Upload
shaukat-ali
View
415
Download
0
Embed Size (px)
Citation preview
By: Shaukat Ali
MITF12M011
USAGE-BASED READING
USAGE-BASED READING• Over View• Effectiveness is not measured in number of faults
• Faults are not equally likely
• Faults are ranked according to user’s prioritized use cases
• Principal Idea• To focus the reading effort on detecting the most critical
faults in the inspected object
• Focuses the reading effort guided by a prioritized use case model
TYPES OF UBR CONTROLS• Ranked Based• A reviewer using ranked-based reading
follows the use cases in the order in which they appear in the ranked use case document.
• Time Controlled• Time-controlled reading adds a time budget to
each use case in order to force a reviewer to utilize a specific use case the specified time.
• Use existing use cases as guide
• Improve efficiency by use case guides
• Scenarios are based on project and can not apply on other projects
• Prioritize use cases once and use throughout the project
• Develop non existing use cases actively
• Improve efficiency by overlapping faults
• A reading technique, can be applied for every documented artifact of each project
COMPARISON B/W PBR AND UBRPerspective Based Reading Usage-Based Reading
EXPERIMENT PREPARATION• The experiment is based on an experimental package
developed at Lund University
• Subjects acting in the experiment
•Reviewers• Inspection Material•Roles and Activities•Fault Classification
Reviewers
• 23 fourth-year software engineering master’s students at Blekinge Institute of Technology in Sweden.
• students have extensive experience from software development.
• The customers for these projects are normally people in industry
• The experiment was a mandatory part of a course in verification and validation
• The course included lectures and assignments
• The objective of the experiment is to empirical study in software verification and validation
Inspection Material
• Four document •Requirements document (English)•design document (specification and description
language)•use case document•one checklist
Roles and Activities• Development, designed Car management system by 6 people paying following roles
• Developer • One person was responsible for the development
• One person developed the textual requirements
• One person developed the checklist
• Prioritization of use cases• Three people prioritized the use cases independently (LHP)
• person combined the priorities and ordered the prioritized use cases (mean)
• Classification of faults
• Two people classified the faults independently
• One person combined the categorized faults
Fault Classification
• Faults are classified into 3 classes
• Class A:• Contain most important and frequently used use cases e.g
Login
• Class B:
• Contain important but rarely used use cases e.g Logout
• Class C:
• Contain not use cases e.g “A sflag is missing but is documented in design ”