An efficient unstructured big data analysis method for enhancing performance using machine learning algorithm
An efficient unstructured big data analysis method for enhancing performance using machine learning algorithm
复制标题
DOI：
10.1109/iccpct.2015.7159492
复制
发表时间：
2015-03
期刊：
影响因子：
--
通讯作者：
A. K. Reshmy;D. Paulraj
中科院分区：
文献类型：
--
作者：
A. K. Reshmy;D. Paulraj
在当今世界，数据挖掘技术在所有主要的工程领域都占有重要的地位。处理非结构化大数据是这个时代的重要任务。目前，使并行处理技术的最大优势和稳定和连续地从各种来源发送或接收的大量数据的快速检查任务变得流行或常规。大数据分析作业被分割成较小的作业，并通过并行处理架构在数十、数百或数千个产品服务器上运行。这有助于保持数据中心的成本效益，并有助于以有效的方式轻松处理大量工作。在本文中，提出的解决方案采取在线消费者购买。该在线系统拥有无与伦比的在线消费者购买行为数据库，可以从其1亿客户账户中挖掘。他们使用所有这1亿客户的客户点击流数据和历史购买数据，并在定制网页上向每个用户显示个性化结果。为了提高大数据的性能，使用机器学习方法，即K-最近邻算法来支持进行良好的分析。Hadoop模拟器就是用来解决这类任务的。
In this modern world, data mining technology holds an essential position in all the major Engineering fields. Handling of Unstructured Big Data is an essential task of this era. At present, making the maximum advantage of parallel processing know-hows and the task of rapid examination of huge data steadily and continuously transmitted or received from various sources is becoming popular or conventional. The big data analytics job is fragmented into smaller jobs and ran over tens, hundreds or thousands of product servers by the parallel processing architecture. This helps in maintaining the data center cost efficient and facilitates easy handling of the enormous work in an efficient way. In this paper, proposed solution takes online consumer purchase. The online system has unrivalled bank of data on online consumer purchasing behavior that can be mined from its 100 million customers accounts. They use customer click-stream data and historical purchase data of all those 100 million customers and each user is shown personalized results on customized web pages. For improving Big Data performance the Machine Learning Method i.e. K-Nearest Neighbour algorithm used to support to take good analysis. Hadoop simulator is used to solve this kind of task.