Επεξεργασία Skyline επερωτήσεων σε πολυδιάστατα δεδομένα στο υπολογιστικό νέφος
Μπισιώτη, Αρετή Δ.
SubjectApache Hadoop ; MapReduce ; Electronic data processing -- Distributed processing ; Cloud computing
This thesis refers to and aims at the “Processing of Skyline Queries over Multidimensional Data in Cloud Computing”. In the first part of the thesis, Cloud Computing is introduced along with the necessity of Skyline Query in terms of the processing of Multidimensional Data. Next, we deal with the presentation of Hadoop-Map Reduce framework - on which the current project is based- as well as with the description of the phases of the parallel processing of Multidimensional Data by Skyline Queries. In the next chapter, the focus is on the goal setting of the thesis and on the designing of the code with regard to architecture and classes. In particular, the three methods of space partitioning that is Grid Partitioning, Angle-based Partitioning, Partitioning via Hyper plane Projection, are presented as well as the way of Skyline Computation. In the following chapter, a description of the environment, where the experimental analysis was carried out, takes place. This analysis consists of three scenarios, which are different when it comes to the number of the data to be processed and their distribution and the number of partitions to be created. There is also, a graphic representation of the experiment results. Finally, conclusions of the study are made as well as suggestions for future research.