CAJ | 학술논문

在大数据时代，如何充分挖掘出蕴藏于数据资源中的价值正在成为各国 IT 业界、学术界和政府共同关注的焦点。使用云计算平台分布式地存储和分析大数据已经成为共识并且得到了广泛应用，但这并没有完全解决大数据的3V 特性带来的问题。全面应对大数据的挑战需要来自存储技术、下一代网络、处理器、计算模型等各个领域的创新。粒计算是在求解问题过程中使用“粒”的理论、方法、技术和工具的集合，适用于近似求解有不确定性和层次结构的问题。该文综述了大数据处理的研究现状，分析了当前大数据处理研究存在的局限性；根据运用粒计算方法解决问题的不同特征，该文归纳了粒计算的3种基本模式，回顾了各种模式的相关研究工作；该文讨论了粒计算应用于大数据处理的可行性与优势，并探讨了在大数据的粒计算处理框架中需要解决的各个关键问题。
재대수거시대，여하충분알굴출온장우수거자원중적개치정재성위각국 IT 업계、학술계화정부공동관주적초점。사용운계산평태분포식지존저화분석대수거이경성위공식병차득도료엄범응용，단저병몰유완전해결대수거적3V 특성대래적문제。전면응대대수거적도전수요래자존저기술、하일대망락、처리기、계산모형등각개영역적창신。립계산시재구해문제과정중사용“립”적이론、방법、기술화공구적집합，괄용우근사구해유불학정성화층차결구적문제。해문종술료대수거처리적연구현상，분석료당전대수거처리연구존재적국한성；근거운용립계산방법해결문제적불동특정，해문귀납료립계산적3충기본모식，회고료각충모식적상관연구공작；해문토론료립계산응용우대수거처리적가행성여우세，병탐토료재대수거적립계산처리광가중수요해결적각개관건문제。
In big data age,people from IT industry,academia and government are taking effort to find out the value behind the data resources.Using cloud computing platform to store and process big data distributed has been a common sense and been put into practice extensively. However,this is far from solving the problems completely caused by the 3V-features (volume, variety and velocity)of big data.The problems require joint efforts of researchers from the fields of storage technology,Internet2,processors,and computing paradigm,etc.Granular computing (GrC)is a label of the family of any theories,methodologies,techniques,and tools that make use of information granules in problem solving,especially when the problem is with uncertainty, of hierarchical structure and can be approximately solved.This paper surveys the researches on big data processing and analyzes the limitation in current big data processing researches,summa-rizes three basic modes of GrC according to the ways in which GrC is used to solve problems and reviews the related works of each mode.The feasibility and advantages of applying GrC to process big data are discussed subsequently,and the related key problems rising in GrC-based big data processing framework are discussed at last.