September 14, 2015 | Written by: Anna Thomas
Categorized: Research | Social Business
Share this post:
The BBC Make It Digital tour is all about digital creativity and aims to inform and inspire people around the UK, both children and adults alike. When the tour visited Cardiff as part of the Harbour Festival I was delighted that Cardiff University would be showcasing “SHERLOCK”, a new kind of digital assistant that is based on some of our collaborative research with the University over the past few years under the ongoing International Technology Alliance (ITA) research programme.
Our main collaborator, Professor Alun Preece from Cardiff University, took the stage on Sunday and Monday to introduce the SHERLOCK assistant that they have been building and explain how human-machine conversation could help in everyday life and also in emergency or disaster situations. In our joint research we are investigating ways to create a seamless and intuitive environment in which normal untrained people can interact with a system: asking questions or providing information. Think of it as an attempt to “harness the power of people” but in a richer and more powerful and flexible way than traditional crowd-sourcing. We propose that it is possible to treat groups of willing participants as a kind of human sensing system if these sorts of capabilities are readily available and easy to use.
To help demonstrate the idea Alun and his colleagues had designed a crowd-sourcing game where the audience had to provide answers to a variety of “Dr Who” and “Sherlock” TV Series questions. The inspiration for these topics is that Cardiff is a regular location for filming of these and other BBC TV series. This game is a variation on some of our earlier research experiments in which we have enlisted the help of students at the University to collaboratively answer a series of “whodunit” style questions in a controlled environment and other work where we analysed Twitter conversations during the NATO summit in Newport in 2014.
I joined in the exercise and found that my Dr Who knowledge was far poorer than I imagined, but that was ok because I took the opportunity to visit the excellent Dr Who experience after the event and managed to fill in those gaps in my knowledge!
The experiment worked really well on both days and the audience managed to contribute a lot of information into the environment. The folks at Cardiff University are analysing the data now and I’m looking forward to carrying on with our collaborative research until May 2016 to see where else we can apply this interesting approach towards making agile human-machine teams a reality.
If you are interested in more details about the research we’ve been doing you can see a variety of our publications here and here.