Expert systems and test-plan generation.
Item
-
Title
-
Expert systems and test-plan generation.
-
Identifier
-
AAI9020794
-
identifier
-
9020794
-
Creator
-
Papavasiliou, Ioannis.
-
Contributor
-
Adviser: Howard Rubin
-
Date
-
1990
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Computer Science | Artificial Intelligence
-
Abstract
-
The purpose of this work is the design of an expert system that supports the automation of the contents of the test plan, improving in this way the productivity, quality and reliability of software, which subsequently reduce the software cost. Chapter I gives some background of the problems that currently exist in the development of software systems and drive the software costs up. To overcome these problems, we propose a solution: to automate the labor intensive functions in the software development process using expert systems, and restrict our work in constructing the test plan using this approach. In Chapter II the scope of the system is defined. It accepts as input the flow graph of a program written in a simple subset of PASCAL and produces as output the number and definitions of test cases, the test coverage, the number of predicted defects and the estimated reliability. In Chapter III we discuss several expert systems that are currently supporting software development, and we also summarize the related work on software testing.;Chapter IV is the core of this work. The design of the system is described step by step. The system uses the data selection criterion "all-uses", from the family of criteria proposed by Rapps and Weyuker, to select the paths to be tested and then forms the path constraints by using symbolic evaluation. The implementation of the first part, i.e., the selection of the test paths, and some actual results are given in the appendixes. Chapter V is a brief discussion on the possible future research directions for this project.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
