Observing Social Machines Part 1: What to Observe?
1. David De Roure
Observing Social Machines Part 1
What to Observe?
Clare Hooper
Megan Meredith-Lobay
Kevin Page
Ségolène Tarte
Don Cruickshank
Cat De Roure
13. What to observe? Logs
Analytics
Data findings
e.g. Success rate
of transcription
Social sciences
Qualitative study
Motivation
Individual and
group
Mixed methods
Differences in
technique and scale
Unlikely to be an simple
transferable metric
18. • Are the tree, bot and/or dating site Social Machines?
• What are their trajectories?
• Cyberphysical scenario involving machine-to-machine
communication without human mediation
• Illustrates automatic assembly – unintended but
purposeful
• Glimpse of APIs and the service-oriented ecosystem
• Bot detection algorithm illustrates observation
mechanism (human / automated?)
• Machines impersonating people; e.g. people can buy
twitter followers, how do they know they’re not bots?
The Lessons of the Raspberry Tree
21. Identify Ecosystems
Where Social Machines are
1. Interacting and competing with others
2. Being designed, born and co-evolving
3. Variable in size, purpose, lifetime and intent
4. Reflecting the trends towards cyber-physical
and machine-to-machine systems
22. • The constituent Social Machines and their trajectories
• Technologies, humans and their interfaces, including
intersection with physical world
• The design processes, and how they correlate with
successful machines
• Ground rules leading to emergent behaviour
– rules by which people abide
– rules encoded in design
– part of community conduct
– grounded in how other Social Machines behave
Analyse
23. • Observing Social
Machines Part 2:
How to Observe?
• Toolkit approach
• Embrace other
observatories
• Instrument the
ecosystem
Future Work
24.
25. • You are all observers…
• Go forth, engage with the machines
• Design new ones!
• The observatory is really a laboratory
• Share the (methodological) toolkit
• Report your findings in these workshops
Your mission should you choose to
accept it…