Browsing by Author "Nurkey U."

Now showing 1 - 1 of 1
  • Thumbnail Image
    ItemOpen Access
    Comparison of Big Data analytic tools
    (2019) Nurkey U.
    A huge repository of petabytes of data is generated each day from modern information systems and digital technologies such as scientific data analysis, social media data mining, recommendation systems, analysis on web service logs and Internet of Things along with cloud computing. The data has a huge power to directly guide us to knowledge detection. Examination of these enormous information requires a great deal of undertakings at different dimensions to extricate knowledge for making decisions by Big Data analyzing tools, and picking the correct tool requires an inside and out learning about the abilities of each. Along with these, big data examination and exploring various tools for it is a current area of research and development. This thesis is an effort to present the basic understanding of BIG DATA is and its usefulness to an organization from the performance perspective alongside with integrating it to the curriculum of higher education. The basic objective is to explore platforms for analyzing big data and compare them in different perspectives. Additionally, it will open a new horizon for researchers to develop the solution, based on potential impact of big data challenges. This creates the need to incorporate the investigation of Big Data an- alyzing systems as a major aspect of the computing curriculum. First experiment consists from examples of analytic problems that can be solved as introduction into big data projects on Apache’s tools by demonstrating how each type of system can be integrated into education via sample datasets and data analysis tasks, also analyzing their results from educational perspectives. Second experiment is conducted to compare Hadoop, Spark and Pig , as a major and modern tools in big data analytic - on iteration of supporting task, computing time for each task on each tool and Input/Output data access with real life data. Mentioned tools were chosen due to their popularity for analyzing big data. Results of this research show that various tasks require different tools and there is no all-in-one solution. Any big data problems stand in need developers to use proper tool to make job done in a way better and quicker.