TwigPipe: A pipelining holistic solution of Twig Query

With XML plays an increasingly important role in data representation and exchange, Twig Query has attracted significant attentions in recent years. Various approaches have been proposed to improve the performance of Twig Query. However, those existing algorithms are either designed for single-core processor architecture, or suffering from expensive cost of pre-or post-processing. In this paper, we propose a premiere pipelining holistic solution of Twig Query, TwigPipe, which introduces a novel data-partitioning strategy and a querying algorithm. Experimental results show that TwigPipe successfully reduces memory requirement and achieves performance improvement.