Thesis Defense - Anıl Doğru (MSCS)

Anıl Doğru – M.Sc. Computer Science

Assist. Prof. Reyhan Aydoğan – Advisor

Date: 22.08.2022

Time: 14.00

Location: AB1 412

“Interpreting Human Arguments in Human Agent Negotiations”

Assist. Prof. Reyhan Aydoğan, Özyeğin University

Prof. Olcay Taner Yıldız, Özyeğin University

Associate Professor Arzucan Özgür, Boğaziçi University

Abstract:

Autonomous negotiating agents, which can interact with other agents, aim to solve decision-making problems involving participants with conflicting interests. Designing agents capable of negotiating with human partners requires considering some factors such as expressing the emotional state and arguments. For this purpose, we introduce a taxonomy of argument types capturing human arguments during the negotiation and propose an argument-based automated negotiating agent that can extract human arguments from a chat-based environment by using a BERT-based classifier. Consequently, the proposed agent is able to understand the received arguments and adapt its strategy with respect to them while negotiating with their human counterparts. We conducted human-agent negotiation experiments to construct a negotiation corpus. According to the experimental results, the proposed BERT-based classifier successfully extracted the arguments from the given text for the argument type taxonomy. Moreover, we observed that the proposed agent adapts its strategy based on these factors during a negotiation.

Bio:

Anıl Doğru is a member of OzU Artificial Intelligence Lab. He received his bachelor’s degree at Özyeğin University Computer Sciece Department in 2014. After receiving his Bsc degree in 2019, he started his master’s studies in the same department and institution. He worked both as a research assistant and teaching assistant in Özyeğin University Artificial Intelligence Laboratory under the supervision of Asst. Prof. Reyhan Aydoğan.