NoTube Archive Browser Experiment 1

This is an experiment for people inside the BBC. You are welcome to share links, but only to BBC colleagues (sorry!)

Introduction

We were inspired by an XKCD cartoon to ask: can we make a large video collection interesting enough so that people keep browsing rather than give up? Many people find choosing in large collections difficult, and 'what's popular' and 'what's new' mean that many items are never viewed.

This prototype is based on metadata about approximately 23,000 programmes from the BBC archives. It uses a similarity match between different programmes to suggest related programmes to you. If you have a Redux account (BBC staff only) you can then watch the videos on your playlist.

For this test we are specifically interested in how people browse through a large video collection, so we have deliberately removed the search box.

About NoTube

NoTube is a three year EU-funded project about Web and TV. There's more information on our website. The BBC is one of thirteen partners in the project. Within the BBC, NoTube is part of the Audience Experience section in R&D.

Instructions

Please spend as much or as little time as you like over the next seven days to browse through this video collection and add programmes that you'd like to watch to your playlist using the [+] link beside each programme. If you get stuck, click on 'shuffle' to get a new random set of programmes.

You can use this application as many times as you want during the course of the week. You can access the playlist you make until the end of the NoTube project (March 2012) - please bookmark the permalink for the playlist if you want to do this (under 'playlist').

At the end of the seven days we would like you to fill in a questionnaire to let us know if you found the application useful.

Problems

Please email us if you come across any bugs or would like to ask us any questions (vicky.buser1 at bbc.co.uk or libby.miller at bbc.co.uk). It helps if you send us your user id and current url (go to 'report bug' for an easy to copy and paste version of these pieces of information).

There are a few known problems:

Testing

This is an evaluation so we are testing different configurations for different people at random. We want to test two specific things:

You will get therefore one of four different options:

We suspect that the last one - random false, categories true - will be the most interesting version but we really don't know, which is why we want to test it. We will mail round a link to this version, and to the results of the research, after the end of the test.

Who gets picked for which group is completely random.

Technical details

Privacy

Cookies

We place two cookies on your computer: one that stores your random user_id (which you can see in 'report bug') and one that stores the id of the last programme you visited, so you will be taken back to it. These both last seven days. After that time you'll get prompted for your consent again, and you'll be assigned a new user id. The playlist you make is keyed off your user_id, so after seven days this will change. You can still access the old one, but you need to make a note of the permalink. Removing the cookies resets the programme, though the backend still keeps a record of your playlist, as long as you know your user id.

Your email

We ask for your email because we want to be able to contact the people who participate in this initial study for a follow up. You are of course perfectly free to refuse, and you can still use the system without entering your email. We will destroy any data with emails in it at the end of the project or before then if you ask us to.

Tracking

We track the links you click in the application in order to find out which kinds of links are interesting, and so help improve the application in the future. We are also interested in the size of playlists you generate and the length of time and the path you take through the application. No personally identifying information is stored with this tracking data. We plan to use this data for statistical analysis in our report on this experiment (this will be a public document). If you choose to participate in a follow-up study, we will use the data to create personal summary profiles of your interests.

Deleting data

You can remove items from your playlist at any time. You are welcome to ask us to remove any other data at any time (and we will!).