Survey on incremental and iterative models in big data mining environment

摘要

It has become increasingly popular to mine big data in order to gain insights to help business decisions or to provide more desirable personalized, higher quality services. They usually include data sets with sizes beyond the ability of commonly used software tools to retrieve, manage, and process data within an adequate elapsed time. So there is big demand for distributed computing framework. As new data and updates are constantly arriving, the results of data mining applications become incomplete over time. In such situations it is desirable to periodically refresh the mined data in order to keep it up-to-date. This paper describes the existing approaches to big data mining which uses these frameworks in an incremental approach that saves and reuses the previous states of computations. It also explores several enhancements introduced in this same framework with iterative mapping characteristics. Gaps in the current methods are identified in this literature review.