Christina Chavez

On the modular representation of architectural aspects

An architectural aspect is a concern that cuts across architecture modularity units and cannot be effectively modularized using the given abstractions of conventional Architecture Description Languages (ADLs). Dealing with crosscutting concerns is not a trivial task since they affect each other and the base architectural decomposition in multiple heterogeneous ways. The lack of ADL support for modularly representing such aspectual heterogeneous influences leads to a number of architectural breakdowns, such as increased maintenance overhead, reduced reuse capability, and architectural erosion over the lifetime of a system. On the other hand, software architects should not be burdened with a plethora of new ADL abstractions directly derived from aspect-oriented implementation techniques. However, most aspect-oriented ADLs rely on a heavyweight approach that mirrors programming languages concepts at the architectural level. In addition, they do not naturally support heterogeneous architectural aspects and proper resolution of aspect interactions. This paper presents AspectualACME, a simple and seamless extension of the ACME ADL to support the modular representation of architectural aspects and their multiple composition forms. AspectualACME promotes a natural blending of aspects and architectural abstractions by employing a special kind of architectural connector, called Aspectual Connector, to encapsulate aspect-component connection details. We have evaluated the applicability and scalability of the AspectualACME features in the context of three case studies from different application domains.

Separation of Concerns in Multi-agent Systems: An Empirical Study

With multi-agent systems (MASs) growing in size and complexity, the separation of their concerns throughout the different development phases is crucial to MAS engineers. Separation of concerns is a well-known principle in software engineering to achieve improved reusability and maintainability of complex software. Hence it is necessary to investigate systematically whether ions from object-oriented (OO) software engineering are able to isolate explicitly MAS concerns. This paper presents an empirical study that evaluates the degree to which abstractions associated with two OO techniques enable modularization of MAS concerns. The selected techniques involve basic OO ions, such as classes and objects, and more sophisticated ones, namely design patterns and aspects. The gathered results shown that the use of aspects allowed the construction of a MAS with a significant improvement in the modularization of different concerns. Also, the use of aspects resulted in: (i) fewer lines of code, (ii) fewer design and implementation components, and (iii) lower coupling between the components. However, the aspect-oriented approach produced lower cohesion in the MAS components. Finally, an important finding of this empirical study is that aspects supported a better alignment with higher-level abstractions from agent-oriented design models.

Taming agents and objects in software engineering

Agent-based software engineering has been proposed in addition to object-oriented software engineering as a means of mastering the complexity associated with the development of large-scale distributed systems. However, there is still a poor understanding of the interplay between the notions of agents and objects from a software engineering perspective. Moreover, the many facets of agent-based software engineering are rarely used in the various phases of the software development lifecycle because of the lack of a comprehensive framework to provide the software designers with a clear understanding of the use of these two key abstractions. In this context, this paper presents TAO, an evolving innovative conceptual framework based on agent and object abstractions, which are the foundations for modeling large-scale software systems. The conceptual framework allows for the characterization of large-scale software systems as organizations of passive components, the objects, and autonomous components, the agents, with each of these elements playing roles to interact with each other and to coordinate their actions in order to fulfill system goals.

Engineering multi-agent systems with aspects and patterns

Objects and agents are software engineering abstractions that have many common concerns. However, agents are more complex entities since they encompass additional concerns: their state is driven by beliefs, goals, capabilities and plans, and their behavior is composed of a number of agency properties such as autonomy, adaptation, interaction, learning, mobility, and collaboration. A multi-agent system usually incorporates multiple objects and types of agents, with each agent type addressing distinct agency concerns. These agency concerns typically overlap and interact with each other, and so a disciplined scheme for composition is required. In this context, this paper presents and compares an aspect-based proposal with a new pattern-based proposal for building multi-agent software. Both proposals have the following goals: (i) minimize the misalignments between high-level agent models and object-oriented designs, (ii) promote the separation of agency concerns, (in) provide explicit support for disciplined composition of agency concerns in complex software agents, (iv) incorporate flexible facilities to build different types of software agents, and (v) allow the production of multi-agent software systems that are easy to understand, maintain and reuse. We demonstrate the applicability of the two proposals through the Portalware system, a Web-based environment f or the development of e-commerce portals.

A Study of the Relationships between Source Code Metrics and Attractiveness in Free Software Projects

A significant number of Free Software projects has been widely used and considered successful. However, there is an even larger number of them that cannot overcome the initial step towards building an active community of users and developers. In this study, we investigated whether there are relationships between source code metrics and attractiveness, i.e., the ability of a project to attract users and developers. To verify these relationships, we analyzed 6,773 Free Software projects from the SourceForge.net repository. The results indicated that attractiveness is indeed correlated to some source code metrics. This suggests that measurable attributes of the project source code somehow affect the decision to contribute and adopt a Free Software project. The findings described in this study show that it is relevant for project leaders to monitor source code quality, most specifically a few objective metrics, since these can have a positive influence in their chances of forming a community of contributors and users around the software, enabling further enhancement in its quality.

An Empirical Study on the Structural Complexity Introduced by Core and Peripheral Developers in Free Software Projects

Background: Several factors may impact the process of software maintenance and evolution of free software projects, including structural complexity and lack of control over its contributors. Structural complexity, an architectural concern, makes software projects more difficult to understand, and consequently more difficult to maintain and evolve. The contributors in a free software project exhibit different levels of participation in the project, and can be categorized as core and peripheral developers. Research aim: This research aims at characterising the changes made to the source code of 7 web server projects written in C with respect to the amount of structural complexity added or removed and the developer level of participation. Method: We performed a observational study with historical data collected from the version control repositories of those projects, recording structural complexity information for each change as well as identifying each change as performed by a core or a peripheral developer. Results and conclusions: We have found that core developers introduce less structural complexity than peripheral developers in general, and that in the case of complexity-reducing activities, core developers remove more structural complexity than peripheral developers. These results demonstrate the importance of having a stable and healthy core team to the sustainability of free software projects.

On the modularity assessment of aspect-oriented multiagent architectures: a quantitative study

A number of concerns in Multiagent Systems (MAS) have a broadly-scoped impact on the system architectural decomposition, which in turn hinder the design of modular MAS architectures. Typical examples of crosscutting concerns in MAS architectures include learning, mobility, coordination, and autonomy. Nowadays there are some architectural proposals that envisage an emerging aspect-oriented architectural pattern as a potential solution to address modularity shortcomings of conventional architectural patterns for MAS designs. However, little effort has been dedicated to effectively assess when and which of these emerging and traditional architectural solutions promote in fact superior modularity in the presence of crosscutting MAS concerns. This paper presents a quantitative comparison between aspect-oriented and conventional MAS architectures. Our analysis evaluates how the architectures under comparison support the promotion of enhanced modularity in the presence of architectural crosscutting concerns in MAS design. Our evaluation used two medium-sized MAS applications and was centred on fundamental modularity attributes.

Crosscutting interfaces for aspect-oriented modeling

Aspect-oriented software development promotes improved separation of concerns by introducing a new modular unit, called aspect, for the modularization of crosscutting concerns. As a new kind of modular unit, aspects should have explicit interfaces that describe the way they interact with the rest of the system and how they affect other modules. This interaction can be homogeneous or heterogeneous. In this paper, we present crosscutting interfaces as a conceptual tool for dealing with the complexity of heterogeneous aspects at the design level. Crosscutting interfaces have been incorporated by the aSideML modeling language in order to enhance aspect description at the design level. Moreover, we present a modeling notation for the description of architecture-level aspects that also supports the explicit representation of crosscutting interfaces. Finally, we present two large-scale case studies we have performed using this modeling language that support our arguments in favor of crosscutting interfaces.

Aspects in agent-oriented software engineering: lessons learned

Several concerns in the development of multi-agent systems (MASs) cannot be represented in a modular fashion. In general, they inherently affect several system modules and cannot be explicitly captured based on existing software engineering abstractions. These crosscutting concerns encompass internal agent properties and systemic properties, such as learning, code mobility, error handling, and context-awareness. In this context, it is important to systematically verify whether emerging development paradigms support improved modularization of the crosscutting concerns relative to MASs. This paper reports some lessons learned based on our experience in using aspect-oriented techniques and methods to address these problems. In the light of these lessons, related work and a set of future research directions are also discussed.

Enhancing agent-oriented models with aspects

The modular representation of concerns is imperative in the design modeling of multi-agent systems (MASs). However, MAS designers typically face a number of concerns that inherently affect several system agents and their respective internal elements, such as actions and goals. These crosscutting concerns encompass both internal and systemic properties, such as learning, mobility, error handling, and security. Without an explicit modeling of such MAS properties, designers can not properly communicate and reason about them and their broadly-scoped effects. The result is a multitude of agent-oriented design breakdowns: poor traceability from requirements to design models, hindering of reuse opportunities, and decreased evolvability. This paper presents an enhancement of an existing agent-oriented modeling language with aspects. We illustrate the use of our approach through some classical examples of crosscutting concerns in agent-oriented design models.