Research Information System
Thesis Type: Postgraduate
Institution Of The Thesis: Gazi Üniversitesi, Fen Bilimleri Enstitüsü, Turkey
Approval Date: 2018
Student: DUYGU NAZİFE ZARALI
Supervisor: HACER KARACAN
Abstract:Intelligent systems refer broadly to computer embedded or controlled systems, machines and devices that possess a certain degree of intelligence with the capacity to gather and analyze data and communicate with other systems. Developments of intelligent technologies that allow farmers to manage farms with ease has introduced the smart farming concept. These technologies take over the labor of dairy farming and reduce the need for human-animal interactions. As implementation of these technologies on farms, farm data grow in quantity and scope and farming processes become increasingly data-driven and data-enabled. Therefore, big data analysis is being used to provide predictive insights in farming operations, drive real-time operational decisions. The analysis of big data requires algorithms to handle more varied and complex structures with difficulties in storing, analyzing and visualizing. Different storage and processing methods are implemented to meet the requirement of processing large-volume, growing datasets. Apache Spark is an effective distributed data processing engine that makes big data processing easier and faster. Spark has advanced in-memory programming model and upper-level libraries for scalable machine learning, graph analysis, streaming and structured data processing. In this thesis, large scale data obtained from intelligent farms, are analyzed and PrefixSpan, a sequential pattern mining algorithm, is implemented using Apache Spark machine learning library. Based on the analysis of the past data, the main sources of the problems could be predicted and the possible problems that may arise can be eliminated. With this analysis, it can be possible to minimize significant costs by early detection of failures that may occur in systems and management of maintenance processes accordingly.