Software analytics: Difference between revisions

Revision as of 19:28, 2 June 2021

Revision as of 14:03, 13 May 2022

Line 1:

{{Expert needed|Software|reason=it appears to misrepresent the history of software analytics|date=December 2014}}

'''Software analytics''' is the [[analytics]] specific to the ___domain of [[software system]]s taking into account [[source code]], static and dynamic characteristics (e.g., [[software metric]]s) as well as related processes of their [[software development|development]] and [[software evolution|evolution]]. It aims at describing, monitoring, predicting, and improving the efficiency and effectivityeffectiveness of [[software engineering]] throughout the [[software lifecycle]], in particular during [[software development]] and [[software maintenance]]. The data collection is typically done by mining [[software repository|software repositories]], but can also be achieved by collecting user actions or production data. One avenue for using the collected data is to augment the [[integrated development environment]]s (IDEs) with data-driven features.<ref name="Bruch2010">{{cite journal|last1=Bruch|first1=Marcel|last2=Bodden|first2=Eric|last3=Monperrus|first3=Martin|last4=Mezini|first4=Mira|title=IDE 2.0: Collective Intelligence in Software Development|year=2010|url=https://hal.archives-ouvertes.fr/hal-01575346/document|doi=10.1145/1882362.1882374}}</ref>

== Definitions ==

Line 15:

== Approach ==

Methods, techniques, and tools of software analytics typically rely on gathering, measuring, analyzing, and visualizing information found in the manifold data sources stored in software development environments and ecosystems. Software systems are well suited for applying analytics because, on the one hand, mostly formalized and precise data is available and, on the other hand, software systemsystems are extremely difficult to manage ---in a nutshell: "software projects are highly measurable, but often unpredictable."<ref name="ICSESEIP12InfoNeeds"/>

Core data sources include [[source code]], "check-ins, work items, bug reports and test executions [...] recorded in software repositories such as CVS, Subversion, GIT, and Bugzilla."<ref>Harald Gall, Tim Menzies, [[Laurie Williams (software engineer)|Laurie Williams]], and Thomas Zimmerman. "Software Development Analytics". Dagstuhl Reports, Vol. 4, Issue 6, pp. 64-83.</ref> [[telemetry | Telemetry data]] as well as execution traces or logs can also be taken into account.

Automated analysis, massive data, and systematic reasoning support decision -making at almost all levels. In general, key technologies employed by software analytics include analytical technologies such as [[machine learning]], [[data mining]], [[statistics]], [[pattern recognition]], [[information visualization]] as well as large-scale data computing & processing. For example, software analytics tools allow users to map derived analysis results by means of [[software map]]s, which support interactively exploring system artifacts and correlated software metrics. There are also software analytics tools using analytical technologies on top of [[software quality]] models in [[agile software development]] companies, which supportssupport assessing software qualities (e.g., reliability), and derivederiving actions for their improvement.<ref>{{Cite journal|last=Martínez-Fernández|first=Silverio|last2=Vollmer|first2=Anna Maria|last3=Jedlitschka|first3=Andreas|last4=Franch|first4=Xavier|last5=Lopez|first5=Lidia|last6=Ram|first6=Prabhat|last7=Rodriguez|first7=Pilar|last8=Aaramaa|first8=Sanja|last9=Bagnato|first9=Alessandra|date=2019|title=Continuously assessing and improving software quality with software analytics tools: a case study|journal=IEEE Access|volume=7|pages=68219–68239|doi=10.1109/ACCESS.2019.2917403|issn=2169-3536|url=https://upcommons.upc.edu/bitstream/2117/133374/1/FINAL-Access-Paper-preprint.pdf|doi-access=free}}</ref>

== History ==

Line 26:

In May 2009, software analytics was first coined and proposed when Dongmei Zhang founded the Software Analytics Group (SA) at Microsoft Research Asia (MSRA). The term has become well known in the [[software engineering]] research community after a series of tutorials and talks on software analytics were given by Zhang and her colleagues, in collaboration with Tao Xie from North Carolina State University, at software engineering conferences including a tutorial at the IEEE/ACM International Conference on Automated Software Engineering (ASE 2011),<ref name="ASE2011" /> a talk at the International Workshop on Machine Learning Technologies in Software Engineering (MALETS 2011),<ref name="MALETS" /> a tutorial and a keynote talk given by Zhang at the IEEE-CS Conference on Software Engineering Education and Training,<ref name="CSEETKeynote" /><ref name="CSEETTutorial" /> a tutorial at the International Conference on Software Engineering - Software Engineering in Practice Track,<ref name="ICSETutorial" /> and a keynote talk given by Zhang at the Working Conference on Mining Software Repositories.<ref name="MSRKeynote" />

In November 2010, Software Development Analytics (Software Analytics with a focus on Software Development) was proposed by Thomas Zimmermann and his colleagues at the Empirical Software Engineering Group (ESE) at Microsoft Research Redmond in their FoSER 2010 paper.<ref name="FoSER" /> A goldfish bowl panel on software development analytics was organized by Zimmermann and [[Tim Menzies]] from West Virginia University at the International Conference on Software Engineering, Software Engineering in Practice trackTrack.<ref name="Goldfish" />

==See also==