| IEICE Technical Committee Submission System Conference Paper's Information |
Online Proceedings [Sign in] Tech. Rep. Archives |
| Go Top Page | Go Previous | [Japanese] / [English] |
| Paper Abstract and Keywords | ||
| Presentation | 2024-03-15 15:15 Bayesian Statistical Analysis of Commit History Data Using WBIC for OSS Evolution Analysis Toru Sugiyama, Takako Nakatani (OUJ)  KBSE2023-90 |
|
| Abstract | (in Japanese) | (See Japanese page) |
| (in English) | This paper utilizes the latest advancements in computational Bayesian statistics to analyze the commit histories of Open Source Software (OSS) and elucidate their statistical characteristics. Risks associated with the use of OSS, such as project forks and disappearances, are identified as significant factors that can cause major fluctuations or even halt the evolution of OSS. Since the evolution of OSS is fundamentally driven by commits from developers, understanding their nature is essential. Therefore, we have employed the Widely Applicable Bayesian Information Criterion (WBIC) as a versatile and effective measure for evaluating models suited to the data characteristics and conducted a statistical analysis of OSS commit histories. Our case study revealed that OSS commit histories follow a power-law distribution, specifically a Zipf distribution, and further satisfied the condition of lacking mean and variance, as substantiated by WBIC. Furthermore, given that commit histories lack mean and variance, it became clear that instead of analyzing the collective behavior of developers through models representing these probability distributions, focusing on individual developers through methods akin to agent-based simulations is more appropriate for addressing this challenge. | |
| Keyword | (in Japanese) | (See Japanese page) |
| (in English) | Open Source Software / Repository Mining / Model Selection / WBIC / Software Evolution / / / | |
| Reference Info. | IEICE Tech. Rep., vol. 123, no. 443, KBSE2023-90, pp. 138-142, March 2024. | |
| Paper # | KBSE2023-90 | |
| Date of Issue | 2024-03-07 (KBSE) | |
| ISSN | Online edition: ISSN 2432-6380 | |
| Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) | |
| Download PDF | KBSE2023-90 |
|
| Conference Information | |
| Committee | KBSE |
| Conference Date | 2024-03-14 - 2024-03-15 |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Okinawa Prefectual General Welfare Center |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Paper Information | |
| Registration To | KBSE |
| Conference Code | 2024-03-KBSE |
| Language | Japanese |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Bayesian Statistical Analysis of Commit History Data Using WBIC for OSS Evolution Analysis |
| Sub Title (in English) | |
| Keyword(1) | Open Source Software |
| Keyword(2) | Repository Mining |
| Keyword(3) | Model Selection |
| Keyword(4) | WBIC |
| Keyword(5) | Software Evolution |
| Keyword(6) | |
| Keyword(7) | |
| Keyword(8) | |
| 1st Author's Name | Toru Sugiyama |
| 1st Author's Affiliation | The Open Univ. of Japan (OUJ) |
| 2nd Author's Name | Takako Nakatani |
| 2nd Author's Affiliation | The Open Univ. of Japan (OUJ) |
| Speaker | Author-1 |
| Date Time | 2024-03-15 15:15:00 |
| Presentation Time | 25 minutes |
| Registration for | KBSE |
| Paper # | KBSE2023-90 |
| Volume (vol) | vol.123 |
| Number (no) | no.443 |
| Page | pp.138-142 |
| #Pages | 5 |
| Date of Issue | 2024-03-07 (KBSE) |
[Return to Top Page]