chapter 9: tapping the crowd for fast innovation isto sipilÄ
TRANSCRIPT
Chapter 9:Tapping the Crowd for Fast InnovationISTO SIPILÄ
Collective Intelligence Other term crowdsourcing
Deriving data, knowledge and insight from large groups connected online
Many varieties of collective intelligence
Three approaches that use Open Data in different ways are examined in the book Open innovation through collaboration (discussed in the next
chapter)
The Match.com model
The Data Hive
The Match.com Model Search for a small number of experts who have the right
experience, skills and interests to solve a problem
Someone with a problem to solve or data to analyze try to reach as large crowd as possible to find the few right people
If you understand a problem well enough to publish it and you put the right inducements around the system then innovators from all over the world can work with it
Often new perspective and outside thinking is the key to identifying new patterns and correlations
Examples InnoCentive: works with companies and organizations, runs contests,
largely focused on science and technical problems, offers prizes ranging from 500 dollars to over a million
Kaggle: presents challenges for data scientists, predicted what level of IT access different employees would need based on their jobs
TopCoder: another community for data problems, worked with Harvard Medical School, solution to the tough gene-sequencing problem
Washington University: tried to understand the structure of a virus, released their data on the foldit.com site, solution in three weeks
Peer-to-Patent: commenting and studying patent applications
Stack Exchange (started as Stack Overflow), Wikipedia etc.
The Data Hive Volunteers are doing routine work to analyze or improve Open
Data
Each individual does small pieces of work that contribute to the solution
Model has been applied to government data and scientific data with some striking results
Sometimes untrained amateurs see important things that experts miss (Galaxy Zoo example)
Examples NASA: invited people to help identify planetary systems in space
telescope images
Zooniverse: an international hub where anyone can help solve large-scale scientific puzzles, over 800 000 people registered Galaxy Zoo: participant looked at an image and said what type of
galaxy it is or is it a star, an artifact or more than one galaxy
CrowdCrafting.org: Zooniverse kind of approach, presents a range of challenges in areas like basic science, linguistics and the analysis of social media
SkyTruth: uses the crowd to analyze Open Data in satellite images to keep a collective eye on the environmental impact of corporate activities
How to Make People to Participate Mass data-hive projects that pay people to participate in dull
routine tasks have had limited success
The nature of the task and the community engaged in it seems to determine dedication and performance
Research done to InnoCentive says there are three factors motivating problem solvers: They want to take on projects that will have an impact
They want to be part of a group of elite problem solvers
They want whatever inducement is being offered, intrinsic reward or extrinsic rewards like money and recognition
Question
What is the best way to motivate people to take part on collective mind projects?