Kamer Kaya

Task assignment in heterogeneous computing systems

The problem of task assignment in heterogeneous computing systems has been studied for many years with many variations. We consider the version in which communicating tasks are to be assigned to heterogeneous processors with identical communication links to minimize the sum of the total execution and communication costs. Our contributions are three fold: a task clustering method which takes the execution times of the tasks into account; two metrics to determine the order in which tasks are assigned to the processors; a refinement heuristic which improves a given assignment. We use these three methods to obtain a family of task assignment algorithms including multilevel ones that apply clustering and refinement heuristics repeatedly. We have implemented eight existing algorithms to test the proposed methods. Our refinement algorithm improves the solutions of the existing algorithms by up to 15% and the proposed algorithms obtain better solutions than these refined solutions.

Performance Evaluation of Sparse Matrix Multiplication Kernels on Intel Xeon Phi

Intel Xeon Phi is a recently released high-performance coprocessor which features 61 cores each supporting 4 hardware threads with 512-bit wide SIMD registers achieving a peak theoretical performance of 1Tflop/s in double precision. Its design differs from classical modern processors; it comes with a large number of cores, the 4-way hyperthreading capability allows many applications to saturate the massive memory bandwidth, and its large SIMD capabilities allow to reach high computation throughput. The core of many scientific applications involves the multiplication of a large, sparse matrix with a single or multiple dense vectors which are not compute-bound but memory-bound. In this paper, we investigate the performance of the Xeon Phi coprocessor for these sparse linear algebra kernels. We highlight the important hardware details and show that Xeon Phi’s sparse kernel performance is very promising and even better than that of cutting-edge CPUs and GPUs.

Threshold cryptography based on Asmuth-Bloom secret sharing

In this paper, we investigate how threshold cryptography can be conducted with the Asmuth-Bloom secret sharing scheme and present three novel function sharing schemes for RSA, ElGamal and Paillier cryptosystems. To the best of our knowledge, these are the first provably secure threshold cryptosystems realized using the Asmuth-Bloom secret sharing. Proposed schemes are comparable in performance to earlier proposals in threshold cryptography.

Integrated data placement and task assignment for scientific workflows in clouds

We consider the problem of optimizing the execution of data-intensive scientific workflows in the Cloud. We address the problem under the following scenario. The tasks of the workflows communicate through files; the output of a task is used by another task as an input file and if these tasks are assigned on different execution sites, a file transfer is necessary. The output files are to be stored at a site. Each execution site is to be assigned a certain percentage of the files and tasks. These percentages, called target weights, are pre-determined and reflect either user preferences or the storage capacity and computing power of the sites. The aim is to place the data files into and assign the tasks to the execution sites so as to reduce the cost associated with the file transfers, while complying with the target weights. To do this, we model the workflow as a hypergraph and with a hypergraph-partitioning-based formulation, we propose a heuristic which generates data placement and task assignment schemes simultaneously. We report simulation results on a number of real-life and synthetically generated scientific workflows. Our results show that the proposed heuristic is fast, and can find mappings and assignments which reduce file transfers, while respecting the target weights.

Betweenness centrality on GPUs and heterogeneous architectures

The betweenness centrality metric has always been intriguing for graph analyses and used in various applications. Yet, it is one of the most computationally expensive kernels in graph mining. In this work, we investigate a set of techniques to make the betweenness centrality computations faster on GPUs as well as on heterogeneous CPU/GPU architectures. Our techniques are based on virtualization of the vertices with high degree, strided access to adjacency lists, removal of the vertices with degree 1, and graph ordering. By combining these techniques within a fine-grain parallelism, we reduced the computation time on GPUs significantly for a set of social networks. On CPUs, which can usually have access to a large amount of memory, we used a coarse-grain parallelism. We showed that heterogeneous computing, i.e., using both architectures at the same time, is a promising solution for betweenness centrality. Experimental results show that the proposed techniques can be a great arsenal to reduce the centrality computation time for networks. In particular, it reduces the computation time of a 234 million edges graph from more than 4 months to less than 12 days.

Design, implementation, and analysis of maximum transversal algorithms

We report on careful implementations of seven algorithms for solving the problem of finding a maximum transversal of a sparse matrix. We analyze the algorithms and discuss the design choices. To the best of our knowledge, this is the most comprehensive comparison of maximum transversal algorithms based on augmenting paths. Previous papers with the same objective either do not have all the algorithms discussed in this article or they used nonuniform implementations from different researchers. We use a common base to implement all of the algorithms and compare their relative performance on a wide range of graphs and matrices. We systematize, develop, and use several ideas for enhancing performance. One of these ideas improves the performance of one of the existing algorithms in most cases, sometimes significantly. So much so that we use this as the eighth algorithm in comparisons.

Diversified recommendation on graphs: pitfalls, measures, and algorithms

Result diversification has gained a lot of attention as a way to answer ambiguous queries and to tackle the redundancy problem in the results. In the last decade, diversification has been applied on or integrated into the process of PageRank- or eigenvector-based methods that run on various graphs, including social networks, collaboration networks in academia, web and product co-purchasing graphs. For these applications, the diversification problem is usually addressed as a bicriteria objective optimization problem of relevance and diversity. However, such an approach is questionable since a query-oblivious diversification algorithm that recommends most of its results without even considering the query may perform the best on these commonly used measures. In this paper, we show the deficiencies of popular evaluation techniques of diversification methods, and investigate multiple relevance and diversity measures to understand whether they have any correlations. Next, we propose a novel measure called expanded relevance which combines both relevance and diversity into a single function in order to measure the coverage of the relevant part of the graph. We also present a new greedy diversification algorithm called BestCoverage, which optimizes the expanded relevance of the result set with (1-1/e)-approximation. With a rigorous experimentation on graphs from various applications, we show that the proposed method is efficient and effective for many use cases.

Heuristics for scheduling file-sharing tasks on heterogeneous systems with distributed repositories

We consider the problem of scheduling an application on a computing system consisting of heterogeneous processors and data repositories. The application consists of a large number of file-sharing otherwise independent tasks. The files initially reside on the repositories. The processors and the repositories are connected through a heterogeneous interconnection network. Our aim is to assign the tasks to the processors, to schedule the file transfers from the repositories, and to schedule the executions of tasks on each processor in such a way that the turnaround time is minimized. We propose a heuristic composed of three phases: initial task assignment, task assignment refinement, and execution ordering. We experimentally compare the proposed heuristics with three well-known heuristics on a large number of problem instances. The proposed heuristic runs considerably faster than the existing heuristics and obtains 10-14% better turnaround times than the best of the three existing heuristics.

Diversifying Citation Recommendations

Literature search is one of the most important steps of academic research. With more than 100,000 papers published each year just in computer science, performing a complete literature search becomes a Herculean task. Some of the existing approaches and tools for literature search cannot compete with the characteristics of today’s literature, and they suffer from ambiguity and homonymy. Techniques based on citation information are more robust to the mentioned issues. Thus, we recently built a Web service called the advisor, which provides personalized recommendations to researchers based on their papers of interest. Since most recommendation methods may return redundant results, diversifying the results of the search process is necessary to increase the amount of information that one can reach via an automated search. This article targets the problem of result diversification in citation-based bibliographic search, assuming that the citation graph itself is the only information available and no categories or intents are known. The contribution of this work is threefold. We survey various random walk--based diversification methods and enhance them with the direction awareness property to allow users to reach either old, foundational (possibly well-cited and well-known) research papers or recent (most likely less-known) ones. Next, we propose a set of novel algorithms based on vertex selection and query refinement. A set of experiments with various evaluation criteria shows that the proposed γ-RLM algorithm performs better than the existing approaches and is suitable for real-time bibliographic search in practice.

A Verifiable Secret Sharing Scheme Based on the Chinese Remainder Theorem

In this paper, we investigate how to achieve verifiable secret sharing (VSS) schemes by using the Chinese Remainder Theorem (CRT). We first show that two schemes proposed earlier are not secure by an attack where the dealer is able to distribute inconsistent shares to the users. Then we propose a new VSS scheme based on the CRT and prove its security. Using the proposed VSS scheme, we develop a joint random secret sharing (JRSS) protocol, which, to the best of our knowledge, is the first JRSS protocol based on the CRT.