In recent years, there have been incidents which decreased people's trust in some organizations and authorities responsible for ratings and accreditation. For a few prominent examples, there was a security breach at Equifax (2017), misconduct was found in the Standard & Poor's Ratings Services (2015), and the Accrediting Council for Independent Colleges and Schools (2022) validated some of the low-performing schools as delivering higher standards than they actually were. A natural solution to these types of issues is to decentralize the relevant trust management processes using blockchain technologies. The research problems which are tackled in this thesis consider the issue of trust in reputation for assessment and review credibility at different angles, in the context of blockchain applications. We first explored the following questions. How can we trust courses in one college to provide students with the type and level of knowledge which is needed in a specific workplace? Micro-accreditation on a blockchain was our solution, including using a peer-review system to determine the rigor of a course (through a consensus). Rigor is the level of difficulty in regard to a student's expected level of knowledge. Currently, we make assumptions about the quality and rigor of what is learned, but …

Computer vision technologies are rapidly improving and becoming more important in disaster response. The majority of disaster description techniques now focus either on identify objects or categorize disasters. In this study, we trained multiple deep neural networks on low-altitude imagery with highly imbalanced and noisy labels. We utilize labeled images from the LADI dataset to formulate a solution for general problem in disaster classification and object detection. Our research integrated and developed multiple deep learning models that does the object detection task as well as the disaster scene classification task. Our solution is competitive in the TRECVID Disaster Scene Description and Indexing (DSDI) task, demonstrating that it is comparable to other suggested approaches in retrieving disaster-related video clips.

In this research, three aspects of data quality with regard to artifical intelligence (AI) have been investigated: detection of misleading fake data, especially deepfakes, data scarcity, and data insufficiency, especially how much training data is required for an AI application. Different application domains where the selected aspects pose issues have been chosen. To address the issues of data privacy, security, and regulation, these solutions are targeted for edge devices. In Chapter 3, two solutions have been proposed that aim to preempt such misleading deepfake videos and images on social media. These solutions are deployable at edge devices. In Chapter 4, a deepfake resilient digital ID system has been described. Another data quality aspect, data scarcity, has been addressed in Chapter 5. One of such agricultural problems is estimating crop damage due to natural disasters. Data insufficiency is another aspect of data quality. The amount of data required to achieve acceptable accuracy in a machine learning (ML) model has been studied in Chapter 6. As the data scarcity problem is studied in the agriculture domain, a similar scenario—plant disease detection and damage estimation—has been chosen for this verification. This research aims to provide ML or deep learning (DL)-based methods to solve …

Non-rigid point set registration of significantly uneven deformations is a challenging problem for many applications such as pose estimation, three-dimensional object reconstruction, human movement tracking. In this dissertation, we present a novel probabilistic non-rigid registration method to align point sets with significantly uneven deformations by enforcing constraints from corresponding key points and preserving local neighborhood structures. The registration method is treated as a density estimation problem. Incorporating correspondence among key points regulates the optimization process for large, uneven deformations. In addition, by leveraging neighborhood embedding using Stochastic Neighbor Embedding (SNE) as well as an alternative means based on Locally Linear Embedding (LLE), our method penalizes the incoherent transformation and hence preserves the local structure of point sets. Also, our method detects key points in the point sets based on geodesic distance. Correspondences are established using a new cluster-based, region-aware feature descriptor. This feature descriptor encodes the association of a cluster to the left-right (symmetry) or upper-lower regions of the point sets. We conducted comparison studies using public point sets and our Human point sets. Our experimental results demonstrate that our proposed method successfully reduced the registration error by at least 42.2% in contrast to the state-of-the-art method. Especially, our method …

The vehicular community is moving towards a whole new paradigm with the advancement of new technology. Vehicular communication not only supports safety services but also provides non-safety services like navigation support, toll collection, web browsing, media streaming, etc. The existing communication frameworks like Dedicated Short Range Communication (DSRC) and Cellular V2X (C-V2X) might not meet the required capacity in the coming days. So, the vehicular community needs to adopt new technologies and upgrade the existing communication frameworks so that it can fulfill the desired expectations. Therefore, an increment in reliability and data rate is required. Multiple Input Multiple Output (MIMO), 5G New Radio, Low Density Parity Check (LDPC) Code, and Massive MIMO signal detection and equalization algorithms are the latest addition to the 5G wireless communication domain. These technologies have the potential to make the existing V2X communication framework more robust. As a result, more reliability and throughput can be achieved. This work demonstrates these technologies' compatibility and positive impact on existing V2X communication standard.

This dissertation focuses on a novel area of secure data management referred to as data cooperatives. A data cooperative solution promises its users better protection and control of their personal data as compared to the traditional way of their handling by the data collectors (such as governments, big data companies, and others). However, despite the many interesting benefits that the data cooperative approach tends to provide its users, it suffers from a few challenges hindering its development, adoption, and widespread use among data providers and consumers. To address these issues, we have divided this dissertation into two parts. In the first part, we identify the existing challenges and propose and implement a decentralized architecture built atop a blockchain system. Our solution leverages the inherent decentralized, tamper-resistant, and security properties of the blockchain. The implementation of our system was carried out on an existing blockchain test network, Ropsten, and our results show that blockchain is an efficient and scalable platform for the development of a decentralized data cooperative solution. In the second part of this work, we further addressed the existing challenges and the limitations of the implementation from the first part of our work. In particular, we addressed inclusivity---a core …

There is an increased use of block-based programming environments in K-12 education and computing outreach activities to introduce novices to programming and computational thinking skills. However, despite their appealing design that allows students to focus on concepts rather than syntax, block-based programming by design is inaccessible to people with visual impairments and people who cannot use the mouse. In addition to this inaccessibility, little is known about the instructional experiences of students with visual impairments on current block-based programming environments. This dissertation addresses this gap by (1) investigating the challenges that students with visual impairments face on current block-based programming environments and (2) exploring ways in which we can use the keyboard and the screen reader to create block-based code. Through formal survey and interview studies with teachers of students with visual impairments and students with visual impairments, we identify several challenges faced by students with visual impairments on block-based programming environments. Using the knowledge of these challenges and building on prior work, we explore how to leverage the keyboard and the screen reader to improve the accessibility of block-based programming environments through a prototype of an accessible block-based programming library. In this dissertation, our empirical evaluations demonstrate that people …

In this dissertation, I start with an analysis of negation in eleven benchmark corpora covering six Natural Language Understanding (NLU) tasks. With a thorough investigation, I first show that (a) these benchmarks contain fewer negations compared to general-purpose English and (b) the few negations they contain are often unimportant. Further, my empirical studies demonstrate that state-of-the-art transformers trained using these corpora obtain substantially worse results with the instances that contain negation, especially if the negations are important. Second, I investigate whether translating negation is also an issue for modern machine translation (MT) systems. My studies find that indeed the presence of negation can significantly impact translation quality, in some cases resulting in reductions of over 60%. In light of these findings, I investigate strategies to better understand the semantics of negation. I start with identifying the focus of negation. I develop a neural model that takes into account the scope of negation, context from neighboring sentences, or both. My best proposed system obtains an accuracy improvement of 7.4% over prior work. Further, I analyze the main error categories of the systems through a detailed error analysis. Next, I explore more practical ways to understand the semantics of negation. I consider …

CPU-memory bottleneck is a widely recognized problem. It is known that majority of high performance computing (HPC) database systems are configured with large memories and dedicated to process specific workloads like weather prediction, molecular dynamic simulations etc. My research on optimal address mapping improves the memory performance by increasing the channel and bank level parallelism. In an another research direction, I proposed and evaluated adaptive page migration techniques that obviates the need for offline analysis of an application to determine page migration strategies. Furthermore, I explored different migration strategies like reverse migration, sub page migration that I found to be beneficial depending on the application behavior. Ideally, page migration strategies redirect the demand memory traffic to faster memory to improve the memory performance. In my third contribution, I worked and evaluated a memory-side accelerator to assist the main computational core in locating the non-zero elements of a sparse matrix that are typically used in scientific, machine learning workloads on a low-power embedded system configuration. Thus my contributions narrow the speed-gap by improving the latency and/or bandwidth between CPU and memory.

Recently, disaster scene description and indexing challenges attract the attention of researchers. In this dissertation, we solve a disaster-related multi-labeling task using a newly developed Low Altitude Disaster Imagery dataset. In the first task, we realize video content by selecting a set of summary key frames to represent the video sequence. Through inter-frame differences, the key frames are generated. The key frame extraction of disaster-related video clips is a powerful tool that can efficiently convert video data into image-level data, reduce the requirements for the extraction environment and improve the applicable environment. In the second, we propose a novel application of using deep learning methods on low altitude disaster video feature recognition. Supervised learning-based deep-learning approaches are effective in disaster-related features recognition via foreground object detection and background classification. Performed dataset validation, our model generalized well and improved performance by optimizing the YOLOv3 model and combining it with Resnet50. The comprehensive models showed more efficient and effective than those in prior published works. In the third task, we optimize the whole scene labeling classification by pruning the lightweight model MobileNetV3, which shows superior generalizability and can disaster features recognition from a disaster-related dataset be accomplished efficiently to assist disaster recovery.

The main research objective of this dissertation is to understand the sensing and communication challenges to achieving cooperative perception among autonomous vehicles, and then, using the insights gained, guide the design of the suitable format of data to be exchanged, reliable and efficient data fusion algorithms on vehicles. By understanding what and how data are exchanged among autonomous vehicles, from a machine learning perspective, it is possible to realize precise cooperative perception on autonomous vehicles, enabling massive amounts of sensor information to be shared amongst vehicles. I first discuss the trustworthy perception information sharing on connected and autonomous vehicles. Then how to achieve effective cooperative perception on autonomous vehicles via exchanging feature maps among vehicles is discussed in the following. In the last methodology part, I propose a set of mechanisms to improve the solution proposed before, i.e., reducing the amount of data transmitted in the network to achieve an efficient cooperative perception. The effectiveness and efficiency of our mechanism is analyzed and discussed.

This dissertation presents novel approaches to test context aware applications that suffer from a cost prohibitive number of context and GUI events and event combinations. The contributions of this work to test context aware applications under test include: (1) a real-world context events dataset from 82 Android users over a 30-day period, (2) applications of Markov models, Closed Sequential Pattern Mining (CloSPAN), Deep Neural Networks- Long Short Term Memory (LSTM) and Gated Recurrent Units (GRU), and Conditional Random Fields (CRF) applied to predict context patterns, (3) data driven test case generation techniques that insert events at the beginning of each test case in a round-robin manner, iterate through multiple context events at the beginning of each test case in a round-robin manner, and interleave real-world context event sequences and GUI events, and (4) systematically interleaving context with a combinatorial-based approach. The results of our empirical studies indicate (1) CRF outperforms other models thereby predicting context events with F1 score of about 60% for our dataset, (2) the ISFreqOne that iterates over context events at the beginning of each test case in a round-robin manner as well as interleaves real-world context event sequences and GUI events at an interval one achieves …

NAND flash-based solid state drives (SSDs) have been widely adopted in data centers and high performance computing (HPC) systems due to their better performance compared with hard disk drives. However, little is known about the reliability characteristics of SSDs in production systems. Existing works that study the statistical distributions of SSD failures in the field lack insights into distinct characteristics of SSDs. In this dissertation, I explore the SSD-specific SMART (Self-Monitoring, Analysis, and Reporting Technology) attributes and conduct in-depth analysis of SSD reliability in a production environment with a focus on the unique error types and health dynamics. QLC SSD delivers better performance in a cost-effective way. I study QLC SSDs in terms of their architecture and performance. In addition, I apply thermal stress tests to QLC SSDs and quantify their performance degradation processes. Various types of big data and machine learning workloads have been executed on SSDs under varying temperatures. The SSD throughput and application performance are analyzed and characterized.

Cyber physical systems (CPS) are an integration of computational and physical processes, where the cyber components monitor and control physical processes. Cyber-attacks largely target the cyber components with the intention of disrupting the functionality of the components in the physical domain. This dissertation explores the role of semantic inference in understanding such attacks and building resilient CPS systems. To that end, we present SIMON, an ontological design and verification framework that captures the intricate relationship(s) between cyber and physical components in CPS by leveraging several standard ontologies and extending the NIST CPS framework for the purpose of eliciting trustworthy requirements, assigning responsibilities and roles to CPS functionalities, and validating that the trustworthy requirements are met by the designed system. We demonstrate the capabilities of SIMON using two case studies – a vehicle to infrastructure (V2I) safety application and an additive manufacturing (AM) printer. In addition, we also present a taxonomy to capture threat feeds specific to the AM domain.

This work introduces three new coverage criteria based on combinatorial-based event and element sequences that occur in the mobile environment. The novel combinatorial-based criteria are used to reduce, prioritize, and generate test suites for mobile applications. The combinatorial-based criteria include unique coverage of events and elements with different respects to ordering. For instance, consider the coverage of a pair of events, e1 and e2. The least strict criterion, Combinatorial Coverage (CCov), counts the combination of these two events in a test case without respect to the order in which the events occur. That is, the combination (e1, e2) is the same as (e2, e1). The second criterion, Sequence-Based Combinatorial Coverage (SCov), considers the order of occurrence within a test case. Sequences (e1, ..., e2) and (e2,..., e1) are different sequences. The third and strictest criterion is Consecutive-Sequence Combinatorial Coverage (CSCov), which counts adjacent sequences of consecutive pairs. The sequence (e1, e2) is only counted if e1 immediately occurs before e2. The first contribution uses the novel combinatorial-based criteria for the purpose of test suite reduction. Empirical studies reveal that the criteria, when used with event sequences and sequences of size t=2, reduce the test suites by 22.8%-61.3% while the reduced …

High-performance computing (HPC) systems continue growing in both scale and complexity. These large-scale, heterogeneous systems generate tens of millions of log messages every day. Effective log analysis for understanding system behaviors and identifying system anomalies and failures is highly challenging. Existing log analysis approaches use line-by-line message processing. They are not effective for discovering subtle behavior patterns and their transitions, and thus may overlook some critical anomalies. In this dissertation research, I propose a system log event block detection (SLEBD) method which can extract the log messages that belong to a component or system event into an event block (EB) accurately and automatically. At the event level, we can discover new event patterns, the evolution of system behavior, and the interaction among different system components. To find critical event sequences, existing sequence mining methods are mostly based on the a priori algorithm which is compute-intensive and runs for a long time. I develop a novel, topology-aware sequence mining (TSM) algorithm which is efficient to generate sequence patterns from the extracted event block lists. I also train a long short-term memory (LSTM) model to cluster sequences before specific events. With the generated sequence pattern and trained LSTM model, we can predict …

This dissertation emphasizes the security aspects of real-time media. The problems of existing real-time media protections are identified in this research, and viable solutions are proposed. First, the security of real-time media depends on the Secure Real-time Transport Protocol (SRTP) mechanism. We identified drawbacks of the existing SRTP Systems, which use symmetric key encryption schemes, which can be exploited by attackers. Elliptic Curve Cryptography (ECC), an asymmetric key cryptography scheme, is proposed to resolve these problems. Second, the ECC encryption scheme is based on elliptic curves. This dissertation explores the weaknesses of a widely used elliptic curve in terms of security and describes a more secure elliptic curve suitable for real-time media protection. Eighteen elliptic curves had been tested in a real-time video transmission system, and fifteen elliptic curves had been tested in a real-time audio transmission system. Based on the performance, X9.62 standard 256-bit prime curve, NIST-recommended 256-bit prime curves, and Brainpool 256-bit prime curves were found to be suitable for real-time audio encryption. Again, X9.62 standard 256-bit prime and 272-bit binary curves, and NIST-recommended 256-bit prime curves were found to be suitable for real-time video encryption.The weaknesses of NIST-recommended elliptic curves are discussed and a more secure new …

The solutions presented in this dissertation describe a new paradigm in which we shepherd these network security protocols through atmosphere transitions, offering new ways to analyze and monitor the state of the protocol. The approach involves identifying a protocols transitional weaknesses through adaption of formal models, measuring the weakness as it exists in the wild by statically analyzing applications, and show how to use network traffic analysis to monitor protocol implementations going into the future. Throughout the effort, we follow the popular Open Authorization protocol in its attempts to apply its web-based roots to a mobile atmosphere. To pinpoint protocol deficiencies, we first adapt a well regarded formal analysis and show it insufficient in the characterization of mobile applications, tying its transitional weaknesses to implementation issues and delivering a reanalysis of the proof. We then measure the prevalence of this weakness by statically analyzing over 11,000 Android applications. While looking through source code, we develop new methods to find sensitive protocol information, overcome hurdles like obfuscation, and provide interfaces for later modeling, all while achieving a false positive rate of below 10 percent. We then use network analysis to detect and verify application implementations. By collecting network traffic from Android …

This dissertation presents spatial partitioning algorithms to solve location-allocation problems. Location-allocations problems pertain to both the selection of facilities to serve demand at demand points and the assignment of demand points to the selected or known facilities. In the first part of this dissertation, we focus on the well known and well-researched location-allocation problem, the "p-median problem", which is a distance-based location-allocation problem that involves selection and allocation of p facilities for n demand points. We evaluate the performance of existing p-median heuristic algorithms and investigate the impact of the scale of the problem, and the spatial distribution of demand points on the performance of these algorithms. Based on the results from this comparative study, we present guidelines for location analysts to aid them in selecting the best heuristic and corresponding parameters depending on the problem at hand. Additionally, we found that existing heuristic algorithms are not suitable for solving large-scale p-median problems in a reasonable amount of time. We present a density-based decomposition methodology to solve large-scale p-median problems efficiently. This algorithm identifies dense clusters in the region and uses a MapReduce procedure to select facilities in the clustered regions independently and combine the solutions from the subproblems. Lastly, …

Resilient networks have the ability to provide the desired level of service, despite challenges such as malicious attacks and misconfigurations. The primary goal of this dissertation is to be able to provide uninterrupted network services in the face of an attack or any failures. This dissertation attempts to apply control system theory techniques with a focus on system identification and closed-loop feedback control. It explores the benefits of system identification technique in designing and validating the model for the complex and dynamic networks. Further, this dissertation focuses on designing robust feedback control mechanisms that are both scalable and effective in real-time. It focuses on employing dynamic and predictive control approaches to reduce the impact of an attack on network services. The closed-loop feedback control mechanisms tackle this issue by degrading the network services gracefully to an acceptable level and then stabilizing the network in real-time (less than 50 seconds). Employing these feedback mechanisms also provide the ability to automatically configure the settings such that the QoS metrics of the network is consistent with those specified in the service level agreements.

This research methodology isolates coding properties and identifies the probability of security vulnerabilities using machine learning and historical data. Several approaches characterize the effectiveness of detecting security-related bugs that manifest as vulnerabilities, but none utilize vulnerability patch information. The main contribution of this research is a framework to analyze LLVM Intermediate Representation Code and merging core source code representations using source code properties. This research is beneficial because it allows source programs to be transformed into a graphical form and users can extract specific code properties related to vulnerable functions. The result is an improved approach to detect, identify, and track software system vulnerabilities based on a performance evaluation. The methodology uses historical function level vulnerability information, unique feature extraction techniques, a novel code property graph, and learning algorithms to minimize the amount of end user domain knowledge necessary to detect vulnerabilities in applications. The analysis shows approximately 99% precision and recall to detect known vulnerabilities in the National Institute of Standards and Technology (NIST) Software Assurance Metrics and Tool Evaluation (SAMATE) project. Furthermore, 72% percent of the historical vulnerabilities in the OpenSSL testing environment were detected using a linear support vector classifier (SVC) model.

Software defects can frequently be traced to poorly-specified requirements. Many software teams manage their requirements using tools such as checklists and databases, which lack a formal semantic mapping to system behavior. Such a mapping can be especially helpful for safety-critical systems. Another limitation of many requirements analysis methods is that much of the analysis must still be done manually. We propose techniques that automate portions of the requirements analysis process, as well as clarify the syntax and semantics of requirements models using a variety of methods, including machine learning tools and our own tool, VeriCCM. The machine learning tools used help us identify potential model elements and verify their correctness. VeriCCM, a formalized extension of the causal component model (CCM), uses formal methods to ensure that requirements are well-formed, as well as providing the beginnings of a full formal semantics. We also explore the use of statecharts to identify potential abnormal behaviors from a given set of requirements. At each stage, we perform empirical studies to evaluate the effectiveness of our proposed approaches.

Malfunctions on loom machines are the main causes of faulty fabric production. An on-loom fabric inspection system is a real-time monitoring device that enables immediate defect detection for human intervention. This dissertation presented a solution for the on-loom fabric defect inspection, including the new hardware design—the configurable contact image sensor (CIS) module—for on-loom fabric scanning and the defect detection algorithms. The main contributions of this work include (1) creating a configurable CIS module adaptable to a loom width, which brings CIS unique features, such as sub-millimeter resolution, compact size, short working distance and low cost, to the fabric defect inspection system, (2) designing a two-level hardware architecture that can be efficiently deployed in a weaving factory with hundreds of looms, (3) developing a two-level inspecting scheme, with which the initial defect screening is performed on the Raspberry Pi and the intensive defect verification is processed on the cloud server, (4) introducing the novel pairwise-potential activation layer to a convolutional neural network that leads to high accuracies of defect segmentation on fabrics with fine and imbalanced structures, (5) achieving a real-time defect detection that allows a possible defect to be examined multiple times, and (6) implementing a new color segmentation technique …

A malicious actor often relies on security vulnerabilities of IT systems to launch a cyber attack. Most cloud services are supported by an orchestration of large and complex systems which are prone to vulnerabilities, making threat assessment very challenging. In this research, I developed formal and practical ontology-based techniques that enable automated evaluation of a cloud system's security threats. I use an architecture for threat assessment of cloud systems that leverages a dynamically generated ontology knowledge base. I created an ontology model and represented the components of a cloud system. These ontologies are designed for a set of domains that covers some cloud's aspects and information technology products' cyber threat data. The inputs to our architecture are the configurations of cloud assets and components specification (which encompass the desired assessment procedures) and the outputs are actionable threat assessment results. The focus of this work is on ways of enumerating, assessing, and mitigating emerging cyber security threats. A research toolkit system has been developed to evaluate our architecture. We expect our techniques to be leveraged by any cloud provider or consumer in closing the gap of identifying and remediating known or impending security threats facing their cloud's assets.