TIME-BASED SLOPE ONE ALGORITHM

摘要：

As time goes by, users rating scale or interests may change in real life.Due to that, temporal dynamics becomes a research hotspot in the field of recommender systems based on collaborative filtering.Among many collaborative filtering algorithms, slope one algorithm has been applied widely in practical life for its many merits such as simplicity, efficiency.Therefore, this paper presents time-based slope one algorithm, considering temporal information via instance weighting based on slope one algorithm, and experiments with MovieLens data.The results demonstrate that there is much improvement in accuracy compared to the traditional slope one algorithm, while still keeping the numerous advantages derived from the traditional slope one algorithm.

关键词： Collaborative filtering(CF) Slope one Temporal information

作者: SHAOJUAN LIN XIAOPENG TAO TAO YU

作者单位: School of Computer Science,Fudan University

会议类型: 国际会议

会议名称: 3rd International Conference on Mechanical and Electrical Technology(ICMET2011) (2011第三届机械与电气技术国际会议)

会议地点: 大连

会议语种:英文

页码: 165-172

在线出版日期: 2011-08-26（万方平台首次上网日期，不代表论文的发表时间）