الفهرس | Only 14 pages are availabe for public view |
Abstract IoT technologies and applications are facing many challenges and issues which requires great efforts. One of these challenges is data quality and uncertainty, and since that IoT is considered to be a massive quantity of heterogeneous networked embedded devices that generate big data. So a huge quantity of data is being gathered by many organizations, communities and in a continuous raise. So, it is very complex to compute and analyze such a massive data. And as data volume increases the data noise, inconsistency and redundancy increases within data and causes paramount issues for IoT technologies. So this thesis introduces a new model named NRDD-DBSCAN which based on DBSCAN algorithm and using resilient distributed datasets (RDDs), NRDD-DBSCAN will be used to enhance and improve the data quality of IoT technologies by detecting the outliers that exists in IoT data. NRDD-DBSCAN is applied on three types of datasets, and these three types are 2-dimension, 3dimension and 22-dimension datasets, and the results were conducted, analyzed and compared with the results of previous models to prove the efficiency of the proposed model. |