The Data Team are coming to the end of their third release of the year and have started to plan out their upcoming Hackweek. Hackweeks are a great way for teams to work together on a shared problem, and to develop new areas of research. This time the Data team will be having a look at , a powerful and much discussed new tool for the automatic generation of language. The team think it could have a number of potential use-cases for the ±«Óãtv, ranging from the automated description of archive programmes through to the summarisation of news articles.
The team have also been beavering away on their respective workstreams - speech-to-text, recommendations, sentiment, natural language processing and the DSRP initiative. Of particular note this sprint are Chris’s experiments with the Mentions tagger tool, which is now working significantly better than our longstanding workhorse, Starfruit. Ben has also been doing some very useful background work in getting our development environment up, alongside his work helping Joanna out with the data analysis of her Sentiment survey. Finally, Fearn has hit the ground running on her graduate project, getting stuck into the challenge of debugging and training herself up on the many and varied joys of applied machine learning.
For Interaction and Prototyping, March was a busy month organising and running the testing of our prototype with external participants in a remote scenario. The team had a good break leading to Easter and now back in top gear with the analysis of the testing insights already underway.
In the Internet and Society team, Tristan, Miranda and Libby ran a workshop exploring ways to explain AI for NTAIL, , David having redone the materials in Google Slides to make them more accessible to participants without Miro access. Tristan's also been doing his 'understanding machine learning' talk a few times to various groups inside the ±«Óãtv.
David, Alicia and BenH have been focused on the last pieces of "tiffling" and testing for the ±«Óãtv Introducing prototype. It looks great, and is now with the Mix curators and other Introducing producers.
Alicia's been figuring out how display her chatbot survey results and is writing a blog post about it. Chris has been looking into FLoC. Libby has been writing "Futures" docs and helping Galen start up his project; Galen's been getting to grips with machine learning.