Automatic Traceability Maintenance via Machine Learning Classification

Abstract

Previous studies have shown that software traceability, the ability to link together related artifacts from different sources within a project (e.g., source code, use cases, documentation, etc.), improves project outcomes by assisting developers and other stakeholders with common tasks such as impact analysis, concept location, etc. Establishing traceability links in a software system is an important and costly task, but only half the struggle. As the project undergoes maintenance and evolution, new artifacts are added and existing ones are changed, resulting in outdated traceability information. Therefore, specific steps need to be taken to make sure that traceability links are maintained in tandem with the rest of the project.In this paper we address this problem and propose a novel approach called TRAIL for maintaining traceability information in a system. The novelty of TRAIL stands in the fact that it leverages previously captured knowledge about project traceability to train a machine learning classifier which can then be used to derive new traceability links and update existing ones. We evaluated TRAIL on 11 commonly used traceability datasets from six software systems and compared it to seven popular Information Retrieval (IR) techniques including the most common approaches used in previous work. The results indicate that TRAIL outperforms all IR approaches in terms of precision, recall, and F-score.

RVprio: A tool for prioritizing runtime verification violations

“…The research on using machine learning in software engineering has a rich history. Examples of software engineering problems to which machine learning was applied include (1) predicting fault-prone or costly-to-maintain software system components based on historical information [80][81][82], (2) classifying field executions as passing or failing runs [83,84], (3) duplicate bug report detection [85][86][87], (4) bug localization [88,89], (5) code search, code completion, code mining, code clone detection and code synthesis [90][91][92][93][94][95][96], (6) learning how to apply patches [97,98], (7) prioritizing test programs for compilers [99], (8) selecting and prioritizing test cases [100], (9) establishing traceability links between artefacts of the system [101], (10) statically detecting flaky tests [102], and (11) classifying warnings from SCA tools [44,45,65,[103][104][105][106][107][108][109][110][111][112][113][114]…”

Graph queries for analyzing the coverage of requirements by test cases

“…The problem of measuring similarity or relevance between software artifacts have also been studied elsewhere, especially in the context of automatic construction of traceability links between artifacts. Examples of this line of studies include, for example, [8,9,2,10,11].…”

Enhancing Automated Software Traceability by Transfer Learning from Open-World Data

“…Given the high cost and effort required to manually create and maintain trace links during the software development process, researchers have proposed various solutions for generating links automatically [11], [12], [13]. Classical information retrieval (IR) solutions, such as the Vector Space Model (VSM) [14], Latent Dirichlet Analysis (LDA) [15], [16], and Latent Semantic Indexing (LSI) [17] have been explored in-depth over the past decade, but have met a glassceiling in terms of achievable accuracy, with basic machine learning (ML) approaches [18], [19], [20], [21] suffering from similar fates. The primary impedance is caused by their lack of semantic analysis and of the textual artifacts.…”