Recommending content

A user device for recommending content to a user including:

• display means arranged to present content to a user;
• a user interface arranged to receive one of at least a first input and a second input in relation to content presented to a user, the display means being arranged to display the presented content in response to a first input;
• and, section means arranged, in response to a second input, to select fresh content for presentation, each content having a probability associated therewith, the selection means being arranged to select content in dependence on the associated probability of that content.

Method for recommending content items to users, said method using a user interface comprising at least two options, one option allowing users to approve a content item and the other option allowing users to discard a content item, wherein a new content item is presented to the users when they select option and discard a content item, wherein said method uses probabilistic rules for recommending content items to users, in that said probabilistic rules for recommending content items to users are updated when a user selects option and approves a content item, with the approved content item's probability being increased and in that the magnitude of the increase is evaluated as a function of previous recommendation probabilities and the number of content items viewed and discarded by said user before it approved said content item.

Method for recommending content items to a user, including:

• (i) receiving one of at least a first input and a second input from a user in relation to content presented to the user;
• (ii) in response to a first input, rendering the presented content, or in response to a second input, selecting fresh content for presentation;
• and, (iii) repeating steps (i) and (ii) until a first input is received, wherein content has a probability of being selected associated therewith, and content is selected in dependence on the associated probability, the probability being increased in response to a first input, wherein the increase is determined at least in part on a measure of a predicted reduction in user satisfaction associated with an additional second input.