arXiv subjects

Kihwan Han

Publications and source records attributed to Kihwan Han.

3 recordsLinked to original sources

arXiv2026

Expert-validated STEM QA

Recent advancements in AI are helping scientists achieve breakthroughs in fields such as mathematics, medicine, and materials sciences. New evaluation datasets for AI models contribute to such advancement in AI. In the STEM domain, frontier models have consumed most of the available online data, creating the need for human-created datasets that codify the knowledge of leading experts in the domain. There are several STEM datasets available for the research community in this field. However, there are some gaps in these datasets, leaving room for improvement. Examples of gaps include (1) saturation in model performance on these datasets, leaving no head-room for meaningful evaluations, (2) skewed taxonomy distributions, (3) multiple choice question format that is misaligned with how scientists use AI in the real world, and (4) inaccurate answers and rationales partially led by a contest-based data collection and a time-bound review process. In this study, we present 'Expert-validated STEM QA', a high-quality, expert-validated STEM dataset (N=398) in Physics, Chemistry, Biology, and Mathematics, created by 241 domain experts. We (1) carefully designed a taxonomy with balanced distribution, (2) vetted question contributors with quality-driven incentive, (3) conducted multiple rounds of reviews with revisions validated by domain experts based on consensus, and (4) created the dataset in verifiable question and answer format. Our study demonstrated low performance ($<25\%$) of frontier AI models on the dataset as a benchmark. Post-training on a separate, private version of the dataset (N=2,000) increased performance of the open source model by $15\%$ relative to the baseline model (p=0.045) on the STEM subset of HLE-verified dataset, indicating potential utility of the dataset for model training. We have open-sourced a portion of our dataset for the AI research community.

cs.AI↗

arXiv2024

Applying graph neural network to SupplyGraph for supply chain network

Supply chain networks describe interactions between products, manufacture facilities, storages in the context of supply and demand of the products. Supply chain data are inherently under graph structure; thus, it can be fertile ground for applications of graph neural network (GNN). Very recently, supply chain dataset, SupplyGraph, has been released to the public. Though the SupplyGraph dataset is valuable given scarcity of publicly available data, there was less clarity on description of the dataset, data quality assurance process, and hyperparameters of the selected models. Further, for generalizability of findings, it would be more convincing to present the findings by performing statistical analyses on the distribution of errors rather than showing the average value of the errors. Therefore, this study assessed the supply chain dataset, SupplyGraph, with better clarity on analyses processes, data quality assurance, machine learning (ML) model specifications. After data quality assurance procedures, this study compared performance of Multilayer Perceptions (MLP), Graph Convolution Network (GCN), and Graph Attention Network (GAT) on a demanding forecasting task while matching hyperparameters as feasible as possible. The analyses revealed that GAT performed best, followed by GCN and MLP. Those performance improvements were statistically significant at $\alpha = 0.05$ after correction for multiple comparisons. This study also discussed several considerations in applying GNN to supply chain networks. The current study reinforces the previous study in supply chain benchmark dataset with respect to description of the dataset and methodology, so that the future research in applications of GNN to supply chain becomes more reproducible.

cs.LG↗

arXiv2023

A Comparative Study of Text Embedding Models for Semantic Text Similarity in Bug Reports

Bug reports are an essential aspect of software development, and it is crucial to identify and resolve them quickly to ensure the consistent functioning of software systems. Retrieving similar bug reports from an existing database can help reduce the time and effort required to resolve bugs. In this paper, we compared the effectiveness of semantic textual similarity methods for retrieving similar bug reports based on a similarity score. We explored several embedding models such as TF-IDF (Baseline), FastText, Gensim, BERT, and ADA. We used the Software Defects Data containing bug reports for various software projects to evaluate the performance of these models. Our experimental results showed that BERT generally outperformed the rest of the models regarding recall, followed by ADA, Gensim, FastText, and TFIDF. Our study provides insights into the effectiveness of different embedding methods for retrieving similar bug reports and highlights the impact of selecting the appropriate one for this task. Our code is available on GitHub.

cs.SE↗