IRFS Weeknotes #137
These are weekly notes from the Internet Research & Future Services team in BBC R&D where we share what we do. We work in the open, using technology and design to make new things on the internet. You can follow us on Twitter at @bbcirfs
I have the privilege of writing the first IRFS weeknotes on our new website. This may mean that you are reading these weeknotes a few weeks late, but such is the price of progress.
This week: A new person joins the team, we do some hardware hacking, plan some projects and bring a wealth of radio content to Snippets.
First off, a huge welcome to our latest team member Thomas Parisot. Thomas brings an incredible amount of experience in front-end web development to the team and has already started to make an impact fixing bugs in Snippets and sharing his thoughts on our team mailing list. We're delighted to have him on board.
This week, the team has been working on a couple of hardware-based projects. Jasmine has continued work on a functioning prototype of a key fob which can be used as a standalone button to interact with radio programmes. She's built a fob with an accelerometer which uses Androino to control an android app using bluetooth. The accelerometer can be used to detect a number of states, including "freefall" which is then transmitted to the phone for display. A discussion on how the accelerometer could be used to detect weightlessness ensued on our internal mailing list. In the interests of reproducible research, Yves conducted a free-fall experiment of his own. He's now in the market for a new phone.
Meanwhile, Andrew N, Dan and I spent a bit of time working on our port of Go Free Range's little printer to the Raspberry Pi. Dan and Andrew paired on a node.js application to handle the printing and I added some features to an application to allow radio producers to print out incoming tweets to their shows.
We're also planning a physical prototyping day to explore some of these ideas together as a team, Andrew has been organising the event and getting some equipment together. More on that soon!
The Snippets team have been making headway getting all of the Redux radio archive into the Snippet interface. Matt has been indexing all the content and working on ways to keep the 800,000 links between PIPs and Redux accurate and up-to-date. James, with Yves's help, has been commissioning servers using Puppet to start extracting metadata from the archive content using some of the technology developed as part of the ABC-IP project. He's also been developing APIs to find appropriate images. Meanwhile Chris F and Andrew W have been working on a new snipping interface for audio-only content.
In the metadata team Denise has continued her work on mood segmentation using some ground truth data gathered as part of some user trials. Yves has been wrapping up an ABC-IP deliverable. His main contribution to this part of the work has been a new API called "Mango". It uses Named Entity Recognition to extract key terms from text documents and then attempts to match and disambiguate them against DBpedia topics. According to Yves, "It seems to work OK in some cases, but gets really confused when the right disambiguation for a particular topic is not available in DBpedia". Jana is working on a paper for a special session on semantic media for WIAMIS, and on her tech-note about a Matlab/Octave machine learning toolbox she's been developing.
Elsewhere, Michael attended an Music Open Data event and blogged about it , while the World Service Archive team welcomed Michael Magruder, an artist from King's College London, to discuss a collaborative project looking at the artistic possibilities for some of the data we've collected as part of the project. Vicky and Libby have been thinking about set-top boxes for children and Tristan has been contributing to a paper on "citizen social science".
Olivier's week was taken up with work on the new R&D website, which, if you're reading this, was very successful!