Although several machine learning driven solutions are deemed to be effective at detecting data breaches, the recent proliferation in data breach incidents resulting from cyber attacks on computer networks demands an updated, thorough analysis of Machine Learning (ML) based data breach countermeasures to identify research gaps and guide future studies. In view of this, this study employs a systematic approach and draws insight from 89 research articles to classify machine learning based data breach countermeasures using eight criteria namely learning tasks, learning classifiers, datasets, feature engineering methods, multimodal approaches, pre-training approaches and performance. In classifying the studies, we: (a) propose a taxonomy of feature extraction and representation to classify studies using ten sub-criteria, (b) classify multimodal machine learning approaches used in the studies into three fusion sub-criteria: namely early fusion, intermediate fusion and late fusion, (c) show a comparison of studies based on pre-training techniques employed such as pre-text objective, learning model and data used in pre-training, (d) classify the datasets used in the study evaluation into two categories: real dataset and simulated dataset and (e) evaluate studies by detection performance and effectiveness against data breaches on unknown and obfuscated network traffic. To aid the literature identification, we analyse forty recent incidents and obtain prevalent cyber attack vectors of data breaches, which we present as the general workflow for data breaches due to cyber attacks. Finally, we highlight the research issues associated with existing ML-based data breach countermeasures and recommend future research directions.

Cyber-attacks on industrial applications, specifically, phishing and web attacks are the most common data breach vectors and, as such, have attracted significant research attention. To mitigate these types of attacks, many countermeasures based on machine learning (ML) have been proposed. Although ML-based countermeasures are reported to yield satisfactory detection performance on phishing and web attacks, they often require massive amounts of manually labelled email and web request data to build these countermeasures. The manual generation of labels, however, can be laborious, error-prone and infeasible to scale. To cope with the evolution of web attacks and phishing emails, methods which exploit the vast volumes of unlabelled email texts and web request data should be adopted. Recent studies have primarily employed sequential models such as BERT, to learn semantic contextual features from unlabelled email and HTTP request text. In this study, we take a step further by extracting complementary features from unlabelled email and web request data represented in two-dimensional structures. Our method applies computer vision-based transformations to these structured representations and employing a self-supervise learning approach, we pre-train a convolutional neural network model to recognise these transformations, enabling the model to learn syntactic structural features from unlabelled data. We then adopt concatenation as our ensemble strategy to combine contextual and syntactic features, yielding robust representation of email and HTTP request text, which we leverage in a downstream fully connected neural network model for phishing and web attack detection. Extensive experiments conducted on three phishing datasets including Nazario, Enron and Subhadeep’s email corpus, as well as the benchmark SR-BH web attack dataset show that, the proposed method outperforms baseline sequential models developed in this work, which rely solely on contextual representations for detecting phishing and web attacks.

The purpose of this paper is to analyse the effects of predatory approach words in the detection of cyberbullying and to propose a mechanism of generating a dictionary of such approach words. The research incorporates analysis of chat logs from convicted felons, to generate a dictionary of sexual approach words. By analysing data across multiple social networks, the study demonstrates the usefulness of such a dictionary of approach words in detection of online predatory behaviour through machine learning algorithms. It also shows the difference between the nature of contents across specific social network platforms. The proposed solution to detect cyberbullying and the domain of approach words are scalable to fit real-life social media, which can have a positive impact on the overall health of online social networks. Different types of cyberbullying have different characteristics. However, existing cyberbullying detection works are not targeted towards any of these specific types. This research is tailored to focus on sexual harassment type of cyberbullying and proposes a novel dictionary of approach words. Since cyberbullying is a growing threat to the mental health and intellectual development of adolescents in the society, models targeted towards the detection of specific type of online bullying or predation should be encouraged among social network researchers.

Over the last decade, the proliferation of the Internet of Things (IoT) has produced an overwhelming flow of continuous streaming data. A massive amount of IoT data will be generated in the future. Therefore, it is necessary to create more sophisticated frameworks to measure IoT data quality, considering relevant attributes such as the freshness, reliability and trustworthiness of IoT data. Existing data freshness models and frameworks mostly depend on the timestamp. However, the frequency of IoT data (e.g., data generated by sensors which is measured per millisecond or minute) needs to be considered, that is, IoT data can change frequently. We introduce a new model for measuring IoT data freshness. In our model, we define unreliable IoT data and discard them while considering fresh data. We introduce a formal approach to IoT data freshness including the underlying concepts and definitions. Using this formal approach, we propose an algorithm for the numerical calculation of the freshness attributes. We conduct several sets of experiments and demonstrate the feasibility of the proposed framework by quantifying the performance of the freshness measurement algorithm. We also demonstrate the capability of the framework to capture freshly generated IoT data through a software prototype and several case studies. Finally, we provide a roadmap for future research considering other IoT data quality attributes, such as reliability and trustworthiness.

RFID technology has gained popularity to address localization problem in the manufacturing shopfloor by tracking the manufacturing object location to increase the production's efficiency. However, the signals (data) used for localization task is not easy to analyze because it is generated from the nonstationary environment. It also continuously arrive over time and yields the large-volume of data. Therefore, an advanced big data analytic is required to overcome this problem. We propose a distributed big data analytic framework based on PANFIS (Scalable PANFIS), where PANFIS is an evolving algorithm which has capability to learn data stream in the single pass mode. Scalable PANFIS can learn big data stream by processing many chunks/partitions of data stream. Scalable PANFIS is also equipped with rule' structure merging to eliminate the redundancy among rules. Scalable PANFIS is validated by measuring its performance against single PANFIS and other Spark's scalable machine learning algorithms. The result shows that Scalable PANFIS performs running time more than 20 times faster than single PANFIS. The rule merging process in Scalable PANFIS shows that there is no significant reduction of accuracy in classification task with 96.67 percent of accuracy in comparison with single PANFIS of 98.71 percent. Scalable PANFIS also generally outperforms some Spark MLib machine learnings to classify RFID data with the comparable speed in running time.

A domain ontology is an ontology that describes the fundamental knowledge of the domain, including the domain vocabulary, concepts, taxonomy, relations, properties, constraints, and axioms. The creation of a domain ontology may be done manually from scratch by domain experts or translated from existing knowledge sources. Our earlier work in [4] presented a framework on building an ontology by matching and merging existing domain ontologies. In this paper, we expand the framework to include different types of community knowledge representations in a common ontology. A common ontology is a domain ontology that is developed from community knowledge by gathering their commonality. To build a common ontology, we standardize different community data format using a schema mediation notation. We apply rule-based mapping and information extraction methodology, and we perform matching, clustering and merging, to collect common knowledge together. The evaluation shows that our proposed framework can build a valid and rich common ontology.

Ontology is widely used as a mean to represent and share common concepts and knowledge from a particular domain or specialisation. As a knowledge representation, the knowledge within an ontology must be able to evolve along with the recent changes and updates within the community practice. In this paper, we propose a new Ontology-based Information Extraction (OBIE) system that extends existing systems in order to enrich and validate an ontology. Our model enables the ontology to find related recent knowledge in the domain from communities, by exploiting their underlying knowledge as keywords. The knowledge extraction process uses ontology-based and pattern-based information extraction technique. Not only the extracted knowledge enriches the ontology, it also validates contradictory instance-related statements within the ontology that is no longer relevant to recent practices. We determine a confidence value during the enrichment and validation process to ensure the stability of the enriched ontology. We implement the model and present a case study in herbal medicine domain. The result of the enrichment and validation process shows promising results. Moreover, we analyse how our proposed model contributes to the achievement of a richer and stable ontology.

This paper presents an object-based method for indexing recursive structured XML data and process branched queries efficiently. The proposed method is called Object-based Twig Query processing for Recursive data (OTQℜ). It is an extended approach of our existing work in [1] in order to handle recursion in XML data. Our motivation of extending OTQ to OTQℜ is to support many applications that require recursive data structure to be fully functional. OTQℜ is proposed to utilize semantics of XML data to efficiently process branched queries on recursive XML data. The experiments and evaluation are presented to cover variant evaluating points and the efficiency of our approach.

The extensible Mark up Language (XML) is a meta language that is widely used to provide a non-proprietary universal format for sharing hierarchical data among different software systems and application domains. Moreover, many organizations and content providers have been publishing and sharing their information through XML and its standard schemas. In this context, it is extremely important when designing new schemas or enhancing current ones, there is a mechanism to ensure that the schemas will be well-designed versions. In this paper, we develop a tool that aids schema developers and standard groups to track XML schema changes, log them, and help in the enhancement of a particular schema version. We develop a schema monitoring tool called XSM, which efficiently stores and retrieves versioned XSDs and evaluates them based on the quality indicators defined for this purpose. The quality of delta changes in the schema versions is examined through a set of synthetic XSDs.

Ontology is used widely as a knowledge representation form. In the recent years, there is substantial growth of ontology utilization and development in different application domains that requires standardization and semantic interoperability. The increasing number in ontology development leads to overlapping in domain ontologies. Our approach called CODE (Common Ontology Development) is created to address the need for knowledge integration and reuse from distributed knowledge across multiple underlying ontologies. CODE is fully automated and it is a complete framework that aims to preserve source ontology knowledge. It uses natural language processing and scenario-based rules to ensure the accuracy of the integration mechanism. In addition, CODE can be used to merge more than two ontologies at the same time. To evaluate knowledge preservation and correctness in the proposed approach, SPARQL queries are used on the new integrated ontology and the results are verified against the source ontologies.

XML Schema standards often undergo several revisions to fit application requirements and business demands. In order to be successful, the development process of such standards must be collaborative allowing multiple users to work on the same schema. In this editing environment, the ability to merge branched versions of the schema is significant in certain situations. Using conventional three-way XML merging tools is not suitable for the purpose of merging XML Schema because the tree model of XML Schema is different from that of XML document.

The increased deployment of the XML-based standard for representation and exchange in multi-disciplinary domains has enforced the need for a more effective way to deal with XML query processing. Since very limited attention has been given to the semantic nature of the XML data being processed, this paper focuses on a technique for XML query optimization, called Object-based Twig Query (OTQ), to utilize the semantic structure of the data being queried to process twig queries. A twig query, which is a type of query with multiple branches, requires complex processing due to the joins between multiple paths. Outperforms object-based data partitioning, which aims at leveraging the notion of frequently-accessed data subsets and putting these subsets together into adjacent partitions. It evaluates branched queries through two essential components: (i) OTQ indexing, which uses an object-based connection to construct its indices i.e. Schema index and Data index, and (ii) OTQ processing to produce the final results in optimal time. At the end of this paper, a set of experimental results for the proposed approach on arange of real and synthetic XML data, as well as a comparative study of a similar work in the area, is presented to demonstrate the effectiveness of OTQ optimization.

Due to the growing use of XML data format in global information, an effective XML data management system is needed. An Enabled XML DB is one of the recent widely accepted approaches to store XML documents. This ability coupled with the increase use of XML data in different areas have triggered the need for a better method to structure a large data in order to improve query performance. Issues concerning the ways to efficiently partition large XML documents into a more manageable form are yet to be addressed. At the same time, it is essential to ensure that the partitioning method maintains the preservation of XML data hierarchical structure. For this reason, this paper introduces OXDP that structures large XML data logically by partitioning them into object based XML components. An evaluation is shown to demonstrate the effectiveness of OXDP in XML partitioning which subsequently has the potential of improving query performance in Enabled XML DB environments.

Security is considered to be one of the most critical aspects in a cloud computing environment due to the sensitive and important information stored in the cloud for users. Users are wondering about attacks on the integrity and the availability of their data in the cloud from malicious insiders and outsiders, and from any collateral damage of cloud services. These issues are extremely significant but there is still much room for security research in cloud computing. This paper focuses more on the issues related to the data security and privacy aspects in cloud computing, such as data integrity, data intrusion, service availability. It proposes a Multi-clouds Database Model (MCDB) which is based on Multi-clouds service providers instead of using single cloud service provider such as in Amazon cloud service. In addition, it will discuss and present the architecture of the proposed MCDB model and describe its components and layers. The results and implementation for the new proposed model will be analyzed, in relation to addressing the security factors in cloud computing, such as data integrity, data intrusion, and service availability.

Currently, the rapid development of XML databases, peer-to-peer architecture and Web services has encouraged changes in distributed computing and databases. AXML systems, which are extensions of XML, have been developed to exploit the computation powers of the three technologies XML, P2P and Web services. AXML systems promise many advantages for data exchange between database systems such as fresh, dynamic data and materialized data on demand. However, several problems in data exchange between AXML peers are of concern. This paper will focus on improvements to enhance the mechanism to facilitate AXML data exchange, including changes in AXML representations and mechanisms to control data exchange between peers.

The area of XML database outsourcing, whereby the data owner enlists an external service provider to manage the storage and retrieval of their database, has been of increasing interest in recent years due to the relatively inexpensive nature of hardware/bandwidth, compared to the higher expense of in-house expert staff/software. As such it has become increasingly practical to use outsourced database solutions. However, as the service provider may not be fully trusted, XML database outsourcing introduces several security concerns that are new or more complex than those encountered in traditional database implementations. These include: data confidentiality, privacy, secure auditing, query assurance and secure and efficient storage. Of particular importance due to its relevance to most outsourced database models is query assurance - ensuring the database responds correctly to queries. In this paper, we propose the use of temporary time stamps and hash granularity to increase the efficiency of query assurance. This approach is tested against real datasets of varying type and size. Further, we consider how best to create time stamps and the issues associated with expiring versus distributed time stamp models.

The need to extract knowledge from web data warehousing just 'in-time' for decision making has increased significantly. An efficient system that can generate up-to-date analysis and decision making of the ever changing web-based information will play a very important role in the current global market and society. The new era of business intelligence and web databases brings in new research and development issues whereby the efficient integration of various web data is needed and timely analysis of data resources are vital. Web Data Warehousing is a growing area that addresses the need for an efficient web data summary to support decision making and ensure the quality of web data analysis.

With increasing demands for a proper and efficient XML data storage, XML-enabled database (XEnDB) has emerged as one of the popular answers. It claims to combine the strengths and limit the shortcomings of the traditional database management systems and native XML database. The implication is more research need to be done for this database family. This paper focuses on the XML update management in XEnDB. Our aim is to preserve the conceptual semantic constraints in XML data during update operations. The constraints are classified and represented in SQL/XML schema. Then, we propose the update methodology that utilizes the proposed schema and implement the method in one of the current XEnDB products

XML data can be stored in different types of databases including object-relational databases (ORDB). Using ORDB, we get the benefit of relational maturity and the richness of object-oriented modeling. One modeling concept that can be captured is composition hierarchy, which is a special type of relationship that shows an exclusive existence-dependent "part-of" relationship. This type of relationship frequently occurs in XML data, yet very often when the data is stored in a database repository, the "part-of" relationship is either flattened or split into an entirely separate table. In this paper we propose a model to preserve composition type in XML data into ORDB using the concept of row types. We use the Semantic Network diagram to represent the composition hierarchy in XML data. The composition hierarchy is divided into three types, namely single row composition, multi rows composition, and multi level composition. Each of these composition types will then be transformed into storage in an ORDB environment.