NoTube blog – the future of television is social and semantic

How willing are you to share data about what you watch on TV?

Posted in Beancounter by vickybuser on December 7, 2009

I’m working on NoTube as part of the BBC team, and together with Lora from VU, I’m looking into various potential privacy issues. In particular, as someone with a background in user experience, I’m interested to find out how willing people are to share information online about what they *actually* watch on TV – i.e. the type of attention data that can be counted with a Beancounter.

For example: you’ve spent years carefully cultivating and managing an online persona which reflects you in the best light – and you choose to post and tweet only about TV programmes which reinforce this persona. Given all this effort, do you really want your online contacts/friends to know that you’re also a secret fan of several ‘trashy’ TV shows – or this information that you’d rather keep to yourself?

On the other hand, how curious would you be to find out what your friends are watching (with their permission of course)? How about if you could get programme recommendations based on their tastes and preferences, combined with yours? For example, if you discovered that ten of your friends watched a programme about a subject you’re interested in that you hadn’t, might that influence you to watch it too? Yes? But how might your friends feel about sharing their TV viewing data with you?

Is what you choose to watch on TV in the privacy of your own sitting room at the end of a long day more intimately personal somehow than, say, the music you listen to, the DVDs you rent or the books you buy? And, if so, might this be because we tend to revert to ‘couch potato’ mode when we’re feeling tired and bored and need passive entertainment – when we want to indulge in the guilty pleasures of less highbrow content than we might pursue in other (more public) contexts?

The ability to share information online about programmes you’ve watched is already being made possible by experimental services such as 4iP’s Test Tube Telly and Whomwah’s twitter bot which sends details about what you are listening to and watching to the @radioandtvbot account on Twitter. Similarly Boxee allows you to match up your feed output to a twitter stream.

Do people have reservations about sharing this kind of data? If so, would the trade-off of potentially interesting programme recommendations based on friends’ viewing behaviours be worth any perceived risk of exposure? These are some questions I hope to answer during the course of this project.

I’d be interested to hear your thoughts. Please leave any comments on the blog.

Would you like to know more? [*]

Posted in Thinking Out Loud by libbymiller on December 1, 2009

I’m a representative of the BBC as one of the ‘Usecase’ partners in NoTube, and it’s my responsibility to draft a document describing one of the prototypes we want to make, combining components created by all the workpackages. We will demonstrate what we build at our project review next May, so it’s important to get it right, and having such a distributed team, clear communication of ideas inside the project is important too.

Here’s the basic idea, as described by Dan Brickley in our deliverable document draft:

Although there may never be a unified global network for TV content, there can and should be a unified global network of TV meta-content, by which we mean informally the overlay of annotations, ratings, comments, descriptions, related links and tags that can provide new means for users to inform, educate and entertain themselves online.

I’ll let Dan expand on his thoughts about this ‘NoTube Network’ in a later post, but my job is to propose a way we can show these ideas in action. Here is my current thinking, embodied in two simple scenarios:

Scenario 1: Recommendations for me on my TV

Jana wants to see recommendations based on her social activity on her TV when she gets home at night. She talks a lot on twitter and facebook about what she watches in the context of her online social life, and so do her friends, and doesn’t she see why she should have to explicitly tell any system what her preferences are. She wants to see recommendations clearly featured on the user interface of her set top box.

Scenario 2: Would you like to know more?

While watching TV, Frederik would sometimes like more information about a programme. He’d like to be able to ‘bookmark’ a programme to come back to it later and find out more about it. He doesn’t necessarily want to have his laptop open all the time during this and neither does he want to interfere with the playing of the programme too much as he often watches TV with other people in the same room.

The picture below shows the sorts of physical objects a person might want to use to fulfil these usecases, including TV, an open source set top box, a remote and a laptop – not just using their own laptop for everything, but in the context of the way most people still watch TV.

What the user sees

These are scenarios that will only be of immediate relevance to a relatively unusual group of people at the current time: those who commonly watch TV on their media centre, and who watch a lot of on-demand media and who heavily use social networking sites. It’s important to relate what we make to the majority of users as well, especially from my point of view given the inclusive nature of the BBC and the way it is funded.

In Scenario 2, wanting to know something more about a programme and remembering it for later is the key idea, a desire common to many TV watchers, regardless of the technologies they use. In Scenario 1, the filtering interesting programmes and bringing them to the user’s attention is the most important aspect, and one becoming increasingly relevant to all TV watchers as channels proliferate. Using a setup that replicates at least the feeling of watching TV in a living room allows us to address the most common way of watching TV and also to potentially look at face-to-face communal TV watching.

APIs, formats and applications

The APIs and data formats – the components that allow us to make the prototype but also specifically allow us to filter an EPG, export activity feeds, control a television and so on – are the pieces that can be reused in other user situations and can form part of the ‘NoTube Network’.

If you want to see more detail about the architecture and the prototype, you can see the requirements and description as a work in progress in the NoTube Deliverables GIT.

[*] “Would you like to know more…” is an interactive TV user interface component from the film “Starship Troopers”