コロキアムB発表

日時: 6月11日(金)3限(13:30~15:00)


会場: L1

司会: 高橋 慧智
MAEPRASART VITTUNYUTA M, 1回目発表 ソフトウェア工学 松本 健一, 飯田 元, 石尾 隆, 畑 秀明, Raula G. Kula

title: Who Maintains a Dependency? A Study of Developer dependency updates in the npm Ecosystem.  

abstract: Package dependencies are increasingly popular among software projects. The dependencies provide features that allow developers to reuse and build upon others’ packages. Dependencies also come with withdrawal, such as conflicts in version inconsistency, defects, or vulnerabilities. Therefore, it is important for developers to always keep dependencies as fresh as possible. Prior works show that developers struggle and need much effort to maintain dependencies. As a result, most of the dependencies are outdated and delay in updating. When these developers disappear, who else should be the person to maintain the dependency. In this study, we want to define the developer’s skill responsible for maintaining dependency, including migration by mining through the repository’s history in member of npm packages. 

language of the presentation: English 

 
SAKULNIWAT TATTIYA M, 1回目発表 ソフトウェア工学 松本 健一, 飯田 元, 石尾 隆, 畑 秀明, Raula G. Kula

title: Do Expert Programmers Tend to Accept Idiomatic Code? A Study of PyPI Maintainers 

abstract: Python is still the most popular programming language for people to learn, with interest rising by 27% last year from 2020. As with any other community, Pythonic code (idiomatic code) is claimed to be widely adopted by mainly experienced Python developers. With the growing usage of third-party libraries, the PyPI ecosystem has also grown with usage, especially with Machine learning libraries. Objectives: In this research, we present the study protocols to investigate the writing of three pythonic code (i.e., with open, if-statement, and enumerate idioms) written by the maintainers of PyPI packages. The preliminary results show that there is the existence of both pythonic and non-pythonic ways of writing code. Consequently, the results lead us to the objectives to see if the maintainers of PyPI packages, which are coding experts, more likely to accept code that contains Pythonic code than other type of codes. Method: Our execution plan includes a sanity check and a quantitative approach. First, we explore the prevalence of both Pythonic and non-Pythonic code writing. Then, we explore the writing of Pythonic code by maintainers. Finally, we investigate whether or not writing Pythonic code has more acceptance from the maintainers. Our end-goal is to understand and advice on the impact of writing Pythonic code by maintainers and their packages. Limitations: We also discuss the following limitations of our study: a generalization of our experiment and results, the detection accuracy of the tool, and threats to defining a maintainer.  

language of the presentation: English  

 
WATTANAKRIENGKRAI SUPATSARA M, 2回目発表 ソフトウェア工学 松本 健一, 飯田 元, 石尾 隆, 畑 秀明, Kula Raula Gaikovina
title: Traceability of linkage between Academic Papers and Software
abstract: Traceability between published scientific breakthroughs and their implementation is essential, especially in the case of open-source scientific software which implements bleeding-edge science in its code. However, aligning the link between GitHub repositories and academic papers can prove difficult, and the current practice of establishing and maintaining such links remains unknown. This paper investigates the role of academic paper references contained in these repositories. We conduct a large-scale study of 20 thousand GitHub repositories that make references to academic papers. We use a mixed-methods approach to identify public access, traceability and evolutionary aspects of the links. Although referencing a paper is not typical, we find that a vast majority of referenced academic papers are public access. These repositories tend to be affiliated with academic communities. More than half of the papers do not link back to any repository. We find that academic papers from top-tier SE venues are not likely to reference a repository, but when they do, they usually link to a GitHub software repository.
language of the presentation: English
 
ISLAM SYFUL D, 中間発表 ソフトウェア工学 松本 健一, 飯田 元, 石尾 隆, Raula Gaikovina Kula
title: Using Stack Overflow to Understand Developer Discussions about Software Package Managers
abstract: A package manager (PM) is crucial to most technology stacks, acting as a broker to ensure that a verified dependency package is correctly installed, configured, or removed from an application. Diversity in technology stacks has led to dozens of PMs with various features. While recent studies have shown that developers struggle to migrate their dependencies, the common assumption is that PMs broker dependencies without any issues. In this research, we explore issues faced by developers when using PMs through an empirical study of content on Stack Overflow (SO). Our task is divided two-fold. First, we characterize the package manager developer's experience. The analysis results indicate that specific features of package managers impact developer's experience. Second, we investigate the package usage experience of developers. Our investigation of the npm ecosystem shows that developers experience different types of errors when using third-party packages. We also observed that popular packages are less likely to be discussed in Stack Overflow. Finally, analysis on accepted answers related to package usage shows that answers containing usage examples and execute commands are helpful to solve package usage-related errors.
language of the presentation: ***English***
発表題目: *** この部分を発表題目に ***
発表概要: *** この部分を発表概要に ***
 

会場: L2

司会: 磯山 直也
CHEN ZHENG D, 中間発表 計算システムズ生物学 金谷 重彦, 峠 隆之, 小野 直亮, Md. Altaf-ul-Amin, 黄 銘
title:EEG-Based Physiological Interpretable Sleep Stage Scoring System
abstract:Sleep is a complex human function which is characterized by a sequence of alterations in the brain, muscle, eye, heart, and respiratory activity. The gold standard of sleep construction is re-defined as five different stages, i.e., wake, rapid eye movement (REM), or non-REM where the non-REM stage can be further divided into N1, N2, and N3 according to the American Academy of Sleep Medicine (AASM). The gold-standard in-lab multi-lead EEG capturing from the overnight polysomnography (PSG) still requires manual scoring by sleep experts. Inspired by the Transformer scaling successes in classification tasks, we experiment with applying a standard Transformer directly to the largest public EEG database for sleep scoring. Here, we segment the spectrogram of each epoch into the temporal patches, in order to capture the transient characteristics from the 1-second frequency band. Afterward, a sequence of linear embeddings of these patches is fed into the Transformer expecting to parallel find the global dependencies among the durations.
language of the presentation: English
 
ZHANG YONGXIN M, 1回目発表 計算システムズ生物学 金谷 重彦, 峠 隆之, MD.Altaf-Ul-Amin, 小野 直亮, 黄 銘
HAR

title: Wearable multi-sensor based Human Activity Recognition.

abstract: In the current background, people pay more and more attention to health, medical care, many health problems can be expressed through the daily activities of the human body, therefore, the study of human activity recognition has also made considerable development. The result of simple human activities recognition have been very accurate. So this time I set the goal for the recognition of complex activities, as well as the activity recognition, on the basis of analysis of the activity of the pattern in a period of time. We collected human daily activities in real life by ourselves, and utilize a hierarchical model to process time-series data, Finally, the model is evaluated by analysis and comparison. Finally, the model is evaluated by analysis and comparison with base model.*** 

language of the presentation: *** English *** 

 
ZHU GUANGXIAN M, 1回目発表 計算システムズ生物学 金谷 重彦, 峠 隆之, MD.Altaf-Ul-Amin, 小野 直亮, 黄 銘

title: *** Generation and Classification According to the Stage of pancreatic cancer precursor lesions *** 

abstract: *** Pancreatic cancer is an intractable cancer and its prognosis is poor after surgery. The reason is that patients are usually not diagnosed in early stage. Studies have shown that PanIN is a precursor lesion of pancreatic cancer, and PanIN has four grades according to severity. It is of great medical significance to distinguish PanIN grade accurately. At present, there are not many labeled data in PanIN, so it is not suitable for effective supervised machine learning. Therefore, this study aims to use InfoGAN to generate different grades of PanIN images as a data enhancement. It is expected to enhance the classifier by using the synthesized images. As a new attempt to generate image in the field of pancreatic cancer, this study will provide theoretical support for future researches. *** 

language of the presentation: *** English *** 

 
中前 美祐 M, 1回目発表 数理情報学 池田 和司☆, 金谷 重彦, 川鍋 一晃(客員教授), 森本 淳(客員准教授), 吉本 潤一郎, 福嶋 誠

title: Comparing the impact of regulations on people’s mental health during the covid-19 epidemic  

abstract: Since the outbreak of COVID-19, regulations such as lockdown have been enforced with various scale and strictness in each country or region to prevent the spread of infection. While regulations are somewhat effective in preventing the spread of infection, they have a negative impact on people’s mental health. Studies on this fact have been conducted within each country or region, but there are not many studies that have made comparisons between countries or regions. In addition, few studies have investigated the effects of strict regulations on mental health. From these situations, detailed research is essential to control the covid-19 epidemic while maintaining people’s good mental health. In this study, I first compare the degree of lockdown regulation and mental health in Japan and Taiwan, which has implemented different policies in spite of similar ethnicity, to clarify the mental health effects of the degree of regulation and other social factors. 

language of the presentation: Japanese