Citation Context

...airwise Distance Calculation Calculating similarity or dissimilarity between each element of a data set with each element in another data set is a common problem and is generally known as an All-pairs=-=[22]-=- problem. The application we have selected calculates the Smith Waterman Gotoh(SW-G)[23] distance (say ...

Citation Context

...the authors compare the performance with Hadoop for tera-sort application. Sphere stores intermediate data on files, and hence is susceptible to higher overheads for iterative applications. All-Paris =-=[26]-=- is an abstraction that can be used to solve a common problem of comparing all the elements in a data set with all the elements in another data set by applying a given function. This problem can be im...

Citation Context

...milar documents, set-similarity joins [6] for efficient string similarity computation in databases, pairwise distance computation [16] for clustering complex objects, and all-pairs matrix computation =-=[17]-=- for scientific computing. All approaches follow a similar idea like ER using blocking: One or more signatures (e.g., tokens or terms) are generated per object (e.g., document) to avoid the computatio...

Citation Context

...cution in distributed environments, e.g., MapReduce [29] and MasterSlave [28]. We are extending this generic higher-order actor set to support more data-parallel execution patterns, such as All-Pairs =-=[18]-=-, and Match and CoGroup in PACT [5]. These higherorder actors will reused to build domain-specific bioActors. • bioKepler: The bioKepler module contains a specialized set of actors, namely bioActors, ...

Citation Context

...t the extreme case where the locality issinfinitely large, there would be only one file on one computesnode, eventually all the tasks need to be run on that node.s2) All-Pairs in BiometricssAll-Pairs =-=[31]-=- is a common benchmark for data-intensivesapplications that describes the behavior of a new function onssets A and sets B. For example, in Biometrics, it is verysimportant to find out the covariance o...

Citation Context

...on for the many subsets of the input records. The workflow of the three distributed grouped aggregation approaches is shown in Figure 2. 3.1.Pleasingly Parallel Application The Alu clustering problem =-=[11]-=- [12] is one of the most challenging problems when sequencing clustering because Alus represent the largest repeat families in the human genome. About one million copies of the Alu sequence exist in t...

Citation Context

.... Pairwise sequence alignment (PSA) does all-pair sequence alignment over a given sequence dataset, which is usually in FASTA format, where the result is generated as an all-pair dissimilarity matrix =-=[16]-=-. There are many pairwise sequence alignment algorithms, such as SmithWaterman-GOTOH (SWG) [17] and Needleman-Wunsch [18]. In this particular pipeline, we use SWG algorithm for the each pair’s sequenc...

Citation Context

...pute half of the result matrix. Previous researchers have studied All-Pairs theoretically [28] and on small clusters [4]. Our contribution is to scale the problem up to hundreds of nodes in the cloud =-=[16]-=-. As with the previous abstraction, the user provides a “function” in the form of a program that compares two input files. The data sets A and B are text files listing the remaining files to be compar...