Aulti-actor simulation of organisations--论文代写范文精选

动力社会科学采用MAS创建社会模拟，为了阐明组织和社会现象。胃镜理论问题存在于社会科学的高龄化，缺乏理解或解释。此外,莫斯指出,避开形式主义盛行的社会科学。下面的论文代写范文进行论述。

Introduction
  M ULTI-AGENT SYSTEMS (MAS) is a eld of research that is concerned with the design, analysis and implementation of computational decision and simulation systemsa collection of agents that work in conjunction with each otherto enhance one's understanding of how people cooperate and coordinate their actions to solve problems. Upon its foundation in the mid-nineties, MAS propagated in the elds of cognitive science and social science, and was adopted, more specically, in cognitive modelling and modelling of multi-agent interaction as in, for instance, social simulation (Conte & Gilbert, 1995; Troitzsch, 1997). However, there has not been enough interaction between these two elds, and tools that try to connect both elds have not been sufciently developed (Sun, 2006b). In our research, we will address this problem and make an attempt to bring both eldscognitive science and social sciencecloser together. This chapter is structured as follows. Section 1.1 discusses the motivation for selecting organisational, social and cognitive theories and the need for the development of a cognitive agent-based computational social simulation model. Section 1.2 presents the research background of the dissertation. In section 1.3, we will elaborate on the methodology applied in this research. Next, the research questions and objectives are outlined in section 1.4 and we end this chapter with section 1.5 that provides an outline of the dissertation. 

  1.1 Motivation Social sciences have adopted MAS for the creation of social simulations in order to shed light on organisational and sociological phenomena. However, as Castelfranchi (2001) points out, a signicant theoretical problem exists in the eld of social sciences; there is a lack of understanding or explanation of unconscious, unplanned forms of cooperation among intentional agents. Moreover, Moss (2006) states that science that eschews observation in favour of formalism is rife in social science. His claim is that agent-based models can capture independent validation by . . . ensuring that the specication is well veried with respect to formal models from cognitive science understanding that those formal models are themselves well validated experimentally and observationally (ibid., p. 398). 

  According to Castelfranchi (2001), Articial Intelligence (AI) and, in particular, MASthe cognitive modelling of agents and its learning aspectscan contribute to the simulation of articial societies1 . On the other hand, cognitive science needs Multi-Agent Systems, social simulation, and social sciences in general (Sun, 2006b). There is a real need for new complex cognitive models that take into account social factors in cognition. In other words, there is a shortage of good ideas and theories that address socio-cultural concepts/signs/symbols within social structures from a cognitive standpoint (Sun, 2001). Cognitive agent-based social simulations have adopted MAS to give insights into sociallevel phenomena based on the individual agent's actions, i.e. the cognitive level. 

  The aim of this dissertation is to design and implement a cognitive agentbased computational social simulation model based on a selection of social and cognitive theories in an attempt to satisfy the need for a complex cognitivesocial model. In comparison to models emanating from traditional social (and economic) sciences, cognitive-social models enable one to gather, organise, and interpret observations and data more directly and test various factors more thoroughly (Sun, 2006b). Furthermore, such models are validated by incorporating behavioural descriptions and previously validated concepts from cognitive science and social psychology at the micro level, and by capturing the statistical output of (longitudinal) studies in the eld of social science at the macro level (Moss, 2006). 

  The motivation for this research is twofold: (1) to reveal the constituents of MAS and the social cognitive agent (cf. Conte & Castelfranchi, 1995a) based on theoretical considerations to (2) construct a simulation model that plausibly explains the interactive social behaviour of agents in a physical and socially situated environment (cf. Gilbert & Troitzsch, 1999). The aim is to relate the behaviour of individuals (micro level) that form a group or an organisation to the behaviour and performance of a group or organisation as a whole (Alexander & Giesen, 1987; Van den Broek, 2001). This rests on the `methodological individualism' assumption that supports MAS methodology: all description and explanation of social phenomena should ultimately be in terms of individuals, their properties, and their interrelations in terms of these properties (Franssen, 1997) (Van den Broek, 2001, p. 26).

  Similarly, it is established in social constructivism (Mead, 1934) that society and all its related issues, such as money and marriage, are products of (symbolic) social interaction between individuals. These interactions are caused by actions of individuals as part of a group. Accordingly, the individual and its interactions with others are the objects of study in this dissertation. Semiotics, as a study of signs and sign systems, is also addressed in this dissertation and is concerned with the `symbolic interaction' of individuals and the (shared) meaning that they assign to constructs (i.e. semiotic codes) which they create through semiosis2 . 

In MAS, it is common to dene varying levels of description; for instance from the biological, cognitive, rational to the social level of description (Newell, 1990). Social and economic sciences often operate at the social level. This is acceptable so far as the properties of actors at lower levels are assumed to be constant (Gilbert, 2006). However, when the analysis of lower levels constrains that at higher levels, or when there is a need to model the social and cognitive levels, then a `mixed-level' analysis is necessary (Sun, 2006b). Under that circumstance, descriptions at one level can be applied to other levels (Gilbert, 2006).

  A complete design and implementation of a MAS has to incorporate the basic aspects of all the levels listed above. As signs and symbols are common to all levels, MAS and its agents should support the processing of signs and symbols. The implication for the overall system is that: 
  1. The actor should be able to process and create signs and symbols, i.e. a cognitive plausible agent is a requirement. 
  2. The environment should support: (a) Physical objects that represent signs and symbols. (b) A communication and social environment that provides an infrastructure for the transfer of signs and symbols and allows actors to interact in a coordinated manner. 
  3. The actor must have the capabilities to (socially) construct the world and use a common medium, or language, to exchange, negotiate and reach agreements about (social) constructs with other actors. A simulation with a MAS can provide (1) valuable insights into and description of organisational behaviour, cognitive and social phenomena (for instance, emergence of organisations), (2) locate errors and gaps in verbal theories, (3) demonstrate which theoretical propositions are logically consistent, (4) predict the state of an agent or system in the future, and (5) discover, formalise and test (new) theories based on the simulation outcomes (Carley, Prietula, & Lin, 1998; Gilbert & Troitzsch, 1999).
  In this dissertation, we will provide a selection of theories that are applicable for the design of a MAS. We will construct a model of a MAS, named MRS (Multi-RBot System) and a cognitive and social actor, named RBot, by drawing on these theories. An overview of these theories is provided in the following section.(论文代写)


  51Due网站原创范文除特殊说明外一切图文著作权归51Due所有；未经51Due官方授权谢绝任何用途转载或刊发于媒体。如发生侵犯著作权现象，51Due保留一切法律追诉权。
  更多论文代写范文欢迎访问我们主页 www.51due.com 当然有论文代写需求可以和我们24小时在线客服 QQ:800020041 联系交流。-X(论文代写)