Patches as an expressive medium for exploratory multi‐agent modelling
Patches as an expressive medium for exploratory multi‐agent modelling
复制标题
补丁作为探索性多智能体建模的表达媒介
DOI：
10.1111/bjet.13087
复制
发表时间：
2021
影响因子：
6.6
通讯作者：
Brady, Corey
中科院分区：
文献类型：
--
作者：
Brady, Corey
本文概述了建构主义研究和建模实践的潜在收获，通过认识到补丁的力量-NetLogo建模环境中的一个不起眼的计算实体，已经被其更受欢迎的同事海龟所掩盖。为了将这一机会与背景联系起来，我描述了建构主义建模是如何通过推广依赖于学习者与代理人的认同的学习形式而蓬勃发展的。我认为补丁是这种多主体建模传统中被忽视的主体类型，而学习者采用补丁视角支持探索性造型和审美表达的可能性一直没有得到充分的研究。尽管如此，我展示了学习者有多种有效的方式--无论是单独的还是集体的--来认同补丁。我描述了正在进行的研究，这些研究表明，对补丁采取美学方法如何潜在地支持个人和群体通过多智能体建模和关于多智能体建模的强大形式的学习。实践者注意到关于这个主题的已知情况海龟(在Logo中的可移动智能体和从Logo派生的建构主义环境中的可移动智能体)可以是为学习者提供自己产生强大想法的方法的“过渡对象”。这些智能体可以成为强大的“思考对象”，在很大程度上是因为它们鼓励学习者以一种被称为“和谐学习”的学习形式认同它们。利用学习者的审美兴趣的激励性活动可以支持他们的学习多代理建模是基于LOGO的学习环境的强大扩展，其提供对关于复杂系统及其紧急性质的强大想法的访问。在多代理设置中，在多智能体建模的研究中，补丁(NetLogo建模环境中不可移动的智能体)的代表性不足，忽视了学习者采用补丁视角的可能性。审美驱动的补丁方法可以为学生理解其表达价值奠定基础。学习者扮演补丁角色的参与性活动(称为体育卡活动)可以为补丁视角奠定基础。在斑块网格上融合内在和外在视角的参与性活动可以进一步提高学习者对斑块编程的熟练程度和他们对斑块集体表达能力的理解。实践和/或策略的意义平衡海龟和斑块之间的关注点可以丰富基于代理的建模的初学者的建模工具箱。斑块捕捉个体和集体经验的重要方面，因此可以是很好的思考对象，特别是在更大规模的概念化现象时。斑块网格的表达潜力也是计算机科学的一个重要主题(例如，通过2D元胞自动机)。这本身就是一个丰富的学习环境，可以通过物理或虚拟的参与式角色扮演使学习者群体能够访问。此外，物理或虚拟的人-斑块网格可能显示出新的聚合计算属性，这些属性反过来可能成为计算机科学研究的有趣领域。
This paper outlines the potential gains for Constructionist research and praxis in modelling that might be obtained by recognising the power of the Patch—a humble computational being in the NetLogo modelling environment that has been overshadowed by its more popular fellow agent, the Turtle. To contextualise this opportunity, I describe how Constructionist modelling has thrived by promoting forms of learning that rely on learners’ identifying with agents. I argue that patches are a neglected agent type in this multi‐agent modelling tradition, and that the possibilities for learners to adopt the patch perspective in support of exploratory forms of modelling and aesthetic expression have been under‐researched. Nevertheless, I show there are a variety of powerful ways for learners––both individually and in groups––to identify with patches. I describe ongoing research showing how taking an aesthetic approach to patches has the potential to support individuals and groups in powerful forms of learning with and about multi‐agent modelling.Practitioner notesWhat is already known about this topicTurtles (movable agents in Logo and Constructionist environments descended from Logo) can be ‘transitional objects’ that provide learners a way to make powerful ideas their own.These agents can be powerful ‘objects‐to‐think‐with’ in large part because they encourage learners to identify with them in a form of learning known as ‘syntonic learning’.Expressive activities that draw on learners’aestheticinterests can support their learning with and about computational representations.Multi‐agent modelling is a powerful extension of Logo‐based learning environments that provides access to powerful ideas about complex systems and their emergent properties.In the multi‐agent setting, individual learners and/or groups of learners can identify syntonically with agents to provide entry points for reasoning about complexity.What this paper addsPatches (non‐movable agents in the NetLogo modelling environment) are under‐represented in the research on multi‐agent modelling, and the potential for learners to adopt the patches’ perspective has been neglected.An aesthetically driven approach to patches can ground students’ understanding of their expressive value.Participatory activities in which learners play the role of patches (called ‘Stadium Card’ activities) can ground the patch perspective, so that learners can achieve a form of syntonicity and/or collectively adopt the perspective of patches in the aggregate.Participatory activities that blend intrinsic and extrinsic perspectives on the patch grid can further enhance learners’ facility with programming for patches and their understanding of patches’ collective expressive power.Implications for practice and/or policyBalancing the focus between turtles and patches can enrich the modelling toolbox of learners new to agent‐based modelling.Patchesdocapture important aspects of individual and collective experience, and so can be good objects‐to‐think‐with, especially when conceptualising phenomena at a larger scale.The expressive potential of the patch grid is an important topic for computer science as well (eg, through 2D cellular automata). This is a rich context for learning in itself, which can be made accessible to groups of learners through physical or virtual participatory role‐play.Moreover, physical or virtual grids of people‐patches may exhibit novel aggregate computational properties that could in turn become interesting areas for research in computer science.