This year IBM Emerging Technologies UK submitted 10 out of 86 entries for TechConnect. TechConnect is a Europe wide internal IBM annual competition for individuals showcasing their work to senior technical leaders and decision makers.
This years winning entry, “Showcasing our academic research”, was one of the entries from the UK Emerging Technology Team. It is a web application which visualises a large data set of academic research papers. The Science Library uses a rich knowledge base generated from the 10 year NIS ITA collaborative research programme.
The Science Library creates a rich social network visualising documents, authors, organisations, venues and topics and the relationships between them to allow for easy navigation between views. Below shows a visualisation of the author Dave Braines and his co-authors.
We also measure impact of individual users and organisations through data such as citation count and h-index.
The Science Library has been demonstrated at high profile US and UK events and has received great reviews from the clients. IBM research, a client and a large US research programme are in the process of creating their own versions with their own data. Creating your own version requires the simple step of transforming your data into Controlled English using the model we’ve created. To make the data curation as easy as possible we can generate the knowledge base from an Excel spreadsheet This also enables non-technical staff to create and review the data in a familiar setting.
This project uses a Controlled English (CE) model to describe the underlying data. Controlled English is a human-friendly language used to enable non-technical users to define their semantic models in a language that is familiar to them already. CE is an example of a Controlled Natural Language, a subset of English, that is both human-readable and directly machine parseable. It is designed for the expression of everyday knowledge, concepts, and reasoning.
An environment called the CE Store is available on Github for building rich semantic CE models in a very agile way. It provides the powerful capabilities promised by semantic knowledge-base environments, but without the technical complexity involved in building them.
In addition to our winning entry we had a really strong set of submissions from across the Emerging Technologies UK team showcasing the breadth and impact of some of the work we are involved with. For example, our ongoing research and development with IBM research in the area of Fully Homomorphic Encryption, exciting work with embedded cognitive agents in the world of Minecraft, and some interesting ideas about harnessing the power of IoT to better enable power distribution via the UK national grid.