Predicting build outcome with developer interaction in Jazz

Author

Schröter, Adrian

Author_Institution

Univ. of Victoria, Victoria, BC, Canada

Volume

2

fYear

2010

fDate

2-8 May 2010

Firstpage

511

Lastpage

512

Abstract

Investigating the human aspect of software development is becoming prominent in current research. Studies found that the misalignment between the social and technical dimensions of software work leads to losses in developer productivity and defects. We use the technical and social dependencies among pairs of developers to predict the success of a software build. Using the IBM Jazz™ data we found information about developers and their social and technical relation can build a powerful predictor for the success of a software build. Investigating human aspects of software development is becoming prominent in current research. High misalignment between the social and technical dimensions of software work lowers productivity and quality.

Keywords

software metrics; software quality; IBM Jazz data; build outcome; developer defects; developer interaction; developer productivity; human aspect; social dependency; social dimensions; software build; software development; software productivity; software quality; technical dependency; technical dimensions; Data mining; Humans; Programming; Social network services; Software; Software engineering; Software measurement;

fLanguage

English

Publisher

ieee

Conference_Titel

Software Engineering, 2010 ACM/IEEE 32nd International Conference on

Conference_Location

Cape Town

ISSN

0270-5257

Print_ISBN

978-1-60558-719-6

Type

conf

DOI

10.1145/1810295.1810456

Filename

6062122