The datasets used in our preprent titled "Towards Effective Personalized Service QoS Prediction from the Perspective of Multi-Task Learning".

3Attributes_Context.tsv is the original version, which consists of 11400 records. Each record includes the fields of UserID,ServiceID,ResponseTime,Throughput, Reliability,UserRegion,UserAS,UserSubnet,UserIP,UserGP,ServiceRegion ServiceAS,ServiceSubnet,ServiceIP,ServiceGP,..., small-3t.csv is processed.

small-3t.csv is the refined and enhanced version really used in our paper.