Efficient Processing of Area Skyline Query in MapReduce Framework

In many application domains, it is badly needed to select a good location on a map. Selecting a location which is close to desirable facilities like nearby university or college areas or train/bus station but far away from the undesirable facilities like competitor cafe, noise sources and so on can be considered as a crucial problem in data mining research. A vast amount of information needed to be processed to answer such a query. The skyline query, which is one of the most powerful queries for selecting the interesting dataset, can be used to find some suitable space. Perhaps few researchers have already proposed a variant of the skyline query known as Area skyline query. However, the computation of area skyline query is a time-consuming process. Even if the dominance check using any efficient skyline query algorithms are not so suitable for selecting a good location when the area size is large. In this research, we have proposed an alternative way to select some good locations using the conventional skyline query in MapReduce architecture. MapReduce, proposed by Google Inc, is a programming model and an associated implementation for processing and generating big data sets with a parallel, distributed algorithm on a cluster. Hadoop, the open source implementation of Google's MapReduce framework, is our implementation framework. A series of experiments and comparisons with conventional area skyline query to check the efficiency of our proposed mechanism have been done.