Show simple item record

dc.contributor.authorManjunath, Kavitha
dc.description.abstractRequirements engineering is the first and perhaps the most important phase of software life cycle. Software faults committed during the requirements development, if left undetected can affect downstream activities. While previous research has developed fault-detection techniques, they are inadequate because they lack an understanding of root cause of faults to be able to avoid its future occurrences. Our research is aimed at helping software engineers understand human errors (i.e., the root cause) that cause faults that are inserted into Software Requirements Specification (SRS). This can help software engineers become more aware and reduce the likelihood of committing human errors during requirements development. Using a retrospective error abstraction approach, this paper reports the results from an industrial study wherein software engineers were trained on human errors followed by their application of human error abstraction (based on fifteen requirement faults supplied to them) and recommendations on prevention techniques for those abstracted errors.en_US
dc.publisherNorth Dakota State Universityen_US
dc.rightsNDSU Policy 190.6.2
dc.titleUnderstanding Human Errors to Improve Requirements Qualityen_US
dc.typeMaster's paperen_US
dc.date.accessioned2018-05-14T21:11:17Z
dc.date.available2018-05-14T21:11:17Z
dc.date.issued2018
dc.identifier.urihttps://hdl.handle.net/10365/28129
dc.subject.lcshRequirements engineering.en_US
dc.subject.lcshSoftware failures.en_US
dc.subject.lcshComputer software -- Human factors.en_US
dc.subject.lcshHuman-computer interaction.en_US
dc.subject.lcshSoftware engineering.en_US
dc.rights.urihttps://www.ndsu.edu/fileadmin/policy/190.pdf
ndsu.degreeMaster of Science (MS)en_US
ndsu.collegeEngineeringen_US
ndsu.departmentComputer Scienceen_US
ndsu.programSoftware Engineeringen_US
ndsu.advisorWalia, Gursimran


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record