Bilgi Yönetimi

Yazarlar: Ertuğrul AKTAN

Konular:Bilgi, Belge Yönetimi

DOI:10.33721/by.403010

Anahtar Kelimeler:Big data,Big data analytics,Hadoop,Map-Reduce,Big data security

Özet: Volume and variety of data have been increasing globally in an unprecedented rate throughout human history. People begin to create data even in their daily activities thanks to the advent of internet technologies and social media on every stage of our lives and especially our mobile phones. While operated manually in old days, many devices today are referred as intelligent, and almost all of them produce sensor data. Producing such an intense and different data from various sources reveals a new concept: “Big Data”. Big data is a collection of structured, semi-structured and unstructured data generated in high volume, velocity and variety. Numerous industrial areas are sources of big data by means of generating new data or digitizing the existing data. Information is an essential production factor of the organizations to gain competitive advantage. In this respect, the fact that big data is an input for information, and therefore affecting decision-making processes, increases the importance of big data analytics. Deriving economic value from big data requires processing validated data by advanced analytical methods. At present, big data and analytics are used in many areas from economic and commercial activities to public administration, from national security to scientific research. In today’s world where 2.5 exabytes (1 exabyte = 1,073,741,824 gigabytes) of data is produced in a day, it deems impossible to perform big data analytics with obsolete relational database and structured query languages. In order to process big data, advanced analytic methods operating on distributed file systems such as Hadoop and Map-Reduce are foremostly used in addition to Hive, Hcatalog, Hbase, MPP (Massively Parallel Processing), PIG, Mahout, NoSQL and Cassandra. Despite it’s benefits, big data analytics can be a reason of some security vulnerabilities jeopardizing the privacy of personal information. In this study; big data phenomenon along with its components and sources has been explained, advantages of big data in some application areas have been focused, big data analytics processes and the operation of Hadoop software architecture, which runs the Map-Reduce model on the distributed file system, have been examined. Furthermore, besides the advantages offered by big data, the security problems it creates have also been examined, and in this context some security measures to be taken have been evaluated.

---