A Scalable General Purpose System for Large-Scale Graph Processing

Graph analytics is an important and computationally demanding class of data analytics. It is essential to balance scalability, ease-of-use and high performance in large scale graph analytics. As such, it is necessary to hide the complexity of parallelism, data distribution and memory locality behind an abstract interface. The aim of this work is to build a scalable graph analytics framework that does not demand significant parallel programming experience based on NUMA-awareness.
The realization of such a system faces two key problems:
(i)~how to develop a scale-free parallel programming framework that scales efficiently across NUMA domains; (ii)~how to efficiently apply graph partitioning in order to create separate and largely independent work items that can be distributed among threads.

Real-time analytics on large dynamic graphs

In today's fast-paced and interconnected digital world, the data generated by an increasing number of applications is being modeled as dynamic graphs. The graph structure encodes relationships among data items, while the structural changes to the graphs as well as the continuous stream of information produced by the entities in these graphs make them dynamic in nature. Examples include social networks where users post status updates, images, videos, etc.; phone call networks where nodes may send text messages or place phone calls; road traffic networks where the traffic behavior of the road segments changes constantly, and so on. There is a tremendous value in storing, managing, and analyzing such dynamic graphs and deriving meaningful insights in real-time. However, a majority of the work in graph analytics assumes a static setting, and there is a lack of systematic study of the various dynamic scenarios, the complexity they impose on the analysis tasks, and the challenges in building efficient systems that can support such tasks at a large scale. In this dissertation, I design a unified streaming graph data management framework, and develop prototype systems to support increasingly complex tasks on dynamic graphs. In the first part, I focus on the management and querying of distributed graph data. I develop a hybrid replication policy that monitors the read-write frequencies of the nodes to decide dynamically what data to replicate, and whether to do eager or lazy replication in order to minimize network communication and support low-latency querying. In the second part, I study parallel execution of continuous neighborhood-driven aggregates, where each node aggregates the information generated in its neighborhoods. I build my system around the notion of an aggregation overlay graph, a pre-compiled data structure that enables sharing of partial aggregates across different queries, and also allows partial pre-computation of the aggregates to minimize the query latencies and increase throughput. Finally, I extend the framework to support continuous detection and analysis of activity-based subgraphs, where subgraphs could be specified using both graph structure as well as activity conditions on the nodes. The query specification tasks in my system are expressed using a set of active structural primitives, which allows the query evaluator to use a set of novel optimization techniques, thereby achieving high throughput. Overall, in this dissertation, I define and investigate a set of novel tasks on dynamic graphs, design scalable optimization techniques, build prototype systems, and show the effectiveness of the proposed techniques through extensive evaluation using large-scale real and synthetic datasets.

BUILDING EFFICIENT AND COST-EFFECTIVE CLOUD-BASED BIG DATA MANAGEMENT SYSTEMS

In today’s big data world, data is being produced in massive volumes, at great velocity and from a variety of different sources such as mobile devices, sensors, a plethora of small devices hooked to the internet (Internet of Things), social networks, communication networks and many others. Interactive querying and large-scale analytics are being increasingly used to derive value out of this big data. A large portion of this data is being stored and processed in the Cloud due the several advantages provided by the Cloud such as scalability, elasticity, availability, low cost of ownership and the overall economies of scale. There is thus, a growing need for large-scale cloud-based data management systems that can support real-time ingest, storage and processing of large volumes of heterogeneous data. However, in the pay-as-you-go Cloud environment, the cost of analytics can grow linearly with the time and resources required. Reducing the cost of data analytics in the Cloud thus remains a primary challenge. In my dissertation research, I have focused on building efficient and cost-effective cloud-based data management systems for different application domains that are predominant in cloud computing environments. In the first part of my dissertation, I address the problem of reducing the cost of transactional workloads on relational databases to support database-as-a-service in the Cloud. The primary challenges in supporting such workloads include choosing how to partition the data across a large number of machines, minimizing the number of distributed transactions, providing high data availability, and tolerating failures gracefully. I have designed, built and evaluated SWORD, an end-to-end scalable online transaction processing system, that utilizes workload-aware data placement and replication to minimize the number of distributed transactions that incorporates a suite of novel techniques to significantly reduce the overheads incurred both during the initial placement of data, and during query execution at runtime. In the second part of my dissertation, I focus on sampling-based progressive analytics as a means to reduce the cost of data analytics in the relational domain. Sampling has been traditionally used by data scientists to get progressive answers to complex analytical tasks over large volumes of data. Typically, this involves manually extracting samples of increasing data size (progressive samples) for exploratory querying. This provides the data scientists with user control, repeatable semantics, and result provenance. However, such solutions result in tedious workflows that preclude the reuse of work across samples. On the other hand, existing approximate query processing systems report early results, but do not offer the above benefits for complex ad-hoc queries. I propose a new progressive data-parallel computation framework, NOW!, that provides support for progressive analytics over big data. In particular, NOW! enables progressive relational (SQL) query support in the Cloud using unique progress semantics that allow efficient and deterministic query processing over samples providing meaningful early results and provenance to data scientists. NOW! enables the provision of early results using significantly fewer resources thereby enabling a substantial reduction in the cost incurred during such analytics. Finally, I propose NSCALE, a system for efficient and cost-effective complex analytics on large-scale graph-structured data in the Cloud. The system is based on the key observation that a wide range of complex analysis tasks over graph data require processing and reasoning about a large number of multi-hop neighborhoods or subgraphs in the graph; examples include ego network analysis, motif counting in biological networks, finding social circles in social networks, personalized recommendations, link prediction, etc. These tasks are not well served by existing vertex-centric graph processing frameworks whose computation and execution models limit the user program to directly access the state of a single vertex, resulting in high execution overheads. Further, the lack of support for extracting the relevant portions of the graph that are of interest to an analysis task and loading it onto distributed memory leads to poor scalability. NSCALE allows users to write programs at the level of neighborhoods or subgraphs rather than at the level of vertices, and to declaratively specify the subgraphs of interest. It enables the efficient distributed execution of these neighborhood-centric complex analysis tasks over largescale graphs, while minimizing resource consumption and communication cost, thereby substantially reducing the overall cost of graph data analytics in the Cloud. The results of our extensive experimental evaluation of these prototypes with several real-world data sets and applications validate the effectiveness of our techniques which provide orders-of-magnitude reductions in the overheads of distributed data querying and analysis in the Cloud.

ANALYSIS AND VISUALIZATION OF FLOW FIELDS USING INFORMATION-THEORETIC TECHNIQUES AND GRAPH-BASED REPRESENTATIONS

Three-dimensional flow visualization plays an essential role in many areas of science and engineering, such as aero- and hydro-dynamical systems which dominate various physical and natural phenomena. For popular methods such as the streamline visualization to be effective, they should capture the underlying flow features while facilitating user observation and understanding of the flow field in a clear manner. My research mainly focuses on the analysis and visualization of flow fields using various techniques, e.g. information-theoretic techniques and graph-based representations. Since the streamline visualization is a popular technique in flow field visualization, how to select good streamlines to capture flow patterns and how to pick good viewpoints to observe flow fields become critical. We treat streamline selection and viewpoint selection as symmetric problems and solve them simultaneously using the dual information channel [81]. To the best of my knowledge, this is the first attempt in flow visualization to combine these two selection problems in a unified approach. This work selects streamline in a view-independent manner and the selected streamlines will not change for all viewpoints. My another work [56] uses an information-theoretic approach to evaluate the importance of each streamline under various sample viewpoints and presents a solution for view-dependent streamline selection that guarantees coherent streamline update when the view changes gradually. When projecting 3D streamlines to 2D images for viewing, occlusion and clutter become inevitable. To address this challenge, we design FlowGraph [57, 58], a novel compound graph representation that organizes field line clusters and spatiotemporal regions hierarchically for occlusion-free and controllable visual exploration. We enable observation and exploration of the relationships among field line clusters, spatiotemporal regions and their interconnection in the transformed space. Most viewpoint selection methods only consider the external viewpoints outside of the flow field. This will not convey a clear observation when the flow field is clutter on the boundary side. Therefore, we propose a new way to explore flow fields by selecting several internal viewpoints around the flow features inside of the flow field and then generating a B-Spline curve path traversing these viewpoints to provide users with closeup views of the flow field for detailed observation of hidden or occluded internal flow features [54]. This work is also extended to deal with unsteady flow fields. Besides flow field visualization, some other topics relevant to visualization also attract my attention. In iGraph [31], we leverage a distributed system along with a tiled display wall to provide users with high-resolution visual analytics of big image and text collections in real time. Developing pedagogical visualization tools forms my other research focus. Since most cryptography algorithms use sophisticated mathematics, it is difficult for beginners to understand both what the algorithm does and how the algorithm does that. Therefore, we develop a set of visualization tools to provide users with an intuitive way to learn and understand these algorithms.

Benchmarking of distributed computing engines spark and GraphLab for big data analytics

In this paper we evaluate and compare two representativeand popular distributed processing engines for large scalebig data analytics, Spark and graph based engine GraphLab. Wedesign a benchmark suite including representative algorithmsand datasets to compare the performances of the computingengines, from performance aspects of running time, memory andCPU usage, network and I/O overhead. The benchmark suite istested on both local computer cluster and virtual machines oncloud. By varying the number of computers and memory weexamine the scalability of the computing engines with increasingcomputing resources (such as CPU and memory). We also runcross-evaluation of generic and graph based analytic algorithmsover graph processing and generic platforms to identify thepotential performance degradation if only one processing engineis available. It is observed that both computing engines showgood scalability with increase of computing resources. WhileGraphLab largely outperforms Spark for graph algorithms, ithas close running time performance as Spark for non-graphalgorithms. Additionally the running time with Spark for graphalgorithms over cloud virtual machines is observed to increaseby almost 100% compared to over local computer clusters.

Adaptive edge analytics for distributed networked control of water systems

Over the last decade, there has been a trend where water utility companies aim to make water distribution networks more intelligent in order to improve their quality of service, reduce water waste, minimize maintenance costs etc., by incorporating IoT technologies. Current state of the art solutions use expensive power hungry deployments to monitor and transmit water network states periodically in order to detect anomalous behaviors such as water leakage and bursts. However, more than 97% of water network assets are remote away from power and are often in geographically remote underpopulated areas, facts that make current approaches unsuitable for next generation more dynamic adaptive water networks. Battery-driven wireless sensor/actuator based solutions are theoretically the perfect choice to support next generation water distribution. In this paper, we present an end-to-end water leak localization system, which exploits edge processing and enables the use of battery-driven sensor nodes. Our system combines a lightweight edge anomaly detection algorithm based on compression rates and an efficient localization algorithm based on graph theory. The edge anomaly detection and localization elements of the systems produce a timely and accurate localization result and reduce the communication by 99% compared to the traditional periodic communication. We evaluated our schemes by deploying non-intrusive sensors measuring vibrational data on a real-world water test rig that have had controlled leakage and burst scenarios implemented.

Multi-vehicles interaction graph model for cooperative collision warning system

Cooperative collision warning system for road vehicles, enabled by recent advances in positioning systems and wireless communication technologies, can potentially reduce traffic accident significantly. To improve the system, we propose a graph model to represent interactions between multiple road vehicles in a specific region and at a specific time. Given a list of vehicles in vicinity, we can generate the interaction graph using several rules that consider vehicle's properties such as position, speed, heading, etc. Safety applications can use the model to improve emergency warning accuracy and optimize wireless channel usage. The model allows us to develop some congestion control strategies for an efficient multi-hop broadcast protocol.

Accurate silhouette segmentation using motion detection and graph cuts

Acquiring accurate silhouettes has many applications in computer vision. This is usually done through motion detection, or a simple background subtraction under highly controlled environments (i.e. chroma-key backgrounds). Lighting and contrast issues in typical outdoor or office environments make accurate segmentation very difficult in these scenes. In this paper, gradients are used in conjunction with intensity and colour to provide a robust segmentation of motion, after which graph cuts are utilised to refine the segmentation. The results presented using the ETISEO database demonstrate that an improved segmentation is achieved through the combined use of motion detection and graph cuts, particularly in complex scenes.

Accurate silhouettes for surveillance : improved motion segmentation using graph cuts

Silhouettes are common features used by many applications in computer vision. For many of these algorithms to perform optimally, accurately segmenting the objects of interest from the background to extract the silhouettes is essential. Motion segmentation is a popular technique to segment moving objects from the background, however such algorithms can be prone to poor segmentation, particularly in noisy or low contrast conditions. In this paper, the work of [3] combining motion detection with graph cuts, is extended into two novel implementations that aim to allow greater uncertainty in the output of the motion segmentation, providing a less restricted input to the graph cut algorithm. The proposed algorithms are evaluated on a portion of the ETISEO dataset using hand segmented ground truth data, and an improvement in performance over the motion segmentation alone and the baseline system of [3] is shown.

Improved algebraic cryptanalysis of QUAD, Bivium and Trivium via graph partitioning on equation systems

We present a novel approach for preprocessing systems of polynomial equations via graph partitioning. The variable-sharing graph of a system of polynomial equations is defined. If such graph is disconnected, then the corresponding system of equations can be split into smaller ones that can be solved individually. This can provide a tremendous speed-up in computing the solution to the system, but is unlikely to occur either randomly or in applications. However, by deleting certain vertices on the graph, the variable-sharing graph could be disconnected in a balanced fashion, and in turn the system of polynomial equations would be separated into smaller systems of near-equal sizes. In graph theory terms, this process is equivalent to finding balanced vertex partitions with minimum-weight vertex separators. The techniques of finding these vertex partitions are discussed, and experiments are performed to evaluate its practicality for general graphs and systems of polynomial equations. Applications of this approach in algebraic cryptanalysis on symmetric ciphers are presented: For the QUAD family of stream ciphers, we show how a malicious party can manufacture conforming systems that can be easily broken. For the stream ciphers Bivium and Trivium, we nachieve significant speedups in algebraic attacks against them, mainly in a partial key guess scenario. In each of these cases, the systems of polynomial equations involved are well-suited to our graph partitioning method. These results may open a new avenue for evaluating the security of symmetric ciphers against algebraic attacks.

An improved genetic algorithm and graph theory based method for optimal sectionalizer switch placement in distribution networks with DG

In this paper a new graph-theory and improved genetic algorithm based practical method is employed to solve the optimal sectionalizer switch placement problem. The proposed method determines the best locations of sectionalizer switching devices in distribution networks considering the effects of presence of distributed generation (DG) in fitness functions and other optimization constraints, providing the maximum number of costumers to be supplied by distributed generation sources in islanded distribution systems after possible faults. The proposed method is simulated and tested on several distribution test systems in both cases of with DG and non DG situations. The results of the simulations validate the proposed method for switch placement of the distribution network in the presence of distributed generation.

Graph rigidity for near-coplanar structure from motion

Recent algorithms for monocular motion capture (MoCap) estimate weak-perspective camera matrices between images using a small subset of approximately-rigid points on the human body (i.e. the torso and hip). A problem with this approach, however, is that these points are often close to coplanar, causing canonical linear factorisation algorithms for rigid structure from motion (SFM) to become extremely sensitive to noise. In this paper, we propose an alternative solution to weak-perspective SFM based on a convex relaxation of graph rigidity. We demonstrate the success of our algorithm on both synthetic and real world data, allowing for much improved solutions to marker less MoCap problems on human bodies. Finally, we propose an approach to solve the two-fold ambiguity over bone direction using a k-nearest neighbour kernel density estimator.

Graph-based concept weighting for medical information retrieval

This paper presents a graph-based method to weight medical concepts in documents for the purposes of information retrieval. Medical concepts are extracted from free-text documents using a state-of-the-art technique that maps n-grams to concepts from the SNOMED CT medical ontology. In our graph-based concept representation, concepts are vertices in a graph built from a document, edges represent associations between concepts. This representation naturally captures dependencies between concepts, an important requirement for interpreting medical text, and a feature lacking in bag-of-words representations. We apply existing graph-based term weighting methods to weight medical concepts. Using concepts rather than terms addresses vocabulary mismatch as well as encapsulates terms belonging to a single medical entity into a single concept. In addition, we further extend previous graph-based approaches by injecting domain knowledge that estimates the importance of a concept within the global medical domain. Retrieval experiments on the TREC Medical Records collection show our method outperforms both term and concept baselines. More generally, this work provides a means of integrating background knowledge contained in medical ontologies into data-driven information retrieval approaches.

Towards persistent indoor appearance-based localization, mapping and navigation using CAT-Graph

The challenge of persistent appearance-based navigation and mapping is to develop an autonomous robotic vision system that can simultaneously localize, map and navigate over the lifetime of the robot. However, the computation time and memory requirements of current appearance-based methods typically scale not only with the size of the environment but also with the operation time of the platform; also, repeated revisits to locations will develop multiple competing representations which reduce recall performance. In this paper we present a solution to the persistent localization, mapping and global path planning problem in the context of a delivery robot in an office environment over a one-week period. Using a graphical appearance-based SLAM algorithm, CAT-Graph, we demonstrate constant time and memory loop closure detection with minimal degradation during repeated revisits to locations, along with topological path planning that improves over time without using a global metric representation. We compare the localization performance of CAT-Graph to openFABMAP, an appearance-only SLAM algorithm, and the path planning performance to occupancy-grid based metric SLAM. We discuss the limitations of the algorithm with regard to environment change over time and illustrate how the topological graph representation can be coupled with local movement behaviors for persistent autonomous robot navigation.

Expander graph based key distribution mechanisms in wireless sensor networks

Secure communications between large number of sensor nodes that are randomly scattered over a hostile territory, necessitate efficient key distribution schemes. However, due to limited resources at sensor nodes such schemes cannot be based on post deployment computations. Instead, pairwise (symmetric) keys are required to be pre-distributed by assigning a list of keys, (a.k.a. key-chain), to each sensor node. If a pair of nodes does not have a common key after deployment then they must find a key-path with secured links. The objective is to minimize the keychain size while (i) maximizing pairwise key sharing probability and resilience, and (ii) minimizing average key-path length. This paper presents a deterministic key distribution scheme based on Expander Graphs. It shows how to map the parameters (e.g., degree, expansion, and diameter) of a Ramanujan Expander Graph to the desired properties of a key distribution scheme for a physical network topology.