Log Summarisation for Defect Evolution Analysis

arXiv:2403.08358v1 Announce Type: cross
Abstract: Log analysis and monitoring are essential aspects in software maintenance and identifying defects. In particular, the temporal nature and vast size of log data leads to an interesting and important research question: How can logs be summarised and monitored over time? While this has been a fundamental topic of research in the software engineering community, work has typically focused on heuristic-, syntax-, or static-based methods. In this work, we suggest an online semantic-based clustering approach …

abstract analysis arxiv cs.cl cs.se data defects evolution leads log data logs maintenance monitoring nature question research software temporal type vast