| 研究課題/領域番号 |
17F17308
|
|---|
| 研究機関 | 東京大学 |
|---|
研究代表者 |
チン ユ 東京大学, 大学院新領域創成科学研究科, 教授 (00272394)
|
|---|
| 研究分担者 |
GU JIE 東京大学, 新領域創成科学研究科, 外国人特別研究員
|
|---|
| 研究期間 (年度) |
2017-10-13 – 2020-03-31
|
| キーワード | agent-based modeling / behavioral experiment / knowledge management / complexity and evolution / risk and uncertainty |
|---|
| 研究実績の概要 |
The ultimate goal is to establish a new field of study on evolutionary knowledge management (KM) with the advancement of agent-based modeling (ABM) and behavioral experiments. Several model extensions were implemented, e.g.one on KM under turbulent and stable environment, one with monetary and social incentives and dilemmas, and one on KM impact to the macroeconomic outcomes. With deep understanding of human agents' independent and interdependent behavior from the empirical experiments, computer agents were modeled in the simulation emulating the human reality. Then, large scale and long term results were produced by the ABM which also exhibit the inter-temporal dynamics of the social interactions. With enhancement of the simulation, practical and effective administrative policy for optimization can be suggested and tested.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
Although the research project involves intensive labor and computer effort, the output has been delivered as planned, which includs a book chapter of Advances in Intelligent Systems and Computing on Springer publisher, a conference paper in the Proceedings of IKMAP2018, a poster and oral presentation at 9th International Conference on Complex Systems, a conference paper in the Proceedings of JSAI2019. Not only in academia, several industry-collaborated projects are also carried out. A side project on developing an Agent-Based Model of Tokenomics with X-order Lab in Shanghai, and a project on cryptocurrency ecology with a wealth management in HK, and a stakeholder conflict resolution ABM with KTH university Sweden. These efforts will also extend academic value of complex system simulation.
|
| 今後の研究の推進方策 |
Much future work is to be done. First is to improve, re-evaluate, deepen and consolidate the developed agent-based model. Hence, further theory and inside can be derived. Also, some system exhibited phenomena are not yet elucidated, such as phase transition and boundary conditions. Therefore, when doing the econophsical analysis, model revision or recheck the computing codes are needed. The integration of ABM and behavioral experiments is advantageous but difficult. Human vs. Computer agents' decision-making needs to be further explored. Additionally, many ongoing work needs to be published as journal papers. The research execution has been carried out smoothly, but the paper publication has been relatively slow. Hence, the productivity of publication needs to be improved as well.
|
