| MAIPRADIT ARNAN | M, 2回目発表 | モバイルコンピューティング | 伊藤 実, 安本 慶一, 柴田 直樹, Juntao Gao, 川上 朋也 |
|
Title: Adaptive Traffic Control Algorithm Based on Back-Pressure and Q-Learning
Abstract: Nowadays traffic congestion has increasingly been a significant problem, which results in longer travel time and aggravates air pollution. Back-pressure based traffic control algorithms have been shown to be effective in reducing traffic congestion. However, those available work control traffic are based on either inaccurate traffic information or local traffic information, which causes inefficient traffic scheduling. In this paper, we propose an adaptive traffic control algorithm based on back-pressure and Q-learning, which can efficiently reduce congestion. Our algorithm controls traffic based on accurate real- time traffic information and global traffic information learned by Q-learning. As verified by simulation, our algorithm can significantly reduce traffic congestion: reducing average vehicle traveling time by at most 36% as compared with state-of-the-art algorithm under the tested scenario. Language of the presentation: English | |||
| NGUYEN QUYNH MAI | M, 2回目発表 | 知能コミュニケーション | 中村 哲, 安本 慶一, 吉野 幸一郎 |
|
Title: Natural Langauge Generation for Smart Assistant System
Abstract: Smart assistant system with natural language interface, such as spoken dialogue systems (SDS), can help users to solve complex tasks have become an emerging research topic in artificial intelligence and natural language processing areas. As the endpoint of interacting with users, Natural Language Generation (NLG) plays a critical role in the system and has a significant impact on user’s impression of the system, where its task is to convert a meaning representation (MR) produced by the dialogue manager into one or more sentences in a natural language. It is traditionally divided into two subtasks: sentence planning, which decides the overall sentence structure, and surface realization, determining the exact word forms and flattening the sentence structure into a string (Reitier and Dale, 2000). In this study, we address the task of generating textual descriptions about tourist spots to support user decision making. Our automatic evaluation showed that our generation model produces fluent text overall, enhances the quality of textual description in the tourist domain. In term of human evaluation, our NLG system could support users during their trip. Language of the presentation: English | |||
| NUKULKIT PONGBHOP | M, 2回目発表 | ソフトウェア工学 | 松本 健一, 飯田 元, 石尾 隆, Raula G. Kula |
|
Title: Code to Doc: An Exploratory Study into How npm GitHub README Files Updates With Source Code
abstract: GitHub opensource software project using a README.md file to show important information such as usage examples, installation associated to their software project. Related works shows that developers often struggle to write documentation. Moreover, Opensource survey also reported that software documentation is highly valued but frequently overlooked and developers complained that most documentation either incomplete or outdated. Our goal is to understand the relation of README file and source code. We investigate through pull-request of five npm GitHub project which are 'fullpage.js', 'gulp', 'json-server', 'pm2', and 'reveal.js'. Results shows that README file most contents changes by source code are 'usage', option', and 'installation'. In additions, we also found that README file content change are affected by adding new feature into source code. However, the results also report that developers still inconstant to update README file when they update their projects. This study help new comer or existing developers to understand README file and source code update and also help them in term of document update decision making when they update their project. language of the presentation: *** English *** | |||
| NUGROHO YUSUF SULISTYO | D, 中間発表 | 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: Software Analytics for New Insights – Analyses of Code Histories, Code Comments, Forums in Software Development
abstract: Software analytics aims to obtain insightful and actionable information. Insightful information conveys knowledge that is meaningful and useful for practitioners to perform such specific tasks. Software analytics is employed to improve development productivity, software quality, and user experience. Thus, it focuses on software development, systems, and users. The data collection is typically done by mining software repositories, but can also be collected from users’ action or production data. In our research, we conducted 3 analyses: application of different diff algorithms in git, academic paper citations in code comments, and the impact of forums in Eclipse community. However, in this presentation, we only discuss the first analysis: how different are different diff in git? In software development, mining code changes from software repositories is a common and basic task. Users can collect the data from a software repository (e.g. Git repository) in different format according to their purposes by applying different options in git command. Git also offers four diff algorithms to identify the code changes. Without specifying the algorithm, Myers is used as the default. However, different algorithms may produce different diff results. Thus, this can have an impact on empirical studies. We applied two diff algorithms; Myers and Histogram in our three applications: collect metrics, SZZ, and get patches. Our findings show that there are differences between Myers and Histogram in generating the code changes based on the number and the location of lines. From our manual analysis, we recommend using the Histogram algorithm when mining Git repositories to extract code changes. language of the presentation: English | |||
| RUANGWAN SHADE | D, 中間発表 | ソフトウェア工学 | 松本 健一, 飯田 元, 石尾 隆, Raula G. Kula |
|
Title: Toward a Better Understanding of Developer Participation in Open Source Software Development Processes
Abstract: Developer participation is the main factor that drives Open Source Software (OSS) development processes. Modern Code Review (MCR) is one of the OSS development processes and it plays a key role in software quality practices. In MCR process, a new patch (i.e., a set of code changes) is encouraged to be examined by reviewers in order to identify weaknesses in source code prior to an integration into main software repositories. To mitigate the risk of having future defects, prior work suggests that MCR should be performed with sufficient review participation. Indeed, recent work shows that a low number of participated reviewers is associated with poor software quality. However, there is a likely case that a new patch still suffers from poor review participation even though reviewers were invited. Hence, in this paper, we set out to investigate the factors that are associated with the participation decision of an invited reviewer. Through a case study of 230,090 patches spread across the Android, LibreOffice, OpenStack and Qt systems, we find that (1) 16%-66% of patches have at least one invited reviewer who did not respond to the review invitation; (2) human factors play an important role in predicting whether or not an invited reviewer will participate in a review; (3) a review participation rate of an invited reviewers and code authoring experience of an invited reviewer are highly associated with the participation decision of an invited reviewer. These results can help practitioners better understand about how human factors associate with the participation decision of reviewers and serve as guidelines for inviting reviewers, leading to a better inviting decision and a better reviewer participation. Language of the presentation: English | |||