IMPLEMENTASI BIG DATA MENGGUNAKAN MODEL PEMROGRAMAN MAPREDUCE DAN KMEANS CLUSTERING

Big data refers to very large and complex data sets, which are difficult to process using traditional methods of data processing. This data can come from various sources, such as social media, IoT (Internet of Things) devices, online transactions, and more. This very rapid development of data makes organizations look for methods to store and process data. Big data technology is a solution for storing data and also being able to process that data. The use of MapReduce is very important in big data processing, especially when the data that must be processed is very large and spread across many machines or servers. MapReduce is a programming model used to process and produce output from large amounts of data in a parallel and distributed manner. MapReduce allows dividing large tasks into small parts that can be processed on many machines or nodes. Hadoop is a framework that can store data on a large scale without paying attention to the structure of the data. To handle large amounts of data or big data, various technologies and techniques are needed that are more sophisticated than traditional data processing. The main aim of this research is to harness the enormous potential of existing data to create deeper insights, improve operational efficiency, and support smarter decision making. Another goal is also to propose and implement a Hadoop MapReduce based system. By utilizing the power of Hadoop MapReduce, it is hoped that the implementation of big data will have various significant benefits in many sectors so that it can ultimately encourage better performance and success in various aspects of business or operations. The results and conclusions from research on implementing K-Means in a big data environment using the MapReduce programming model are that the implementation of K-Means on liver disorder data using the Mahout library can run correctly. K-Means computation using the Mahout library produces the same output of centroid data items as is also proven by manual calculations.