| 井ノ口 輝 | 1751012: M, 1回目発表 | ソフトウェア工学 | 松本 健一 |
|
title: Analysis of Programming Skill Improvement in Online Judges
abstract: Online judge system is a system that provides programming problems and an environment for compiling and testing submitted source codes. The systems store histories of executions and users can see them. Learners can improve their programming skills by addressing problems in an online judge system. However, since there is no guideline for addressing various problems, it is not easy for learners to improve their skills by themselves efficiently. In this research, we analyze the histories stored in Aizu online judge (AOJ) to find succeeded patterns in sequential problem trials for skill improvement. language of the presentation: Japanese 発表題目:オンラインジャッジにおけるプログラミングスキル上達の分析 発表概要: プログラミングの学習にオンラインジャッジシステム（OJS）が利用されている．OJSはプログラミングの問題を提示し，解答に対する正誤判定をするシステムである．学習者は，難易度や用いるアルゴリズムが異なる様々な問題に取り組み，プログラミングを学習する．しかし，学習者の取り組み方は様々であるため，学習方法によっては十分な学習効果が得られないことがある．これまでの履歴は蓄積されているため，過去の履歴から成長に役立つ学習方法であるパターンを見つけることができれば有用である．本研究では，OJSであるAizu Online Judge（AOJ）に蓄積された学習履歴から，個人の学習パターンを導出し，学習パターンと成長の関係を明らかにする．これにより，どのようなOJSでの学習パターンが成長に繋がるかが明らかになると期待できる． | |||
| 上田 裕己 | 1751017: M, 1回目発表 | ソフトウェア工学 | 松本 健一 |
|
title: How is the IF Statement Fixed through Code Review?
abstract: In software development process, code review is key to ensuring the absence of software defects. The developers who are called reviewers will verify source code quality issues such as code convention issues. This study aims to find the hidden code review policy in a specific project. To achieve, goal to understand how a code will be fixed conditional statement issues based on reviewer's feedback. We conduct an empirical study to understand "if" conditional statement that is one of most frequently changes after review. Also, we analyze changes of the "if" conditional statements that are requested to be reviewed, and that are implemented after review. In this presentation, I will introduce code review problem and our future review model. language of the presentation: Japanese | |||
| 桂川 大輝 | 1751029: M, 1回目発表 | ソフトウェア工学 | 松本 健一 |
|
title: The Impact of Using Functional Categories in Library Recommendation abstract: Most software projects today rely on third party libraries for effective software development like saving the cost of implementing from scratch. Library repositories like the Maven Central allows Java developers access to thousands of Open Source Software (OSS) libraries. However, when selecting new libraries, developers are reported to be overwhelmed by the massive size and diversity of available libraries. We propose using functional categories (i.e., grouping of libraries that perform similar tasks) can assist with this issue. In this presentation, we investigate the impact of functional categories to assist with library recommendation. language of the presentation: Japanese ￣: *** ¬￣ *** ￣×: *** ¬￣× *** | |||
| 田内 遥夏 | 1751059: M, 1回目発表 | ソフトウェア工学 | 松本 健一 |
|
title: Classifying Self-admitted Technical Debts using N-gram IDF
abstract: Technical debt is a metaphor for non-optimal solutions in software development. Although it can accelerate development in the short term, it causes issues in software maintenance in the long term. Self-admitted technical debts (SATD) are one of technical debts that are intentionally introduced into source code along with comments, such as “FIXME” and “Hacky.” Previous study presented a technique of detecting design or requirement SATD using NLP. In this research, we apply N-gram IDF, which can extract useful N-gram features in comments. language of the presentation: Japanese 発表題目: N-gram IDFを用いたSelf-admitted Technical Debtsの分類 発表概要: ソフトウェア開発には，短期的な目標を達成するための次善の策のことを表現する比喩として，技術的負債というものがある．技術的負債が蓄積すると，開発の遅延やバグの温床になってしまう．そのため，自動的な発見・分類によって技術的負債を早期に解決していくことが望まれる．特に，ソースコード中で開発者によるコメントを伴って残される技術的負債を自覚的な技術的負債と呼ぶ．本研究ではSATDの自動発見・分類のため，ソースコード中のコメントをN-gram IDF により特徴量化して分類する方法を提案する．提案手法は個々のSATDに含まれる特徴的なN-gramを抽出し，そのパターンを解析することにより，先行研究よりも高い精度でのSATD分類の実現が期待される． | |||
| 吉上 康平 | 1751122: M, 1回目発表 | ソフトウェア工学 | 松本 健一 |
|
title: An empirical study of clone-and-own practices in OSS development
abstract: In industrial software development, code reusing in product line engineering, called clone-and-own, is typically adopted. Although there are benefits in clone-and-own practices, such as simplicity, availability and independence of developers, it is discouraged since it makes maintenance complex and difficult. Recently, it is reported that clone-and-own reuses also appear in open source software (OSS) development. In this research, we aim to clarify the distributions of clone-and-own reuses in existing OSS projects. We will investigate large-scale code repositories using a file-level similarity measurement technique. In this presentation, we report the pilot study with JUnit repository. language of the presentation: Japanese | |||
| ELIZALDE ZAPATA RODRIGO | 1751132: M, 1回目発表 | ソフトウェア工学 | 松本 健一 |
|
title: LibViews - An information Visualization Application for Third-Party Libraries on Software Projects abstract: Software third party libraries are very common nowadays. Libraries allow developers to include functions that otherwise will take a lot of time to create themselves. Developers may not be aware of included vulnerability issues or bugs to libraries dependencies. But not everything is lost, dependencies often update to fix these issues or new and useful functions. However, developers do not update their dependencies. Related tools such as LibViews[1] allow better understanding of libraries among the project, For my research, I would like to propose an approach to aid developers in the the update decision. Through a visualization, we propose to allow the developer to compare the differences in the methods and components between the different versions of a dependency, to identify whether or not a new version will affect the their code directly without complications. language of the presentation: English | |||
| 真鍋 優 | 1751113: M, 1回目発表 | 大規模システム管理 | 笠原 正治 |
|
title: On designing a cost driven Hadoop scheduler in a cloud environment used by multiple users. abstract: Hadoop is an open source software framework for distributed file systems and parallel processing. It allows users to handle massive amount of data with a large compute cluster. To avoid spending much costs of storing big data and operating such clusters, users desire to share them. However, it is assumed that default schedulers, such a FIFO scheduler and Fair scheduler, are used by a small number of users at one time. In previous research, Dynamic Proportional Share Scheduling was proposed. It allows many users to simultaneously use a Hadoop cluster by their bids. The research did not investigate for datacenter pricing models, so we are trying to devise it. I will explain Dynamic Proportional Share Scheduling and the progress of our research in this presentation. language of the presentation: Japanese | |||
| 木下 裕太 | 1751037: M, 1回目発表 | 情報基盤システム学 | 藤川 和利 |
|
title:(Paper Introduction) Improving datacenter performance and robustness with multipath TCP
abstract: The latest large-scale data centers offer higher aggregate bandwidth and robustness by creating multiple paths in the core of the network. To utilize this bandwidth requires different flows take different paths, which poses a challenge. In this Paper, they propose using Multipath TCP as a replacement for TCP in such data centers, as it can effectively and seamlessly use available bandwidth, giving improved throughput and better fairness on many topologies. language of the presentation: Japanese | |||
| 寺井 元基 | 1751073: M, 1回目発表 | 情報基盤システム学 | 藤川 和利 |
|
title: [Paper Introduction] iFrame: Dynamic Indoor Map Construction through Automatic Mobile Sensing
abstract:Map data is used in various applications. There is not much indoor map data, it is difficult to obtain. In this presentation, in order to solve this problem, introduce iFrame which is a method of dynamically creating an indoor map using smartphone, bluetooth, wi - fi and so on. Implemented the proposed method, experimented and verified accuracy. language of the presentation: Japanese | |||
| 中島 颯人 | 1751079: M, 1回目発表 | 情報基盤システム学 | 藤川 和利 |
|
title: (Paper Introduction) Globally Synchronized Time via Datacenter Networks
abstract: In this presentation, we present Datacenter Time Protocol (DTP), a clock synchronization protocol. DTP does not use packets at all and it is able to achieve nanosecond precision. DTP uses the physical layer of network devices to implement a decentralized clock synchronization protocol. DTP has two strongth. (1)DTP eliminates most non-deterministic elements in clock synchronization protocols. (2)DTP has virtually zero overhead because DTP does not interfere into layers 2 or higher layers. We demonstrate that the precision provided by DTP is bounded by 25.6 nanoseconds for directly connected nodes. The precision is hundreds of nanoseconds when a network is idle and can decrease to hundreds of microseconds when a network is heavily congested. language of the presentation: Japanese | |||
| 仲野 弘一 | 1751083: M, 1回目発表 | 情報基盤システム学 | 藤川 和利 |
|
title: [Paper Introduction] A Distributed Blockchains-Based Secure SDN Architecture for IoT Networks
abstract: The rapid increase in the number and diversity of smart devices connected to the Internet has raised the issues of flexibility, efficiency, availability, security, and scalability. In this presentation, in order to solve these problem, I introduce high performance and secure IoT distributed network work using block chain technology to SDN (Software Defined Network). The proposed method has been shown to be scalable and safe compared to the existing network architecture. language of the presentation: Japanese | |||