Vol. 52, No. 7,
Jul. 2025
Digital Library
Optimizing HBM-PIM Throughput through DRAM ACT and PRE Hiding
http://doi.org/10.5626/JOK.2025.52.7.557
Modern applications, such as Large Language Models (LLMs), increasingly demand high memory bandwidth, which is difficult to meet using conventional memory devices alone. The insufficiency in memory bandwidth causes significant performance bottlenecks, as excessive time is spent on data transfers between the host processor and memory. Processing in memory (PIM) architectures address this challenge by placing processing units (PUs) near memory banks, offloading tasks from the host and leveraging DRAM internal bandwidth. HBM-PIM is one of the PIM devices developed in practice, which features one PU per two banks and enables parallel operations across all PUs. In this paper, we conduct an in-depth analysis of HBM-PIM operations, taking into account the DRAM microarchitecture. A detailed examination of HBM-PIM's microarchitecture reveals that its characteristics are not fully exploited. Based on this insight, we propose optimization techniques that leverage the structural features of HBM-PIM and can be implemented without hardware modification. By modifying the order of instructions, adjusting data mapping, and loosening memory barriers, we minimize latency caused by DRAM row conflicts and improve the performance of HBM-PIM. Our optimizations yield average performance improvements of 1.15×, 1.43×, and 1.29× for GEMV, ADD/MUL, and ReLU operations, respectively.
An effective Seed Selection Method for Maximizing the Performance of Symbolic Execution
http://doi.org/10.5626/JOK.2025.52.7.578
Symbolic execution is a promising software testing technique that aims to maximize the coverage of executed code areas of program under test by effectively generating test cases. A well-known challenge in symbolic execution is the high cost associated with solving path conditions. One solution to this challenge is to utilize generated test cases as seed inputs for a symbolic execution tool, thereby reducing the number of required solver calls. The effectiveness of this solution depends on the selection of appropriate test cases as seed inputs. This paper proposes a method to enhance the performance of symbolic execution by grouping the generated test cases into clusters, identifying the most promising cluster and selecting the most potential seed input from within it. Experimentally, the proposed method achieved 42.0% more branch coverage on average than traditional symbolic execution tools without seed inputs.
Aspect-Based Comparative Summarization with Large Language Model
http://doi.org/10.5626/JOK.2025.52.7.579
This paper proposes an aspect-based comparative summarization method to generate summary comparisons between two items based on their reviews, aiming to assist users in making informed decisions. Given the reviews of two items, aspects are dynamically generated from each review using a large language model. To identify common aspects for comparison, the generated aspect lists of both items are merged. The review sentences of each item are classified into the most relevant aspects, and then the summarization process removes redundant and unnecessary information. Subsequently, an abstract summary is generated for each common aspect to capture the overall content of the reviews. Experiments were conducted in the domains of hotels, electronic devices, and furniture, comparing human-written summaries with system-generated ones. The proposed method demonstrated superior summarization performance compared to existing comparison models.
Aspect-Based Comparative Summarization with Large Language Model
Yunseok Kang, Jaeseok Lee, Jaewoong Choi, Jaekoo Lee
http://doi.org/10.5626/JOK.2025.52.7.587
We proposed a strategy to mitigate the VRAM(Video Random Access Memory) shortage problem encountered when applying 3D Gaussian Splatting in large-scale 3D mapping environments derived from drone footage. To efficiently manage large scale scenes, we partitioned input data, optimized each partition independently, and subsequently merged optimized scenes. Additionally, we introduced a technique to augment point data by considering specific characteristics of drone-captured footage during the optimization process. As a result, our method reduced VRAM usage by two-thirds compared to previous studies, while achieving a 2.5% average improvement in quality as measured by PSNR(Peak Signal-to-Noise Ratio). Our approach emphasizes enhancing the accuracy and quality of 3D reconstructions while minimizing VRAM consumption.
Time Series Data Imbalance Resolution Techniques for Early Prediction
Eungseon An, Taehyoung Kwon, Doguk Kim
http://doi.org/10.5626/JOK.2025.52.7.593
Time series forecasting is a critical task that involves analyzing observed time series data to predict future values. However, when dealing with imbalanced data, model performance can degrade, leading to biased predictions. Although recent studies have explored various deep learning techniques and data augmentation methods, many fail to address challenges posed by data imbalance and the intrinsic characteristics of time series data simultaneously, leaving underlying issues unresolved. This study proposed a novel approach that could leverage temporal patterns to generate synthetic samples and extend the scope of early prediction. By identifying key moments that could effectively distinguish between positive and negative classes, our method enhanced the ability to predict further into the future. The method proposed in this study demonstrated superior performance to existing methods and proved the feasibility of early prediction for longer time lags.
A Pretrained Model-Based Approach to Improve Generalization Performance for ADMET Prediction of Drug Candidates
http://doi.org/10.5626/JOK.2025.52.7.601
Accurate prediction of Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET) properties plays an important role in reducing clinical trial failure rates and lowering drug development costs. In this study, we propose a novel method to improve ADMET prediction performance for drug candidate compounds by integrating molecular embeddings from a graph transformer model with pretrained embeddings from a UniMol model. The proposed model can capture bond type information from molecular graph structures, generating chemically refined representations, while leveraging UniMol’s pretrained 3D embeddings to effectively learn spatial molecular characteristics. Through this, the model is designed to address the problem of data scarcity and enhance the generalization performance. In this study, we conducted prediction experiments on 10 ADMET properties. The experiment results demonstrated that our proposed model outperformed existing methods and that the prediction accuracy for ADMET properties could be improved by effectively integrating atomic bond information and 3D structures.
A Study on Buffer Management for Read-Once Pages Using a Read Buffer
Seongjae Moon, Sang-Won Lee, Young Ik Eom
http://doi.org/10.5626/JOK.2025.52.7.611
The relational database systems store pages in main memory to minimize storage access and improve transaction throughput. However, read-once pages, referenced only once before eviction, may force dirty pages to be flushed, reducing the page hit ratio. In SSD-based DBMS, read-once pages cause I/O serialization, forcing faster read operations to wait for slower write operations to complete. We analyze the I/O serialization caused by read-once pages and their characteristics, and propose a buffer management scheme to isolate read-once pages by using a read buffer. In the TPC-C benchmark, dirty pages evicted due to read-once pages account for about 8.9% of all flush operations. By isolating read-once pages into the read buffer, we reduced single page flushes by 56% and the page miss ratio in the normal buffer by 32% while increasing transaction throughput by 8% compared to Vanilla MySQL.
Enhancing LLM-based Zero-Shot Conversational Recommendation via Reasoning Path
Heejin Kook, Seongmin Park, Jongwuk Lee
http://doi.org/10.5626/JOK.2025.52.7.617
Conversational recommender systems provide personalized recommendations through bi-directional interactions with users. Traditional conversational recommender systems rely on external knowledge, such as knowledge graphs, to effectively capture user preferences. While recent rapid advancement of large language models has enabled zero-shot recommendations, challenges remain in understanding users' implicit preferences and designing optimal reasoning paths. To address these limitations, this study investigates the importance of appropriate reasoning path construction in zero-shot based conversational recommender systems and explores the potential of using a new approach based on this foundation. The proposed framework consists of two stages: (1) comprehensively extracting both explicit and implicit preferences from conversational context, and (2) constructing reasoning trees to select optimal reasoning paths based on these preferences. Experimental results on benchmark datasets INSPIRED and ReDial show that our proposed method achieves up to 11.77% improvement in Recall@10 compared to existing zero-shot methods, It even outperforms some learning-based models.
Effective Importance-Based Entity Grouping Method in Continual Graph Embedding
http://doi.org/10.5626/JOK.2025.52.7.627
This study proposed a novel approach to improving entity importance evaluation in continual graph embeddings by incorporating edge betweenness centrality as a weighting factor in a Weighted PageRank algorithm. By normalizing and integrating betweenness centrality, the proposed method effectively propagated entity importance while accounting for the significance of information flow through edges. Experimental results demonstrated significant performance improvements in MRR and Hit@N metrics across various datasets using the proposed method compared to existing methods. Notably, the proposed method showed enhanced learning performance after the initial snapshot in scenarios where new entities and relationships were continuously added. These findings highlight the effectiveness of leveraging edge centrality in promoting efficient and accurate learning in continual knowledge graph embeddings.
Search
Journal of KIISE
- ISSN : 2383-630X(Print)
- ISSN : 2383-6296(Electronic)
- KCI Accredited Journal
Editorial Office
- Tel. +82-2-588-9240
- Fax. +82-2-521-1352
- E-mail. chwoo@kiise.or.kr