GB2438645A

GB2438645A - System for content item recommendation

Info

Publication number: GB2438645A
Application number: GB0610621A
Authority: GB
Inventors: Nicolas Lhuillier; Makram Bouzid; Kevin Christopher Mercer; Jerome Picault
Original assignee: Motorola Inc
Current assignee: Motorola Solutions Inc
Priority date: 2006-05-30
Filing date: 2006-05-30
Publication date: 2007-12-05
Also published as: GB0610621D0

Abstract

An apparatus for content item recommendation comprises a grouping processor (105) which groups user ratings for content items into rating groups in response to a content item match criterion. Each user rating comprises content item description data and preference data. A content item processor (109) receives content item data for a plurality of content items and a first recommendation processor (107) generates a set of content item recommendations for each rating group in response to the content item data and user ratings of that rating group. A second recommendation processor (111) generates a single set of content item recommendations for a first user by combining the first sets of content item recommendations of the rating groups in response a user preference profile for the first user. The user preference profile comprises a user preference indication for each rating group. The invention may allow efficient generation of recommendations for e.g. television programmes.

Description

METHOD AND APPARATUS FOR CONTENT ITEM RECOMMENDATION

Field of the Invention

The invention relates to recommendation of content items and in particular, but not exclusively, to recommendation of television or radio programmes.

Background of the Invention

In recent years, the availability and provision of multimedia and entertainment content has increased substantially. For example, the number of available television and radio channels has grown considerably and the popularity of the Internet has provided new content distribution means. Consequently, users are increasingly provided with a plethora of different types of content from different sources. In order to identify and select the desired content, the user must typically process large amounts of information which can be very cumbersome and impractical.

Accordingly, significant resources have been invested in research into techniques and algorithms that may provide an improved user experience and assist a user in identifying and selecting content.

For example, Digital Video Recorders (DVR5) or Personal Video Recorders (PVRs) have become increasingly popular and are increasingly replacing conventional Video Cassette Recorders (VCRs) as the preferred choice for recording television broadcasts. Such DVRS (in the following the term DVR is used to denote both DVRs and PVR5) are typically based on storing the recorded television programs in a digital format on a hard disk or optical disc. Furthermore, DVR5 can be used both for analogue television transmissions (in which case a conversion to a digital format is performed as part of the recording process) as well as for digital television transmissions (in which case the digital television data can be stored directly).

Increasingly, devices, such as televisions or DVR5 provide new and enhanced functions and features which provide an improved user experience. For example, televisions or DVR5 can comprise functionality for providing recommendations of television programs to the user. More specifically, such devices can comprise functionality for monitoring the viewing/recording preferences of a user. These preferences can be stored in a user preference profile and subsequently can be used to autonomously select and recommend suitable television programs for viewing or recording. E.g. a DVR may automatically record programs which are then recommended to the user, for example by inclusion of the automatically recorded programs in a listing of all the programs recorded by the DVR.

Such functionality may substantially improve the user experience. Indeed, with hundreds of broadcast channels diffusing thousands of television programs per day, the user may quickly become overwhelmed by the offering and therefore may not fully benefit from the availability of content.

Furthermore, the task of identifying and selecting suitable CMLO3689EV content becomes increasingly difficult and time-consuming.

The ability of devices to provide recommendations of television programs of potential interest to the user substantially facilitates this process.

In order to enhance the user experience, it is advantageous to personalise the recommendations to the individual user.

In this context, a recommendation consists in predicting how much a user may like a particular content item and recommending it if it is considered of sufficient interest.

The process of generating recommendations requires that user preferences have been captured so that they can be used as input by the prediction algorithm.

There are two main techniques used to collect user preferences. The first approach is to explicitly obtain user preferences by the user(s) manually inputting their preferences, for example by manually providing feedback on content items that the user(s) particularly liked or disliked. The other approach is to implicitly obtain user preferences by the system monitoring user actions to infer their preferences.

Although these techniques may be suitable for many single-user environments, they are not particularly well suited to many other environments or to multi-user environments.

For example, most of the known recommendation approaches are not ideal in the context of television viewing. A television or video recorder, such as specifically a DVR, is commonly a multi-user device and the activity of watching television is characterised by being a low effort and highly passive CMLO3689EV activity. In this context, although users ask for individual recommendations, creating individual user profiles tends not to be easy or effective.

Specifically, explicit elicitation of preferences is not effective as it is difficult for users to precisely describe their tastes. Furthermore, the user will typically consider it cumbersome and tedious to manually initialise and maintain a user preference profile.

Explicit feedback on programmes is impractical in user environments as it requires the user to be identified before the programme feedback can be recorded in order to allow the system to differentiate between the preferences of the different users.

Also, implicit learning of preferences tends not to be effective as current users would need to be automatically identified and in addition implicit learning does not work well in contexts such as radio or television since the radio or television is often used as a background medium and therefore may play programmes that are not of interest to the user(s).

Known recommendation systems accordingly tend to be inflexible and/or require a significant manual involvement of the user(s). Furthermore, conventional recommenders tend to be complex and especially require complex algorithms for manipulating user rating inputs to generate personalised content item recommendations, especially in multi user environments.

CMLO3689EV Therefore, an improved system for content item recommendation would be advantageous. In particular, a system allowing an improved user experience, increased flexibility, reduced complexity, improved suitability for multi-user environments, reduced need for user inputs and/or improved performance would be advantageous.

Summary of the Invention

Accordingly, the Invention seeks to preferably mitigate, alleviate or eliminate one or more of the above mentioned disadvantages singly or in any combination.

According to a first aspect of the invention there is provided an apparatus for content item recommendation, the apparatus comprising: grouping means for grouping user ratings for content items into rating groups in response to a content item match criterion, each user rating comprising content item description data and preference data; means for receiving content item data for a plurality of content items; first recommendation means for generating a set of content item recommendations for each rating group in response to the content item data and user ratings of that rating group; and second recommendation means for generating a single set of content item recommendations for a first user by combining the first sets of content item recommendations of the rating groups in response a user preference profile for the first user, the user preference profile comprising a user preference indication for each rating group.

CMLO36B9EV The user ratings may e.g. comprise user ratings for a plurality of users and may be anonymous user ratings. In particular, the user ratings may be user ratings which comprise no identity information of the originating user(s) of the user ratings.

The invention may allow an improved recommendation of content items. Specifically, the invention may e.g. provide increased flexibility and/or reduced complexity of the recommendation. For example, the invention may allow content recommendation in multi-user systems which can be personalised for an individual user without requiring that the user ratings are correlated to specific users or are individually manipulated for each user.

The invention may allow a facilitated operation and/or improved user experience. For example, the invention may allow a flexible and personalised recommendation without requiring substantial involvement by the individual users.

The invention allows a simple personalised user preference profile to be used to personalise recommendations in a multi-user environment and allows some common processing for a plurality of users.

The content items may specifically be television programmes or radio programmes. The apparatus may specifically be a television, a DVR or a media server.

According to an optional feature of the invention, the first recommendation means is arranged to match a first content item of the plurality of content items to at least a first CMLO3689EV rating group and to assign a preference value for the first content item in response to the preference data of the first rating group. The second recommendation means may be arranged to modify a user preference indication for the first rating group in response to the preference value for the first rating group.

This may allow a facilitated implementation while providing accurate personalised recommendations. In particular, it may allow both group preferences and individual preferences to be taken into account in the generation of recommendations.

According to an optional feature of the invention, the second recommendation means is arranged to combine the sets of content item recommendations by selecting content items in response to the preference indications for each rating group.

This may allow a facilitated implementation while providing accurate personalised recommendations.

According to an optional feature of the invention, the second recommendation means is arranged to select a number of content items included in the single set of content item recommendations for a given rating group in response to the user preference indication of the given rating group.

This may allow a facilitated implementation while providing accurate personalised recommendations. Specifically, the feature may allow low complexity, personalised recommendation and/or may allow a variety of content items to be recommended.

CMLO3689EV According to an optional feature of the invention, the apparatus further comprises means for receiving a user preference feedback for the first user for a first content item of the single set of content item recommendations; and means for modifying a user preference indication for a ratings group to which the first content items belongs in response to the user preference feedback.

This may allow improved recommendations and may in particular allow an adaptation of the user preference profile to the user's preferences. The feature may allow low complexity and facilitated maintenance of the user preference profile. The user preference indication may for example be modified to increase the preference or decrease the preference and may specifically be modified to prevent any content items from the set of content item recommendations for the specific rating group to be included in the final single set of content item recommendations.

According to an optional feature of the invention, the apparatus further comprises means for receiving an identification of the first user and wherein the second recommendation means is arranged to retrieve the user preference profile stored for the identification.

This may allow improved recommendations and in particular the apparatus may provide an improved user experience in a multi-user environment. The user preference profiles for a plurality of users may be stored and a single user identification provided at the time of requesting a CMLO3689EV recommendation is sufficient to allow a personalised recommendation list to be generated for that user.

According to an optional feature of the invention, the storing means comprises a default user preference profile and the second recommendation means is arranged to retrieve the default user preference profile if no user preference profile is stored for the first user.

The default user preference profile may specifically correspond to an equal preference for each rating group and/or may correspond to an average user preference profile of the stored user preference profiles.

This may allow an improved user experience and may in particular allow generation of recommendations for new users. The feature may furthermore facilitate the initialisation and adaptation for new users.

According to another aspect of the invention, there is provided a method of content item recommendation, the method comprising: grouping user ratings for content items into rating groups in response to a content item match criterion, each user rating comprising content item description data and a preference data; receiving content item data for a plurality of content items; generating a set of content item recommendations for each rating group in response to the content item data and user ratings of that rating group; and generating a single set of content item recommendations for a first user by combining the first sets of content item recommendations of the rating groups in response a user preference profile for the first user, the user preference CMLO3689EV profile comprising a user preference indication for each rating group.

These and other aspects, features and advantages of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.

Brief Description of the Drawings

Embodiments of the invention will be described, by way of example only, with reference to the drawings, in which FIG. 1 is an illustration of an example of a device for content item recommendation in accordance with some embodiments of the invention; FIG. 2 illustrates an example of data being processed in a device for content item recommendation in accordance with some embodiments of the invention; and CMLO3689EV FIG. 3 illustrates an example of a method of content item recommendation in accordance with some embodiments of the invention.

Detailed Description of Some Embodiments of the Invention The following description focuses on embodiments of the invention applicable to a recommendation system for television programmes. However, it will be appreciated that the invention is not limited to this application but may be applied to many other content items including for example radio programmes, audiovisual files, music files etc. FIG. 1 is an illustration of a device for making content item recommendations. The device may for example be a DVR or a television.

The device of FIG. 1 comprises functionality for recommending content items to a user. For example, the device may recommend upcoming television programmes to the user of the device. The device uses an approach for generating recommendations which is highly flexible yet allows a high degree of personalisation. Specifically, the device uses a two-stage approach wherein user ratings are first clustered and wherein personalisation is then achieved in response to user preferences for the clusters of content items. This approach may provide an efficient implementation with high flexibility and is in particularly useful for multi-user environments wherein, possibly anonymous, user ratings may be received from a plurality of users.

CMLO3689EV The device comprises a user input 101 that can receive manual inputs from one or more users. Specifically the user input 101 can receive feedback of the user preferences for various content items. As an example, a user watching or playing back a specific television programme can manually input how he rates the program.

The user input 101 is coupled to a user rating store 103.

When a user preference is received from the user input 101, a user rating record comprising the user preference measure and contents item data describing the contents are stored in the user rating store 103.

The user rating record can for example store the user preference as a number between 1 and 10, the genre of the television programme, the title of the television programme, the duration of the television programme, people involved in the television programme (such as actors or directors) etc. In the example, the device is a multi-user device that may be used by many different users. Furthermore, the user preferences are inputted without any identification of the specific user that is providing the data. Accordingly, the user rating records stored in the user rating store 103 are anonymous user ratings and the records do not comprise any information of the identity of the user who provided the input. Hence, it is not feasible to generate content item recommendations which are personalised to an individual user based only on the stored user ratings. Rather, such an approach will only be able to provide personalised recommendations for the group of users using the device.

CMLO3689EV A user rating store 103 is coupled to a grouping processor which is arranged to cluster or group user rating records into groups of user ratings. The grouping of the user rating records is performed in response to a content item match criterion which may be any suitable match criterion that allows a grouping of content items into groups having desirable common characteristics. The match criterion may be a simple similarity criterion for specific characteristics of the user ratings or may e.g. be a complex clustering algorithm.

For example, the content item match criterion may require that a content characteristic, such as a genre or actor, is the same for all the content items in a given group.

Additionally or alternatively, the content item match criterion may require that user preferences for content items in the same group are the same or similar. For example, the grouping processor 105 can generate groups as content items corresponding to for example movies the users like, movies the users do not like, actors the users like, actors the users do not like, etc. In more complex embodiments, the grouping processor 105 may for example group the content items by using a clustering algorithm such as a k-means or isodata clustering algorithm.

A k-means clustering algorithm initially defines k clusters with given initial parameters. The user rating records are then matched to the k clusters. The parameters for each cluster are then recalculated based on the user rating records that have been assigned to each cluster. The algorithm then proceeds to reallocate the user rating CMLO3689EV records to the k clusters in response to the updated parameters for the clusters. If these operations are iterated a sufficient number of times, the clustering converges resulting in k groups of content items having similar properties.

The devise furthermore comprises a first recommendation processor 107 which is coupled to the grouping processor 105. In addition the first recommendation processor is coupled to a content item processor 109. The content item processor 109 receives information of various content items which are eligible to be recommended to a user.

For example, the content item processor 109 can be provided with information of the television programmes that are to be received within a given time interval. Specifically the content item processor 109 can receive an Electronic Programme Guide (EPG) that indicates the television programmes that will be transmitted in, say, the next week.

In addition to the time and titles of the television programmes, the EPG can contain further meta-data such as an indication of the genre, actors, directors etc. As another example, the content item processor 109 may alternatively or additionally be provided with information of television programmes that has been recorded by e.g. a DVR.

The first recommendation processor 107 is arranged to generate recommendations for each of the user rating groups which were determined by the grouping processor 105. Thus, the first recommendation processor 107 processes each user rating group independently of the other user rating groups.

CMLO3689EV For each user rating group, a list of recommendations is generated.

Specifically, for each user rating group, the first recommendation processor 107 compares each of the potential content items from the content item processor 109 to the characteristics of the user rating group. If the match is sufficiently close, the content item is considered to belong to this group and is accordingly considered to have a rating that can be determined from the user ratings of the group.

As a simple example, for a given user rating group a user preference value can be set to correspond to the average of all the user preference values for the user rating records associated with group. Thus, if the content item is found to match a group, it is included in the list of recommendations for that group and is assigned the rating of the group.

Thus, the first recommendation processor generates a number of recommendation lists with each list comprising a number of content items that are considered to have characteristics matching the group.

Hence, when generating recommendations, the device retrieves the list of content available for the time period being considered (for instance via the EPG) and uses the groups to compute recommendations. For each piece of content, this is done by determining the closest group (e.g. using a similarity or distance function) and computing the recommendations for that group using a content matching algorithm and the programme ratings of this group. This process results in obtaining one list of recommendations per CMLO3689EV group. These lists may e.g. be sorted first by the confidence level of the prediction, for instance using a threshold, and then by the actual value of the prediction.

The first recommendation processor 107 is coupled to a second recommendation processor 111. The second recommendation processor 111 is arranged to generate a single list of recommended content items by combining the list of content items generated by the first recommendation processor 107. The second recommendation processor 111 is furthermore coupled to a user preference profile store 113.

The user preference profile store 113 stores individual profiles for the individual users of the device.

Prior to the generation of recommendations for a specific user, the user authenticates with the device in order to be identified. The identification may for example be by clicking on a dedicated button on a remote control, selecting an appropriate icon or a menu on the television screen, entering an identity code or any other identification mechanism (e.g. biometric).

The second recommendation processor 111 receives the identification information and in response it retrieves the user preference profile for the current user from the user preference profile store 113. Thus, if the user already has an individual profile (for instance created at bootstrap or during previous usage of the system), the second recommendation processor 111 obtains this profile.

The user preference profile comprises preference data indicative of an individual user preference for the user CMLO3689EV rating groups. The user preference profile is thus associated with the user rating groups and does not indicate specific preferences for individual content items, or individual content item characteristics. Rather, it provides a user preference indication for each rating group.

The user preference profile can for example indicate that a specific user rating group is rated highly by the user whereas another user rating group is not rated very highly.

The combination of the recommendation lists for the different groups is then performed taking this rating into account.

As an example, the second recommendation processor 111 selects content items from the individual lists to be included in the final list depending on the ratings which is assigned to the content items of the group lists.

As a specific example, the second recommendation processor 111 can select a number of content items from each group list where the number of content items that are selected from each group depends on the rating of that group in the user preference profile. For example, if the user preference profile indicates a high preference a first number of content items is selected (e.g. five content items may be selected), if the user preference profile indicates a lower preference a lower number of content items is selected (e.g. three content items may be selected), if the user preference profile indicates an even lower preference an even lower number of content items is selected (e.g. one content items may be selected). If the user preference profile indicates a dislike for the group, no content items are selected.

CMLO3689EV In some embodiments, the user preference profile may be used to more gradually bias the different rating groups.

Specifically, a rating or preference for the group can be determined in response to the preference indication in the user preference profile and the rating determined from the content items in the rating group. For example, the user preference indication of the user preference profile for a given user rating group may be used to bias the rating or ratings determined from the user rating records by the first recommendation means. Specifically, if the user preference profile indicates a preference for a given group or cluster, the rating(s) of this group may be increased, whereas if the user preference profile indicates a negative preference for a given group or cluster, the rating(s) of this group may be decreased. The modified ratings can then be used to select the number of content items taken from each group.

The selection of the content items for the single list may for example be by selecting the highest rated content items of each list and/or the content items that most closely match the group characteristics. In some embodiments, the single lists may have an equal rating for all recommended content items and the content items for the final list may simply be randomly selected.

The second recommendation processor 111 is coupled to a recommendation output 115 which is fed the personalised list of recommended content items. The recommendation output 115 can output this list to the user in any suitable form. For example, for a television application, the recommendation output 115 can display a list of upcoming television CMLO3689EV programs that are considered of particular interest for the currently authenticated user. Likewise, for a DVR the recommendation output 115 can present a list of recommended television programmes to the user and/or can automatically record the recommended content items.

FIG. 2 illustrates an example of how data may be processed by the device of FIG. 1.

The device receives content item data 201 for example in the form of an EPG. The content item data 201 is compared to the rating groups 203 to generate a plurality of lists 205. In the specific example, one list of recommendations 205 is generated for each rating group. In the example, a first list corresponds to recommended sports programs, a second list corresponds to recommended soaps and a third list corresponds to recommended movies. The individual lists 205 are then processed with reference to the specific individualised user preference profile 207 that is associated with the user for which the recommendations are generated. As a result, a single list of recommendations 209 is generated. This listis highly personalised for the individual user although it is predominantly based on anonymous user ratings from a plurality of users. In the example, the user preference profile 207 may indicate that the user has a high preference for sports programs, does not like soaps, and has a medium preference for movies.

Accordingly, the final list has a high number of sport programmes, a lower number of movies and no soaps recommended.

CMLO3689EV In the example, the generation of recommended content item lists for each user rating group is based entirely on information which is not specific to the individual user for which the recommendation list is generated but rather is associated with the entire group of users that use the device. Thus, the second recommendation processor 111 uses an individual user preference profile to modify the user group preferences and characteristics to more closely adapt to that of the individual user. However, this processing is performed without requiring that all the data is processed or sorted for the individual user. Thus, a low complexity, flexible and accurate content item recommendation is achieved. Specifically, in some embodiments, the grouping of user ratings and the content item recommendations for each user rating group may be common for all users with the only individual personalisation being introduced by the second recommendation processor 111 in response to the user preference profile.

The above examples assumed that a user preference profile was available for the first user. However, if a recommendation list is generated for a user which has not identified himself or for which there is no individual user preference profile stored, a default user preference profile may be used. This default user preference profile can specifically correspond to a user preference profile which has an equal preference for all rating groups. For example, this may lead to an equal number of content items being selected from each user rating group. Thus, by using a default user preference profile, a content item recommendation list is generated that conforms to the preferences of the user group as a whole.

CMLO3689EV The device may furthermore be capable of dynamically modifying the user preference profile to more accurately reflect the individual user's preferences.

Specifically the user input 101 can receive user preference feedback for the specific content items that are recommended to the user. The user may e.g. indicate that he would like more content items recommended which are similar to a specific content item, that he would like fewer content items recommended which are similar to a specific content item or that he does not want any content items recommended which are similar to a specific content item. In response, the device may modify the individual preferences stored in the user preference profile.

Specifically, if the user indicates that he particularly likes a first content item, the user preference indication in the user preference profile for the group to which this content item belongs is increased. Similarly, if the user indicates that he does not particularly like a first content item, the user preference indication in the user preference profile for the group to which this content item belongs is decreased. If the user indicates a specific dislike for a first content item, the device may indicate that this rating group should be banned and no content items of the list for this rating group should be included in the final list of recommendations.

Such adaptation may also allow an efficient and user-friendly initialisation for a new user. Thus, whenever a recommendation of content items is requested, the device can CMLO3689EV retrieve the user preference profile stored for that user and can store the modified user preference profile that takes into account the user preference feedback for the recommendation list. When a user is provided with a recommendation for the first time, a default user preference profile is used and this is modified in accordance with the user preference feedback from the user. The modified user preference profile is then stored. This process is repeated every time a new recommendation list is generated and accordingly the user preference profile increasingly accurately reflects the user's preferences.

In some embodiments, the grouping of the processing of the user ratings may be performed every time a new recommendation is generated. However, in other embodiments, the grouping processor 105 can use groupings from previous recommendations. Specifically, the grouping processor 105 may select all the new user ratings which have not before been included in a grouping process and may update the groupings to include these. For example, the grouping processor 105 can simply detect the closest user rating group for each new user rating and can include the new user rating in this group.

In some embodiments, the user rating groups are stored for each individual user and the grouping into the user rating groups may be different for different users. In such an example, after providing feedback on the recommended list, the user may decide to store the final configuration of user rating groups for later reuse. In such a case, the new groups can replace the previously stored groupings.

CMLO3689EV FIG. 3 illustrates an example of a method of content item recommendation in accordance with some embodiments of the invention.

The method initiates in step 301 wherein user ratings for content items are grouped into rating groups in response to a content item match criterion. Each user rating comprises content item description data and a preference data.

Step 301 is followed by step 303 wherein content item data is received for a plurality of content items.

Step 303 is followed by step 305 wherein a set of content item recommendations is generated for each rating group in response to the content item data and user ratings of that rating group.

Step 305 is followed by step 307 wherein a single set of content item recommendations is generated for a first user by combining the first sets of content item recommendations of the rating groups in response a user preference profile for the first user. The user preference profile comprises a user preference indication for each rating group.

It will be appreciated that the above description for clarity has described embodiments of the invention with reference to different functional units and processors.

However, it will be apparent that any suitable distribution of functionality between different functional units or processors may be used without detracting from the invention. For example, functionality illustrated to be performed by separate processors or controllers may be CMLO3689EV performed by the same processor or controllers. Hence, references to specific functional units are only to be seen as references to suitable means for providing the described functionality rather than indicative of a strict logical or physical structure or organization.

The invention can be implemented in any suitable form including hardware, software, firmware or any combination of these. The invention may optionally be implemented at least partly as computer software running on one or more data processors and/or digital signal processors. The elements and components of an embodiment of the invention may be physically, functionally and logically implemented in any suitable way. Indeed the functionality may be implemented in a single unit, in a plurality of units or as part of other functional units. As such, the invention may be implemented in a single unit or may be physically and functionally distributed between different units and processors.

Although the present invention has been described in connection with some embodiments, it is not intended to be limited to the specific form set forth herein. Rather, the scope of the present invention is limited only by the accompanying claims. Additionally, although a feature may appear to be described in connection with particular embodiments, one skilled in the art would recognize that various features of the described embodiments may be combined in accordance with the invention. In the claims, the term comprising does not exclude the presence of other elements or steps.

CMLO3689EV Furthermore, although individually listed, a plurality of means, elements or method steps may be implemented by e.g. a single unit or processor. Additionally, although individual features may be included in different claims, these may possibly be advantageously combined, and the inclusion in different claims does not imply that a combination of features is not feasible and/or advantageous. Also the inclusion of a feature in one category of claims does not imply a limitation to this category but rather indicates that the feature is equally applicable to other claim categories as appropriate. Furthermore, the order of features in the claims does not imply any specific order in which the features must be worked and in particular the order of individual steps in a method claim does not imply that the steps must be performed in this order. Rather, the steps may be performed in any suitable order.

CMLO3689EV

Claims

CLAIMS

1. An apparatus for content item recommendation, the apparatus comprising: grouping means for grouping user ratings for content items into rating groups in response to a content item match criterion, each user rating comprising content item

description data and preference data;

means for receiving content item data for a plurality of content items; first recommendation means for generating a set of content item recommendations for each rating group in response to the content item data and user ratings of that rating group; and second recommendation means for generating a single set of content item recommendations for a first user by combining the first sets of content item recommendations of the rating groups in response a user preference profile for the first user, the user preference profile comprising a user preference indication for each rating group.

2. The apparatus of claim 1 wherein the user ratings comprise user ratings for a plurality of users.

3. The apparatus of claim 2 wherein the user ratings are anonymous.

4. The apparatus of any of the previous claims arranged to generate content item recommendations for a plurality of users and wherein the rating groups are common to a plurality of users CMLO3689EV 5. The apparatus of any of the previous claims wherein the first recommendation means is arranged to match a first content item of the plurality of content items to at least a first rating group and to assign a preference value for the first content item in response to the preference data of the first rating group.

6. The apparatus of claim 5 wherein the second recommendation means is arranged to modify a user preference indication for the first rating group in response to the preference value for the first rating group.

7. The apparatus of any previous claim wherein the second recommendation means is arranged to combine the sets of content item recommendations by selecting content items in response to the preference indications for each rating group.

8. The apparatus of claim 7 wherein the second recommendation means is arranged to select a number of content items included in the single set of content item recommendations for a given rating group in response to the user preference indication of the given rating group.

9. The apparatus of any previous claim wherein the content item match criterion comprises a content match criterion.

10. The apparatus of any previous claim wherein the content item match criterion comprises a user preference indication match criterion.

11. The apparatus of any previous claim further comprising: CMLU3689EV means for receiving a request for content item recommendations and in response determining if any user ratings have been generated since a previous grouping; and wherein the grouping means is arranged to update the grouping of user ratings to include the user ratings generated since the previous grouping.

12. The apparatus of any previous claim comprising: means for storing the rating groups and wherein the first and second recommendation means is arranged to use rating groups from a previous generation of user recommendations.

13. The apparatus of any previous claim further comprising means for receiving a user preference feedback for the first user for a first content item of the single set of content item recommendations; and means for modifying a user preference indication for a ratings group to which the first content items belongs in response to the user preference feedback.

14. The apparatus of any previous claim further comprising storing means for storing the user preference profile and wherein the second recommendation means is arranged to retrieve a user preference profile from a previous generation of content item recommendations.

15. The apparatus of claim 14 further comprising means for receiving an identification of the first user and wherein the second recommendation means is arranged to retrieve the user preference profile stored for the identification.

CMLO3689EV 16. The apparatus of claim 14 or 15 wherein the storing means comprises a default user preference profile and the second recommendation means is arranged to retrieve the default user preference profile if no user preference profile is stored for the first user.

17. The apparatus of claim 16 wherein the default user preference profile corresponds to an equal preference for each rating group.

18. The apparatus of any previous claim wherein the means for receiving content item data is arranged to extract at least some of the content item data from an Electronic Programme Guide.

19. A method of content item recommendation, the method comprising: grouping user ratings for content items into rating groups in response to a content item match criterion, each user rating comprising content item description data and a preference data; receiving content item data for a plurality of content items; generating a set of content item recommendations for each rating group in response to the content item data and user ratings of that rating group; and generating a single set of content item recommendations for a first user by combining the first sets of content item recommendations of the rating groups in response a user preference profile for the first user, the user preference profile comprising a user preference indication for each rating group.

CMLO3689EV 20. A computer program product enabling the carrying out of a method according to claim 19.

CMLO3689EV