MIIC Final Presentation
Hyunwoo Park
Goal
Framework
Results
Conclusion
Motivational Barriers
Non-motivational Barriers
Not-invented-here
Hoarding
Search problem
Transfer problem
How can we mitigate
search and transfer problems
in a systematic way?
Network structural approach
Q1. How can we measure
the collaboration performance
of a network?
Q2. How can constraints of
collaboration network optimization
be quantitatively modeled?
Q3. How can we optimize a network
and how much is the performance
improved by the optimization?
1. Measurement Development
SPI = Search Performance Index
TPI = Transfer Performance Index
2. Defining Constraints
Fixed total strength of the network
Fixed total number of links in the network
1) "link-based algorithm"
Moving edges to find
higher performance index
2) "strength-based algorithm"
Redistributing strengths
between two edges
3. How to Optimize
By link-based algorithm
1) One strong cluster as information intermediary emerged.
2) When maximizing Search Performance Index,
the central cluster got strong pointings from all units.
By strength-based algorithm
Strength becomes polarized.
How much were the performances improved?
Search
Transfer
Implications
1) Search and transfer problems can be improved simultaneously.
It's not a perfect trade-off problem.
2) Have a central unit or cluster playing a role of information intermediary.
3) Make strong ties stronger, weak ties weaker.
Do not waste organizational effort to strengthen weak ties.
Rather, focus on the strong ones.More presentations by Hyunwoo Park
American Universities in 1950s
Hyunwoo Park on
Class Presentation for Transforming Academia: Aftermath of World War II, Impact of the GI Bill on American Universities
Popular presentations
Trello Architecture
Brett Kiefer on
This is the visual part of a talk I gave on the trello.com architecture at the MongoDB user group on 18 Jan 2012. Blog post ...
More popular prezis in Explore>