Understanding Intelligence Rooted in Coincidence

Play Video
Entry for:
Image1467764285?1467764285  180 Seconds of Science





1. Summary

May people act constructively as a collective without being conscious about it?

Nowadays we can witness numerous situations where humans contribute to collective problem solving and sense making by publicly sharing information enabled by means of the Internet and the World Wide Web. Examples include natural disasters (e.g. the Haiti earthquake) or political crises (e.g. the Kenyan election) but also sports events, election campaigns and scientific discoveries.

Studying those exemplars shows that even though there is some common topic or goal hovering above the information sharing activities of the individuals (e.g. coordinating help in disaster response or optimising travel routes of people being affected by traffic disruptions), people are not necessarily talking with each other or along social ties. Instead we can see an upgraded importance of information broadcasting and coincidence, especially when the time to make decisions is limited (e.g. in life threatening situations).

It was the swiss psychiatrist C. G. Jung - amongst others - who did some early and influential reasoning about “temporally coincident occurrences of acausal events”. Revisiting Jung's ideas in the context of the Internet and the World Wide Web may help us to create meaning out of coincidence and capture a kind of implicit algorithm or digital subconscious, that is the substrate of the accumulated information sharing behaviour of individuals.

To understand and formalise this phenomenon is the purpose of my research. It involves investigating the fundamental relationship of time and information as well as the question where information actually resides in a multi-dimensional space.

Recent debates about our increasingly selective view to information - a result of content personalisation in search results and social network timelines due to the commercial interest of the social network providers - are an additional motivation for this kind of data analysis, which aims at avoiding the amplification of some biases that lie in digital social structures (e.g. gender imbalances but also deliberate spread of misinformation).


Luczak-Roesch, M., Tinati, R. and Shadbolt, N., 2015, May. When resources collide: Towards a theory of coincidence in information spaces. In Proceedings of the 24th International Conference on World Wide Web (pp. 1137-1142). ACM.

Luczak-Roesch, M., Tinati, R., Van Kleek, M. and Shadbolt, N., 2015, August. From coincidence to purposeful flow? properties of transcendental information cascades. In 2015 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) (pp. 633-638). IEEE.

Luczak-Roesch, M., Tinati, R., O'Hara, K. and Shadbolt, N., 2015, February. Socio-technical computation. In Proceedings of the 18th ACM Conference Companion on Computer Supported Cooperative Work & Social Computing(pp. 139-142). ACM.

Tinati, R., Luczak-Roesch, M. and Hall, W., 2016, April. Finding Structure in Wikipedia Edit Activity: An Information Cascade Approach. In Proceedings of the 25th International Conference Companion on World Wide Web (pp. 1007-1012). International World Wide Web Conferences Steering Committee.

Luczak-Roesch, M., Tinati, R., Aljaloud, S., Hall, W. and Shadbolt, N., 2016, June. A Universal Socio-Technical Computing Machine. In International Conference on Web Engineering (pp. 559-562). Springer International Publishing.

2. Do you have a video hashtag for sharing via twitter?
3. Please share your social media handles (optional)

You find me on twitter as @mluczak but also on LinkedIn.

4. Do you want to sign up to EMCR Forum mailing list?


5. DO you want to sign up to RSNZECR mailing list?


Voters Map


Ben McNeil
over 2 years ago

Fascinating! Seems like an immense challenge, I'm assuming there's enough data but is there enough computational power to do this?

Markus Luczak-Roesch
over 2 years ago

Thanks, Ben. Your remark is absolutely valid. If we are successful in mapping the entire Web and start running historic analysis on that, the computational capabilities may be challenged. But this vision to capture, analyse and preserve the entire content of the World Wide Web is nothing we can claim exclusively, so pushing towards that boundary seems to be a worthy challenge we should take.

The computational costs of data analysis are usually a game of tradeoffs. Transcendental Information Cascades can bring the early costs for the creation of the model down, because they are defined as a very simple network model, which can even be created incrementally at low computational costs and thus can be applied to real-time analysis use cases. Furthermore, the method does not rely on a huge amount of system-specific context features, which is not only more privacy preserving (we are interested in publicly accessible data, not private data) but also what we expect to help us being more resilient against (or at least more transparent about) biases that result from the data that usually makes up this context (e.g. system-specific social networks).

And then there are also application areas for this kind of data analysis that are of much smaller scale than mapping the entire World Wide Web.


Markus Luczak-Roesch is a Senior Lecturer in the School of Information Management at the Victoria University of Wellington. Markus' research is focused on the fundamental relationship between time ...