The computation of skyline points has become a particularly interesting topic in recent years because of its application in multi-criteria decision-making systems. Though many e cient algorithms have been designed, several important issues related to this problem still persist. In particular, most algorithms are ine cient when applied to high-dimensional datasets. The increase in the size of the skyline is a consequence of the increase in the number of dimensions. Given that most operational databases are very large, existing skyline algorithms perform poorly and, with a view to avoid this in- e ciency, other algorithms have sacri ced the accuracy of results by removing from the dataset some dimensions quali ed as less signi cant. Our research intends to propose a method to compute skylines for high-dimensional databases with large cardinality that delivers accurate results and fast processing of the data. Following our objective, we have studied the application of parallel programming techniques and the bene ts provided by GPGPU development frameworks. We have found that simplicity is better when implementing parallelism and our proposed algorithm scans the data avoiding the creation of data structures and making extensive use of the functionalities provided by the GPGPU framework to reduce computing time. Furthermore, our implementation bene ts from hardware advantages delivered by GPU devices as the texture memory space. Nevertheless, the GPGPU framework guarantees portability to our implementation. Besides cardinality and dimensionality, the correlation coe cient becomes a decisive factor at the characterization of data. Therefore, we have designed a group of tests using datasets with di erent levels of correlation in order to evaluate our algorithm's performance.
|Publication status||Unpublished - 2014|