An Effective Grouping Method for Unstructured Data Based on Swift
Miaomiao Dai, Dongjie Zhu · 2018
Unstructured data is one of the most prominent buzzwords of this era, and from the business to the personal computer, unstructured data is ubiquitous and growing exponentially, managing these data and improving the access performance of these unstructured data is Critical. This paper is based on Swift which is the object storage service of open source cloud computing platform OpenStack, with the help of object storage framework, the use of grouping-based machine learning technology and the corresponding prefetching cache strategy to improve the access performance of unstructured data. Experimentally verify the performance improvement of the proposed method with respect to memory consumption, cache hit ratio, and latency of requests. Experimental results show that the proposed method can effectively reduce the cache consumption and request delay time and can greatly improve the cache hit rate.