Biblio

Website phishing continues to persist as one of the most important security threats of the modern Internet era. A major concern has been that machine learning based approaches, which have been the cornerstones of deployed phishing detection solutions, have not been able to adapt to the evolving nature of the phishing attacks. To create updated machine learning models, the collection of a sufficient corpus of real-time phishing data has always been a challenging problem as most phishing websites are short-lived. In this work, for the first time, we address these important concerns and describe an adaptive phishing detection solution that is able to adapt to changes in phishing attacks. Our solution has two major contributions. First, our solution allows for multiple organizations to collaborate in a privacy preserving manner and generate a robust machine learning model for phishing detection. Second, our solution is designed to be flexible in order to adapt to the novel phishing features introduced by attackers. Our solution not only allows for incorporating novel features into the existing machine learning model, but also can help, to a certain extent, the “unlearning” of existing features that have become obsolete in current phishing attacks. We evaluated our approach on a large real-world data collected over a period of six months. Our results achieve a high true positive rate of 97 %, which is on par with existing state-of-the art centralized solutions. Importantly, our results demonstrate that, a machine learning model can incorporate new features while selectively “unlearning” the older obsolete features.

Logic locking is a holistic design-for-trust technique that aims to protect the design intellectual property (IP) from untrustworthy entities throughout the supply chain. Functional and structural analysis-based attacks successfully circumvent state-of-the-art, provably secure logic locking (PSLL) techniques. However, such attacks are not holistic and target specific implementations of PSLL. Automating the detection and subsequent removal of protection logic added by PSLL while accounting for all possible variations is an open research problem. In this paper, we propose GNNUnlock, the first-of-its-kind oracle-less machine learning-based attack on PSLL that can identify any desired protection logic without focusing on a specific syntactic topology. The key is to leverage a well-trained graph neural network (GNN) to identify all the gates in a given locked netlist that belong to the targeted protection logic, without requiring an oracle. This approach fits perfectly with the targeted problem since a circuit is a graph with an inherent structure and the protection logic is a sub-graph of nodes (gates) with specific and common characteristics. GNNs are powerful in capturing the nodes' neighborhood properties, facilitating the detection of the protection logic. To rectify any misclassifications induced by the GNN, we additionally propose a connectivity analysis-based post-processing algorithm to successfully remove the predicted protection logic, thereby retrieving the original design. Our extensive experimental evaluation demonstrates that GNNUnlock is 99.24% - 100% successful in breaking various benchmarks locked using stripped-functionality logic locking [1], tenacious and traceless logic locking [2], and Anti-SAT [3]. Our proposed post-processing enhances the detection accuracy, reaching 100% for all of our tested locked benchmarks. Analysis of the results corroborates that GNNUnlock is powerful enough to break the considered schemes under different parameters, synthesis settings, and technology nodes. The evaluation further shows that GNNUnlock successfully breaks corner cases where even the most advanced state-of-the-art attacks [4], [5] fail. We also open source our attack framework [6].

Internet of Things (IoT) is defined as the connection between places and physical objects (i.e., things) over the internet/network via smart computing devices. IoT is a rapidly emerging paradigm that now encompasses almost every aspect of our modern life. As such, it is crucial to ensure IoT devices follow strict security requirements. At the same time, the prevalence of IoT devices offers developers a chance to design and develop Machine Learning (ML)-based intelligent software systems using their IoT devices. However, given the diversity of IoT devices, IoT developers may find it challenging to introduce appropriate security and ML techniques into their devices. Traditionally, we learn about the IoT ecosystem/problems by conducting surveys of IoT developers/practitioners. Another way to learn is by analyzing IoT developer discussions in popular online developer forums like Stack Overflow (SO). However, we are aware of no such studies that focused on IoT developers’ security and ML-related discussions in SO. This paper offers the results of preliminary study of IoT developer discussions in SO. First, we collect around 53K IoT posts (questions + accepted answers) from SO. Second, we tokenize each post into sentences. Third, we automatically identify sentences containing security and ML-related discussions. We find around 12% of sentences contain security discussions, while around 0.12% sentences contain ML-related discussions. There is no overlap between security and ML-related discussions, i.e., IoT developers discussing security requirements did not discuss ML requirements and vice versa. We find that IoT developers discussing security issues frequently inquired about how the shared data can be stored, shared, and transferred securely across IoT devices and users. We also find that IoT developers are interested to adopt deep neural network-based ML models into their IoT devices, but they find it challenging to accommodate those into their resource-constrained IoT devices. Our findings offer implications for IoT vendors and researchers to develop and design novel techniques for improved security and ML adoption into IoT devices.

The intelligent cyber-physical system (CPS) has been applied in various fields, covering multiple critical infras-tructures and human daily life support areas. CPS Security is a major concern and of critical importance, especially the security of the intelligent control component. Side-channel analysis (SCA) is the common threat exploiting the weaknesses in system operation to extract information of the intelligent CPS. However, existing literature lacks the systematic theo-retical analysis of the side-channel attacks on the intelligent CPS, without the ability to quantify and measure the leaked information. To address these issues, we propose the SCA-based model extraction attack on intelligent CPS. First, we design an efficient and novel SCA-based model extraction framework, including the threat model, hierarchical attack process, and the multiple micro-space parallel search enabled weight extraction algorithm. Secondly, an information theory-empowered analy-sis model for side-channel attacks on intelligent CPS is built. We propose a mutual information-based quantification method and derive the capacity of side-channel attacks on intelligent CPS, formulating the amount of information leakage through side channels. Thirdly, we develop the theoretical bounds of the leaked information over multiple attack queries based on the data processing inequality and properties of entropy. These convergence bounds provide theoretical means to estimate the amount of information leaked. Finally, experimental evaluation, including real-world experiments, demonstrates the effective-ness of the proposed SCA-based model extraction algorithm and the information theory-based analysis method in intelligent CPS.

Industrial Control Systems (ICSs) are used in several domains such as Transportation, Manufacturing, Defense and Power Generation and Distribution. ICSs deal with complex physical systems in order to achieve an industrial purpose with operational safety. Security has not been taken into account by design in these systems that makes them vulnerable to cyberattacks.In this paper, we rely on existing public ICS datasets as well as on the existing literature of Machine Learning (ML) applications for anomaly detection in ICSs in order to improve detection scores. To perform this purpose, we propose a systematic framework, relying on established ML algorithms and suitable data preprocessing methods, which allows us to quickly get efficient, and surprisingly, better results than the literature. Finally, some recommendations for future public ICS dataset generations end this paper, which would be fruitful for improving future attack detection models and then protect new ICSs designed in the next future.

The number of organisational cybersecurity threats continues to increase every year as technology advances. All too often, organisations assume that implementing systems security measures like firewalls and anti-virus software will eradicate cyber threats. However, even the most robust security systems are vulnerable to threats. As advanced as machine learning cybersecurity technology is becoming, it cannot be solely relied upon to solve cyber threats. There are other forces that contribute to these threats that are many-a-times out of an organisation's control i.e., human behaviour. This research article aims to create an understanding of the trends in key cybersecurity management issues that have developed in the past five years in relation to human behaviour and machine learning. The methodology adopted to guide the synthesis of this review was a systematic literature review. The guidelines for conducting the review are presented in the review approach. The key cybersecurity management issues highlighted by the research includes risky security behaviours demonstrated by employees, social engineering, the current limitations present in machine learning insider threat detection, machine learning enhanced cyber threats, and the underinvestment challenges faced in the cybersecurity domain.

The critical nature of smart grids (SGs) attracts various network attacks and malicious manipulations. Existent SG solutions are less capable of ensuring secure and trustworthy operation. This is due to the large-scale nature of SGs and reliance on network protocols for trust management. A particular example of such severe attacks is the false data injection (FDI). FDI refers to a network attack, where meters' measurements are manipulated before being reported in such a way that the energy system takes flawed decisions. In this paper, we exploit the secure nature of blockchains to construct a data management framework based on public blockchain. Our framework enables trustworthy data storage, verification, and exchange between SG components and decision-makers. Our proposed system enables miners to invest their computational power to verify blockchain transactions in a fully distributed manner. The mining logic employs machine learning (ML) techniques to identify the locations of compromised meters in the network, which are responsible for generating FDI attacks. In return, miners receive virtual credit, which may be used to pay their electric bills. Our design circumvents single points of failure and intentional FDI attempts. Our numerical results compare the accuracy of three different ML-based mining logic techniques in two scenarios: focused and distributed FDI attacks for different attack levels. Finally, we proposed a majority-decision mining technique for the practical case of an unknown FDI attack level.

Malware detection is an important area of cyber security. Computer systems rely on malware detection applications to prevent malware attacks from succeeding. Malware detection is not a straightforward task, as new variants of malware are generated at an increasing rate. Machine learning (ML) has been utilised to generate predictive classification models to identify new malware variants which conventional malware detection methods may not detect. Machine learning, has however, been found to be vulnerable to different types of adversarial attacks, in which an attacker is able to negatively affect the classification ability of the ML model. Several defensive measures to prevent adversarial poisoning attacks have been developed, but they often rely on the use of a trusted clean dataset to help identify and remove adversarial examples from the training dataset. The defence in this paper does not require a trusted clean dataset, but instead, identifies intentional false negatives (zero day malware classified as benign) at the testing stage by examining the activation weights of the ML model. The defence was able to identify 94.07% of the successful targeted poisoning attacks.

The presence of experts, such as a data protection officer (DPO) and a privacy engineer is essential in Privacy Requirements Engineering. This task is carried out in various forms including threat modeling and privacy impact assessment. The knowledge required for performing privacy threat modeling can be a serious challenge for a novice privacy engineer. We aim to bridge this gap by developing an automated approach via machine learning that is able to detect privacy-related entities in the user stories. The relevant entities include (1) the Data Subject, (2) the Processing, and (3) the Personal Data entities. We use a state-of-the-art Named Entity Recognition (NER) model along with contextual embedding techniques. We argue that an automated approach can assist agile teams in performing privacy requirements engineering techniques such as threat modeling, which requires a holistic understanding of how personally identifiable information is used in a system. In comparison to other domain-specific NER models, our approach achieves a reasonably good performance in terms of precision and recall.

Differential privacy enables general statistical analysis of data with formal guarantees of privacy protection at the individual level. Tools that assist data analysts with utilizing differential privacy have frequently taken the form of programming languages and libraries. However, many existing programming languages designed for compositional verification of differential privacy impose significant burden on the programmer (in the form of complex type annotations). Supplementary library support for privacy analysis built on top of existing general-purpose languages has been more usable, but incapable of pervasive end-to-end enforcement of sensitivity analysis and privacy composition. We introduce DDuo, a dynamic analysis for enforcing differential privacy. DDuo is usable by non-experts: its analysis is automatic and it requires no additional type annotations. DDuo can be implemented as a library for existing programming languages; we present a reference implementation in Python which features moderate runtime overheads on realistic workloads. We include support for several data types, distance metrics and operations which are commonly used in modern machine learning programs. We also provide initial support for tracking the sensitivity of data transformations in popular Python libraries for data analysis. We formalize the novel core of the DDuo system and prove it sound for sensitivity analysis via a logical relation for metric preservation. We also illustrate DDuo's usability and flexibility through various case studies which implement state-of-the-art machine learning algorithms.

PCIe (Peripheral Component Interconnect express) protocol is the de facto protocol to bridge CPU and peripheral devices like GPU, NIC, and SSD drive. There is an increasing demand to install more peripheral devices on a single machine, but the PCIe interfaces offered by Intel CPUs are fixed. To resolve such contention, PCIe switch, PCH (Platform Controller Hub), or virtualization cards are installed on the machine to allow multiple devices to share a PCIe interface. Congestion happens when the collective PCIe traffic from the devices overwhelm the PCIe link capacity, and transmission delay is then introduced.In this work, we found the PCIe delay not only harms device performance but also leaks sensitive information about a user who uses the machine. In particular, as user’s activities might trigger data movement over PCIe (e.g., between CPU and GPU), by measuring PCIe congestion, an adversary accessing another device can infer the victim’s secret indirectly. Therefore, the delay resulted from I/O congestion can be exploited as a side-channel. We demonstrate the threat from PCIe congestion through 2 attack scenarios and 4 victim settings. Specifically, an attacker can learn the workload of a GPU in a remote server by probing a RDMA NIC that shares the same PCIe switch and measuring the delays. Based on the measurement, the attacker is able to know the keystroke timings of the victim, what webpage is rendered on the GPU, and what machine-learning model is running on the GPU. Besides, when the victim is using a low-speed device, e.g., an Ethernet NIC, an attacker controlling an NVMe SSD can launch a similar attack when they share a PCH or virtualization card. The evaluation result shows our attack can achieve high accuracy (e.g., 96.31% accuracy in inferring webpage visited by a victim).

A text is a meaningful source of information. Capturing the right patterns in written text gives metrics to measure and infer to what extent this text belongs or is relevant to a specific author. This research aims to introduce a new feature that goes more in deep in the language structure. The feature introduced is based on an attempt to differentiate stylistic changes among authors according to the different sentence structure each author uses. The study showed the effect of introducing this new feature to machine learning models to enhance their performance. It was found that the prediction of authors was enhanced by adding sentence structure as an additional feature as the f1\_scores increased by 0.3% and when normalizing the data and adding the feature it increased by 5%.

Social bots are computer algorithms able to produce content and interact with other users on social media autonomously, trying to emulate and possibly influence humans’ behavior. Indeed, bots are largely employed for malicious purposes, like spreading disinformation and conditioning electoral campaigns. Nowadays, bots’ capability of emulating human behaviors has become increasingly sophisticated, making their detection harder. In this paper, we aim at recognizing bot-driven accounts by evaluating the consistency of users’ writing style over time. In particular, we leverage the intuition that while bots compose posts according to fairly deterministic processes, humans are influenced by subjective factors (e.g., emotions) that can alter their writing style. To verify this assumption, by using stylistic consistency indicators, we characterize the writing style of more than 12,000 among bot-driven and human-operated Twitter accounts and find that statistically significant differences can be observed between the different types of users. Thus, we evaluate the effectiveness of different machine learning (ML) algorithms based on stylistic consistency features in discerning between human-operated and bot-driven Twitter accounts and show that the experimented ML algorithms can achieve high performance (i.e., F-measure values up to 98%) in social bot detection tasks.

Code authorship identification can assist in identifying creators of malware, identifying plagiarism, and giving insights in copyright infringement cases. Taking inspiration from facial recognition work, we apply recent advances in metric learning to the problem of authorship identification and verification. The metric learning approach makes it possible to measure similarity in the learned embedding space. Access to a discriminative similarity measure allows for the estimation of probability distributions that facilitate open-set classification and verification. We extend our analysis to verification based on sets of files, a previously unexplored problem domain in large-scale author identification. On closed-set tasks we achieve competitive accuracies, but do not improve on the state of the art.

The problem of authorship attribution has applications from literary studies (such as the great Shakespeare/Marlowe debates) to counter-intelligence. The field of stylometry aims to offer quantitative results for authorship attribution. In this paper, we present a combination of stylometric techniques using machine learning. An implementation of the system is used to analyse chat logs and attempts to construct a stylometric model for users within the presented chat system. This allows for the authorship attribution of other works they may write under different names or within different communication systems. This implementation demonstrates accuracy of up to 84 % across the dataset, a full 34 % increase against a random-choice control baseline.

Identification of source code authorship is vital for attribution. In this work, a machine learning framework is described to identify source code authorship. The framework integrates the features extracted using natural language processing based approaches and abstract syntax tree of the code. We evaluate the methodology on Google Code Jam dataset. We present the performance measures of the logistic regression and deep learning on the dataset.

Incast traffic is a many-to-one communication pattern used in many applications, including distributed storage, web-search with partition/aggregation design pattern, and MapReduce, commonly in data centers. It is generally composed of short-lived flows that may be queued behind large flows' packets in congested switches where performance degradation is observed. Smart buffering at the switch level is sensed to mitigate this issue by automatically and dynamically adapting to traffic conditions changes in the highly dynamic data center environment. But for this dynamic and smart buffer management to become effectively beneficial for all the traffic, and especially for incast the most critical one, incast performance models that provide insights on how various factors affect it are needed. The literature lacks these types of models. The existing ones are analytical models, which are either tightly coupled with a particular protocol version or specific to certain empirical data. Motivated by this observation, we propose a machine-learning-based incast performance inference. With this prediction capability, smart buffering scheme or other QoS optimization algorithms could anticipate and efficiently optimize system parameters adjustment to achieve optimal performance. Since applying machine learning to networks managed in a distributed fashion is hard, the prediction mechanism will be deployed on an SDN control plane. We could then take advantage of SDN's centralized global view, its telemetry capabilities, and its management flexibility.

Background: Most of the people are not medically qualified for studying or understanding the extremity of their diseases or symptoms. This is the place where natural language processing plays a vital role in healthcare. These chatbots collect patients' health data and depending on the data, these chatbot give more relevant data to patients regarding their body conditions and recommending further steps also. Purposes: In the medical field, AI powered healthcare chatbots are beneficial for assisting patients and guiding them in getting the most relevant assistance. Chatbots are more useful for online search that users or patients go through when patients want to know for their health symptoms. Methods: In this study, the health assistant system was developed using Dialogflow application programming interface (API) which is a Google's Natural language processing powered algorithm and the same is deployed on google assistant, telegram, slack, Facebook messenger, and website and mobile app. With this web application, a user can make health requests/queries via text message and might also get relevant health suggestions/recommendations through it. Results: This chatbot acts like an informative and conversational chatbot. This chatbot provides medical knowledge such as disease symptoms and treatments. Storing patients personal and medical information in a database for further analysis of the patients and patients get real time suggestions from doctors. Conclusion: In the healthcare sector AI-powered applications have seen a remarkable spike in recent days. This covid crisis changed the whole healthcare system upside down. So this NLP powered chatbot system reduced office waiting, saving money, time and energy. Patients might be getting medical knowledge and assisting ourselves within their own time and place.

Graphs have proved to be a promising data structure to solve complex problems in various domains. Graphs store data in an associative manner which is analogous to the manner in which humans store memories in the brain. Generathe chatbots lack the ability to recall details revealed by the user in long conversations. To solve this problem, we have used graph-based memory to recall-related conversations from the past. Thus, providing context feature derived from query systems to generative systems such as OpenAI GPT. Using graphs to detect important details from the past reduces the total amount of processing done by the neural network. As there is no need to keep on passingthe entire history of the conversation. Instead, we pass only the last few pairs of utterances and the related details from the graph. This paper deploys this system and also demonstrates the ability to deploy such systems in real-world applications. Through the effective usage of knowledge graphs, the system is able to reduce the time complexity from O(n) to O(1) as compared to similar non-graph based implementations of transfer learning- based conversational agents.

The CAN (Controller Area Network) bus, which transmits and receives ECU control information in vehicle, has a critical risk of external intrusion because there is no standardized security system. Recently, the need for IDS (Intrusion Detection System) to detect external intrusion of CAN bus is increasing, and high accuracy and real-time processing for intrusion detection are required. In this paper, we propose Hybrid MR (Machine learning and Ruleset) -IDS based on machine learning and ruleset to improve IDS performance. For high accuracy and detection rate, feature engineering was conducted based on the characteristics of the CAN bus, and the generated features were used in detection step. The proposed Hybrid MR-IDS can cope to various attack patterns that have not been learned in previous, as well as the learned attack patterns by using both advantages of rule set and machine learning. In addition, by collecting CAN data from an actual vehicle in driving and stop state, five attack scenarios including physical effects during all driving cycle are generated. Finally, the Hybrid MR-IDS proposed in this paper shows an average of 99% performance based on F1-score.

There are problems in which the use of linear regression is not sufficiently justified. In these cases, fuzzy linear regression can be used as a modeling tool. The problem of constructing a fuzzy linear regression can usually be reduced to a linear programming problem. One of the features of the resulting linear programming problem is that it uses a relatively large number of constraints in the form of inequalities with a relatively small number of variables. It is known that the problem of constructing a fuzzy linear regression is reduced to the problem of linear programming. If the user does not have enough computing power the resulting problem can be transferred to the cloud server. Two approaches are used for the confidential transfer of the problem to the server: the approach based on cryptographic encryption, and the transformational approach. The paper describes a protocol based on the transformational approach that allows for secure outsourcing of fuzzy linear regression.

Making use of spectrally diverse communications links to re-route traffic in response to dynamic environments to manage network bottlenecks has become essential in order to guarantee message delivery across heterogeneous networks. We propose an innovative, proactive Congestion Aware Intent-Based Routing (CONAIR) architecture that can select among available communication link resources based on quality of service (QoS) metrics to support continuous information exchange between networked participants. The CONAIR architecture utilizes a Network Controller (NC) and artificial intelligence (AI) to re-route traffic based on traffic priority, fundamental to increasing end user quality of experience (QoE) and mission effectiveness. The CONAIR architecture provides network behavior prediction, and can mitigate congestion prior to its occurrence unlike traditional static routing techniques, e.g. Open Shortest Path First (OSPF), which are prone to congestion due to infrequent routing table updates. Modeling and simulation (M&S) was performed on a multi-hop network in order to characterize the resiliency and scalability benefits of CONAIR over OSPF routing-based frameworks. Results demonstrate that for varying traffic profiles, packet loss and end-to-end latency is minimized.

Phasor measurement units (PMUs) are used in power grids across North America to measure the amplitude, phase, and frequency of an alternating voltage or current. PMU's use the IEEE C37.118 protocol to send telemetry to phasor data collectors (PDC) and human machine interface (HMI) workstations in a control center. However, the C37.118 protocol utilizes the internet protocol stack without any authentication mechanism. This means that the protocol is vulnerable to false data injection (FDI) and false command injection (FCI). In order to study different scenarios in which C37.118 protocol's integrity and confidentiality can be compromised, we created a testbed that emulates a C37.118 communication network. In this testbed we conduct FCI and FDI attacks on real-time C37.118 data packets using a packet manipulation tool called Scapy. Using this platform, we generated C37.118 FCI and FDI datasets which are processed by multi-label machine learning classifier algorithms, such as Decision Tree (DT), k-Nearest Neighbor (kNN), and Naive Bayes (NB), to find out how effective machine learning can be at detecting such attacks. Our results show that the DT classifier had the best precision and recall rate.

Power grids are the most extensive man-made systems that are difficult to control and monitor. With the development of conventional power grids and moving toward smart grids, power systems have undergone vast changes since they use the Internet to transmit information and control commands to different parts of the power system. Due to the use of the Internet as a basic infrastructure for smart grids, attackers can sabotage the communication networks and alter the measurements. Due to the complexity of the smart grids, it is difficult for the network operator to detect such cyber-attacks. The attackers can implement the attack in a manner that conventional Bad Data detection (BDD) systems cannot detect since it may not violate the physical laws of the power system. This paper uses the cross wavelet transform (XWT) to detect stealth false data injections attacks (FDIAs) against state estimation (SE) systems. XWT can capture the coherency between measurements of adjacent buses and represent it in time and frequency space. Then, we train a machine learning classification algorithm to distinguish attacked measurements from normal measurements by applying a feature extraction technique.

Generative Adversarial Networks (GANs) are machine learning methods that are used in many important and novel applications. For example, in imaging science, GANs are effectively utilized in generating image datasets, photographs of human faces, image and video captioning, image-to-image translation, text-to-image translation, video prediction, and 3D object generation to name a few. In this paper, we discuss how GANs can be used to create an artificial world. More specifically, we discuss how GANs help to describe an image utilizing image/video captioning methods and how to translate the image to a new image using image-to-image translation frameworks in a theme we desire. We articulate how GANs impact creating a customized world.