2022 Fiscal Year Research-status Report
Proposal on discussion element-specific deliberation based on autonomous agent facilitation for consensus building in complex society
Project/Area Number |
22K17948
|
Research Institution | Kyoto University |
Principal Investigator |
SAHAB Sofia 京都大学, 情報学研究科, 特定助教 (00892928)
|
Project Period (FY) |
2022-04-01 – 2025-03-31
|
Keywords | AI agent policy / Conversational agent / Online discussion / IBIS framework |
Outline of Annual Research Achievements |
In order to support diverse facilitation, I aimed to develop a discussion element-specific facilitation policy based on the Issue-Based Information System (IBIS). In the first term of research, I conducted literature studies on facilitation policy and proposed element-specific facilitation policy. In the second term, two types of element-specific facilitation policy proposed and completed the policy configurations of agents. The first proposed policy defines the agent behavior to interact with the author of proposed ideated solutions using predefined threshold while targeting proposed solutions. The proposed policy aims to question proposed ideated solution in order to raise the possibility of problems and challenges that lie within the ideated solution to form a diverse consensus. The second configured policy adopts idea-seeking policy using predefined threshold while targeting posted issues and challenges in order to promote ideated solutions for the posted issues and challenges. In order to evaluate the proposed approach, a set of experiments has been conducted. The experimental results with first policy showed that the agent had capable to stimulate discussion elements in line with proposed policy objectives. The results have been submitted to international conferences for publication.
|
Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
I have proposed and configured the discussion element-specific facilitation policy on conversational agents using the IBIS system, as a suitable format for critically facilitating problem-solving conversation. In order to evaluate the first proposed policy, conducted social experiment in real-world. The agent’s performance was submitted to two international conferences and I have prepared to submit it in an academic journal. Furthermore, I also start preparing to conduct a set of experiments to evaluate the second proposed policy efficiency. The works are going as well as I expected.
|
Strategy for Future Research Activity |
I plan to conduct a set of experiments to evaluate the second proposed policy efficiency. Additionally, in order to evaluate the performance of the proposed facilitation policies with existing policy, some cross-country social and controlled experiment will be required to explore the generalizability and applicability of the research findings in different cultural or contextual settings, and based on the experiment’s results, I can adjust design parameters of the conversational agent facilitation policies. The agent’s performance will be further analyzed and summarized to submit in an international conference or journal.
|
Causes of Carryover |
Travel expenses: Incurred for domestic and international conferences/workshops/seminars, including registration fees, travel costs, accommodation, and related expenses, to disseminate research outcomes, foster collaborations, and expand research impact. Labor costs: Incurred for research assistants for agent development and other research-related tasks to ensure smooth progress and timely completion of research milestones. Subject recruiting agencies, article publishing charges, statistical software, computer and monitor, and consumables: Expenses for hiring subject recruiting agencies for cross-country experiments, publishing research outcomes, upgrading statistical software, purchasing necessary equipment, and procuring consumables for group work, agent development, and research-related processes.
|