaborative filtering algorithm. The experimental results show that the new algorithm is more accurate than traditional collaborative filtering algorithm.