Introducing%20Empirical%20Software%20Engineering%20into%20Japanese%20Industry - PowerPoint PPT Presentation

About This Presentation
Title:

Introducing%20Empirical%20Software%20Engineering%20into%20Japanese%20Industry

Description:

Real field data were collected from some Japanese software development companies. ... efforts to complete the ongoing project based on the similar past projects. ... – PowerPoint PPT presentation

Number of Views:47
Avg rating:3.0/5.0
Slides: 7
Provided by: mediaOs
Category:

less

Transcript and Presenter's Notes

Title: Introducing%20Empirical%20Software%20Engineering%20into%20Japanese%20Industry


1
Introducing Empirical Software Engineering into
Japanese Industry
  • Naoki Ohsugi
  • Nara Institute of Science and Technology

2
EASE (Empirical Approach to Software Engineering)
Project
  • Goal of the project is to activate knowledge flow
    between industry and academia.
  • Real field data
  • On-the-spot knowledge
  • Data analysis methods
  • Scientific expertise

Software industry
EASE project
Academia
Industrial staffs
Core universities
Core companies
Academic staffs
Associated universities
Associatedcompanies
  • Discovered improvement point
  • Revealed know-how
  • Objective evidence for evaluating validity of
    theories

3
Current Status of EASE Project
  • Real field data were collected from some Japanese
    software development companies.
  • Survey of the following data analysis methods was
    conducted.
  • Case-Based Reasoning (CBR)
  • Multi-Dimensional Scaling (MDS)
  • Association Analysis
  • Parallel Coordinate Plot (PCP)
  • Mahalanobis-Taguchi System (MTS)
  • Analysis of the data with the methods is in
    progress.

4
Example of Analysis MethodCase-Based Reasoning
(CBR)
  • Evaluating similarities between an ongoing
    project and past projects.
  • Estimating efforts to complete the ongoing
    project based on the similar past projects.

Estimated test cost 40.25
Similarity 0.94
Similarity 0.87
Similarity -0.77
5
Example of Analysis MethodMulti-Dimensional
Scaling (MDS)
  • Evaluating similarities among projects in the
    same manner with CBR.
  • Plotting the projects as dots on a
    two-dimensional coordinate system.
  • where, the distance between dots depends on the
    similarity between projects.

R
D
J
Y
H
Project A
S
C
I
M
A
O
X
W
G
F
K
E
B
Q
U
V
N
Z
T
L
P
6
Future Work
  • Field data collection and survey of analysis
    methods should be continued.
  • Know-how on method of the data collection and the
    analysis should be accumulated.
  • Rewards to the industry and academia have to be
    returned.
  • The journal paper can be rewards to the academia?
  • What is the rewards to the industry? Debrief
    reports of the analysis?
Write a Comment
User Comments (0)
About PowerShow.com