Pek van Andel’s talk on “Anatomy of the Unsought Finding”
The experience of `accidental’ discovery and acquisition of information generally known as Serendipity refers to `accidentally’ bumping into (new, true, useful, or personal interest-related) information, initially not looked for. Serendipity also plays an important role in the emerging field of data science by enhancing information retrieval and by promoting unexpected knowledge discovery. The World Wide Web has provided a global information space comprising billions of connected documents. “The unexpected connection is more powerful than one that is obvious”, as aptly asserted by Haraclitus in 500 BC. However, most of the existing centralized ‘nearest neighbor’ search approaches on the Web, such as Google, although very useful in finding explicitly relevant results, are killing serendipity by excessively limiting the encountering of unexpected information.
On the other hand, the ever-growing amount of Linked Data publicly accessible and distributed on the Web increases the likelihood that some of the data, which will make an impact in our professional or private lives will come to us by chance—without searching it initially.
`Unsought discoveries’ most often take place in the context of browsing unbounded data spaces; people immerse themselves in the items that interest them, meandering from topic to topic, and so on and so forth (i.e., the ‘Follow-Your-Nose‘ method to traverse the given semantic links from a resource) while concurrently remarking interesting and informative information en route.
Therefore, flexible and intuitive browsing user interfaces (UIs) which support serendipity triggers, can increase the likelihood of accidental knowledge discovery on Linked Open Data (LOD).
SMS (Semantically Mapping Science) platform within the RISIS project provides a novel UI for browsing heterogeneous data scattered over multiple distributed knowledge graphs. One of the features integrated into our SMS platform is the capability to foster serendipitous knowledge discovery. In order to realize this feature in the SMS platform, we collaborated with dr. Pek van Andel from the university medical center Groningen, a well-known ‘serendipitiologist’ who helped us to add some serendipity-fostering patterns into the SMS linked data browser.
On Friday 19 May 2017, Pek van Andel gave a lecture at Vrije Universiteit Amsterdam on the “Anatomy of the Unsought Finding. Serendipity: Origin, History, Domains, Traditions, Appearances, Patterns and Programmability“.
During his talk, he presented the process of serendipity together with a set of patterns which can foster serendipity.
Some of these patterns are already supported by the SMS faceted browser:
– Make surprising observations more noticeable.
– Facilitate the explanation of surprising observations.
– Allow sharing of surprising observations among multiple users.
– Make errors in data more visible in order to detect successful errors easier.
– Enable reasoning by analogy.
– Allow inversion and contrast.
– Support randomization and disturbance.
– Support extending the memory of user by invoking provocative reminders.
– Allow monitoring of side-effects when interacting with data.
– Support detection and investigation of by-products.
– Support background knowledge and user contextualization.
– Support both convergent and divergent information behavior.