Automating Root Cause Analysis via Machine Learning in Agile Software Testing Environments

Abstract

We apply machine learning to automate the root cause analysis in agile software testing environments. In particular, we extract relevant features from raw log data after interviewing testing engineers (human experts). Initial efforts are put into clustering the unlabeled data, and despite obtaining weak correlations between several clusters and failure root causes, the vagueness in the rest of the clusters leads to the consideration of labeling. A new round of interviews with the testing engineers leads to the definition of five ground-truth categories. Using manually labeled data, we train artificial neural networks that either classify the data or pre-process it for clustering. The resulting method achieves an accuracy of 88.9%. The methodology of this paper serves as a prototype or baseline approach for the extraction of expert knowledge and its adaptation to machine learning techniques for root cause analysis in agile environments.

Mutant SF3B1 promotes malignancy in PDAC

“…The gene annotation used was GENCODE v.m25. For gene expression quantification, we used a custom script, available at GitHub (copy archived at Kahles, 2021 ); commit hash d074114f1d0a9f518c9cd039f68de0cdf8d583ff. SplAdder v.2.2 ( Kahles et al, 2016 ) was run to build splicing graphs and determine splice events.…”

Machine Learning models to predict Agile Methodology adoption

“…Indeed, past studies have successfully utilized ML within the context of Agile software development. Kahles and others [11] applied ML to automate the root cause analysis in Agile software testing environments. The study was able to produce an ML model that could achieve a prediction accuracy of 88.9% by using artificial neural networks to either classify or pre-process the data for clustering, using manually labelled data.…”

Enhancing Continuous Integration Systems with Assisted Root Cause Analysis

“…While failure categorization might seem similar to anomaly detection in plain texts, we have separated the two fields as they are two distinct parts of our research. Due to the amount of workloads we are investigating on a regular basis, we cannot rely on manual techniques such as qualitative interviews and labeling of data [16] or discarding workflows that include error messages that cannot be classified [17].…”