Sonia Frutos

SONIA: a methodology for natural agent development

Agent-Oriented Software Engineering has emerged as a powerful engineering discipline that can deal with the complexity of todays software systems (primarily in distributed and open environments) better than other more traditional approaches. However, AOSE does not provide a software development process that naturally leads, if the problem so requires, to an agent architecture. Current agent development methodologies have two separate drawbacks. One is that development processes tend to target an agent organization, which is not necessarily always the best structure, as of the requirements definition stage. The other is that the identification and design of agents are complex, and designer experience plays an essential role in their definition. In this paper, we present the SONIA methodology (Set of mOdels for a Natural Identification of Agents) in an attempt to solve these problems. Based on a generic problem-independent analysis and a bottom-up agent identification process, SONIA naturally outputs an agent-based system.

Towards a Natural Agent Paradigm Development Methodology

It is indisputable that software development using agents and, more specifically, the multi-agent systems concept has greater potential for dealing with complex problems than other more traditional approaches. The agent paradigm is acquiring the status of an engineering discipline and gradually leaving the laboratory and moving into industry. However, it has two major omissions: it is missing an agent modeling language and a consolidated development process such as the object paradigm now has. Although we do not provide a definitive answer to this question in this paper, we do try to help to solve the problem as it relates to the agent-oriented development process by considering what features an agent-based development methodology should have, pointing out the omissions of current methodologies and presenting the SONIA methodology that includes the required features.

A formal framework for interaction protocol engineering

This paper presents a formal framework devised to support interaction protocol (IP) engineering. The proposed framework is organized into three views that consider all the stages of a protocol engineering process, i.e. the design, specification, validation, implementation and management of IPs. The modeling view allows visual IP design. The specification view automatically outputs, from the design, the syntactic specification of the IPs in a declarative-type language called ACSL, which improves IP publication, localization and machine learning by agents. Finally, the implementation view provides a formal operational semantics for the ACSL language. This semantics allows protocol property verification and eases automatic code generation from the ACSL specification for the purpose of simulating code execution at design time, as well as improving and assuring correct IP compliance at run time.

A generic blind user interface model

This paper presents BLIMD, a generic model to represent the process of interaction between blind users (or users with impaired vision) and the computer. This model is based on studies about interfaces for the blind and prototype programs developed at our research lab. Recent advances in graphical interfaces and blind technology are included to get a robust and flexible model. This model has been adopted and approved by the Spanish National Organisation of the Blind.

Adapting computer-human interaction in real time

There are various types of application-user interactions with different needs. It is almost impossible to define a communication process that covers all communication needs. The solution would be to redefine the dialog process (in run time) to adapt to people with specific characteristics, allowing a user to manage the application easily. We are proposing an independent module that could generate the process of interaction in real time and the communication process from an interpreted script.

The synthesis stage in the software agent development process

In most existing software agents methodologies, system analysis is dependent on an agent-oriented, object-oriented or knowledge-based design paradigm. This simplifies the complex transformation of the conceptual model produced during analysis into the physical model output at design time. We, like other authors, believe that the conceptual model has to be conceived as independent of the design paradigm and that the physical model should be driven by the solution, both models leading to very different conceptions of the problem. In this paper we present the SONIA agents development methodology that includes a transitional synthesis stage between analysis and architectural design that mends the break between the construction of the two models.

Engineering agent conversations with the DIALOG framework

This paper presents the rationale behind DIALOG: a formal framework for interaction protocol (IP) modeling that considers all the stages of a protocol engineering process, i.e. the design, specification, validation, implementation and management of IPs. DIALOG is organized into three views. The modeling view allows visual IP design. The specification view automatically outputs, from the design, the syntactic specification of the IPs in a declarative-type language called ACSL. This improves IP publication, localization and communication on the Web, as well as IP machine learning by agents. Finally, the implementation view provides a formal structural operational semantics (SOS) for the ACSL language. The paper focuses on the developed SOS, and shows how this semantics allows protocol property verification and eases automatic rule-based code generation from an ACSL specification for the purpose of simulating IP code execution at design time, as well as improving and assuring correct IP compliance at run time.

Information agents for optimal repurposing and personalization of web contents in semantics-aware ubiquitous and mobile computing environments

Web contents repurposing and personalization is becoming crucial for enabling ubiquitous Web access from a wide range of mobile devices under varying conditions that may depend on device capabilities, network connectivity, navigation context, user preferences, user disabilities and existing social conventions. Semantic annotations can provide additional information so that a content adaptation engine, based on the holistic integration of both Information Agents and Semantic Web technologies, can make better decisions, leading to optimal results in terms of legibility and usability. Bearing this in mind, this paper presents the rationale behind MorfeoSMC: an open source mobility platform that enables the development of semantics-aware mobile applications and services in order to provide improved Web accessibility and increase social inclusion. In particular, the paper focuses on how MorfeoSMC information agents tackle the use of semantic markup in the information rendered for users through the mobility platform and as part of a user-interest profile-aware and navigation context-aware Web content adaptation process. It also presents an innovative semantic matching framework that is at the core of this semantics-aware Web content adaptation process.