Format: On Campus
Number of Students: 1
Duration of Project: 24.06.2024 – 19.07.2024
Project Supervisor: Assoc. Prof. Reyhan Aydoğan
Research Areas: Artificial Intelligence, Explainable AI, Recommender Systems
Daily Supervisor: Berkecan Koçyiğit

Interactive Intelligent Systems Laboratory

We design, develop, and analyze agent technologies that integrate different aspects of intelligence, such as reasoning, decision-making, and learning. We aim to support human decision-makers in complex and dynamic environments, which also require the design of effective human-computer interaction.

Our primary expertise is designing and developing agent-based negotiation approaches. Recently, we have focused on human-agent negotiations where a robot or a Web bot negotiates with its human counterpart on a particular topic (e.g., resource allocation, travel arrangement, etc.). The main challenges are natural language processing, emotion recognition, argument generation, and designing bidding strategies. In addition, we are working on personalized food recommendation systems involving agent-human explainability.

You can find the research projects we worked on in the following link:

Interactive Intelligent Systems Laboratory Projects

About the Project: Enhancing Interaction for Negotiation and Recommender Systems Using Large Language Models

Project Description:

The "Expectation" project at Özyeğin University's Interactive Intelligent Systems Lab is dedicated to advancing the development of explainable AI systems and multi-agent negotiation. This summer, we are seeking a dedicated high school student to contribute to our efforts in enhancing interaction and data labeling within our negotiation and recommender systems using Large Language Models (LLMs) such as ChatGPT.

Project Objectives:

1. Interaction Enhancement: The primary objective of this project is to enhance the interaction capabilities of our negotiation and recommender systems. The student will collaborate with our research team to integrate LLMs for improving the natural language processing capabilities, enabling more natural and effective interactions.
2. Data Labeling: The student will also assist in labeling data using LLMs to improve the accuracy and relevance of our systems. This will involve generating contextually relevant data to augment our existing datasets and refining data labeling processes.

Project Tasks:

• Exploring the integration of Large Language Models for improving interaction in negotiation and recommender systems.
• Assisting in the labeling of data using LLMs to enhance the datasets.
• Reviewing and refining interaction strategies based on user feedback and domain knowledge.
• Generating synthetic data to supplement the existing dataset.
• Evaluating the impact of LLM integration and data labeling on the effectiveness of the negotiation and recommender systems.

Deliverables:

• Enhanced interaction models with improved natural language processing capabilities.
• Annotated dataset enriched with contextually relevant data generated by LLMs.
• Report summarizing the methodology, findings, and implications of LLM integration and data labeling.
• Presentation of project outcomes to the research team.

Benefits for the Student:

• Hands-on experience with Large Language Models and data labeling.
• Exposure to cutting-edge techniques in natural language processing and AI research.
• Collaboration with experienced researchers in the field of explainable AI systems.
• Contribution to a real-world research project with potential for publication and impact.

Requirements:

• Fluent in English
• Familiarity with API’s
• Basic Python programming language knowledge

Our applications are closed. Thank you for your intense interest.