Search

The ever-increasing popularity of recommendation systems allows users to find appropriate services without excessive effort. However, due to the unstable and complex network environment, the historical behavior data of users are quite sparse in most cases. The inherent drawbacks render preference prediction infeasible for cold-start users and have become a crucial issue to be resolved in recommendation systems. To deal with the problems, we first present a Trust-based Collaborative Filtering (TbCF) algorithm to perform basic rating prediction in a manner consistent with the existing CF methods. Then, we propose the Hybrid Collaborative Filtering Recommendation approach with User-Item-Trust Records (UIThybrid), a novel approach that incorporates user trust into the existing CF-based methods in a harmonious way to supplement rating information. UIThybrid employs multiple perspectives to extract proper services and achieves a good tradeoff between the robustness, accuracy, and diversity of the recommendation. We conduct extensive real-world experiments on the Epinions data set to demonstrate the feasibility and efficiency of UIThybrid.

Many research areas depend on group anomaly detection. The use of group anomaly detection can maintain and provide security and privacy to the data involved. This research attempts to solve the deficiency of the existing literature in outlier detection thus a novel hybrid framework to identify group anomaly detection from sequence data is proposed in this paper. It proposes two approaches for efficiently solving this problem: i) Hybrid Data Mining-based algorithm, consists of three main phases: first, the clustering algorithm is applied to derive the micro-clusters. Second, the kNN algorithm is applied to each micro-cluster to calculate the candidates of the group's outliers. Third, a pattern mining framework gets applied to the candidates of the group's outliers as a pruning strategy, to generate the groups of outliers, and ii) a GPU-based approach is presented, which benefits from the massively GPU computing to boost the runtime of the hybrid data mining-based algorithm. Extensive experiments were conducted to show the advantages of different sequence databases of our proposed model. Results clearly show the efficiency of a GPU direction when directly compared to a sequential approach by reaching a speedup of 451. In addition, both approaches outperform the baseline methods for group detection.

With the increasing demand of users for personalized social services, social recommendation (SR) has been an important concern in academia. However, current research on SR universally faces two main challenges. On the one hand, SR lacks the considerable ability of robust online data management. On the other hand, SR fails to take the ambiguity of preference feedback into consideration. To bridge these gaps, a deep learning-embedded social Internet of Things (IoT) is proposed for ambiguity-aware SR (SIoT-SR). Specifically, a social IoT architecture is developed for social computing scenarios to guarantee reliable data management. A deep learning-based graph neural network model that can be embedded into the model is proposed as the core algorithm to perform ambiguity-aware SR. This design not only provides proper online data sensing and management but also overcomes the preference ambiguity problem in SR. To evaluate the performance of the proposed SIoT-SR, two real-world datasets are selected to establish experimental scenarios. The method is assessed using three different metrics, selecting five typical methods as benchmarks. The experimental results show that the proposed SIoT-SR performs better than the benchmark methods by at least 10% and has good robustness.

The Internet of Health Things (IoHT) is a concept that describes uniquely identifiable devices connected to the Internet and able to communicate with each other in the medical area. As one of the most important components of smart health monitoring and improvement systems, there are numerous challenges in the IoHT, among which cybersecurity is a major challenge that must be addressed with priority. As a well-received security solution to achieve fine-grained access control, ciphertext-policy weighted attribute-based encryption (CP-WABE) has the potential to ensure data security in the IoHT. However, many issues, such as inflexibility, poor computational capability, and insufficient storage efficiency in attributes comparison, remain. To address these issues, we propose a novel access policy expression method using 0-1 coding technology. Based on this method, a flexible and efficient CP-WABE is constructed for the IoHT. Our scheme supports not only weighted attributes but also any form of comparison of weighted attributes. Furthermore, we use offline/online encryption and outsourced decryption technology to ensure that the scheme can run on an inefficient IoT terminal. Both theoretical and experimental analyses show that our scheme is more efficient and feasible than other schemes. Moreover, security analysis indicates that our scheme achieves security against chosen-plaintext attack.

The Internet of Things (IoT) is made up of billions of physical devices connected to the Internet via networks that perform tasks independently with less human intervention. Such brilliant automation of mundane tasks requires a considerable amount of user data in digital format, which in turn makes IoT networks an open-source of Personally Identifiable Information data for malicious attackers to steal, manipulate and perform nefarious activities. Huge interest has developed over the past years in applying machine learning (ML)-assisted approaches in the IoT security space. However, the assumption in many current works is that big training data is widely available and transferable to the main server because data is born at the edge and is generated continuously by IoT devices. This is to say that classic ML works on the legacy set of entire data located on a central server, which makes it the least preferred option for domains with privacy concerns on user data. To address this issue, we propose federated learning (FL)-based anomaly detection approach to proactively recognize intrusion in IoT networks using decentralized on-device data. Our approach uses federated training rounds on Gated Recurrent Units (GRUs) models and keeps the data intact on local IoT devices by sharing only the learned weights with the central server of the FL. Also, the approach’s ensembler part aggregates the updates from multiple sources to optimize the global ML model’s accuracy. Our experimental results demonstrate that our approach outperforms the classic/centralized machine learning (non-FL) versions in securing the privacy of user data and provides an optimal accuracy rate in attack detection.

The Internet of Things (IoT) and smart cities are flourishing with distributed systems in mobile wireless networks. As a result, an enormous amount of data are being generated for devices at the network edge. This results in privacy concerns, sensor data management issues, and data utilization issues. In this research, we propose a collaborative clustering method where the exchange of raw data is not required. The attention-based model used with a federated learning framework. The edge devices compute the model updates using local data and send them to the server for aggregation. Repetition is performed in multiple rounds until a convergence point reached. The transaction data used to train the attention model that gives a low dimensional embedding. Afterwards, we share the convergence model among the client/stores. Then, efficient pattern mining methods known as a clustering-based dynamic method (CBDM) are applied. For experimentation, we used retail store data to cluster the customer based on purchase behaviour. The proposed clustering method used semantic embedding to extract and then cluster them by discovering relevant patterns. The method achieved the 0.75 ROC values for the random distribution and 0.70 for the fixed distribution. The clustering method can help to reduce communication costs while ensuring privacy.

High-Utility Itemset Mining (HUIM) is considered a major issue in recent decades since it reveals profit strategies for use in industry for decision-making. Most existing works have focused on mining high-utility itemsets from databases showing large amount of patterns; however exact decisions are still challenging to make from that large amounts of discovered knowledge. Closed High-utility itemset mining (CHUIM) provides a smart way to present concise high-utility itemsets that can be more effective for making correct decisions. However, none of the existing works have focused on handling large-scale databases to integrate discovered knowledge from several distributed databases. In this paper, we first present a large-scale information fusion architecture to integrate discovered closed high-utility patterns from several distributed databases. The generic composite model is used to cluster transactions regarding their relevant correlation that can ensure correctness and completeness of the fusion model. The well-known MapReduce framework is then deployed in the developed DFM-Miner algorithm to handle big datasets for information fusion and integration. Experiments are then compared to the state-of-the-art CHUI-Miner and CLS-Miner algorithms for mining closed high-utility patterns and the results indicated that the designed model is well designed for handling large-scale databases with less memory usage. Moreover, the designed MapReduce framework can speed up the mining performance of closed high-utility patterns in the developed fusion system.

This paper develops a new framework called MASAD (Multi-Agents System for Anomaly Detection), a hybrid combination of reinforcement learning, and a multi-agents system to identify abnormal behaviors of microservices in industrial environment settings. A multi-agent system is implemented using reinforcement learning, where each agent learns from the given microservice. Intelligent communication among the different agents is then established to enhance the learning of each agent by considering the experience of the agents of the other microservices of the system. The above setting not only allows to identify local anomalies but global ones from the whole microservices architecture. To show the effectiveness of the framework as proposed, we have gone through a thorough experimental analysis on two microservice architectures (NETFLIX, and LAMP). Results showed that our proposed framework can understand the behavior of microservices, and accurately simulate different interactions in the microservices. Besides, our approach outperforms baseline methods in identifying both local and global outliers.

The concepts of sequential pattern mining have become a growing topic in data mining, finding a home most recently in the Internet of Things (IoT) where large volumes of data are presented by the second for analysis and knowledge extraction. One key topic within the realm of sequential pattern mining in high-utility sequential pattern mining (HUSPM), short form for high-utility sequential pattern mining. HUSPM takes into account the fusion of utility and sequence factors to assist in the determination of sequential patterns of high utility from databases and data sources. That being said, almost all current existing literature focus on only using a single machine to increase mining performance. In this work, we present a four-stage MapReduce framework that is solely based on the well-known Spark platform for use in HUSPM. This framework is shown to create a more efficient and faster mining performance for dealing with large data sets. It consists of four phases such as initialization, mining, updating, and generation phases to handle the big data sets based on the MapReduce framework running on the Spark platform. Experiments indicated that the designed model is capable of handling the very big data sets while state-of-the-art algorithms can only achieve good performance in small data sets.

As the increasing availability and use of dynamic mobile communications, information from an Internet of Things (IoT) subset of devices, known as Internet of Connected Vehicles (IoCV), is collected with a level of uncertainty. To bridge this gap of data analytics, some studies take two factors individually to mine knowledge or information, such as uncertainty and utility as two exemplary factors. However, this approach may cause actual loss of knowledge integrity. In this work, our first result is a knowledge called H igh E xpected U tility S equential P atterns (HEUSPs) that is both novel and also provides an alternative option for knowledge discovery regarding utility and uncertainty factors by a single threshold in IoCV environments. Furthermore, two PUL-Chain and EUL-Chain structures with six pruning methodologies are respectively developed to maintain information that is necessary and reduce the search space for improving mining performance. Our experimental results show both efficiency and strength of the designed algorithm compared to HUS-Span which is considered to be the current standard in utility-oriented sequential pattern mining.

This is the author's version of the work. It is posted here for your personal use. Not for redistribution. The definitive Version of Record was published in ACM Transactions on Internet Technology, Volume 21, Issue 4, https://doi.org/10.1145/3423557. In the ever-growing world, the concepts of High-utility Itemset Mining (HUIM) as well as Frequent Itemset Mining (FIM) are fundamental works in knowledge discovery. Several algorithms have been designed successfully. However, these algorithms only used one factor to estimate an itemset. In the past, skyline pattern mining by considering both aspects of frequency and utility has been extensively discussed. In most cases, however, people tend to focus on purchase quantities of itemsets rather than frequencies. In this article, we propose a new knowledge called skyline quantity-utility pattern (SQUP) to provide better estimations in the decision-making process by considering quantity and utility together. Two algorithms, respectively, called SQU-Miner and SKYQUP are presented to efficiently mine the set of SQUPs. Moreover, the usage of volunteer computing is proposed to show the potential in real supermarket applications. Two new efficient utility-max structures are also mentioned for the reduction of the candidate itemsets, respectively, utilized in SQU-Miner and SKYQUP. These two new utility-max structures are used to store the upper-bound of utility for itemsets under the quantity constraint instead of frequency constraint, and the second proposed utility-max structure moreover applies a recursive updated process to further obtain strict upper-bound of utility. Our in-depth experimental results prove that SKYQUP has stronger performance when a comparison is made to SQU-Miner in terms of memory usage, runtime, and the number of candidates.

Within the current transportation infrastructure, we have seen a steady increase in the use of sensor technologies. These sensors, individually produce large amounts of data that then need to be fused and understood. Data commingling and data integration are difficult tasks when it comes to processing such data centrally, which can require costly hardware and software techniques. Over the past few years, high-utility pattern mining (HUPM) has gained popularity due to its growing capability in identifying useful information and knowledge from stored database data, as compared to the traditional frequent pattern mining. Existing works of HUPM mostly focus on mining the set of HUPs from one data source, which cannot be implemented in real-world scenarios. In this paper, we present a pre-large weighted high-utility pattern (PWHUP) fusion framework for integrating HUPs from different sensed data sources. The proposed PWHUP algorithm considers the size of the data source for discovering more relevant HUPs for integration, which is more applicable to real-life applications and scenarios in transportation and also within other data fusion scenarios. Moreover, the pre-large concept is applied to maintain the suggested pattern for later integration, which greatly improves the effectiveness of the proposed algorithm. Our in-depth experiments show that the designed approach has good performance for knowledge integration and outperforms existing non-integration solutions in precision, recall, and runtime.

With the evolution of the Internet of Things (IoT), smart cities have become the mainstream of urbanization. IoT networks allow distributed smart devices to collect and process data within smart city infrastructure using an open channel, the Internet. Thus, challenges such as centralization, security, privacy (e.g., performing data poisoning and inference attacks), transparency, scalability, and verifiability limits faster adaptations of smart cities. Motivated by the aforementioned discussions, we present a Privacy-Preserving and Secure Framework (PPSF) for IoT-driven smart cities. The proposed PPSF is based on two key mechanisms: a two-level privacy scheme and an intrusion detection scheme. First, in a two-level privacy scheme, a blockchain module is designed to securely transmit the IoT data and Principal Component Analysis (PCA) technique is applied to transform raw IoT information into a new shape. In the intrusion detection scheme, a Gradient Boosting Anomaly Detector (GBAD) is applied for training and evaluating the proposed two-level privacy scheme based on two IoT network datasets, namely ToN-IoT and BoT-IoT. We also suggest a blockchain-InterPlanetary File System (IPFS) integrated Fog-Cloud architecture to deploy the proposed PPSF framework. Experimental results demonstrate the superiority of the PPSF framework over some recent approaches in blockchain and non-blockchain systems.

In this digital age, human dependency on technology in various fields has been increasing tremendously. Torrential amounts of different electronic products are being manufactured daily for everyday use. With this advancement in the world of Internet technology, cybersecurity of software and hardware systems are now prerequisites for major business’ operations. Every technology on the market has multiple vulnerabilities that are exploited by hackers and cyber-criminals daily to manipulate data sometimes for malicious purposes. In any system, the Intrusion Detection System (IDS) is a fundamental component for ensuring the security of devices from digital attacks. Recognition of new developing digital threats is getting harder for existing IDS. Furthermore, advanced frameworks are required for IDS to function both efficiently and effectively. The commonly observed cyber-attacks in the business domain include minor attacks used for stealing private data. This article presents a deep learning methodology for detecting cyber-attacks on the Internet of Things using a Long Short Term Networks classifier. Our extensive experimental testing show an Accuracy of 99.09%, F1-score of 99.46%, and Recall of 99.51%, respectively. A detailed metric representing our results in tabular form was used to compare how our model was better than other state-of-the-art models in detecting cyber-attacks with proficiency.

The essence of “blockchain” is a shared database in which information stored is un-falsifiable, traceable, open, and transparent. Therefore, to improve the security of private information in medical systems, this article uses blockchain technology to design a method to protect private information in medical systems and effectively realize anti-theft control of private information. First, the Patient-oriented Privacy Preserving Access Control model is introduced into the access control process of private information in medical systems. Next, a private information storage platform is built by using blockchain technology, and information transmission is realized using standard cryptographic algorithms. In this process, file authorization contracts are also used to guarantee the security of private information and further prevent theft of medical private information. Our simulation results show that the storage response time of this method is kept below 1,000 ms, and the maximum information throughput rate reaches 550 kbit/s, which indicates that this method has strong performance in information storage and transmission efficiency. Moreover, the reliability and bandwidth utilization of data transmission across domains is higher, so the method has higher information security control performance and superior overall performance.

Copyright infringement adversely affects the interest of copyright holders of images and videos which are uploaded to different websites and peer-to-peer image sharing systems. This paper addresses the problem of detecting copyright infringement so that copyright holders are given due credit for their work. There are several images and videos that are shared every day by millions of users with some amount of modification in images and videos originally uploaded by the copyright holders such as photographers, graphic designers, and video providers. Copyright violators, who are not the original creators of multimedia content modify them using image processing and frame modification techniques such as grayscale conversion, cropping, rotation, frame compression, and frame speed manipulation. Then, upload the tampered images and videos. To address this problem, we propose an IPFS-based (InterPlanetary File System-based) decentralized peer-to-peer image and video sharing platform built on blockchain technology. We use a perceptual hash (pHash) technique to detect copyright violations of multimedia. When multimedia is to be uploaded to the IPFS, the pHash of the same content is determined and checked against existing pHash values in the blockchain network. Similarity with existing pHash values would result in the multimedia being detected as tampered with. Blockchain technology offers the advantage of non-involvement of a third party and consequently the avoidance of a single point of failure.

To process data like text and speech, Natural Language Processing (NLP) is a valuable tool. As on of NLP’s upstream tasks, sequence labeling is a vital part of NLP through techniques like text classification, machine translation, and sentiment analysis. In this paper, our focus is on sequence labeling where we assign semantic labels within input sequences. We present two novel frameworks, namely SA-CRFLV-I and SA-CRFLV-II, that use latent variables within random fields. These frameworks make use of an encoding schema in the form of a latent variable to be able to capture the latent structure in the observed data. SA-CRFLV-I shows the best performance at the sentence level whereas SA-CRFLV-II works best at the word level. In our in-depth experimental results, we compare our frameworks with 4 well-known sequence prediction methodologies which include NER, reference parsing, chunking as well as POS tagging. The proposed frameworks are shown to have better performance in terms of many well-known metrics.

Due to the increasing awareness and use of cloud and edge computing, society and industries are beginning to understand the benefits they can provide. Cloud and Edge are the future of information management, and they have transformed the Internet into an innovative and interactive computing platform. The ultimate goal of edge/cloud computing is to reduce the use of computing resources in the network, as well as support information sharing and intercommunication efforts within the network. Secure edge computing methodologies are applied in both open and heterogeneous network systems to protect them from many potential security threats. However, these approaches only provide passive protection for normal edge computing operations, and fail to address the security measures of several applications, particularly forensics in industrial settings. Forensics applications running on edge computing must be capable of support taking legal action against invaders for malicious damage or information theft. This paper proposes an efficient and reliable forensics framework (ERFF) to address industrial intelligent edge computing critical for the industry 4.0 implementation plan. The proposed ERFF consists of a detective module and validation model, with the detective module responsible for detecting the interaction between the client terminal and the edge resource, which means the investigator is capable of gathering the evidence securely. The security-validation model integrated with ERFF is far safer than sharing common key-based cryptographic approaches. The proposed conceptual framework is tested with Live Digital Forensic Framework for a Cloud (LDF2C), and results are compared with other existing industrial frameworks that fulfill fundamental ISO/IEC 17025 accreditation requirements, including Legal Reliable Forensic Framework (LRFF), Source Identification Network Forensics Framework (SINFF) and Logging Framework for Cloud Computing Forensic (LFCCF)). These frameworks were designed to support the digital forensic requirements of industry and academia, and experimental results validate the effectiveness of the proposed framework from reliability and efficiency perspectives as well as realistic scenarios

In recommender systems, Collaborative Filtering (CF) plays an essential role in promoting recommendation services. The conventional CF approach has limitations, namely data sparsity and cold-start. The matrix decomposition approach is demonstrated to be one of the effective approaches used in developing recommendation systems. This paper presents a new approach that uses CF and Singular Value Decomposition (SVD)[Formula: see text] for implementing a recommendation system. Therefore, this work is an attempt to extend the existing recommendation systems by (i) finding similarity between user and item from rating matrices using cosine similarity; (ii) predicting missing ratings using a matrix decomposition approach, and (iii) recommending top-N user-preferred items. The recommender system’s performance is evaluated considering Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). Performance evaluation is accomplished by comparing the systems developed using CF in combination with six different algorithms, namely SVD, SVD[Formula: see text], Co-Clustering, KNNBasic, KNNBaseline, and KNNWithMeans. We have experimented using MovieLens 100[Formula: see text]K, MovieLens 1[Formula: see text]M, and BookCrossing datasets. The results prove that the proposed approach gives a lesser error rate when cross-validation ([Formula: see text]) is performed. The experimental results show that the lowest error rate is achieved with MovieLens 100[Formula: see text]K dataset ([Formula: see text], [Formula: see text]). The proposed approach also alleviates the sparsity and cold-start problems and recommends the relevant items.

There is currently widespread use of drones and drone technology due to their rising applications that have come into fruition in the military, safety surveillance, agriculture, smart transportation, shipping, and delivery of packages in our Internet-of-Things global landscape. However, there are security-specific challenges with the authentication of drones while airborne. The current authentication approaches, in most drone-based applications, are subject to latency issues in real time with security vulnerabilities for attacks. To address such issues, we introduce a secure authentication model with low latency for drones in smart cities that looks to leverage blockchain technology. We apply a zone-based architecture in a network of drones, and use a customized decentralized consensus, known as drone-based delegated proof of stake (DDPOS), for drones among zones in a smart city that does not require reauthentication. The proposed architecture aims for positive impacts on increased security and reduced latency on the Internet of Drones (IoD). Moreover, we provide an empirical analysis of the proposed architecture compared to other peer models previously proposed for IoD to demonstrate its performance and security authentication capability. The experimental results clearly show that not only does the proposed architecture have low packet loss rate, high throughput, and low end-to-end delay in comparison to peer models but also can detect 97.5% of attacks by malicious drones while airborne.

Privacy in the realm of data mining known as PPDM has become a hot topic in both academic research and industry due to the fact it can discover implicit rules as well as hide sensitive information for data sanitization. Many different algorithms and heuristics have been investigated to hide sensitive information using the act of transaction deletion based on evolutionary computation techniques, but to date, these algorithms only consider a uniform threshold value for sanitization progress. This technique is not applicable in real-world situations, especially for eHealth based medical datasets. For example, a patient can still be identified if he/she has more confidential information (i.e., symptoms) that cause privacy threats and security leakage in medical applications. In this work, we investigate a unique novel methodology to set varied threshold values that lead to varied lengths of sensitive patterns within a Genetic Algorithm (GA)-based framework. As the pattern length increases, a tighter threshold manifests to provide better protection of sensitive information that can avoid individual patients to be identified in eHealth datasets. Two GA-based models are developed for data sanitization using record deletion techniques. The experimental results are conducted and compared with the traditional Evolutionary Computation (EC)-based PPDM approaches and the results showed that the designed methods offer greater protection than previous methods in terms of side effects. Therefore, the designed models are effective to hide sensitive information in medical situations that can be used in real-world scenarios.

Nowadays, existing public key infrastructures (PKIs) certificate authentication suffers from many security failures. Trusted certificate authorities (CAs) can issue a valid certificate for any domain name. Although CA is supposed to be trusted by a client if the certificate issued to the client links to the chain of trust (e.g., root CA or subordinate CA). By compromising any of the latter (e.g., root CAs or subordinate CAs) an attacker can jeopardize the security of the entire system. Moreover, third-party CAs have to be trusted by domain owners. Currently, the trust is not balanced among the entities involved in the certificate authentication and issuance process (i.e., CAs and domain owners). To counter this problem approaches such as Domain authentication name entity (DANE) and Certificate Authority Authorization (CAA) offer additional securities for domain authentication. However, these approaches depend upon DNS/DNSSEC infrastructure which requires complex requirements for deployment as well as the adoption rate has been low. In this paper, we design, implement a robust and scalable domain authentication scheme based on blockchain technology with privacy-preserving features for low-constrained devices (e.g., mobile, browser, and IoT devices). The proposed system records a set of trusted CAs each associated with a specific domain in the blockchain. That is, each CA has to first verify if it is trusted to perform the actual issuance process. We compare our scheme with existing authentication methods and show that it requires less storage capacity and low bandwidth to authenticate certificates than other methods.

Resource allocation and management in an Internet of Things (IoT) paradigm requires precise request and response processing irrespective of its scalability support. Unpredictable traffic patterns and user density demands reliable offloading for handling user request traffic and service response. Considering the need for large-scale IoT in an account of its interoperability and heterogeneous support, this manuscript introduces a response-aware traffic offloading scheme (RTOS) for delay-sensitive user requests. This offloading scheme is supported by a multivariate spline regression machine learning model for classifying traffic for reducing the failure rate. The splines are adaptive based on the classified traffic for performing independent and shared offloading. The computation process for determining the offloading model is inherited from the cyber-physical system (CPS) coupled with the IoT-Cloud architecture. The information from the knowledge base and event logs are exploited for decision-making in employing the offloading method for the classified traffic. The simulation analysis of this scheme shows that it is effective in improving the request processing ratio and reducing processing, response time, and delay. The simulation is performed for the varying user density and traffic flows.

Federated learning (FL) is a new breed of Artificial Intelligence (AI) that builds upon decentralized data and training that brings learning to the edge or directly on-device. FL is a new research area often referred to as a new dawn in AI, is in its infancy, and has not yet gained much trust in the community, mainly because of its (unknown) security and privacy implications. To advance the state of the research in this area and to realize extensive utilization of the FL approach and its mass adoption, its security and privacy concerns must be first identified, evaluated, and documented. FL is preferred in use-cases where security and privacy are the key concerns and having a clear view and understanding of risk factors enable an implementer/adopter of FL to successfully build a secure environment and gives researchers a clear vision on possible research areas. This paper aims to provide a comprehensive study concerning FL’s security and privacy aspects that can help bridge the gap between the current state of federated AI and a future in which mass adoption is possible. We present an illustrative description of approaches and various implementation styles with an examination of the current challenges in FL and establish a detailed review of security and privacy concerns that need to be considered in a thorough and clear context. Findings from our study suggest that overall there are fewer privacy-specific threats associated with FL compared to security threats. The most specific security threats currently are communication bottlenecks, poisoning, and backdoor attacks while inference-based attacks are the most critical to the privacy of FL. We conclude the paper with much needed future research directions to make FL adaptable in realistic scenarios.

Online social networks (OSNs) have become so popular that it has changed the Internet to a more collaborative environment. Now, a third of the world’s population participates in OSNs, forming communities, and producing and consuming media in different ways. The recent boom of artificial intelligence technologies provides new opportunities to help improve the processing and mining of social data. In this article, an algorithm that can detect communities in the OSNs using the concepts of granular computing in rough sets is proposed. In this information model, a social network as a rough set granular social network (RGSN) is modeled. A new community detection algorithm named granular-based community detection (GBCD) is implemented. This article also defines and uses two measures, namely, a granular community factor and an object community factor. The proposed algorithm is evaluated on four real-world data sets as well as computer-generated data sets. The model is compared with other state-of-the-art community detection algorithms for the values of modularity, normalized mutual information (NMI), Omega index, accuracy, specificity, sensitivity, and $F1$ -measure. The cumulative performance of the GBCD algorithm is found to be 3.99, which outperforms other state-of-the-art community detection algorithms.

In recent decades, mobile or the Internet of Thing (IoT) devices are dramatically increasing in many domains and applications. Thus, a massive amount of data is generated and produced. Those collected data contain a large amount of interesting information (i.e., interestingness, weight, frequency, or uncertainty), and most of the existing and generic algorithms in pattern mining only consider the single object and precise data to discover the required information. Meanwhile, since the collected information is huge, and it is necessary to discover meaningful and up-to-date information in a limit and particular time. In this paper, we consider both utility and uncertainty as the majority objects to efficiently mine the interesting high expected utility patterns (HEUPs) in a limit time based on the multi-objective evolutionary framework. The benefits of the designed model (called MOEA-HEUPM) can discover the valuable HEUPs without pre-defined threshold values (i.e., minimum utility and minimum uncertainty) in the uncertain environment. Two encoding methodologies are also considered in the developed MOEA-HEUPM to show its effectiveness. Based on the developed MOEA-HEUPM model, the set of non-dominated HEUPs can be discovered in a limit time for decision-making. Experiments are then conducted to show the effectiveness and efficiency of the designed MOEA-HEUPM model in terms of convergence, hypervolume and number of the discovered patterns compared to the generic approaches.

With the ever-increasing prosperity of web Application Programming Interface (API) sharing platforms, it is becoming an economic and efficient way for software developers to design their interested mashups through web API re-use. Generally, a software developer can browse, evaluate, and select his or her preferred web APIs from the API's sharing platforms to create various mashups with rich functionality. The big volume of candidate APIs places a heavy burden on software developers’ API selection decisions. This, in turn, calls for the support of intelligent API recommender systems. However, existing API recommender systems often face two challenges. First, they focus more on the functional accuracy of APIs while neglecting the APIs’ actual compatibility. This then creates incompatible mashups. Second, they often require software developers to input a set of keywords that can accurately describe the expected functions of the mashup to be developed. This second challenge tests partial developers who have little background knowledge in the fields. To tackle the above-mentioned challenges, in this article we propose a compatibility-aware and text description-driven web API recommendation approach (named WAR text ). WAR text guarantees the compatibility among the recommended APIs by utilizing the APIs’ composition records produced by historical mashup creations. Besides, WAR text entitles a software developer to type a simple text document that describes the expected mashup functions as input. Then through textual description mining, WAR text can precisely capture the developers’ functional requirements and then return a set of APIs with the highest compatibility. Finally, through a real-world mashup dataset ProgrammableWeb, we validate the feasibility of our novel approach.

When focusing on the general area of data mining, high-utility itemset mining (HUIM) can be defined as an offset of frequent itemset mining (FIM). It is known to emphasize more factors critically, which gives HUIM its intrinsic edge. Due to the flourishing development of the IoT technique, the uncertainty patterns mining is also attractive. Potential high-utility itemset mining (PHUIM) is introduced to reveal valuable patterns in an uncertainty database. Unfortunately, even though the previous methods are all very effective and powerful to mine, the potential high-utility itemsets quickly. These algorithms are not specifically designed for a database with an enormous number of records. In the previous methods, uncertainty transaction datasets would be load in the memory ultimately. Usually, several pre-defined operators would be applied to modify the original dataset to reduce the seeking time for scanning the data. However, it is impracticable to apply the same way in a big-data dataset. In this work, a dataset is assumed to be too big to be loaded directly into memory and be duplicated or modified; then, a MapReduce framework is proposed that can be used to handle these types of situations. One of our main objectives is to attempt to reduce the frequency of dataset scans while still maximizing the parallelization of all processes. Through in-depth experimental results, the proposed Hadoop algorithm is shown to perform strongly to mine all of the potential high-utility itemsets in a big-data dataset and shows excellent performance in a Hadoop computing cluster.

Decision-making plays an increasing important role in our digitalized society, and hence the need for trustworthy data. For example, in a traditional trading process, a trusted middle person generally helps to arbitrate any dispute between two or more parties involved in the trading. However, having such a trusted middle person can be labor-intensive, increases transaction costs, and lacks fairness during arbitration due to the subjectivity of the middle person. While we can avoid the need for a middle person using smart contract-based approaches, there are other limitations (e.g., fairness and privacy preservation) that need to be addressed. In this paper, we present a fair scheme for big data exchanging that allows buyers and sellers to autonomously and fairly complete transactions, without involving any third-party middle person. Specifically, our scheme uses smart contracts and oblivious transfer protocol, in combination with our proposed Ether cheque system. Smart contracts are used to achieve transaction fairness, autonomy and trading timing control. The oblivious transfer protocol helps us to preserve the privacy of transactions. We use Ether cheque to improve the fairness of the transaction and make the transaction more convenient. Also, our proposed approach can facilitate the identification of a cheating party.

In advanced health care systems, patients’ medical data can be outsourced to cloud servers to enable remote healthcare service providers to access and analyze patients’ data from any location to provide better treatment. However, outsourcing sensitive medical data makes data owners, i.e., patients, concerned about their privacy because private companies run the cloud service and the data can be accessed by them. Therefore, it is important to encrypt the data in the form of documents before outsourcing them to the cloud in a way that enables a data user, i.e., a doctor, to search over these documents without allowing the cloud provider to learn any private information about patients. Several schemes have been proposed to enable search over encrypted medical cloud data to preserve patient privacy, but the existing schemes suffer from high communication/computation overhead because they are designed for a single-data-owner setting. Moreover, they are not secure against known-plaintext/background and linkability attacks and do not allow doctors to customize their search to avoid downloading irrelevant documents. In this paper, we develop an efficient search scheme over encrypted data for a multi-data-owner setting. To secure our scheme, the cloud server obtains noisy similarity scores and doctors de-noise them to download the most relevant documents. Our scheme enables doctors to prescribe search conditions to customize the search without revealing the conditions to the server. Our formal proof and analysis indicate that our scheme can preserve privacy and is secure against known-plaintext/background and linkability attacks, and the results of extensive experiments demonstrate the efficiency of our scheme compared to the existing works.

Cyberbullying is disturbing and troubling online misconduct. It appears in various forms and is usually in a textual format in most social networks. Intelligent systems are necessary for automated detection of these incidents. Some of the recent experiments have tackled this issue with traditional machine learning models. Most of the models have been applied to one social network at a time. The latest research has seen different models based on deep learning algorithms make an impact on the detection of cyberbullying. These detection mechanisms have resulted in efficient identification of incidences while others have limitations of standard identification versions. This paper performs an empirical analysis to determine the effectiveness and performance of deep learning algorithms in detecting insults in Social Commentary. The following four deep learning models were used for experimental results, namely: Bidirectional Long Short-Term Memory (BLSTM), Gated Recurrent Units (GRU), Long Short-Term Memory (LSTM), and Recurrent Neural Network (RNN). Data pre-processing steps were followed that included text cleaning, tokenization, stemming, Lemmatization, and removal of stop words. After performing data pre-processing, clean textual data is passed to deep learning algorithms for prediction. The results show that the BLSTM model achieved high accuracy and F1-measure scores in comparison to RNN, LSTM, and GRU. Our in-depth results shown which deep learning models can be most effective against cyberbullying when directly compared with others and paves the way for future hybrid technologies that may be employed to combat this serious online issue.

Online social networks (OSNs) have become an integral mode of communication among people and even nonhuman scenarios can also be integrated into OSNs. The ever-growing rise in the popularity of OSNs can be attributed to the rapid growth of Internet technology. OSN becomes the easiest way to broadcast media (news/content) over the Internet. In the wake of emerging technologies, there is dire need to develop methodologies, which can minimize the spread of fake messages or rumors that can harm society in any manner. In this article, a model is proposed to investigate the propagation of such messages currently coined as fake news. The proposed model describes how misinformation gets disseminated among groups with the influence of different misinformation refuting measures. With the onset of the novel coronavirus-19 pandemic, dubbed COVID-19, the propagation of fake news related to the pandemic is higher than ever. In this article, we aim to develop a model that will be able to detect and eliminate fake news from OSNs and help ease some OSN users stress regarding the pandemic. A system of differential equations is used to formulate the model. Its stability and equilibrium are also thoroughly analyzed. The basic reproduction number ( $R_{0}$ ) is obtained which is a significant parameter for the analysis of message spreading in the OSNs. If the value of $R_{0}$ is less than one ( $R_{0} ), then fake message spreading in the online network will not be prominent, otherwise if $R_{0}> 1$ the rumor will persist in the OSN. Real-world trends of misinformation spreading in OSNs are discussed. In addition, the model discusses the controlling mechanism for untrusted message propagation. The proposed model has also been validated through extensive simulation and experimentation.

Image watermarking techniques are used to provide copyright protection and verify ownership of media/entities. This technique refers to the concept of embedding of secret data/information of an owner in a given media/entity for determining any ownership conflicts that can arise. Many watermarking approaches have been offered by various authors in the last few years. However, there are not enough studies and comparisons of watermarking techniques in soft computing environments. Nowadays, soft computing techniques are used to improve the performance of watermarking algorithms. This paper surveys soft computing-based image watermarking for several applications. We first elaborate on novel applications, watermark characteristics and different kinds of watermarking systems. Then, soft computing based watermarking approaches providing robustness, imperceptibility and good embedding capacity are compared systematically. Furthermore, major issues and potential solutions for soft computing-based watermarking are also discussed to encourage further research in this area. Thus, this survey paper will be helpful for researchers to implement an optimized watermarking scheme for several applications.

Children who are left alone in environments such as daycares and creches require constant monitoring and care to protect them from abuse. In this paper, we propose a novel deep learning-based method for predicting the occurrence of abnormal events using footage gathered from networked surveillance systems and notifying users of those events in an Internet of Things (IoT) environment. Sequences of images are converted to still frames and de-blurred using adaptive motion detection techniques. Then, abnormal activities are predicted using random forest differential evolution with kernel density (RFKD), and any abnormal activities that are detected cause signals to be sent to IoT devices via the MQTT protocol. The proposed work consists of a multi-classifier, deep neural network and kernel density functions. The multi-classifier is used for input classifications from the sequence of frames of videos. The deep neural network is used to learn and train the data and kernel density is used clustering and prediction of data. The novelty of the proposed work is in the dynamic nature of activity prediction. Most of the previous work in this research area concentrated on static activity prediction. The proposed work is able to support both static and dynamic activities of daycare environments. In our experimental trials, our novel method’s performance is shown to be superior to that of the ReHAR method.

Nowadays, everyone is surrounded by a large volume of data which is generated from multiple sources. Every day 2.5 quintillion bytes of data are created worldwide. It is a very complicated task to manage and extract useful information from social media platforms. As people are more interested in seeking online health guidance, the medical textual data created by patients, healthcare professionals, and medical staff are expanding fast. In order to address this challenge, an ontology-based approach is used to annotate the medical domain data in this article. We have considered Twitter as our social media platform to filter the medical data from the general category data. In this article, we have used the Open Biomedical Annotator ( OBA ), which is an ontology-based Web service that annotates raw text with biomedical ontology concepts based on their textual metadata. Domain-specific knowledge is gathered by using MedlinePlus Health Topics and National Drug Data File ( NDDF ) ontologies licensed under the U. S. National Library of Medicine ( UMLS ), which will filter out all medical related terms. Furthermore, these medical terms are stored in a medical term database, which will then be forwarded to the UMLS REST Application Programming Interface ( API ) that provides important entities, such as Concept Unique Identifier ( CUIs ), semantic relations, attributes, definitions, and much more under the subset of SNOMEDCT_US which is the most precise and comprehensive health terminology platform in the world.

In the present era, secure data storage for any Internet of Things (IoT) platform is plagued by poor performance of secure read and write operations, which limits the use of data storage security on any IoT platform. Therefore, in this paper, a data storage security method based on double secret key encryption and Hadoop suitable for any IoT platform is proposed. First, the Hadoop deep learning architecture and implementation process are analyzed, and the process of client Kerberos identity authentication in the Hadoop framework is discussed. From this, the current shortcomings of data storage security based on the Hadoop framework are analyzed. The elements of data storage security are also determined. Furthermore, a novel double secret key encryption method for data storage security and to improve the security of stored data itself is introduced. Simultaneously, hash computing is used to improve the read and write performance of data after secure storage. Experimental results clearly show that our proposed method can effectively improve read and write performance of data, and that the performance of data security operations is improved from current standard implementations.

Cyber-attacks are evolving at a disturbing rate. Data breaches, ransomware attacks, crypto-jacking, malware and phishing attacks are now rampant. In this era of cyber warfare, the software industry is also growing with an increasing number of software being used in all domains of life. This evolution has added to the problems of software vendors and users where they have to prevent a wide range of attacks. Existing watermark detection solutions have a low detection rate in the software. In order to address this issue, this paper proposes a novel blind Zero code based Watermark detection approach named KeySplitWatermark, for the protection of software against cyber-attacks. The algorithm adds watermark logically into the code utilizing the inherent properties of code and gives a robust solution. The embedding algorithm uses keywords to make segments of the code to produce a key-dependent on the watermark. The extraction algorithms use this key to remove watermark and detect tampering. When tampering increases to a certain user-defined threshold, the original software code is restored making it resilient against attacks. KeySplitWatermark is evaluated on tampering attacks on three unique samples with two distinct watermarks. The outcomes show that the proposed approach reports promising results against cyber-attacks that are powerful and viable. We compared the performance of our proposal with state-of-the-art works using two different software codes. Our results depict that KeySplitWatermark correctly detects watermarks, resulting in up to 15.95 and 17.43 percent reduction in execution time on given code samples with no increase in program size and independent of watermark size.

Due to digitization, a huge volume of data is being generated across several sectors such as healthcare, production, sales, IoT devices, Web, organizations. Machine learning algorithms are used to uncover patterns among the attributes of this data. Hence, they can be used to make predictions that can be used by medical practitioners and people at managerial level to make executive decisions. Not all the attributes in the datasets generated are important for training the machine learning algorithms. Some attributes might be irrelevant and some might not affect the outcome of the prediction. Ignoring or removing these irrelevant or less important attributes reduces the burden on machine learning algorithms. In this work two of the prominent dimensionality reduction techniques, Linear Discriminant Analysis (LDA) and Principal Component Analysis (PCA) are investigated on four popular Machine Learning (ML) algorithms, Decision Tree Induction, Support Vector Machine (SVM), Naive Bayes Classifier and Random Forest Classifier using publicly available Cardiotocography (CTG) dataset from University of California and Irvine Machine Learning Repository. The experimentation results prove that PCA outperforms LDA in all the measures. Also, the performance of the classifiers, Decision Tree, Random Forest examined is not affected much by using PCA and LDA.To further analyze the performance of PCA and LDA the eperimentation is carried out on Diabetic Retinopathy (DR) and Intrusion Detection System (IDS) datasets. Experimentation results prove that ML algorithms with PCA produce better results when dimensionality of the datasets is high. When dimensionality of datasets is low it is observed that the ML algorithms without dimensionality reduction yields better results.

For the past two decades, most of the people from developing countries are suffering from heart disease. Diagnosing these diseases at earlier stages helps patients reduce the risk of death and also in reducing the cost of treatment. The objective of adaptive genetic algorithm with fuzzy logic (AGAFL) model is to predict heart disease which will help medical practitioners in diagnosing heart disease at early stages. The model consists of the rough sets based heart disease feature selection module and the fuzzy rule based classification module. The generated rules from fuzzy classifiers are optimized by applying the adaptive genetic algorithm. First, important features which effect heart disease are selected by rough set theory. The second step predicts the heart disease using the hybrid AGAFL classifier. The experimentation is performed on the publicly available UCI heart disease datasets. Thorough experimental analysis shows that our approach has outperformed current existing methods.

Emails are being used as a reliable, secure, and formal mode of communication for a long time. With fast and secure communication technologies, reliance on Email has increased as well. The massive increase in email data has led to a big challenge in managing emails. Emails so far can be classified and grouped based on sender, size, and date. However, there is a need to detect and classify emails based on the contents contained therein. Several approaches have been used in the past for content-based classification of emails as Spam or Non-Spam Email. In this paper, we propose a multi-label email classification approach to organize emails. An efficient classification method has been proposed for forensic investigations of massive email data (e.g., a disk image of an email server). This method would help the investigator in Email related crimes investigations. A comparative study of machine learning algorithms identified Logistic Regression as a method that achieves the highest accuracy compared to Naive Bayes, Stochastic Gradient Descent, Random Forest, and Support Vector Machine. Experiments conducted on benchmark data sets depicted that logistic Regression performs best, with an accuracy of 91.9% with bi-gram features.

Canada is one of the largest countries by area in the world. Rural populations of Canada have for decades suffered from lacking the infrastructure necessary to offer technological advancements to its citizens. Finally, in recent years, we have seen our Canadian government take the initiative to ensure our remote citizens have the Internet capabilities needed to operate in our digital world. With advancement also comes challenges, which in the context of this paper involves the additional risk to Canadian citizens to cyber-threats. Through an educational initiative funded Nationally, this paper discusses the efforts to offer cybersecurity education to rural Indigenous citizens in the province of Manitoba, Canada. We discuss in detail our initial progress, recent workshops with results, and the future of this program post-COVID-19.

Malicious websites predominantly promote the growth of criminal activities over the Internet restraining the development of web services. Furthermore, we see different types of devices being equipped with WiFi capabilities, that allow web traffic to pass through the device’s data systems with ease. The proposed framework in the present study analyzes the Uniform Resource Locator (URL) through which malicious users can gain access to the content of the websites. It thus eliminates issues of run-time latency and possibilities of users being subjected to browser oriented vulnerabilities. The primary objective of this paper is to detect malicious links on the web using a machine learning classification technique that would help users defend against cyber-crime attacks and related threats of the real world. This may be helpful in the newly expanding Intelligent Infrastructures, where we see more data availability almost daily. The embedding of malicious URLs is a predominant web threat faced by the Internet community in the present day and age. Attackers falsely claim of being a trustworthy entity and lure users to click on compromised links to extract confidential information, victimizing them towards identity theft. The present work explores the various ways of detecting malicious links from the host-based and lexical features of the URL in order to protect users from being subjected to identity theft attacks.

Mobile cloud computing provides on-demand resources. The architecture of mobile cloud computing is composed of a cluster of mobile devices. It is gaining popularity because of its cost-effectiveness and availability. There are numerous security issues like data breaches due to many data being stored with all of its benefits. According to recent searches, about 70% of the operations are now performed on the cloud. Data loss from mobile devices, unsecured exchange of information through rogue access points are the security threats of mobile cloud computing. Data breaches, account hijacking, denial of services, loss of encryption key are additional security and privacy threats. Examples of mobile cloud applications are Google maps, GMAIL, and Cisco’s WebEx on iPad. The security issues mentioned before in mobile cloud computing are now applying more complicated authentication schemes. We can secure the architecture by integrating a multi-agent system. The simulations used for the analysis are OPNET and SPSS, where OPNET is used to evaluate and develop a network and information security model for cloud computing security, and SPSS be used to build a statistical analysis of how much this is affecting and how much it occurs. In this paper, the protocols to implement different kinds of multi- factor authentication are discussed.

In this research, we propose a collaborative clustering method where the exchange of raw data is not required. The attention-based model is used with a federated learning framework. The edge devices compute the model updates using local data and send them to the server for aggregation. Repetition is performed in multiple rounds until a convergence point is reached. The transaction data is used to train the attention model that gives a low dimensional embedding. Afterward, we share the convergence model among the client/stores. Then, efficient clustering-based dynamic method is then utilized. For experimentation, we used retail store data to cluster the customer based on purchase behavior. The proposed clustering method used semantic embedding to extract centroid and then cluster them by discovering relevant patterns. The method achieved the 0.75 ROC values for the random distribution and 0.70 for the fixed distribution. The clustering method can help to reduce communication costs while ensuring privacy.

Centralized Personal Health Records (PHR) are mutable with compromised security as it may lead to a single point of failure. Confidentiality, protection and security are the common issues in clinical record frameworks. Specific security and protection schemes are being used to secure clinical records. Accordingly, using the Interplanetary File System (IPFS), a decentralized PHR can be maintained to allow patients to access their records without delay. Moreover, a Kademlia-based distributed hash table provides fault tolerance and enables patients to keep track of their medical history. However, a significant issue in IPFS is data availability. It is only available on the web until users or hosts of the network request each peer, later it leads to a permanent loss of data. We propose an architecture that aims to provide faster retrieval and constant PHR availability using Blockchain and IPFS. The results show that an optimal node is selected in each iteration amongst all the available adjacent nodes.

In this paper, we propose a fuzzy classification deep attention-based model that expands emotional lexicons by using linguistic properties of actual patient authored texts. The active learning methods can expand the trained dataset and fuzzy rules over some time. As a result, the model itself can reduce its labeling efforts for mental health application. Thus, the designed model can solve issues related to vocabulary sizes per class, data sources, methods of creation, and create a baseline for human performance levels. This paper also gives fuzzy explainability by visualizing weighted words. Our proposed method uses a similarity-based method that includes a subset of unstructured data as the training set. Next, using an active learning mechanism cycle, our method updates the training model using new training points. This cycle is repeatedly performed until an optimal solution is reached. The designed model also converts all unlabeled texts into the training set. Our in-depth experimental results show that the emotion-based expansion enhances the testing accuracy and helps to build quality rules.