|
Title
|
|
|
|
A domain specific modeling language for semantic web enabled multi-agent systems
| |
|
Author
|
|
|
|
| |
|
Abstract
|
|
|
|
| The need for realization of the many different agent interactions and also agent properties such as autonomy, responsiveness and proactiveness, make development of agent-based software systems more complex than other software systems. Furthermore, the design and implementation of internal agent behavior model and interaction within the agent organizations of Multi-agent Systems (MASs) may become even more complex and difficult to implement especially when the new requirements and interactions for new agent environments such as the Semantic Web are taken into account. In order to deal with the complexities of design and implementation of such MASs, the use of a Domain-specific Modeling Language (DSML) can provide the required abstraction which increases the descriptive power of a MAS metamodel, defines the system semantics, and hence support a more fruitful methodology for the development of MASs. Although agent researchers have a great effort in MAS metamodeling and model-driven MAS development, a significant deficiency exists in current studies when it is considered providing a complete DSML for MASs especially working in Semantic Web environment. In order to help fill the abovementioned gap, a new DSML for MAS, called SEA_ML, is introduced in this study which considers the architectural complexity of design and implementation of MASs. To this end, a domain-specific metamodel is defined that includes new viewpoints (in addition to the classical viewpoints of MASs) which specifically support the development of software agents working within the Semantic Web environment. This new metamodel paves the way for the definition of an abstract syntax and textual/graphical concrete syntaxes for the DSML of agent systems. Achieved DSML syntax is supported with both textual and graphical modeling toolkits. The DSML also supports the interpreter mechanism, so that, the framework can generate the architectural code for different target platforms considering both the MAS and Semantic Web aspects. In the meantime, both of the textual and graphical languages support constraint control to validate the programs/models. In this way, some of the semantic controls are realized in the design time of the program/model and avoids potential errors which lead to more accurate code generation. Considering the formal semantics of the DSML, the formal representation of SEA_ML’s semantics are presented and used for MAS validation. The presented semantics defines both static and dynamic aspects of the interaction between software agents and semantic web services, in addition to the definition of the semantics already required for agent internals and MAS communication. Finally, in this thesis, an evaluation framework and systematic approach are also presented for assessing existing or newly defined DSMLs for MASs. Evaluation of SEA_ML according to this framework and systematic approach has been provided using two evaluator groups. The evaluation is based on a multi case study research approach and provides both qualitative evaluation and quantitative analysis. The lessons learned are discussed considering the adoption of the evaluation approach as well as the SEA_ML for supporting the generation of agent-based systems. Evaluation results show that approximately 85% of the artifacts and 80% of the whole code is generated automatically by using SEA_ML. Also, according to the evaluation results, the total time of development for both groups shows that the group using SEA_ML needs half of the time the other group needs for developing the case studies in average. |
| |
|
Language
|
|
|
|
English
| |
|
Publication
|
|
|
|
Izmir
:
Ege University, Graduate School of Natural and Applied Sciences
,
2016
| |
|
Volume/pages
|
|
|
|
220 p.
| |
|
Note
|
|
|
|
:
Kardaş, Geylani [Supervisor]
| |
|
Full text (publisher's version - intranet only)
|
|
|
|
| |
|