CAJ | 학술논문

随着电子商务的深入发展，基于虚拟资产用户操作数据流立方体的构建技术成为了当前研究热点。文章针对虚拟资产用户操作数据流的特点，首先对数据流进行分组操作，并从中抽取属性集和操作序列集，然后构建流立方体模型。在流立方体构建中采用多路聚集算法，提高聚集性能。在真实数据集上进行实验测试表明，该模型能有效提高异常操作追溯性能。
수착전자상무적심입발전，기우허의자산용호조작수거류립방체적구건기술성위료당전연구열점。문장침대허의자산용호조작수거류적특점，수선대수거류진행분조조작，병종중추취속성집화조작서렬집，연후구건류립방체모형。재류립방체구건중채용다로취집산법，제고취집성능。재진실수거집상진행실험측시표명，해모형능유효제고이상조작추소성능。
With rapid development of E-commerce, The establishment technology of data-stream cube on virtual assets users,operation becoming the current research focus. According to feature of data-stream of virtual assets users,operation, I divide data-stream into groups firstly, and extract at ribute sets and operational sequence sets from data-stream, and build stream-cube model. Multiway gather algorithm is adopt to building stream-cube, which improves performance of gather. Experiment on real data sets proves that the model can improve performance of tracing abnormal operation.