Search

Large networks are useful in a wide range of applications. Sometimes problem instances are composed of billions of entities. Decomposing and analyzing these structures helps us gain new insights about our surroundings. Even if the final application concerns a different problem (such as traversal, finding paths, trees, and flows), decomposing large graphs is often an important subproblem for complexity reduction or parallelization. This report is a summary of discussions that happened at Dagstuhl seminar 23331 on "Recent Trends in Graph Decomposition" and presents currently open problems and future directions in the area of (hyper)graph decomposition.

Graphs are ubiquitous, and they can model unique characteristics and complex relations of real-life systems. Although using machine learning (ML) on graphs is promising, their raw representation is not suitable for ML algorithms. Graph embedding represents each node of a graph as a d-dimensional vector which is more suitable for ML tasks. However, the embedding process is expensive, and CPU-based tools do not scale to real-world graphs. In this work, we present GOSH, a GPU-based tool for embedding large-scale graphs with minimum hardware constraints. GOSH employs a novel graph coarsening algorithm to enhance the impact of updates and minimize the work for embedding. It also incorporates a decomposition schema that enables any arbitrarily large graph to be embedded with a single GPU. As a result, GOSH sets a new state-of-the-art in link prediction both in accuracy and speed, and delivers high-quality embeddings for node classification at a fraction of the time compared to the state-of-the-art. For instance, it can embed a graph with over 65 million vertices and 1.8 billion edges in less than 30 minutes on a single GPU., Comment: 12 pages, 11 tables, 6 figures, submitted for publication at Special Section on Parallel and Distributed Computing Techniques for AI, ML, and DL

There is a plethora of data structures, algorithms, and frameworks dealing with major data-stream problems like estimating the frequency of items, answering set membership, association and multiplicity queries, and several other statistics that can be extracted from voluminous data streams. In this survey, we are focusing on exploring randomized data structures called Bloom Filters. This data structure answers whether an item exists or not in a data stream with a false positive probability fpp. In this survey, many variants of the Bloom filter will be covered by showing the strengths of each structure and its drawbacks i.e. some Bloom filters deal with insertion and deletions and others don't, some variants use the memory efficiently but increase the fpp where others pay the trade-off in the reversed way. Furthermore, in each Bloom filter structure, the false positive probability will be highlighted alongside the most important technical details showing the improvement it is presenting, while the main aim of this work is to provide an overall comparison between the variants of the Bloom filter structure according to the application domain that it fits in.

Influence maximization (IM) is the problem of finding a seed vertex set that maximizes the expected number of vertices influenced under a given diffusion model. Due to the NP-Hardness of finding an optimal seed set, approximation algorithms are frequently used for IM. In this work, we describe a fast, error-adaptive approach that leverages Count-Distinct sketches and hash-based fused sampling. To estimate the number of influenced vertices throughout a diffusion, we use per-vertex Flajolet-Martin sketches where each sketch corresponds to a sampled subgraph. To efficiently simulate the diffusions, the reach-set cardinalities of a single vertex are stored in memory in a consecutive fashion. This allows the proposed algorithm to estimate the number of influenced vertices in a single step for simulations at once. For a faster IM kernel, we rebuild the sketches in parallel only after observing estimation errors above a given threshold. Our experimental results show that the proposed algorithm yields high-quality seed sets while being up to 119x faster than a state-of-the-art approximation algorithm. In addition, it is up to 62x faster than a sketch-based approach while producing seed sets with 3%-12% better influence scores, Comment: 12 pages. Sent to IEEE Transactions on Knowledge and Data Engineering as a regular paper

Many well-known, real-world problems involve dynamic data which describe the relationship among the entities. Hypergraphs are powerful combinatorial structures that are frequently used to model such data. For many of today's data-centric applications, this data is streaming; new items arrive continuously, and the data grows with time. With paradigms such as Internet of Things and Edge Computing, such applications become more natural and more practical. In this work, we assume a streaming model where the data is modeled as a hypergraph, which is generated at the edge. This data then partitioned and sent to remote nodes via an algorithm running on a memory-restricted device such as a single board computer. Such a partitioning is usually performed by taking a connectivity metric into account to minimize the communication cost of later analyses that will be performed in a distributed fashion. Although there are many offline tools that can partition static hypergraphs excellently, algorithms for the streaming settings are rare. We analyze a well-known algorithm from the literature and significantly improve its running time by altering its inner data structure. For instance, on a medium-scale hypergraph, the new algorithm reduces the runtime from 17800 seconds to 10 seconds. We then propose sketch- and hash-based algorithms, as well as ones that can leverage extra memory to store a small portion of the data to enable the refinement of partitioning when possible. We experimentally analyze the performance of these algorithms and report their run times, connectivity metric scores, and memory uses on a high-end server and four different single-board computer architectures., Comment: 8 pages, 6 algorithms, 2 figures, 4 tables, submitted to HPCS'20

A significant portion of the data today, e.g, social networks, web connections, etc., can be modeled by graphs. A proper analysis of graphs with Machine Learning (ML) algorithms has the potential to yield far-reaching insights into many areas of research and industry. However, the irregular structure of graph data constitutes an obstacle for running ML tasks on graphs such as link prediction, node classification, and anomaly detection. Graph embedding is a compute-intensive process of representing graphs as a set of vectors in a d-dimensional space, which in turn makes it amenable to ML tasks. Many approaches have been proposed in the literature to improve the performance of graph embedding, e.g., using distributed algorithms, accelerators, and pre-processing techniques. Graph coarsening, which can be considered a pre-processing step, is a structural approximation of a given, large graph with a smaller one. As the literature suggests, the cost of embedding significantly decreases when coarsening is employed. In this work, we thoroughly analyze the impact of the coarsening quality on the embedding performance both in terms of speed and accuracy. Our experiments with a state-of-the-art, fast graph embedding tool show that there is an interplay between the coarsening decisions taken and the embedding quality., Comment: 10 pages, 6 figures, submitted to 2020 IEEE International Conference on Big Data

In graph embedding, the connectivity information of a graph is used to represent each vertex as a point in a d-dimensional space. Unlike the original, irregular structural information, such a representation can be used for a multitude of machine learning tasks. Although the process is extremely useful in practice, it is indeed expensive and unfortunately, the graphs are becoming larger and harder to embed. Attempts at scaling up the process to larger graphs have been successful but often at a steep price in hardware requirements. We present GOSH, an approach for embedding graphs of arbitrary sizes on a single GPU with minimum constraints. GOSH utilizes a novel graph coarsening approach to compress the graph and minimize the work required for embedding, delivering high-quality embeddings at a fraction of the time compared to the state-of-the-art. In addition to this, it incorporates a decomposition schema that enables any arbitrarily large graph to be embedded using a single GPU with minimum constraints on the memory size. With these techniques, GOSH is able to embed a graph with over 65 million vertices and 1.8 billion edges in less than an hour on a single GPU and obtains a 93% AUCROC for link-prediction which can be increased to 95% by running the tool for 80 minutes., Comment: 11 pages, 4 figures, published in ICPP 2020: The 49th International Conference on Parallel Processing - Edmonton, AB, Canada

Graph coloring is a fundamental problem in combinatorics with many applications in practice. In this problem, the vertices in a given graph must be colored by using the least number of colors in such a way that a vertex has a different color than its neighbors. The problem, as well as its different variants, has been proven to be NP-Hard. Therefore, there are greedy algorithms in the literature aiming to use a small number of colors. These algorithms traverse the vertices and color them one by one. The vertex visit order has a significant impact on the number of colors used. In this work, we investigated if social network analytics metrics can be used to find this order. Our experiments showed that when closeness centrality is used to find vertex visit order, a smaller number of colors is used by the greedy algorithms., Comment: in Turkish language

Influence maximization (IM) is the problem of finding a seed vertex set which is expected to incur the maximum influence spread on a graph. It has various applications in practice such as devising an effective and efficient approach to disseminate information, news or ad within a social network. The problem is shown to be NP-hard and approximation algorithms with provable quality guarantees exist in the literature. However, these algorithms are computationally expensive even for medium-scaled graphs. Furthermore, graph algorithms usually suffer from spatial and temporal irregularities during memory accesses, and this adds an extra cost on top of the already expensive IM kernels. In this work, we leverage fused sampling, memoization, and vectorization to restructure, parallelize and boost their performance on undirected networks. The proposed approach employs a pseudo-random function and performs multiple Monte-Carlo simulations in parallel to exploit the SIMD lanes effectively and efficiently. Besides, it significantly reduces the number of edge traversals, hence the amount of data brought from the memory, which is critical for almost all memory-bound graph kernels. We apply the proposed approach to the traditional MixGreedy algorithm and propose Infuser which is more than 3000 times faster than the traditional greedy approaches and can run on large graphs that have been considered as too large in the literature., Comment: 12 pages, 6 figures. Submitted to IEEE Transactions on Parallel and Distributed Systems

Sketches are probabilistic data structures that can provide approximate results within mathematically proven error bounds while using orders of magnitude less memory than traditional approaches. They are tailored for streaming data analysis on architectures even with limited memory such as single-board computers that are widely exploited for IoT and edge computing. Since these devices offer multiple cores, with efficient parallel sketching schemes, they are able to manage high volumes of data streams. However, since their caches are relatively small, a careful parallelization is required. In this work, we focus on the frequency estimation problem and evaluate the performance of a high-end server, a 4-core Raspberry Pi and an 8-core Odroid. As a sketch, we employed the widely used Count-Min Sketch. To hash the stream in parallel and in a cache-friendly way, we applied a novel tabulation approach and rearranged the auxiliary tables into a single one. To parallelize the process with performance, we modified the workflow and applied a form of buffering between hash computations and sketch updates. Today, many single-board computers have heterogeneous processors in which slow and fast cores are equipped together. To utilize all these cores to their full potential, we proposed a dynamic load-balancing mechanism which significantly increased the performance of frequency estimation., Comment: 12 pages, 4 figures, 3 algorithms, 1 table, submitted to EuroPar'19

In parallel computing, a valid graph coloring yields a lock-free processing of the colored tasks, data points, etc., without expensive synchronization mechanisms. However, coloring is not free and the overhead can be significant. In particular, for the bipartite-graph partial coloring (BGPC) and distance-2 graph coloring (D2GC) problems, which have various use-cases within the scientific computing and numerical optimization domains, the coloring overhead can be in the order of minutes with a single thread for many real-life graphs. In this work, we propose parallel algorithms for bipartite-graph partial coloring on shared-memory architectures. Compared to the existing shared-memory BGPC algorithms, the proposed ones employ greedier and more optimistic techniques that yield a better parallel coloring performance. In particular, on 16 cores, the proposed algorithms perform more than 4x faster than their counterparts in the ColPack library which is, to the best of our knowledge, the only publicly-available coloring library for multicore architectures. In addition to BGPC, the proposed techniques are employed to devise parallel distance-2 graph coloring algorithms and similar performance improvements have been observed. Finally, we propose two costless balancing heuristics for BGPC that can reduce the skewness and imbalance on the cardinality of color sets (almost) for free. The heuristics can also be used for the D2GC problem and in general, they will probably yield a better color-based parallelization performance especially on many-core architectures., Comment: 11 pages

Background: Code cloning - copying and reusing pieces of source code - is a common phenomenon in software development in practice. There have been several empirical studies on the effects of cloning, but there are contradictory results regarding the connection of cloning and faults. Objective: Our aim is to clarify the relationship between code clones and faults. In particular, we focus on inconsistent (or type-3) clones in this work. Method: We conducted a case study with TWT GmbH where we detected the code clones in three Java systems, set them into relation to information from issue tracking and version control and interviewed three key developers. Results: Of the type-3 clones, 17 % contain faults. Developers modified most of the type-3 clones simultaneously and thereby fixed half of the faults in type-3 clones consistently. Type-2 clones with faults all evolved to fixed type-3 clones. Clone length is only weakly correlated with faultiness. Conclusion: There are indications that the developers in two cases have been aware of clones. It might be a reason for the weak relationship between type-3 clones and faults. Hence, it seems important to keep developers aware of clones, potentially with new tool support. Future studies need to investigate if the rate of faults in type-3 clones justifies using them as cues in defect detection., Comment: 11 pages, 0 figures

A recent work of Harn and Fuyou presents the first multilevel (disjunctive) threshold secret sharing scheme based on the Chinese Remainder Theorem. In this work, we first show that the proposed method is not secure and also fails to work with a certain natural setting of the threshold values on compartments. We then propose a secure scheme that works for all threshold settings. In this scheme, we employ a refined version of Asmuth-Bloom secret sharing with a special and generic Asmuth-Bloom sequence called the {\it anchor sequence}. Based on this idea, we also propose the first multilevel conjunctive threshold secret sharing scheme based on the Chinese Remainder Theorem. Lastly, we discuss how the proposed schemes can be used for multilevel threshold function sharing by employing it in a threshold RSA cryptosystem as an example.

We propose HAMSI (Hessian Approximated Multiple Subsets Iteration), which is a provably convergent, second order incremental algorithm for solving large-scale partially separable optimization problems. The algorithm is based on a local quadratic approximation, and hence, allows incorporating curvature information to speed-up the convergence. HAMSI is inherently parallel and it scales nicely with the number of processors. Combined with techniques for effectively utilizing modern parallel computer architectures, we illustrate that the proposed method converges more rapidly than a parallel stochastic gradient descent when both methods are used to solve large-scale matrix factorization problems. This performance gain comes only at the expense of using memory that scales linearly with the total size of the optimization variables. We conclude that HAMSI may be considered as a viable alternative in many large scale problems, where first order methods based on variants of stochastic gradient descent are applicable., Comment: The software is available at https://github.com/spartensor/hamsi-mf

We design, implement, and evaluate GPU-based algorithms for the maximum cardinality matching problem in bipartite graphs. Such algorithms have a variety of applications in computer science, scientific computing, bioinformatics, and other areas. To the best of our knowledge, ours is the first study which focuses on GPU implementation of the maximum cardinality matching algorithms. We compare the proposed algorithms with serial and multicore implementations from the literature on a large set of real-life problems where in majority of the cases one of our GPU-accelerated algorithms is demonstrated to be faster than both the sequential and multicore implementations., Comment: 14 pages, 5 figures

Analyzing networks requires complex algorithms to extract meaningful information. Centrality metrics have shown to be correlated with the importance and loads of the nodes in network traffic. Here, we are interested in the problem of centrality-based network management. The problem has many applications such as verifying the robustness of the networks and controlling or improving the entity dissemination. It can be defined as finding a small set of topological network modifications which yield a desired closeness centrality configuration. As a fundamental building block to tackle that problem, we propose incremental algorithms which efficiently update the closeness centrality values upon changes in network topology, i.e., edge insertions and deletions. Our algorithms are proven to be efficient on many real-life networks, especially on small-world networks, which have a small diameter and a spike-shaped shortest distance distribution. In addition to closeness centrality, they can also be a great arsenal for the shortest-path-based management and analysis of the networks. We experimentally validate the efficiency of our algorithms on large networks and show that they update the closeness centrality values of the temporal DBLP-coauthorship network of 1.2 million users 460 times faster than it would take to compute them from scratch. To the best of our knowledge, this is the first work which can yield practical large-scale network management based on closeness centrality values.

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. Many scientific applications involve operations on large sparse matrices such as linear solvers, eigensolver, and graph mining algorithms. The core of most of these applications involves the multiplication of a large, sparse matrix with a dense vector (SpMV). In this paper, we investigate the performance of the Xeon Phi coprocessor for SpMV. We first provide a comprehensive introduction to this new architecture and analyze its peak performance with a number of micro benchmarks. Although the design of a Xeon Phi core is not much different than those of the cores in modern processors, its large number of cores and hyperthreading capability allow many application to saturate the available memory bandwidth, which is not the case for many cutting-edge processors. Yet, our performance studies show that it is the memory latency not the bandwidth which creates a bottleneck for SpMV on this architecture. Finally, our experiments show that Xeon Phi's sparse kernel performance is very promising and even better than that of cutting-edge general purpose processors and GPUs.

Who is more important in a network? Who controls the flow between the nodes or whose contribution is significant for connections? Centrality metrics play an important role while answering these questions. The betweenness metric is useful for network analysis and implemented in various tools. Since it is one of the most computationally expensive kernels in graph mining, several techniques have been proposed for fast computation of betweenness centrality. In this work, we propose and investigate techniques which compress a network and shatter it into pieces so that the rest of the computation can be handled independently for each piece. Although we designed and tuned the shattering process for betweenness, it can be adapted for other centrality metrics in a straightforward manner. Experimental results show that the proposed techniques can be a great arsenal to reduce the centrality computation time for various types of networks.

Literature search is arguably one of the most important phases of the academic and non-academic research. The increase in the number of published papers each year makes manual search inefficient and furthermore insufficient. Hence, automatized methods such as search engines have been of interest in the last thirty years. Unfortunately, these traditional engines use keyword-based approaches to solve the search problem, but these approaches are prone to ambiguity and synonymy. On the other hand, bibliographic search techniques based only on the citation information are not prone to these problems since they do not consider textual similarity. For many particular research areas and topics, the amount of knowledge to humankind is immense, and obtaining the desired information is as hard as looking for a needle in a haystack. Furthermore, sometimes, what we are looking for is a set of documents where each one is different than the others, but at the same time, as a whole we want them to cover all the important parts of the literature relevant to our search. This paper targets the problem of result diversification in citation-based bibliographic search. It surveys a set of techniques which aim to find a set of papers with satisfactory quality and diversity. We enhance these algorithms with a direction-awareness functionality to allow the users to reach either old, well-cited, well-known research papers or recent, less-known ones. We also propose a set of novel techniques for a better diversification of the results. All the techniques considered are compared by performing a rigorous experimentation. The results show that some of the proposed techniques are very successful in practice while performing a search in a bibliographic database., Comment: 19 pages, manuscript under review

The literature search has always been an important part of an academic research. It greatly helps to improve the quality of the research process and output, and increase the efficiency of the researchers in terms of their novel contribution to science. As the number of published papers increases every year, a manual search becomes more exhaustive even with the help of today's search engines since they are not specialized for this task. In academics, two relevant papers do not always have to share keywords, cite one another, or even be in the same field. Although a well-known paper is usually an easy pray in such a hunt, relevant papers using a different terminology, especially recent ones, are not obvious to the eye. In this work, we propose paper recommendation algorithms by using the citation information among papers. The proposed algorithms are direction aware in the sense that they can be tuned to find either recent or traditional papers. The algorithms require a set of papers as input and recommend a set of related ones. If the user wants to give negative or positive feedback on the suggested paper set, the recommendation is refined. The search process can be easily guided in that sense by relevance feedback. We show that this slight guidance helps the user to reach a desired paper in a more efficient way. We adapt our models and algorithms also for the venue and reviewer recommendation tasks. Accuracy of the models and algorithms is thoroughly evaluated by comparison with multiple baselines and algorithms from the literature in terms of several objectives specific to citation, venue, and reviewer recommendation tasks. All of these algorithms are implemented within a publicly available web-service framework (http://theadvisor.osu.edu/) which currently uses the data from DBLP and CiteSeer to construct the proposed citation graph., Comment: 10 pages, 8 figures, 3 tables