| Organizations exist to solve any number of goals or problems. Typically, an organization is tasked with solving problems an individual cannot resolve without assistance from others. Human organizations serve as an inspiration to create models of organization which can be applied to agents and multiagent systems. Capturing how humans form and interact with a capability-based organization model is a complex task. Applying the same model to software agents requires essentially the same structures and relationships as with human organization models. There have been many research efforts capturing specific functional elements of organization applied to a single problem domain or a similar set of problem domains. In these cases, the research and solution may work and possess functionality to solve the unique problem, but won't necessarily translate into other problem domains or capture the general function of an organization.;The primary goal of my dissertation is to develop and formalize an organization model and transition algorithms, generic in definition but capable of capturing organization dynamics in domains in which intelligent agents act. The model's generic nature will allow it to be applied to any problem where organization is required. The organization model will have structural, state and transitional elements. While the scope of this research will include the establishment and design of structural and state elements, the main scope of work will be in the formal definition, development and evaluation of transition processes and algorithms.;The transition formalization will include a formal transition model and relations, central, distributed and segmented transition algorithms. The formal model and transition algorithms will be implemented for empirical evaluation and analysis. The implemented instances will then be compared for evaluation using several task domains used to validate the model. |
