Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Bug Analysis Towards Bug Resolution Time Prediction

Hasan Yagiz Ozkan, Poul Einer Heegaard, Wolfgang Kellerer, Carmen Mas-Machuca

TL;DR

The paper analyzes Jira bug data to understand and predict bug resolution time in network softwarization projects (ONAP/ONOS and Apache). It develops a data-collection and filtering pipeline, analyzes state-transition flows and time spent in states, and examines how bug attributes affect timing. It then proposes neural-network models to predict exact resolution times and classify bugs as fast or slow, comparing to NB and kNN baselines and achieving favorable results. The findings highlight the practical value of predicting resolution time for workload planning and resource allocation in open-source network software projects, with potential extensions to broader datasets and preprocessing techniques.

Abstract

Bugs are inevitable in software development, and their reporting in open repositories can enhance software transparency and reliability assessment. This study aims to extract information from the issue tracking system Jira and proposes a methodology to estimate resolution time for new bugs. The methodology is applied to network project ONAP, addressing concerns of network operators and manufacturers. This research provides insights into bug resolution times and related aspects in network softwarization projects.

Bug Analysis Towards Bug Resolution Time Prediction

TL;DR

The paper analyzes Jira bug data to understand and predict bug resolution time in network softwarization projects (ONAP/ONOS and Apache). It develops a data-collection and filtering pipeline, analyzes state-transition flows and time spent in states, and examines how bug attributes affect timing. It then proposes neural-network models to predict exact resolution times and classify bugs as fast or slow, comparing to NB and kNN baselines and achieving favorable results. The findings highlight the practical value of predicting resolution time for workload planning and resource allocation in open-source network software projects, with potential extensions to broader datasets and preprocessing techniques.

Abstract

Bugs are inevitable in software development, and their reporting in open repositories can enhance software transparency and reliability assessment. This study aims to extract information from the issue tracking system Jira and proposes a methodology to estimate resolution time for new bugs. The methodology is applied to network project ONAP, addressing concerns of network operators and manufacturers. This research provides insights into bug resolution times and related aspects in network softwarization projects.
Paper Structure (12 sections, 7 figures, 3 tables)

This paper contains 12 sections, 7 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Data repositories
  4. Bug related data
  5. Bug resolution process
  6. Bug Collection and Analysis
  7. Bug collection and filtering
  8. Bug resolution time distribution and common flows in transition workflow
  9. Impact of different bug aspects
  10. Bug Distribution across States over Time
  11. Predicting Resolution Time With Neural Network
  12. Conclusion

Figures (7)

  • Figure 1: Comparison of ONOS, ONAP, and Apache Workflows. The ONOS workflow is fully compliant with the standard Atlassian model. The Apache* workflow applies to the considered projects.
  • Figure 2: The role of the reporter in resolving duration of bugs in ONAP: Insights from the top ten reporters ordered by the median value of priority 2 bugs. The numbers in the x-axis represent the corresponding bug count and the median values are given on top.
  • Figure 3: The role of assignees in resolving duration of bugs in ONAP: Insights from the top ten assignees ordered by the median value of priority 2 bugs. The numbers in the x-axis represent the corresponding bug count, and the mean values are given on top.
  • Figure 4: Impact of assigning a reporter to the bug resolution process. The numbers in the x-axis represent the corresponding bug count, and the median values are given on top.
  • Figure 5: Distribution of the bug states over time.
  • ...and 2 more figures