Adriana Leite

Architectural design of a multi-agent recommender system for the legal domain

Legal information sources are characterized by their growth and dynamism since new laws are written every day. Recommender systems are used as an approach to the information overload problem. Thus they can help professionals of the legal area to deal with legal information sources. This paper describes the architectural design of Infonorma, a multi-agent recommender system for the legal domain. Infonorma monitors a repository of legal normative instruments and classifies them into legal branches. Each user specifies his/her interests for certain legal branches and receives recommendations of instruments they might be interested in. The information source is entirely written according to Semantic Web standards. Infonorma was developed under the guidelines of MAAEM, a software development methodology for multi-agent application engineering.

Application ontology for the development of case-based intrusion detection systems

This work proposes an application ontology who is capable of formally represent the concepts present in the domain of Information Security together with the intrusion detection systems and case-based reasoning. The ontology was evaluated through the development of an IDS capable of detect computers networks attacks and recommend actions to such attacks. The results showed that the developed IDS presented good effectiveness in the detecting attacks, and so it is concluded that the proposed ontology conceptualizes properly the domain concepts and task.

A PROCESS FOR MULTI-AGENT DOMAIN AND APPLICATION ENGINEERING - The Domain Analysis and Application Requirements Engineering Phases

Domain Engineering is a process for the development of a reusable application family in a particular domain problem, and Application Engineering, the one for the construction of a specific application based on the reuse of software artifacts in the application family previously produced in the Domain Engineering process. MADAE-Pro is an ontology-driven process for multi-agent domain and application engineering which promotes the construction and reuse of agent-oriented applications families. This article introduces an overview of MADAE-Pro emphasizing the description of its domain analysis and application requirements engineering phases and showing how software artifacts produced from the first are reused in the last one.

A hybrid and learning agent architecture for network intrusion detection

A hybrid learning software agent for network intrusion detection is proposed.The agent combines a reactive and a case-based reasoning performance components.A learning component allows to improve the performance by learning reactive rules.The performance and effectiveness of the agent were evaluated.Evaluation results show that the agent is well-suited for dynamic environments. Learning is an effective way for automating the adaptation of systems to their environment. This ability is especially relevant in dynamic environments as computer networks where new intrusions are constantly emerging, most of them having similarities and occurring frequently. Traditional intrusion detection systems still have limitations of adaptability because they are just able to detect intrusions previously set in system design. This paper proposes HyLAA a software agent architecture that combines case-based reasoning, reactive behavior and learning. Through its learning mechanism, HyLAA can adapt itself to its environment and identify new intrusions not previously specified in system design. This is done by learning new reactive rules by observing recurrent good solutions to the same perception from the case-based reasoning system, which will be stored in the agent knowledge base. The effectiveness of HyLAA to detect intrusions using case-based reasoning behavior, the accuracy of the classifier learned by the learning component and both the performance and effectiveness of HyLAA to detect intrusions using hybrid behavior with learning and without learning were evaluated, respectively, by conducting four experiments. In the first experiment, HyLAA exhibited good effectiveness to detect intrusions. In the second experiment the classifiers learned by the learning component presented high accuracy. Both the hybrid agent behavior with learning and without learning (third and fourth experiment, respectively) presented greater effectiveness and a balance between performance and effectiveness, but only the hybrid behavior showed better effectiveness and performance as long as the agent learns.

A Reference Architecture of a Hybrid Learning Agent

A software reference architecture specifies a generic architectural solution for the development of specific software architectures. It includes common components to all software architectures and their relationships, a common vocabulary, a mapping methodology for realizing a specific architecture and good design practices. Software agents represent an evolution of traditional software, having the ability to control their own behavior and acting with autonomy. Typically, software agents act reactively, where actions and perceptions are predefined at design time, or in a deliberative way, where the corresponding action for a given perception is found at run time through a process of reasoning. However, to perform better, software agents should act using both forms of behavior with learning abilities in a hybrid way. In this paper, a reference architecture that specifies a generic architectural solution for the development of specific architectures of hybrid learning agents is presented. An example of realization of this architecture in the network intrusion domain is also presented.

A Case-Based Reasoning Architecture of a Hybrid Software Agent

Software agents represent an evolution of traditional software, having the ability to control their own behavior and acting with autonomy. Typically, software agents act reactively, where actions and perceptions are predefined at design time, or in a deliberative manner, where the corresponding action for a given perception is found at run time through a process of reasoning. However, especially in dynamic environments, to perform better, software agents have to act using both forms of behavior and also have learning abilities to follow up changes in the environment. Case-based reasoning is a technique that uses previous experience to solve new problems. This kind of reasoning is particularly appropriate for learning, since learning is already part of the life cycle of case-based reasoning. This paper proposes a hybrid software architecture that combines case-based reasoning, reactive behavior and learning. The main distinguishing feature of the proposal is the learning of reactive behavior, more quickly and efficiently, through continuous interactions of the agent with its environment. Currently, a software agent based on this architecture is being developed in order to evaluate it.

Using Ontologies in Hybrid Software Agent Architectures

Having the properties of autonomy, sociability and learning ability, software agents provide a better approach to the increasing complexity of both software problems and solutions and to support the decision making process. An important decision when developing a software agent is the choice of its internal architecture. Several models of deliberative and reactive architectures have already been proposed. However, approaches of hybrid software architectures that combine deliberative and reactive components, with the advantages of both behaviors, are still an open research topic. This work aims at contributing to the development of complex systems through the proposal of ontology-driven hybrid software agent architecture, by exploring automated reasoning and learning techniques. The architecture is composed of a reactive system and a deliberative one. Through a learning mechanism, the set of inference rules in the ontology of the deliberative system are transformed into reactive rules to be used by the reactive system, thus providing a more effective decision.