EP1891588A1 - Method and apparatus for estimating total interest of a group of users directing to a content - Google Patents

Method and apparatus for estimating total interest of a group of users directing to a content

Info

Publication number
EP1891588A1
EP1891588A1 EP06744970A EP06744970A EP1891588A1 EP 1891588 A1 EP1891588 A1 EP 1891588A1 EP 06744970 A EP06744970 A EP 06744970A EP 06744970 A EP06744970 A EP 06744970A EP 1891588 A1 EP1891588 A1 EP 1891588A1
Authority
EP
European Patent Office
Prior art keywords
directing
users
value
interest
time slice
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
EP06744970A
Other languages
German (de)
French (fr)
Inventor
Xiaowei Philips Electronics China SHI
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Pace Micro Technology PLC
Pace PLC
Koninklijke Philips NV
Original Assignee
Koninklijke Philips Electronics NV
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to CN200510073942 priority Critical
Application filed by Koninklijke Philips Electronics NV filed Critical Koninklijke Philips Electronics NV
Priority to PCT/IB2006/051570 priority patent/WO2006126147A2/en
Publication of EP1891588A1 publication Critical patent/EP1891588A1/en
Withdrawn legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce, e.g. shopping or e-commerce
    • G06Q30/02Marketing, e.g. market research and analysis, surveying, promotions, advertising, buyer profiling, customer management or rewards; Price estimation or determination

Description

METHOD AND APPARATUS FOR ESTIMATING TOTAL INTEREST OF A GROUP
OF USERS DIRECTING TO A CONTENT
FIELD OF THE INVENTION The present invention relates to the information processing technology, particularly to a method and apparatus for estimating total interest of a group of users directing to a content.
BACKGROUND OF THE INVENTION
With the social developments and technological progress, the knowledge people have acquired has been increasing at an exponential pace. With the tremendous amount of information available, methods and apparatuses capable of automatic initial processing of information are urgently needed to free people from their daily drudgery and enable them to more effectively pursue creative labor and better enjoy their life.
Content recommendation is a very hot technology in information processing, and its fundamental principle is to analyze content on the basis of pre-set determination conditions to find out the degree of interest (hereinafter referred to as interest) of particular users/a group of users, and then decide whether to recommend the content to said users/a group of users. This technology has a prospect of wide application in television programs recommendation. For that matter, many researches are now being carried on against this background.
Those to whom content is recommended may be an individual user, or a group of users that made up of at least two users or members. For example, a group of users may be made up of family members or roommates living in the same dormitory. Compared with individual users, a group of users relates to the mutual effect of the members therein; hence, the analysis and evaluation of the interest of a group of users are more complicated than those of an individual user.
In the US Pat Application No. 09/819,440, entitled "Method and Apparatus for Generating Recommendations for a Plurality of Users" and filed on March 28, 2001 has been disclosed a method for determining the results of recommendation to users. The inventors of the patent application are Lalitha Agnihotri and Srinivas Gutta, and the licensee is the Philips Electronics North America Corp. In the embodiments disclosed in said application document, the recommendation results are expressed with the recommendation points or composite recommendation points, derived from computation of preference of each one in a group of users towards the features contained in the information or content. The disclosure of US Pat Application No. 09/819,440 is inserted here as reference.
In a specific embodiment of US Pat Application No. 09/819,440, the method for determining the recommendation point: first obtain the electronic program guide (EPG), namely, the television program forecast information existing in an electronic form. Then, obtain the profile 300 of the member of a group of users containing each viewer's preference towards each feature. After that, if necessary, convert each viewer's preference towards each feature into same numeric rating. The following processing step is to determine, according to the viewer profile described above, the extent to which a program is worth recommending, or the point at which it is recommended to each viewer. After the recommendation points of all the viewers of the group are acquired, combine the recommendation points of all the viewers to come up with the total recommendation points of the group of users, in ways, such as taking the weighted average value or arithmetic average value from the recommendation points of all the viewers. Finally, output the computed recommendation points of the group of users.
To better determine interest of a group of users directing to a content as a whole, i.e. total interest, more factors may be taken into consideration.
SUMMARY OF THE INVENTION
One object of the present invention is to correctly determine total interest of a group of users directing to a content to recommend content to it more individually.
For one aspect, the present invention provides a method for estimating total interest of a group of users directing to a content, wherein the group of users includes at least two members, and the method comprising the steps of: obtaining information relating to available time slice of the content and features contained in the content; obtaining value of each member's priority in different time slice and preference of the member directing to the contained features; and determining the total interest according to the value of each member's priority in the available time slice and preference of the member directing to the contained features.
In one of the embodiments, the value of each member's priority is arranged that value of one of the member's priority is different on at least two time slices, the two time slices is from a plurality of time slices divided by a day. Or the value of each member's priority is arranged that that value of one of the member's priority is different on the same time slice of at least two different days.
In another embodiment, determining step comprises the steps of: determining value of interest of the group of users directing to the features contained in different time slice according to value of each member's priority in different time slice and the preference directing to the contained features; determining the value of interest of the group according to the available time slice and combining the value of interest of the group to acquire the total interest.
For another aspect, the present invention provides a method for recommending content to a group of users, wherein the group of users includes at least two members, and the method comprising the steps of: obtaining information relating to available time slice of the content and features contained in the content; obtaining value of each member's priority in different time slice and preference of the member directing to the contained features; determining the total interest according to the value of each member's priority in the available time slice and preference of the member directing to the contained features and recommending to the group of users according to the total interest.
Still another aspect of the present invention is to provide an apparatus for estimating total interest of a group of users directing to a content, wherein the group of users includes at least two members, comprising: obtaining means for obtaining information relating to the available time slice of the content and features contained in the content; receiving means for receiving value of each member's priority in different time slice and preference of the member directing to the contained features and determining means for determining the total interest according to the value of each member's priority in the available time slice and preference of the member directing to the contained features.
In one of the embodiments, the value of each member's priority is arranged that value of one of the member's priority is different on at least two time slices, the two time slices is from a plurality of time slices divided by a day. Or the value of each member's priority is arranged that that value of one of the member's priority is different on the same time slice of at least two different days. In another embodiment, the determining means comprises: generating means for determining value of interest of the group of users directing to the features contained in different time slice according to value of each member's priority in different time slice and the preference directing to the contained features; searching means for determining the value of interest of the group according to the available time slice; and computer means for combining the value of interest of the group to acquire the total interest.
Still another aspect of the present invention is to provide an apparatus for recommending content to a group of users, wherein the group of users includes at least two members, including: obtaining means for obtaining information relating to available time slice of the content and features contained in the content; receiving means for receiving value of each member's priority in different time slice and preference of the member directing to the contained features; determining means for determining the total interest according to the value of each member's priority in the available time slice and preference of the member directing to the contained features; and recommending means for recommending to the group of users according to the total interest.
Still another aspect of the present invention is to provide a computer program product for estimating total interest of a group of users directing to a content, wherein the group of users includes at least two members, comprising: codes for obtaining information relating to available time slice of the content and features contained in the content; codes for obtaining value of each member's priority in different time slice and preference of the member directing to the contained features; and codes for determining the total interest according to the value of each member's priority in the available time slice and preference of the member directing to the contained features.
The present invention also relates to a memory carrier containing the computer program product.
In the present invention, a way may be used in which priority of each member of the group varies from time to time to adapt to the many states where each group member's influence for content selection is dynamic, so as to recommend content to a group of users more correctly. In one embodiment, a member's priority may be set to be changing from one time slice to another. This implement has the advantage of simplicity and flexibility, and fits the television program interest estimation and recommendation. Besides, interest of the group of users' directing to each feature in every time slice is able to be pre-computed on this basis and it will be used in subsequent computation of the total interest, which reduces the computation expenses, and, thus, simplifies the hardware structure.
These and other objects and advantages of the present invention will be apparent and the present invention will be more fully understood from the following description and claims in conjunction with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
Fig. 1 is a flow diagram illustrating a preferred embodiment according to the method of the present invention for estimating total interest of a group of users directing to a content.
Fig. 2 is a flow diagram illustrating an exemplary embodiment of computing interest of a group of users directing to a feature based on fuzzy logic processing mode.
Figs. 3a, 3b and 3c illustrating the membership functions used in the fuzzy logic processing mode as shown in Fig. 2, corresponding respectively to the preference, priority and interest of a group of users directing to a feature.
Fig. 4 is a flow diagram illustrating another preferred embodiment according to the method of the present invention for estimating total interest of a group of users directing to a content.
Fig. 5 is a flow diagram illustrating a preferred embodiment according to the method of the present invention for recommending content to a group of users.
Fig. 6 is block diagram illustrating an embodiment according to the apparatus of the present invention for estimating total interest of a group of users directing to a content.
Fig. 7a is block diagram illustrating an embodiment according to the apparatus of the present invention for recommending content to a group of users. Fig . 7b is block diagram illustrating another embodiment according to the apparatus of the present invention for recommending content to a group of users.
The same referential numbers in all drawings indicate the similar or identical features or functions.
DETAILED DESCRIPTION OF THE INVENTION
Following is a detailed description of the preferred embodiments of the present invention on the basis of the drawings. Fig. 1 is a flow diagram illustrating a preferred embodiment according to the method of the present invention for estimating total interest of a group of users directing to a content. For example, the content is a television program.
As shown in Fig. 1, in step SI lO, the user's data are obtained, including every group member's preference directing to each feature and the members' priority in selecting television program. The user's data and other user-related data may be stored in a user profile for future use. Preferably, the user profile is data file in a given format. They may be revised in ways of upgrading by program provider on a regular basis, or revision by the group members according to their needs. The following Tables 1 and 2 respectively show the data structures of the two categories of the user data.
Table 1
Table 2
The user data are obtained in various ways. For example, if a set-top box is used to perform the function of total interest computation, the set-top box may obtain the user data by virtue of downloading it from the remote server of a program provider through, say, the cable television network or LAN, while the user data provided in the remote server may be statistic data acquired from the sample of a large number of families; or family members set the user data personally through the input device and store them in devices, such as the set-top box, for calculation of total interest; and family members' information of television viewing is automatically gathered through the monitor device (e.g. video camera and remote control) installed in the vicinity of a television set, and the information is sent to the set-top box, or transmitted to a remote server through network to form the user data via statistic and analysis.
Table 1 contains the data of every group member's preference directing to each feature. The group of users is made up of the father, mother, child 1 and child 2. The television programs viewed contain A-K, summing to 11 features, in which A shows that the programs are in English, or mainly in English; B, C and D show that the programs are relating to computer, mathematics and chemistry respectively; E represents programs of romantic movies; F and G respectively show that the programs are cartoon or television series; H represents news programs; I commercials; J comedic programs; and K sports programs. The numeric values in the Table show a member's preference directing to a particular feature, that is, in the present embodiment, numeric values are used to describe or indicate every group member's preference directing to the features contained in the content. For example, child 1 likes to watch cartoon film; hence his preference directing to it is 0.8. Since he is not interested in television series and news; hence the corresponding values are 0. Besides, child 1 hates to watch commercial advertisement; hence the value of his preference directing to this feature is at -0.6. The other family members may also be given the corresponding values according to their respective preferences.
Table 2 contains every group member's priority data. As the description of the interest computation mode will show, the priority here may manifest a member's effect on total interest of the group of users. Higher priority means the member's relative salient effect on the total interest, or vice versa.
As Table 2 shows, in the example, the television program viewing time is divided into four time slices, and each member's priority in different time slice does not remain constant. Take the time slice of 18:30-19:30 for example, since it is still early, parents generally allow children to watch television, hence the priority values for child 1 and child 2 are both at 1. When late night approaches, the possibility for the children to sit in front of a TV set diminishes, which renders their priorities gradually decrease. Therefore, to an extent, the priority changing from time to time reflects the changing in the viewers of television programs. For another example, every day mother's priority is lower at
18:30-19:30 than the other members (taking the value of 0.1), while higher than the other members (taking the value of 0.7) at the time slice of 20:30 -21:30. The early low priority is due to their being too busy with the familiar chores to watch TV. The later time slice is mother's spare time. To meet her urgent need for watching TV, the priority may be set at a very high value to ensure that she has more say in television program selection.
In the present embodiment, the priority is at different value in different time slice within a day. This mode is relatively suitable for estimation and recommendation of television program interest. Besides, to reflect the effect of the holiday factor on family members' viewing habit, different values are taken for priority change in the time slice in the working days and holidays. It is worth pointing out, however, that other modes that change of priority along with the time may also be used. For example, different priority values are taken by year, month or day to reflect the holiday and seasonal factors, or priority is indicated in the form of a mathematical function changing with the time.
In conclusion, by virtue of reasonably setting the combination of preference and priority values in different time, time varying factors of the objects of the received content are taken into consideration when determining the total interest, so as to facilitate comprehensive, accurate recommendation result. Referring to Fig. 1 again, enter step S 120 and compute interest of a group of users directing to each feature in each time slice (also known as feature Interest in the present invention) after performing step SI lO of obtaining user data. In the following section, the specific mode of computation of the feature Interest will be discussed in detail.
Then enter step S 130, and create a user profile to store interest of the group of users directing to each feature in each time slice computed in said step S 120. Table 3 shows an exemplary structure form of the profile. In Table 3, features A-K have their respective computation values in the four time slices, representing interest of the group of users directing to one of the features in each time slice. Due to the only purpose of representing the profile structure, some blanks are not indicated specific values. What have been shown are the computation values of A, E, and G in the time slice of 20:30-21 :30.
Table 3 Thus, the group of users' profile has been created. It is worth pointing out that total interest of the group of users directing to a program depends on interest of the group users directing to each feature contained in the program; hence, in the following step of computing the total interest, it is enough to take the corresponding values of the interest directing to the feature contained in the program from the group of users' profile, without the need for repeating said steps Sl 10-S130.
The group of users' profile can remain unchanged for a period of time (for example, one week). During the time, whenever performing program recommendation, the group of users' profile may be directly retrieved and used, without the need for repeating said steps S110-S130. For the convenience of the retrieval and use, the group of users' profile may be written in a data file for later use in the step of computing the total interest. When the total interest is computed in the set-top box, the group of users' profile may be retrieved in such a way as being downloaded from a remote server through the network or acquired from the memory contained in the set-top box. It needs to be pointed out that after the user data are updated regularly or randomly to reflect the change in the user data, the group of users' profile should be re-computed using the steps Sl 10-S130.
See Fig. 1, then enter step S 140 and obtain the data relating to the feature contained in a television program and the broadcast time. The data may be stored, in a given data structure, in the provider's database or contained in the EPG transmitted to the group of users to be used by the means for performing the function of obtaining the total interest.
When the means providing these data is physically independent of the means for obtaining the total interest, it is preferable to input all these data into the means for obtaining the total interest for only one time. Suppose that the television program whose total interest should be computed is an original English television series depicting a sentimental love story, thereby the features it contains are A, E and G. Besides, further suppose that the broadcast time of the program is at 20:40-21 :20, which should be in the time slice of 20:30-21 :30.
Then enter step S 150, and search the preference directing to each feature of said television series from the group of users' profile created in step S 130. Here, since the broadcast time falls exactly in the time slice of 20:30-21:30, the corresponding preference data may be directly found from the Table shown in Fig. 3. Suppose that the interest data for features A, E, and G contained in the television program are respectively 0.37, 0.145, and -0.015. The situation most likely to occur is that the broadcast time of a program spans two time slice, say the broadcast time of 20:25-21:15 respectively falls on the two time slice: 19:30-20:30 and 20:30-21:30 shown in fig.2. The matter can be dealt with in ways as follows.
The first way goes like this. The time slice on which the program falls is determined according to the time the program begins to be broadcast. Under the hypothetical situations above, the program is deemed to be in the 19:30-20:30 time slice. The second way is exactly opposite, in which the time slice is determined according to the time the program ends. Then, the program is deemed to be in the 20:30-21:30 time slice. There is still another way, in which the interest directing to each feature is computed on the base of separate time slice according to the duration of the program in the two different time slice. That is, with the duration of the broadcast as the weighting, compute the weighted average value of interest directing to the feature as the final value of interest of the group of users directing to the feature.
Then, enter step S 160, and combine interest of the group of users directing to each feature at the corresponding broadcast time slice as obtained in step S 150 into the total interest directing to the program. For example, the arithmetic average value or the weighted average value of interest of the group of users directing to each feature contained in the program computed according to the following formula may be used as the total interest P:
P = ∑SJ' xWSJ
J=I (1)
Here S l } is interest of the group of users directing to feature j in time slice i, WSj is the weighting of feature j, m is the number of the features of the television program. As for the example taken in relation to S 140, when the arithmetic average value is taken, the total interest is equal to the arithmetic average value of the three interests directing to features A,
E and G, namely (0.37+0.145-0.015)/3~0.043.
For another example, with quite a few features, say 10, the scoring method often used in all sorts of contests or competitions may also be used, in which the average value is computed without considering the features having the maximum and minimum interest values. Any way, a variety of combinations may be used as long as they can reflect how much each feature contributes to the total interest. Then, enter step S 170, and determine whether there is any other television programs requiring computation of the total interest directing to them. If there is, go back to step S140, otherwise terminate the whole process.
Following is a detailed discussion of the mode of computation of the feature interest in step S 120. Apparently, interest of a group of users directing to a feature depends on value of each member's priority in different time slice and preference of the member directing to the contained features; hence, here are many ways to compute preference, or combine preference with priority to obtain such interest data. Below are two examples along the line.
The first one is based on the concept of weighted average. Specifically, here each member's priority in each time slice is uniformly processed. Thus, priority is converted into a weighted value of the member in each time slice. Table 4 illustrates the uniformly processed weighted value of each group member in different time slice. As Table 4 shows, the group of users is made up of three members: the father, mother and a child. In each time slice, the sum of weighted values of each member is always equal to 1. Table 5 shows the preference of each group member directing to features A, B and C. Each item in the Table means the same as that in Table 2 so it is not elaborated further here.
Table 4
Table 5 Suppose that a television program contains feature A, and its broadcast time is 18:30-19:30, then interest of the group of users directing to the feature is computed according to the formula below:
Here S / is interest of the group of users directing to feature j in time slice i, D'k is user k's preference directing to feature j, W'kis the weighting of user k in time slice i, n is the number of the group members, and the product of D'k and W'k represents user k's interest directing to feature j in the time slice i.
As far as the present example is concerned, through querying the Tables 4 and 5, the weightings of the father, mother and child are respectively 0.2, 0 and 0.8 in the time slice 18:30-19:30, their preference directing to feature A are respectively 0.3, 0.5 and 1; hence, their respective interest directing to feature A are 0.06, 0 and 0.8, and the value of interest of the group of users is 0.86.
Another example in which the function of said step S 120 is performed on the basis of fuzzy logic processing.
Fig. 2 is a flow diagram illustrating an exemplary embodiment of computing interest of a group of users directing to a feature based on fuzzy logic processing mode, the circumstances shown in Tables 1 and 2 are taken again for instance.
To compute interest of a group of users in a time slice (e.g. 18:30-19:30) directing to a feature (e.g. A), as Fig. 3 shows, in step S210, input interest of a member (e.g. the father) directing to feature A and his priority in the corresponding time slice. According to Tables 2 and 3, the values are respectively 0.7 and 0.6. Then, Enter step S220, and project the two crisp values of preference and priority input in step S210 into the membership of the fuzzy values using the selected membership function. The form of the membership function depends on specific circumstance of application. For example, as for the exemplary embodiment here, the membership function shown in Figs 3 a and 3b may be used, in which, Fig. 3 a shows the membership function of preference, X-coordinate Q\ stands for preference, ordinate for membership μ, it is possible to use the membership function to project preference into the membership of the three fuzzy values of "dislike", "neutral" and "like". Fig. 3b shows the priority's membership function, X-coordinate e2 represents priority, ordinate represents membership μ. the membership function may be used to project priority into the membership of the three fuzzy values of "subordinate", "normal" and "important".
Then, enter step S230, and deduct by using the predetermined fuzzy logic rules to obtain the fuzzy output. This is a group of membership of different fuzzy values, reflecting the member's (here father's) interest directing to feature A. Fig. 3c illustrates the membership function of the fuzzy output. The X-coordinate α represents the interest directing to the feature attribute, and ordinate represents membership μ. As the Fig.2 shows, the member's interest directing to the feature attribute A is indicated as the membership of several fuzzy values of "dislike much", "dislike", "neutral", "like" and "like much". The specific rules of deduction mainly depend on the characteristic of application circumstance. For instance, in the present exemplary embodiment, the following rules may be adopted:
I. If ei is "dislike", and e2 is "subordinate", then α is "neutral";
Il.If βiis "dislike", and e2 is "normal", then α is "dislike"; III. If e^s "dislike", and e2 is "important", then α is "dislike much";
IV. If eύs "neutral", and e2 is "subordinate", then α is "neutral";
V.If eύs "neutral", and e2 is "normal", then α is "neutral";
VI. If eύs "neutral", and e2 is "important", then α is "neutral";
VII. If eύs "like", and e2 is "subordinate", then α is "neutral"; VIII. If eύs "like", and e2 is "normal", then α is "like";
IX. If eύs "like", and e2 is "important", then α is "like much";
Then, enter step S240, and convert the member's interest directing to feature attribute A as obtained in step S230 into a crisp value S , , that is, the so-called defuzzification processing. Here i is the number of the member. The commonly used defuzzification processing methods include "center-of-gravity", "center-of-maximum" and "mean-of- maximum". When "center-of-gravity" is applied for the process, the following formula may be used for the computation.
Here μ[l] is the height of the output area satisfying deduction rule No. 1, Yj is horizontal coordinate of the center-of-gravity of the output area satisfying deduction rule No. 1, and m is the number of the output area satisfying the rule.
Then, enter step S250, and determine whether Sj of all the members have been computed. If the conditions are not met, then go back to step S210, otherwise enter step
S260.
In step S260, sum up Sj of all group members (e.g. the father, mother and child as in the present exemplary embodiment) to obtain the group's interest directing to a feature in a given time slice, or utilize the average value of S1 of all group members as interest of the group of users directing to a feature in a given time slice.
The flow diagram of Fig. 4 illustrates another exemplary embodiment according to the method of the present invention for estimating total interest of a group of users directing to a content.
As shown in Fig. 4, in step S410, the user data can also be obtained in various ways. The structure of the data may be as shown in Tables 1 and 2 or Tables 4 and 5, but it may also be in other form.
Then, enter step S420, and obtain the data in relation to the feature and broadcast time of a television program. Suppose that the program information taken from the program source database is that the television program contains features B and C, and it is broadcast at 20:30-21:00.
Then, enter step S430, and determine the corresponding priority according to the obtained broadcast time. For example, the priority variation with the time slice is obtained by checking the table.
Then, enter step S440, and compute interest of the group of users directing to each feature in the broadcast time slice according to its all members' preference obtained in step
S410 and the priority determined in step S430. The way of computation of interest of a group of users directing to a feature has been discussed in great detail above, so it is not elaborated here.
Then, enter step S450, and combine interest of the group of users directing to each feature in the broadcast time slice obtained in step S440 into the total interest directing to the program. For the specific way of computation, see the relevant preceding description. Then, enter step S460, and judge whether there is other program directing to which needs to compute the total interest. If it is, then go back to step S420, otherwise terminate the whole process.
When the broadcast time of a television program spans a plurality of time slice in Table 2 or 4, the way mentioned above may be utilized. For instance, suppose using the above first or second way, then, find out the time slice which the time, when the program begins or ends, in step S430, and, correspondingly, compute, in step S440, interest of the group of users directing to each feature in said time slice. Suppose using the third way, then in step S430, find out all time slice which the program's broadcast duration spans, then with the broadcast time in each time slice determined in step S430 as weighting in step S440, compute the weighted average value of the interest directing to the feature in different time slice and use it as the group of users' value of interest directing to the feature.
It is worth pointing out that as an alternative approach, in said step S440, you may compute each group member's interest directing to a program, not interest of the group of users directing to each feature. Correspondingly, in step S450, combine each group member's interest directing to a program into the total interest directing to the program. As the comparison between the two exemplary embodiments described by virtue of Figs. 1 and 4 show, they differ mainly in that the former first creates a group of users' profile (steps S 120 and S 130), which actually defines the mode in which interest of a group of users directing to each feature varies from one time slice to another, then for each specific program, it is only necessary to use the group of users' profile to determine interest of the group of users directing to the features contained in the program in the broadcast time slice (step S 150), while the latter lacks the step to create the group of users' profile, and, instead, for each specific program, the group members' priority of the program in the broadcast time slice is first determined respectively (step S430), and then interest of the group of users values directing to the features contained in the program in the broadcast time slice are computed (step S450).
Using said method for obtaining the total interest makes it possible to more completely and correctly recommend content to the group of users. In the following with the television program taken as example, in virtue of Fig. 5, a preferred embodiment of the method for recommending content to a group of users is described. As Fig. 5 shows, in step S510, obtain total interest of a group of users directing to one or more television programs by using the method of obtaining the total interest in Figs. 1 and 4.
Then, enter step S520, and determine the television program to be recommended to the group of users according to the total interest obtained in step S530. The determination may be made in the following way: compare the total interest directing to each program with a predetermined threshold value. If it is larger than the threshold value, then fill the program into the recommendation list provided for the group of users, otherwise compare and process the total interest directing to the next program. In addition, the total interests directing to all programs may be ordered firstly, then the first n programs in the ordered sequence are filled into the recommendation list. Here n is a predetermined positive integer.
Finally, in step S530, the obtained recommendation list is proposed to the group of users. When the content recommendation is completed on the side of the program provider, the recommendation list may be provided to the group of users' device such as the set-top box or PC by the program provider through the network together with EPG. When it completes on the group of users' side, it may be obtained with the process of the user device for the program information provided by the program provider (e.g. feature contained in, and the broadcast time of, a program).
The following discussion focuses on the exploitation of the apparatus or device for performing said method for obtaining total interest and the method for content recommendation.
Fig. 6 is a block diagram illustrating an embodiment of the apparatus for performing said method for obtaining the total interest. The apparatus 600 comprises an obtaining unit 610, a receiving unit 620, and a determining unit 630. Obtaining unit 610 is for obtaining user data, including the time varying mode of every group member's priority and his preference directing to each feature, which is shown in Tables 1 and 2 and Tables 4 and 5.
For different ways of obtaining user data obtaining unit 610 has different implements. For example, if the user data is obtained through measurement, then, the obtaining unit or device includes the automatic monitoring unit installed on site; if the user data is created by a user himself, devices, such as, the key board, mouse, remote control, voice-input device, can be used as the user data obtaining unit; if the computation of total interest data is done in the group of users' device (e.g. the television set, set-top box or PC) and the user data is provided by the remote terminal computer via network, the module for performing communication function in the group of users' device may be deemed to be the data obtaining unit.
Receiving unit 620 is used to receive program-related information (including the features contained in, and the broadcast time of, program). Of course, in a specific embodiment, receiving unit 620 may be integrated with obtaining unit 610 in the hardware.
Determining unit 630 is used to obtain total interest of the group of users directing to a content, comprising a group of users' profile generating unit 631, a searching unit 632 and a computing unit 633. Apparatus 600 also comprises first memory unit 640, second memory unit 650, and third memory unit 660, wherein first memory unit 640 is connected to obtaining unit 610 and group of users' profile generating unit 631 for storing user data; second memory unit 650 is connected to group of users' profile generating unit 631 for storing group of users' profile data to be used by searching unit 632. It needs to be pointed out that after the user data is updated regularly or randomly, group of users' profile generating unit 631 should re-compute the group of users' profile and update second memory unit 650 with the new computation result. Third memory unit 660 is connected to computation unit 633 for storing the storage result of computing unit 633.
First, second and third memory units 640, 650 and 660 may be volatile memory or non-volatile memory, and may be realized with one memory, when they are actually different storage areas of the memory.
Generating unit 631 takes the user data from first memory unit 640, to determine, according to each group member's priority in each time slice and his preference directing to each feature, interest of the group of users directing to each feature in each time slice. These interest data are stored in second memory unit 650 as user profile data in given format. The specific mode of computation of relevant interest data refers to the preceding detailed description.
Searching unit 632 is connected to receiving unit 620, and searches for corresponding interest data in second memory unit 650 according to the information of the feature contained in, and the broadcast time of, the program provided by receiving unit 620, to obtain interest of the group of users directing to each feature contained in the program in the time slice corresponding to the broadcast time of the program. As for the circumstance where the broadcast duration of a television program spans a plurality of time slice, searching unit 632 performs the search operation depend on different processing modes.
Computing unit 633 is connected to searching unit 632 for evaluating the arithmetic average value or weighted average value of interest of the group of users directing to the features contained in a television program acquired from searching. Said arithmetic average value or weighted average value is stored in third memory unit 660 as total interest of the group of users directing to the program.
Fig. 7a is a block diagram illustrating an embodiment of the apparatus 700a according to the present invention for recommending content to a group of users. Apparatus 700a comprises apparatus 600 and a recommending unit 710a above mentioned.
Recommending unit 710a includes a threshold value comparing unit 711 and a recommendation list memory unit 712.
Threshold value comparing unit 711 is connected to third memory unit 660 in unit
600, and compares the stored total interests with a predetermined threshold value one by one. If the former is larger than the latter, then the corresponding program's id is outputted to recommendation list memory unit 712. The recommendation list may be further displayed to users.
Fig. 7b is a block diagram illustrating another embodiment of the apparatus 700b according to the present invention for recommending content to a group of users. The apparatus 700b shown in this Fig.7b differs from that in Fig. 7a in that in recommending unit 710b, ordering unit 713 replaces the threshold value comparing unit 711.
Ordering unit 713 is connected to third memory unit 660 in apparatus 600. It arranges the stored total interest in order, and outputs the id of first one or several programs which have biggest total interest to recommendation list memory unit 712. It should be understood that all said unit and some or all parts contained therein may also be realized with software.
The present invention may also be realized by computer with applicable program. The code contained in the program equipped for the computer can be provided to a processor to form a machine so that the code performed in the processor can perform the following functions: obtaining the value of each group member's priority in different time slice and his preference directing to the features; obtaining information relating to the content, including the provision time of, and the feature contained in, the content; and determining, according to each group member's priority in the provision time slice and his preference directing to the feature contained in the content, the total interest. The computer program product may be stored in a memory carrier.
In the present invention, the features refers to those features contained in a content that have an effect on users' interest; hence, there are likely unique feature combination for different group of users and content. Take the television program as an example, they may be expressed with the attributes, such as broadcast channel, title, cast of the program and the program genre (e.g. drama, comic play, love story, action-packed movie or sports events), which have an effect on viewer's willingness to watch. It is worth pointing out that how to customize feature combination for group of users and content is not what the present invention is intended to resolve, and has no causality with the technical effect achieved by the present invention; hence the knowledge about above aspect should not constitute any restraints for the claims of the present invention.
While the preceding description takes the television program as an example, the content referred to in the present invention should be understood in broad sense. In fact, it includes all information perceivable by the human organs, such as visual, audio, touch and taste information. Its physical form includes, but is not limited to, various forms of optical, electrical and acoustic signals. Here is another example of the content: tourist promotion material mailed to a group of users. Such material has, as its content features, such attributes as places for a tour, price, and preferential conditions, which have an effect on the recipients' interest. Besides, the group members' priority is likely to vary from time to time. When the Children's Day on June approaches, for instance, children in a family have more say in deciding on where to go for a trip than on the usual days. Similarly, on the Mother's Day, the mother's priority is high and other members' priority low.
It should be understood that those skilled in the art can also make all sorts of substitutions, modifications and changes on the basis of what has been described above.
When such substitutions, modifications and changes fall within the spirit and the scope of the appended claims, they should be covered in the present invention.

Claims

CLAIMS:
1. A method for estimating total interest of a group of users directing to a content, wherein the group of users includes at least two members, and the method comprising the steps of: a. obtaining information relating to available time slice of the content and features contained in the content; b. obtaining value of each member' s priority in different time slice and preference of the member directing to the contained features; and c. determining the total interest according to the value of each member's priority in the available time slice and preference of the member directing to the contained features.
2. A method according to claim 1, wherein the content is a television program.
3. A method according to claim 1, wherein the value of each member's priority is arranged that value of one of the member's priority is different on at least two time slices, the two time slices is from a plurality of time slices divided by a day.
4. A method according to claim 1, wherein the value of each member's priority is arranged that that value of one of the member' s priority is different on the same time slice of at least two different days.
5. A method according to claim 1, wherein the step c comprising the steps of: cl. determining value of interest of the group of users directing to the features contained in different time slice according to value of each member' s priority in different time slice and the preference directing to the contained features; c2. determining the value of interest of the group according to the available time slice; and c3. combining the value of interest of the group to acquire the total interest.
6. A method according to claim 5, wherein value of interest of the group of users directing to the features contained in one time slice is equal to the sum of the products of preference of each member directing to the contained features and the value of each member's priority in the time slice.
7. A method according to claim 5, wherein the step cl comprising: determining the value of interest of the group of users in a fuzzy logic way according to value of each member' s priority in different time slice and the preference directing to the contained features.
8. A method for recommending content to a group of users, wherein the group of users includes at least two members, and the method comprising the steps of: a. obtaining information relating to available time slice of the content and features contained in the content; b. obtaining value of each member' s priority in different time slice and preference of the member directing to the contained features; c. determining the total interest according to the value of each member's priority in the available time slice and preference of the member directing to the contained features; and d. recommending to the group of users according to the total interest.
9. An apparatus for estimating a group of users' total interest directing to a content, wherein the group of users includes at least two members, comprising: obtaining means for obtaining information relating to the available time slice of the content and features contained in the content; receiving means for receiving value of each member's priority in different time slice and preference of the member directing to the contained features; and determining means for determining the total interest according to the value of each member' s priority in the available time slice and preference of the member directing to the contained features.
10. A apparatus according to claim 9, wherein the value of each member's priority is arranged that value of one of the member's priority is different on at least two time slices, the two time slices is from a plurality of time slices divided by a day.
11. A apparatus according to claim 9, wherein the value of each member's priority is arranged that that value of one of the member' s priority is different on the same time slice of at least two different days.
12. A apparatus according to claim 9, wherein the determining means comprising: generating means for determining value of interest of the group of users directing to the features contained in different time slice according to value of each member' s priority in different time slice and the preference directing to the contained features; searching means for determining the value of interest of the group according to the available time slice; and computer means for combining the value of interest of the group to acquire the total interest.
13. An apparatus for recommending content to a group of users, wherein the group of users includes at least two members, including: obtaining means for obtaining information relating to available time slice of the content and features contained in the content; receiving means for receiving value of each member's priority in different time slice and preference of the member directing to the contained features; determining means for determining the total interest according to the value of each member' s priority in the available time slice and preference of the member directing to the contained features; and recommending means for recommending to the group of users according to the total interest.
14. A computer program product for estimating total interest of a group of users directing to a content, wherein the group of users includes at least two members, comprising: codes for obtaining information relating to available time slice of the content and features contained in the content; codes for obtaining value of each member' s priority in different time slice and preference of the member directing to the contained features; and codes for determining the total interest according to the value of each member's priority in the available time slice and preference of the member directing to the contained features.
15. A memory carrier containing the computer program product according to claim 14.
EP06744970A 2005-05-27 2006-05-18 Method and apparatus for estimating total interest of a group of users directing to a content Withdrawn EP1891588A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN200510073942 2005-05-27
PCT/IB2006/051570 WO2006126147A2 (en) 2005-05-27 2006-05-18 Method and apparatus for estimating total interest of a group of users directing to a content

Publications (1)

Publication Number Publication Date
EP1891588A1 true EP1891588A1 (en) 2008-02-27

Family

ID=36791701

Family Applications (1)

Application Number Title Priority Date Filing Date
EP06744970A Withdrawn EP1891588A1 (en) 2005-05-27 2006-05-18 Method and apparatus for estimating total interest of a group of users directing to a content

Country Status (5)

Country Link
EP (1) EP1891588A1 (en)
JP (1) JP2008542870A (en)
KR (1) KR20080021069A (en)
CN (1) CN101874255A (en)
WO (1) WO2006126147A2 (en)

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4433326B2 (en) * 2007-12-04 2010-03-17 ソニー株式会社 Information processing apparatus and method, and program
KR101404010B1 (en) 2008-03-06 2014-06-12 주성엔지니어링(주) Etcher of substrate edge and method of etching substrate edge
JP5337748B2 (en) * 2010-03-09 2013-11-06 日本電信電話株式会社 Information processing apparatus and information processing program
JP5116811B2 (en) * 2010-07-02 2013-01-09 日本電信電話株式会社 Program recommendation device, method and program
JP5508987B2 (en) * 2010-08-13 2014-06-04 日本電信電話株式会社 Provided information selection apparatus, method and program
JP2012222569A (en) * 2011-04-07 2012-11-12 Nippon Telegr & Teleph Corp <Ntt> Broadcast-program recommending device, method and program
CN102957969A (en) * 2012-05-18 2013-03-06 华东师范大学 Device and method for recommending program to IPTV (Internet protocol television) terminal user
CN104035934B (en) * 2013-03-06 2019-01-15 腾讯科技(深圳)有限公司 A kind of method and device that multimedia messages are recommended
CN103297853B (en) * 2013-06-07 2016-04-06 华东师范大学 A kind of IPTV program commending method based on multi-user's context identification
TWI615787B (en) * 2013-11-07 2018-02-21 財團法人資訊工業策進會 Merchandise recommendation system, method and non-transitory computer readable storage medium of the same for multiple users
TWI489725B (en) * 2013-11-07 2015-06-21 Inst Information Industry Apparatus and method for creating a power consumption model and computer program product thereof
KR102232798B1 (en) * 2014-03-18 2021-03-26 에스케이플래닛 주식회사 Service apparatus, user apparatus estimating regions of interest and method therefor, computer readable medium having computer program recorded therefor
CN105163139B (en) * 2014-05-28 2018-06-01 青岛海尔电子有限公司 Information-pushing method, Information Push Server and smart television
CN104735535A (en) * 2015-03-24 2015-06-24 天脉聚源(北京)传媒科技有限公司 Program scoring method and device
US10699181B2 (en) * 2016-12-30 2020-06-30 Google Llc Virtual assistant generation of group recommendations

Also Published As

Publication number Publication date
JP2008542870A (en) 2008-11-27
CN101874255A (en) 2010-10-27
KR20080021069A (en) 2008-03-06
WO2006126147A2 (en) 2006-11-30

Similar Documents

Publication Publication Date Title
WO2006126147A2 (en) Method and apparatus for estimating total interest of a group of users directing to a content
JP6235556B2 (en) Content presentation method, content presentation apparatus, and program
KR101061234B1 (en) Information processing apparatus and method, and recording medium
CN1659882B (en) Method and system for implementing content augmentation of personal profiles
JP4505418B2 (en) Program recommendation device, program recommendation method of program recommendation device, and program
CN102265276B (en) Commending system based on context
JP2005057713A (en) Information processor and information processing method, program, and recording medium
WO2001015449A1 (en) Method and apparatus for creating recommendations from users profile built interactively
US20030131355A1 (en) Program guide system
US6757691B1 (en) Predicting content choices by searching a profile database
JP4358219B2 (en) Program information providing apparatus, program information providing method, and program thereof
CN1720740A (en) Recommendation of video content based on the user profile of users with similar viewing habits
US20020083451A1 (en) User-friendly electronic program guide based on subscriber characterizations
JP2007272451A (en) Recommended program information providing device, recommended program information providing method, and program
JP2005165454A (en) Contents recommending device, contents recommending method and contents recommending program
KR20050106108A (en) Generation of television recommendations via non-categorical information
CN1352776A (en) Smart agent based on habit, statistical inference and psycho-demographic profiling
CN1344465A (en) System and method for tailoring television and/or electronic program guide features, such as advertising
CN1586077A (en) Method and apparatus for recommending items of interest based on preferences of a selected third party
US20070022440A1 (en) Program recommendation via dynamic category creation
KR101054088B1 (en) How to automatically recommend customized IP programs using collaborative filtering
CN106131703A (en) A kind of method of video recommendations and terminal
US20080196064A1 (en) Apparatus and method of providing a recommended broadcast program
JP2006094018A (en) Program recommending apparatus, program recommending method, program, and recoding medium with the program recorded
EP2596464A1 (en) A method and apparatus for creating recommendations for a user

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20071227

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LI LT LU LV MC NL PL PT RO SE SI SK TR

DAX Request for extension of the european patent (deleted)
17Q First examination report despatched

Effective date: 20080908

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: PACE MICROTECHNOLOGY PLC

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: PACE PLC

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION HAS BEEN WITHDRAWN

18W Application withdrawn

Effective date: 20090312