I am a research assistant at USC Information Sciences Institute, where I am advised by Prof. Mayank Kejriwal as part of the Prof. Mayank Kejriwal. I hold a Master’s degree from USC Viterbi School of Engineering, Computer Science where I was fortunate to be advised by Prof. Mohammad Rostami, focusing on Computer Vision and Domain Adaptation.

My research interests include Generative AI, such as Large Language Models (LLMs), and knowledge representation methods, particularly Knowledge Graphs. I am passionate about utilzing these technologies to build AI systems that are not only high-performing but also robust and trustworthy. Additionally, I am interested in Computational Social Science, particularly in using data science techniques to uncover insights and in designing effective visualizations that convey these complex insights clearly and compellingly.

Prior to this, I was a Senior Data Analyst/Scientist at Agoda where I collaborated with business teams to provide data-driven solutions that supported critical business decisions. I also hold a Bachelor’s degree in Mathematics and Computer Science from Chulalongkorn University, where I graduated with a class honor.

Research

My research focuses on evaluating and developing systems that utilize language models to perform various tasks. My research goals revolve around three key objectives:

1. Human-LLM Interaction: Exploring how humans can effectively leverage LLMs and how, in turn, LLMs can benefit from human knowledge intervention.
2. Human Intelligence Structures: Investigating methods to build human intelligence structures, such as Knowledge Graphs and ontologies, to enhance machine learning and language models.
3. Knowledge Graph Applications: Applying Knowledge Graphs to language models in areas such as question answering, commonsense reasoning, and text generation.

A full list of my publications is in this link

News

• 08/2024: Two works are accepted to ASONAM 2024. A main conference paper and our visualization tool (ISAC) (Demo track). See you in Italy!
• 08/2024: Our PMC-dataset is published to Data-in-Brief journal.
• 08/2024: Our paper: “Cost-Efficient Prompt Engineering for Unsupervised Entity Resolution in the Product Matching Domain” is published to Discover AI journal!
• 07/2024: Our paper: What if Red Can Talk? is accepted to ACL 2024 Wordplay workshop!
• 07/2024: Our paper about Multi-Source Data Integration for Segmentation of Unannotated MRI Images is published to IEEE Journal of Biomedical and Health Informatics!
• 05/2024: Our LLMs hallucination ontology HALO is accepted at SPIE2024!