On building a ubiquitous discovery platform

Information is King. It is the primitive tool we have all evolved to use intuitively without a moments thought. It is what guides human perception of the world we live in, our surroundings, how we attain belief, and how we adapt to people we share our lives with. Information is everywhere, and is connectible in so many unimaginable ways. How the human mind infers and unites ideas from information may offer some clues to fathom the subjective emotion of serendipity. Information theory is clearly pertinent and closely aligned to the lines of inquiry within SerenA.

The inception of the SerenA project is auspicious and opportunistic, given the increasing interest and excitement around serendipitous discovery. The fascination of exploring the multi-dimensional space of interdisciplinary research and the occasional, fortuitous, identification of novel research techniques is not new. It is the obligation of the SerenA project to deeply scrutinise the theory of serendipity in unconcealed and novel ways, to advance the wider understanding of chance encounters in the realm of ideas. My role in the project is to interrogate the question of the degree to which serendipitous environments can be engineered, and to explore what part recent technological advancements can play in facilitating a heightened state of serendipitous discovery.

The timeliness of these subtle cognitive connections is often the crucial factor for recognising their significance… the quest of being in the right place, at the right time. The book “Work-Oriented General Education” (Eugene Staley, 1973) succinctly puts it: “the ‘serendipitous person’ is one who is observant and alert to the possible significance of something unexpected”. If SerenA has the ambition of augmenting environments for unexpected and surprising discovery, then a compelling challenge to address is the timeliness and proposition of suggestion.

To address these conundrums, we have been exploring a number of enabling technologies. The two fundamental implementation challenges are: meandering through bewildering quantities of information, finding the critical path for sweet spots of discovery; and when such explorations should occur. At the beginning of this post, I stated that the commencement of the SerenA project is timely. And I certainly meant it. Indeed, some of the empowering technology for this project were nonexistent until just a few years ago. Not until 2007, which marked the birth of the Linking Open Data project , had the scientific community been graced with a rich database of structured information, about everything from research domains, people, organisations, places… Things in the world. This project continues to grow at an expeditious pace, as is neatly depicted in the linked open data cloud diagram. It is a transformational move forward in the realisation of information theory – a truly distributed database of semantic information about unambigious resources. It goes without saying that this mammoth information portal is intrinsically moulded into the mechanics of the SerenA environment.

Through the engagement of willing and open-minded users, SerenA hosts an intangible and autonomous framework in which cognitive connections and new information can be discovered. Triggers for autonomous connection discovery are set in accordance with the user’s location (a city, a museum, a conference etc); their browsing habits, and explicit requests from the system… allowing SerenA to be used whilst socialising, eating, sleeping and travelling to places of interest. SerenA is a willing host to a multi-agent platform, some with more intelligence than others, that interact with the real world (such as the cloud of linked open data), and indeed with one another. Some agents have repetitive and simple tasks, and others are a composition of non-trivial behaviours. Some even represent real people, in the virtual world of agent interaction – mimicking the perceived characteristics and beliefs of the human counterpart. Evolving these user agents in non-obvious and subtle ways is something we hope to soon expand on.

We believe that the combination of semantic and highly structured data, with autonomous agent systems is novel, and yet we also recognise that both have their foundations laid in long standing Good Old Fashioned Artificial Intelligence. We see this only as a glowing endorsement!

This blog post is a high level summary of the technological approaches undertaken from the engineering strand in the SerenA project. Future posts will describe in more pointed detail, some of the methodologies exhibited by the software platforms we have evaluated, and how they assist us in our efforts in building a ubiquitous discovery platform for SerenA.

–
Rob Stewart
@robstewartUK
“Builds stuff”