| 柿本 和希 | D, 中間発表 | ソフトウェア設計学 | 飯田 元, | 松本 健一, | 石濱 直樹, | 高井 利憲 | |
|
title: Research on product quality assurance for safety-critical software systems
abstract: My study aims to product quality assurance for safety-critil systems. Safety standards which applied to software systems are process based, software assurance ensure the target software was developed using the correct processes. Therefore, traditional software quality asuurance is implicit in how product-specific safety requirements are sufficiently identified and met. I present mya latest research to prove the quality of software product. language of the presentation: Japanese | |||||||
| JAISRI PONGCHAI | M, 2回目発表 | ソフトウェア工学 | 松本 健一, | 安本 慶一, | 石尾 隆, | Raula Gaikovina Kula, | 嶋利 一真 |
|
title: A Preliminary Study on Self-Contained Libraries in the NPM Ecosystem
abstract: The widespread of libraries within modern software ecosystems creates complex networks of dependencies. These dependencies are fragile to breakage, outdated, or redundancy, potentially leading to cascading issues in dependent libraries. One mitigation strategy involves reducing dependencies; libraries with zero dependencies become to self-contained. This paper explores the characteristics of self-contained libraries within the NPM ecosystem. Analyzing a dataset of 2763 NPM libraries, we found that 39.49\% are self-contained. Of these self-contained libraries, 40.42\% previously had dependencies that were later removed. This analysis revealed a significant trend of dependency reduction within the NPM ecosystem. The most frequently removed dependency was babel-runtime. Our investigation indicates that the primary reasons for dependency removal are concerns about the performance and the size of the dependency. Our findings illuminate the nature of self-contained libraries and their origins, offering valuable insights to guide software development practices. language of the presentation: English | |||||||
| LERTBANJONGNGAM SILA | M, 1回目発表 | ソフトウェア工学 | 松本 健一, | 安本 慶一, | Raula Gaikovina Kula, | 嶋利 一真 | |
|
title: Using Large Multimodal Models to Analyze Images in Stack Overflow
abstract: My study aims to use Large Multimodal Models to analyze images in Stack Overflow questions to enhance their descriptive quality and improve overall problem resolution. In this presentation, I will talk about two related previous works that show the role of images in Stack Overflow and Show the capability of Large Multimodal Models to detect image glitches These studies mention that developers increasingly use images in their Stack Overflow questions, which makes it more likely to get an accepted answer. However, there is limited research on how these images directly affect question quality and demonstrate the capability of Large Multimodal Models to detect image glitches, showing that Large Multimodal Models have proven effective in interpreting complex images, suggesting their potential utility in platforms like Stack Overflow. Therefore, I aim to use Large Multimodal Models to enhance the descriptive quality of Stack Overflow questions by leveraging images. Currently, I have utilized Stack Overflow data from the SOTorrent dataset as of December 2019 and conducted preliminary experiments using the LLAVA model from Ollama, an open-source Large Multimodal Models software. language of the presentation: English | |||||||