Recursion aware modeling and discovery for hierarchical software event log analysis

Leemans, M Maikel; Aalst, Wil M. P. van der; Brand, Mark van den

doi:10.1109/saner.2018.8330208

Cited by 26 publications

(18 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, we have expanded the scope of the paper to include a new research community that was not yet represented in the initial study: the field of grammar inference. Secondly, we expanded our experimental setup by experimenting on a larger number of datasets and covering a larger set of techniques, additionally covering hidden Markov models [64] and Active LeZi [41] in the machine learning category and in the process mining category adding the Indulpet Miner [51] process discovery algorithm as well as a class of automaton-based prediction techniques. Finally, in an attempt to bring three research communities together and make this manuscript useful for researchers from the machine learning, process mining, as well as grammar inference domains, we have added considerable detail to the description of process-model-based predictions.…”

Section: Introductionmentioning

confidence: 99%

An interdisciplinary comparison of sequence modeling methods for next-element prediction

2020

View full text Add to dashboard Cite

Data of sequential nature arise in many application domains in forms of, e.g. textual data, DNA sequences, and software execution traces. Different research disciplines have developed methods to learn sequence models from such datasets: (i) in the machine learning field methods such as (hidden) Markov models and recurrent neural networks have been developed and successfully applied to a wide-range of tasks, (ii) in process mining process discovery techniques aim to generate human-interpretable descriptive models, and (iii) in the grammar inference field the focus is on finding descriptive models in the form of formal grammars. Despite their different focuses, these fields share a common goal -learning a model that accurately describes the behavior in the underlying data. Those sequence models are generative, i.e, they can predict what elements are likely to occur after a given unfinished sequence. So far, these fields have developed mainly in isolation from each other and no comparison exists. This paper presents an interdisciplinary experimental evaluation that compares sequence modeling techniques on the task of nextelement prediction on four real-life sequence datasets. The results indicate that machine learning techniques that generally have no aim at interpretability in terms of accuracy outperform techniques from the process mining and grammar inference fields that aim to yield interpretable models.

show abstract

Section: Introductionmentioning

confidence: 99%

An interdisciplinary comparison of sequence modeling methods for next-element prediction

2020

View full text Add to dashboard Cite

show abstract

“…Events collected from software in operation (e.g. Java programs) reveals the presence of a hierarchical structure, where methods reside within classes, and classes within packages [118]. The same applies to IDE usage actions where identified menu options and executed commands belong to a specific category of command options built-in the Eclipse framework.…”

Section: Process Discoverymentioning

confidence: 98%

“…Events collected from software in operation (e.g. Java programs) reveals the presence of a hierarchical structure, where methods reside within classes, and classes within packages [118].…”

Section: Process Discoverymentioning

confidence: 99%

Software Development Process Mining: Discovery, Conformance Checking and Enhancement

Caldeira

Abreu

2016

2016 10th International Conference on the Quality of Information and Communications Technology (QUATIC)

View full text Add to dashboard Cite

Context. Modern software projects require the proper allocation of human, technical and financial resources. Very often, project managers make decisions supported only by their personal experience, intuition or simply by mirroring activities performed by others in similar contexts. Most attempts to avoid such practices use models based on lines of code, cyclomatic complexity or effort estimators, thus commonly supported by software repositories which are known to contain several flaws.Objective. Demonstrate the usefulness of process data and mining methods to enhance the software development practices, by assessing efficiency and unveil unknown process insights, thus contributing to the creation of novel models within the software development analytics realm.Method. We mined the development process fragments of multiple developers in three different scenarios by collecting Integrated Development Environment (IDE) events during their development sessions. Furthermore, we used process and text mining to discovery developers' workflows and their fingerprints, respectively.Results. We discovered and modeled with good quality developers' processes during programming sessions based on events extracted from their IDEs. We unveiled insights from coding practices in distinct refactoring tasks, built accurate software complexity forecast models based only on process metrics and setup a method for characterizing coherently developers' behaviors. The latter may ultimately lead to the creation of a catalog of software development process smells.Conclusions. Our approach is agnostic to programming languages, geographic location or development practices, making it suitable for challenging contexts such as in modern global software development projects using either traditional IDEs or sophisticated low/no code platforms.

show abstract

“…An approach for discovering the software architectural model from execution data is described in [29]. Finally, [25] allows to reconstruct the most relevant state-charts and sequence diagram from an instrumented working system. The focus of all these works, however, is software development processes or the understanding of the behaviour of the software, while the focus of this paper is to define a method to identify how people (at an aggregated level) interact with different software artifacts when conducting different comprehension tasks.…”

Section: Process Miningmentioning

confidence: 99%

Mining reading patterns from eye-tracking data: method and demonstration

et al. 2019

View full text Add to dashboard Cite

Understanding how developers interact with different software artifacts when performing comprehension tasks has a potential to improve developers' productivity. In this paper, we propose a method to analyze eye-tracking data using process mining to find distinct reading patterns of how developers interacted with the different artifacts. To validate our approach, we conducted an exploratory study using eyetracking involving 11 participants. We applied our method to investigate how developers interact with different artifacts during domain and code understanding tasks. To contextualize the reading patterns and to better understand the perceived benefits and challenges participants associated with the different artifacts and their choice of reading patterns, we complemented the eye-tracking data with the data obtained from think aloud. The study used behavior driven development (BDD), a development practice that is increasingly used in Agile software development contexts, as a setting. The study shows that our method can be used to explore developers' behavior at an aggregated level and identify behavioral patterns at varying levels of granularity.

show abstract

Recursion aware modeling and discovery for hierarchical software event log analysis

Cited by 26 publications

References 27 publications

An interdisciplinary comparison of sequence modeling methods for next-element prediction

An interdisciplinary comparison of sequence modeling methods for next-element prediction

Software Development Process Mining: Discovery, Conformance Checking and Enhancement

Mining reading patterns from eye-tracking data: method and demonstration

Contact Info

Product

Resources

About