Research Title: A Genetic Based Approach for Reducing Null Values In Object-Oriented Database

Author: Said Ahmad Ammar Ghoul, Published Year: 2010

Faculty: Information Technology

Abstract: In databases, the Null Values considered as one of major problems; where it affects results accuracy and database consistency. This problem affects all databases types: relational, object oriented, object relational, etc. And researchers are unanimous on three reasons that cause null values for attributes: value unknown, value unavailable or inapplicable attribute. But they didn’t observe that the presence of alternative attributes for the same entity may cause null values, which can be solved using the Alternation and the Selection concepts. In this paper we improved the structure of classes used in current object-oriented database model to accept the alternation concept. And we changed the genetic program concept, which is used in a genetic approach to select attributes using rules, to simpler selection technique by improving the structure, without the need to change the depth first strategy for applying the selective inheritance concept. So as result, the null values (by alternation) were removed, and selective inheritance in object-oriented database was introduced.

Keywords: Alternation, Alternatives, Attributes, Classes, Class Hierarchy, Classification, Database, Genetic Approach, Genetic Program, Multiple Inheritance, Null Values, Object Oriented Database, Selection, Selective Inheritance, Syntax Classification.

Research Title: Enhancing the Web Application Security Using Aspect-Oriented Programming

Author: Said Ahmad Ammar Ghoul, Published Year: 2010

Faculty: Information Technology

Abstract: Web applications open the window of security concerns; more used, more attacked. The web security problems can occur in the web server and the data that is on it, or within the information transferred between the web server and the user, or in the end-user's computer and other devices that people use to access the Internet. The AOP approach has already been the subject of several related efforts that addressed the SQL injection and Cross Site Scripting (XSS). These two problems are in the top ten problems for web application security. This paper proposes a solution for a third common problem that was not addressed before which is the broken authentication and session management. The solution consists of an AOP methodology to make web applications more secure. As a result, web applications may be secured and reliable without the need to change their existing code. The security code and the web application code are clearly separated. The security is insured by applying it. In this way the aspect will be able to progress independently. The remote address is used to recognize the real user beside the Session ID.

Keywords: Application Programming Interface, Aspect-Oriented, Broken Authentication and Session Management, Methodology, Web Application, Web Security

Research Title: Bio-inspired Systems – An Integrated Model

Author: Said Ahmad Ammar Ghoul, Published Year: 2010

Faculty: Information Technology

Abstract: The idea of applying the biological principle of natural evolution to artificial systems, introduced more than three decades ago, has seen impressive growth in the past few years. Usually these artificial systems group bio-inspired systems, evolutionary computation, artificial life, etc. This convinced all that nature has much we might be inspired by when designing our systems and the genetics is a promising inspiration. Extensive methodologies and systems, mapping it, were developed. But the mapping of the natural genetics to the artificial one has been however limited. The reasons for this are many, but one of the main problems has always been the lack of a universal platform, and of a proper methodology for such mapping. In this paper we aim to propose firstly a model for such universal architecture, integrating the POE framework components and develop them in more detailed and accurate way. And secondly an bio-inspired systems design methodology guidelines based on this integrated model. We first present the natural genetics process and the POE model. Then the development of the integrated model with its associated design methodology will be introduced and evaluated

Keywords: Artificial systems, Bio-inspired system, epigenesis, genetics, genome, genotype, immune systems, neural networks, ontogeny, phenotype, phylogeny, species

Research Title: Systems Versioning: A Features-Based Meta-modeling Approach

Author: Said Ahmad Ammar Ghoul, Published Year: 2014

ICFSE 2014: International Conference on Forensic Software Engineering, UK

Faculty: Information Technology

Abstract: Systems running these days are huge, complex and exist in many versions. Controlling these versions and tracking their changes became a very hard process as some versions are created using meaningless names or specifications. Many versions of a system are created with no clear difference between them. This leads to mismatching between a user’s request and the version he gets. In this paper, we present a system versions meta-modeling approach that produces versions based on system’s features. This model reduced the number of steps needed to configure a release and gave each version its unique specifications. This approach is applicable for systems that use features in its specification

Keywords: Features, Meta-modeling, Semantic Modeling, SPL, VCS, Versioning.

Research Title: Supporting Aspect-Oriented Paradigm By Bio-inspired Concepts

Author: Said Ahmad Ammar Ghoul, Published Year: 2011

Faculty: Information Technology

Abstract: The idea of applying the biological principle of natural evolution to artificial systems, introduced more than three decades ago, has seen impressive growth in the past few years. Usually these artificial systems group bio-inspired systems, evolutionary computation, artificial life, etc. This convinced all that nature has much we might be inspired by when designing our systems and the genetics is a promising inspiration. Extensive methodologies and systems, mapping it, were developed. But the mapping of the natural genetics to the software design has been however very limited. Actually, intensive efforts are being invested for upgrading programming tools from high level programming languages to design languages, including Aspect-Oriented Paradigm (AOP). The growing interest of AOP and its application in several domains necessitates its enrichment by strong concepts which are intended to support effectively the application domain diversity. This paper proposes such enhancement inspired from natural genetics. We first present Bio-inspired, Aspect-Oriented, and Subject-Oriented software design approaches. Then the enrichment of AOP, induced by a Bio-inspired approach, with its supporting design methodology will be introduced and evaluated.

Keywords: Bio-inspired modelling, Aspect-oriented programming

Research Title: A Methodology for AUML Role Modeling

Author: Said Ahmad Ammar Ghoul, Published Year: 2011

Faculty: Information Technology

Abstract: An agent can play one or more roles. A role is a specific behavior to be played by an agent, defined in terms of permission, responsibilities, activities, and of its interactions with other roles. An agent plays a role by actualizing the behavior in terms of services to be activated and de-activated in dependence of specific pre-condition and post-conditions. So the need to develop a model representing formally role of an agent and its interaction with other agent is valuable. The need for a role model that expresses how agent assumes and change roles is essential and AUML is not completely supportive. The core parts of AUML are interaction protocol diagrams and agent class diagrams, which are extensions of UML's sequence diagrams and class diagrams, respectively. Agents are assigned to roles, belong to classes and an Interaction Protocol diagram shows interactions between these agents roles along a timeline. So, the majority of AUML problems are the lack of: (1) agent role definition methodology, (2) formal semantics to AUML role diagrams, (3) responsibilities (internal) agent role definition (only the external role is defined by sequence diagram), and the agent role control over the time. In this paper, we propose a solution to the above problems. We start by analyzing some significant actual approaches to agent role modeling, we introduce an enhancement to AUML by an agent role definition methodology, and we end by comparing our contribution with similar works.

Keywords: Multi-agent system, Agent role, control role, responsibility role

Research Title: A ROTATIONAL BLENDED LEARNING MODEL: ENHANCEMENT AND QUALITY ASSURANCE

Author: Said Ahmad Ammar Ghoul, Published Year: 2013

Faculty: Information Technology

Abstract: Research on blended learning theory and practice is growing nowadays with a focus on the development, evaluation, and quality assurance of case studies. However, the enhancement of blended learning existing models, the specification of their online parts, and the quality assurance related specifically to them have not received enough attention. This paper presents in detail an enhancement, implementation, and evaluation of a rotational blended learning model. This model was adopted by the UNISCO Avicenna Virtual University. Also, its impact on some quality assurance criteria of the rotational blended learning model was presented. This research is unique in the sense that it specifies clearly the online part of the rotational blended learning model with its impact on the main quality assurance criteria

Keywords: Blended Learning, Rotational Model, Quality Assurance

Research Title: SYSTEMS VARIABILITY MODELING: A TEXTUAL MODEL MIXING CLASS AND FEATURE CONCEPTS

Author: Said Ahmad Ammar Ghoul, Published Year: 2013

Faculty: Information Technology

Abstract: System’s reusability and cost are very important in software product line design area. Developers’ goal is to increase system reusability and decreasing cost and efforts for building components from scratch for each software configuration. This can be reached by developing software product line (SPL). To handle SPL engineering process, several approaches with several techniques were developed. One of these approaches is called separated approach. It requires separating the commonalities and variability for system’s components to allow configuration selection based on user defined features. Textual notationbased approaches have been used for their formal syntax and semantics to represent system features and implementations. But these approaches are still weak in mixing features (conceptual level) and classes (physical level) that guarantee smooth and automatic configuration generation for software releases. The absence of methodology supporting the mixing process is a real weakness. In this paper, we enhanced SPL’s reusability by introducing some meta-features, classified according to their functionalities. As a first consequence, mixing class and feature concepts is supported in a simple way using class interfaces and inherent features for smooth move from feature model to class model. And as a second consequence, the mixing process is supported by a textual design and implementation methodology, mixing class and feature models by combining their concepts in a single language. The supported configuration generation process is simple, coherent, and complete.

Keywords: Class modeling, Configuration, Feature modeling, Mixing class and feature concepts, Software product line design methodology, Variability.

Research Title: A Genetic Methodology for Object Evolution

Author: Said Ahmad Ammar Ghoul, Published Year: 2014

IJSEIA, 8

Faculty: Information Technology

Abstract: One of the major issues of object-oriented programming languages is lack of reclassification mechanisms. Reclassification allows an object to change its class at runtime. Object reclassification is desired in applications whose entities need to change dynamically at runtime. The previous approaches dealt with object reclassification in different ways. But none had approached the notion of real entities evolution, where objects can evolve while belonging to the same class. This had led to a large number of classes most of the times; making the design and the implementation of the intended system more complex. The desired evolution of objects should allow objects to evolve automatically at runtime, without changing their classes membership. In this work, we present a new approach for object evolution; inspired by some genetics concepts. Objects that belong to the same class can change their structure, functions and behaviors at run time automatically, while keeping their membership to the same class. A slight environmental influence on the proposed evolution process is introduced.

Keywords: Bio-inspired modeling, Evolution, Genetic class, View

Research Title: A Bio-Inspired Approach to Selective Inheritance Modeling

Author: Said Ahmad Ammar Ghoul, Published Year: 2014

IJSEIA, 8

Faculty: Information Technology

Abstract: The conventional inheritance concept adopted in the current Object Oriented Programming (OOP) was applied more than three decades ago. It is acting on "is-a" hierarchy model, and has some defects; while OOP is trying to be more close to the real life, it is still far from genetics principles. Adopted inheritance concept means that the child class can inherit, and get everything that is public in the parent class automatically. This process has solved many problems, but it does not simulate what is really happening in our life, where each object can gain just the needed properties from the parent class. So, in simple words, conventional inheritance is not selective and is generating identical objects. One of introduced solutions was the selective inheritance. But, while conventional inheritance acts on “is-a” hierarchy model, the works that introduced the selective inheritance were also done on this model and still suffers from some problems. The inspiration from the “real life” genetics has led us to a selective inheritance acting upon a “Composed by” model rather than the “is-a” model, where the properties are classified into several classes according to "Composed by" relation. This paper proposes a Genetic approach to a new model for selective inheritance that is replacing the “is-a” hierarchy model by a “Composed by” one, which reduces the complexity and makes each object has its own definition. After that, it proposes a language extension to be adopted in any object oriented programming language.

Keywords: Bio-inspired systems modeling, Is-a hierarchy model, composed-by hierarch model, selective inheritance