Department of General Management and Information Systems Prof. Dr. Armin Heinzl Software Outsourcing Decision Aid (SODA): A Requirements based Decision Support Method and Tool Authors: Tommi Kramer & Michael Eschweiler CAISE - June 21, 2013
May 11, 2015
Department of General Management and Information Systems
Prof. Dr. Armin Heinzl
Software Outsourcing Decision Aid (SODA): A Requirements based Decision Support Method and Tool
Authors: Tommi Kramer & Michael Eschweiler
CAISE - June 21, 2013
Outline
• Problem Domain & Motivation
• Research Design
• SODA – A Decision Support Method
– Model Creation Phase
– Model Clustering Phase
– Structural Analysis Phase
• Evaluation
• Summary
2
Problem Domain
• SMEs are inexperienced in software development outsourcing
Where / what / how to outsource?(Klimpke et al. 2011)
• Behavior patterns: – Decisions on a gut level
– Decisions are subjective in nature and people centric
• But, SMEs want to be successful in SDO
3
Objective
Research objective:
Definition of a decision making approach for selective software development outsourcing based on software requirementsdelivering:
• Good clustering quality• Good scalability (low setup costs)• outsourcing success
4
Research Domain
• Applying systems theory and graph theory to existing approaches
• Facing multi-dimensional decision problem with including decision rationales from SE principles (Dibbern et al. 2004, Kramer et al. 2011)
• Focus on selective sourcing of application systems by supporting decision making on component level
5
Research Methodology
• Design Science Research (Hevner et al. 2004, Peffers et al. 2007)
– Graph theory and systems theory deliver requirements for artifact design
– Definition and implementation of a new decision making approach in IS outsourcing as artifact development
– Software development projects with students used for artifact evaluation
6
SODA (1)
• SODA: Software Outsourcing Decision Aid - A decision making method and tool supporting IT project teams in selecting components suitable for outsourcing
• Phase 1: Graph Model Creation– Representing requirements
in a graph
– Nodes: Requirements
– Edge: „similar_to“ or „requires“ relationships
7
SODA (2)
• Phase 2: Graph Model Clustering
– Finding cohesive groups of requirements
– Neither the number of clusters nor the clusters‘ size is known a priori
– Newman algorithm for “community structure detection”(Newman 2006)
8
SODA (3)
• Phase 3: Structural Analysis of requirements
– Modularity
– Coupling and Cohesion
– Requirements Centrality
– Rule-based recommendations
9
SODA
10
PHASE 2
PHASE 1PHASE 3
ResultingDecision Determinants:• Modularity• Cluster Coupling
and Cohesion• Requirements
Centrality
Evaluation• Simulation by using data from four master team projects
developing a software application
• Clustering quality: More interdependencies lead to more coarse-grained partitioning of graph. But cluster quality remains stable!
• Scalability: Higher effort in interdependency definition is not delivering better modularity or clustering quality!
• SDO success: ?
11
Project Require
ments
Interdepen-
dencies
Achievable
Modularity
No. of Clusters in
Optimal Partition
Rand Index
A 45 61 0.71 10 0.80
B 45 43 0.67 8 0.84
C 45 181 0.54 6 0.77
D 46 49 0.65 8 0.82
Conclusion/Contribution
• We apply modularity, clustering & cohesion as well as centrality techniques for requirements analysis to support outsourcing decision making
• Design and development of an appropriate method and tool (scalable and good clustering)
• Contribution to practice– Facilitate decision making for managers in SMEs when
it comes to the question what to outsource and what to realize in-house
– Provide a repeatable and precise method for SDO in order to store decision information
12
References• Dibbern, J., Goles, T., Hirschheim, R., & Jayatilaka, B. (2004). Information Systems Outsourcing: A
Survey and Analysis of the Literature. Communications of the ACM, 35(4), 6-102.
• Hevner, A. R., March, S. T., Park, J., & Sudha, R. (2004). Design Science in Information Systems Research. Management Information Systems Quarterly 28 (1), 75 – 105.
• Klimpke, L., Kramer, T., Betz, S., & Nordheimer, K. Globally Distributed Software Development in Small and Medium-Sized Enterprises in Germany: Reasons, Locations, and Obstacles. In Proceedings of the 19th European Conference on Information Systems (ECIS2011), Helsinki, Finland, 2011
• Kramer, T., Heinzl, A., & Spohrer, K. (2011). Should this Software Component be Developed Inside or Outside our Firm? - A Design Science Perspective on the Sourcing of Application Systems. In J. Kotlarsky, L. P. Willcocks, & O. Ilan (Eds.), New Studies in Global IT and Business Service Outsourcing: 5th Global Scourcing Workshop 2011, Courchevel, France, March 14-17, 2011, Revised Selected Papers (pp. 115-132). Heidelberg, Dordrecht, London, New York: Springer.
• Newman, M. E. J. Modularity and Community Structure in Networks. In Proceedings of the National Academy of Sciences of the United States of America, 2006 (pp. 8577-8582)
• Peffers, K., Tuunanen, T., Rothenberger, M. A., & Chatterjee, S. (2007). A Design Science Research Methodology for Information Systems Research. Journal of Management Information Systems, 24(3), 45 - 78.
14