系统工程理论与实践
繫統工程理論與實踐
계통공정이론여실천
Systems Engineering—Theory & Practice
2014年
2期
382~391
,共null页
可计算组织理论 可计算组织模型 模型验证
可計算組織理論 可計算組織模型 模型驗證
가계산조직이론 가계산조직모형 모형험증
computational organization theory; computational organization model; model validation
模型“有效性”能否得到验证是推动计算组织理论发展的必要条件,也是一大瓶颈.立足于计算组织理论的前沿地位,论文综述并讨论了可计算组织模型的验证难点和验证方法.首先,可计算组织模型验证区别于自然工程模型,也超越了传统实证方法,在验证过程面临诸多主观、客观的难点;其次,发展基于问题域、建模目的、参照系三者内在联系的验证情境,从“度”上考量模型有效性,是可计算组织模型的有效性本质;再次,建立区别于自然工程系统的VV&A验证方法,关注概念效度验证、操作效度验证以及数据效度验证,遵循迭代验证原则,是未来的研究趋势.本课题也列举了相应的验证案例;最后,可计算组织模型验证应该是一个不断增强模型信心的过程,亟待跨领域专家执着于该领域的研究兴趣.
模型“有效性”能否得到驗證是推動計算組織理論髮展的必要條件,也是一大瓶頸.立足于計算組織理論的前沿地位,論文綜述併討論瞭可計算組織模型的驗證難點和驗證方法.首先,可計算組織模型驗證區彆于自然工程模型,也超越瞭傳統實證方法,在驗證過程麵臨諸多主觀、客觀的難點;其次,髮展基于問題域、建模目的、參照繫三者內在聯繫的驗證情境,從“度”上攷量模型有效性,是可計算組織模型的有效性本質;再次,建立區彆于自然工程繫統的VV&A驗證方法,關註概唸效度驗證、操作效度驗證以及數據效度驗證,遵循迭代驗證原則,是未來的研究趨勢.本課題也列舉瞭相應的驗證案例;最後,可計算組織模型驗證應該是一箇不斷增彊模型信心的過程,亟待跨領域專傢執著于該領域的研究興趣.
모형“유효성”능부득도험증시추동계산조직이론발전적필요조건,야시일대병경.립족우계산조직이론적전연지위,논문종술병토론료가계산조직모형적험증난점화험증방법.수선,가계산조직모형험증구별우자연공정모형,야초월료전통실증방법,재험증과정면림제다주관、객관적난점;기차,발전기우문제역、건모목적、삼조계삼자내재련계적험증정경,종“도”상고량모형유효성,시가계산조직모형적유효성본질;재차,건립구별우자연공정계통적VV&A험증방법,관주개념효도험증、조작효도험증이급수거효도험증,준순질대험증원칙,시미래적연구추세.본과제야열거료상응적험증안례;최후,가계산조직모형험증응해시일개불단증강모형신심적과정,극대과영역전가집착우해영역적연구흥취.
Whether the "validity" of a model can be validated is necessary to promote the development of computational organization theory, but is also a major bottleneck. Based on the forward position of computational organization theory, this paper reviewed and discussed the validation difficulties and validation method of computational organization model. First, validation of computational organization ~nodel is different from validation of natural engineering model and is also beyond the traditional empirical method, and faces many subjective and objective difficulties in the validation process; secondly, developing a validation scenario which is based on the internal relation of problem domain, modeling purpose, reference, and considering model validity from "degree" aspect is the essence of computable model validity; thirdly, (i) constructing a validation method which is different from VV&A validation method of natural engineering system, (ii) focusing on the validation of concepts validity, operation validity and data validity, and (iii) following iterative validation principle, are the reseurch trends in the future. This paper also lists some related validation cases; finally, validation of computational organization model should be a process which increases model confidence, and calls for cross-domain experts to insist on the research interest of the field.