"A single ant or bee isn't smart, but their colonies are. The study of swarm intelligence is providing insights that can help humans manage complex systems, from truck routing to military robots." (Peter Miller, 2007)
- Evolution and adaptation of natural systems
- Promoting diversity
- Swarm Intelligence is based on interactions
- Social creatures such as ants, bees, birds and fish are observed
Social creatures have three characteristics that make them successful:
Flexibility (adapt to change)
Robustness (individual failures don't compromise group performance)
Self-organization (activities are neither controlled nor supervised)
Ant Foraging Techniques
- Lay pheromone trails
- Algorithms following the principles of foraging ant behaviour
- Southwest Airlines - routing and cargo operations
- Hewlett Packard (HP) - telecommunications
- efficient routing of phone calls
- dissolving pheremones
- Factory Efficiency
- Dividing Tasks
Predicting Group Behaviour
Integration with Business Intelligence ideas
"The most powerful and fascinating insight from swarm intelligence is that complex collective behaviour can emerge from individuals following simple rules." (Bonabeau & Meyer, 2001)
Three Important Lessons:
1. Unpredictable and often counter-intuitive behaviour can arise from very simple rules
2. A seemingly minor change in the rules can radically alter group behaviour
3. You can MODEL it using constraints
and use the model it to predict group behaviour that emerges, such as:
employee turnover
productivity
loyalty
Advantages of Swarm Intelligence
Small Colonies - Tandem Recruitment
Large Colonies - Mass Recruitment
Medium Colonies - Group Recruitment
- Fast changing and unpredictable environments
- Capital One Philosophy: Employees are primarily responsible to the ideas they have, not to their managers
"Many of our business opportunites are short-lived. We have to move fast to exploit them and move on when they fade." (Bonabeau & Meyer, 2001)
- This system, designed to foster group recruitment, is largely self-organized:
- Good ideas attract others.
- Capital One encourages employees to look for other “food sources” (and rates performance using their employee evaluation system)
- Raiding new markets
Self-organization is being used as a way of exploiting opportunities in volatile, short lived markets, as well as optimizing results within more stable enterprises.
CompanyWay (in Washington) is working with the Bios group to develop a web service that would allow employees to organize and work like swarms of free agents inside their own companies.
- Workers post solution, laying a trail for others to contribute. The resulting work can be integrated with executive decision making tools to give managers control over decision making processes. Rewards in the form of merit points.
- WEB expansion was due to swarm intelligence: a network emerged from local rules of behaviour and interaction
- BMW posts engineering challenges on its website
- Foldit online gamers crack AIDS enzyme puzzle
- XEROX - mass recruitment
- Switzerland - Migros has 2 million stakeholders (who view themselves as part of a swarm)
Swarm businesses, according to Gloor and Cooper in The Perfect Swarm:
- Gain power by giving it away (Amazon, Ebay)
- willing to share with and support the swarm (Open source)
- put the welfare of members of the swarm ahead of making money
Disadvantages of Swarm Intelligence
- Group behaviour can be startling
- People resist and aren't used to self-organizing systems
- Insect/human opposition
- Quantification
- Self-organizing systems are NEW
Example
- Collaborative Innovation Networks (COINs)
- Enron
Conclusion:
Self-organization is still a radical idea. However, swarm intelligence, along with neural networks, text-mining and genetic algorithms will help to form the generations of the Semantic Web that are rapidly approaching, and help us to make some useful decisions with the mass of information and exchangeable data that we are currently able to access. Used correctly, and with a healthy consideration for constraints and anomalies, swarm intelligence will become an inherent part of Business Intelligence Systems, and the harnessing of collective intelligence will prove to be a powerful tool in this data mining world.
BIBLIOGRAPHY
Bonabeau, E., & Meyer, C. (2001). Swarm Intelligence: A whole new way to think about business. Harvard Business Review .
Fisher, L. (2009). The Perfect Swarm: The Science of Complexity in Everyday Life. New York: Basic Books.
Gloor, P. A. (2006). Swarm Creativity: Competitive Advantage Through Collaborative Innovation Networks. Oxford: Oxford University Press.
Gloor, P. A., & Cooper, S. M. (2007). When Swarms go Mad (Egomania at Enron). Chigago: Books 24x7.
Hansen, M. T. (2009). Collaboration, How Leaders Avoid the Traps, Create Unity, and Reap Big Results. Cambridge, MA: Harvard Business Press.
Hinchey, M. G., Sterritt, R., & Rouff, C. (2007). Swarms and Swarm Intelligence. Computer , 111-113.
Linoff, G. S., & Berry, M. J. (2011). Data Mining Techniques: For Marketing, Sales, and Customer Relationship Management, Third Edition. John Wiley & Sons.
Macoveiciue, M., & Stan, C. (2010, November 1st). Nature Inspired Methods for the Semantic Web. Retrieved November 19th, 2011, from slideshare: http://www.slideshare.net/stanconstan/semantic-web-nature/download
Martens, D., Baesens, B., & Fawcett, T. (2011). Editorial survey: swarm intelligence for data mining. Machine Learning , 1-42.
Miller, P. (2007, July). Swarm Theory. Retrieved November 17th, 2011, from National Geographic: http://ngm.nationalgeographic.com/2007/07/swarms/miller-text
Thankyou