Knowledge Sharing Optimization Based on the Game Theory: A Bi-Level Programming Application

• Main Authors: Mohsen Tabatabaei, Abbas Afrazeh, Abbas Seifi
• Format: Article
• Language: fas
• Published: Allameh Tabataba'i University Press 2020-07-01
• Series: Muṭāli̒āt-i Mudīriyyat-i Ṣan̒atī
• Subjects: knowledge sharing bi-level programming game theory motivation-opportunity-ability framework
• Online Access: http://jims.atu.ac.ir/article_11229_d97793e5fa51e351e1fb58d908ad782c.pdf

This paper aims to propose a knowledge sharing optimization model based on the game theory that optimizes both employer and employee(s) decisions simultaneously. This model is a bi-level programming model. The upper-level problem includes employer decision about the reward, and the lower-level problem contains employee(s) decisions about time and effort allocation to knowledge sharing activity. Mathematical formulation of the model designed based on previous literature and in the framework of Motivation-Opportunity-Ability. The proposed bi-level programming model provides a foundation to investigate more different parameters comparing with previous models introduced in the literature. This model considers opportunity and ability factors in addition to the motivation. Also, payoff functions in this model are non-linear and therefore is more consistent with real cases relative to previous linear models. Additionally, this model analyzes the effects of available time as a key factor. The bi-level model coded in GAMS using EMP syntax and solved for a set of randomly generated data using BARON algorithm. Results indicated that the increase of applicability of codified knowledge and impact coefficient of social comparison could improve organizational performance and also save the cost of reward system. Therefore, neglecting these two parameters in designing a reward system could lead to under optimized decision making. This research provides a basis to consider more parameters simultaneously and help to improve organizational decisions. However, based on the results, BARON algorithm is not efficient to solve big problems, so developing a more efficient algorithm is needed.