CN105095442A - Multimedia data recommendation method and device - Google Patents

Multimedia data recommendation method and device Download PDF

Info

Publication number
CN105095442A
CN105095442A CN201510438746.5A CN201510438746A CN105095442A CN 105095442 A CN105095442 A CN 105095442A CN 201510438746 A CN201510438746 A CN 201510438746A CN 105095442 A CN105095442 A CN 105095442A
Authority
CN
China
Prior art keywords
user
medium data
matrix
similarity
mark
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.)
Granted
Application number
CN201510438746.5A
Other languages
Chinese (zh)
Other versions
CN105095442B (en
Inventor
李文强
万艾学
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.)
Hisense Group Co Ltd
Original Assignee
Hisense Group Co Ltd
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
Application filed by Hisense Group Co Ltd filed Critical Hisense Group Co Ltd
Priority to CN201810034582.3A priority Critical patent/CN108108491B/en
Priority to CN201510438746.5A priority patent/CN105095442B/en
Publication of CN105095442A publication Critical patent/CN105095442A/en
Application granted granted Critical
Publication of CN105095442B publication Critical patent/CN105095442B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
    • G06F16/43—Querying
    • G06F16/435—Filtering based on additional data, e.g. user or group profiles

Abstract

The embodiment of the invention provides a multimedia data recommendation method and device, relates to the field of computer technology, and overcomes a problem that a terminal in the prior art can not accurately recommend a video to a user. The method includes: according to attribute information of multimedia data, generating a matrix R and a matrix S corresponding to each user, wherein an element R i j of the matrix R shows that whether a user i watches multimedia data j or not, an element S v u of the matrix S shows that whether multimedia data v watched by the user belongs to a multimedia data type u or not; according to the matrix R, the matrix S corresponding to the first user, and a matrix S corresponding to the other users, calculating a similarity between the first user and the other users; and according to the similarity between the first user and each user of the other user, the number of predetermined similar users, the matrix S, and the number of multimedia data, determining the multimedia data recommended to the first user. The multimedia data recommendation method and device can be applied to the recommendation of the multimedia data.

Description

A kind of recommend method of multi-medium data and device
Technical field
The present invention relates to field of computer technology, particularly relate to a kind of recommend method and device of multi-medium data.
Background technology
Now, in the epoch of this internet high speed development, people are more and more higher for the demand of audiovisual aspect, and video relevant recommendation business can recommend video for user, effectively help user to find demand, advance user for the program request of audio-visual service.In the prior art, collaborative filtering recommending (CollaborativeFiltering, be called for short CF) algorithm is usually used to recommend video for user.
In the prior art; terminal (for intelligent television) according to traditional CF algorithm for user recommend video time; usually clustering algorithm can be utilized according to video generic; respectively the television video stored in the database of terminal background server and the user that watches these videos are carried out cluster; then video classification and video generic belonging to the user after cluster, for user recommends video.But, due in the prior art, terminal when calculating the similarity between user only nationwide examination for graduation qualification consider the viewing behavior of user, and do not consider other aspects, such as, when calculating the similarity between user, according to the similarity degree between the viewed respectively television video source of two users, the similarity between two users can be determined.But; due to a large amount of television video sources usually can be stored in the database of terminal background server; and the number of videos that each user watches is little; make the video similarity of watching between two two users very low; thus cause the discrimination of the user's similarity calculated not high, and then cannot be correct recommend the interested video of user for user.
Summary of the invention
Embodiments of the invention provide a kind of recommend method and device of multi-medium data, and solving terminal of the prior art cannot carry out accurate problem of recommending to user and video.
For achieving the above object, embodiments of the invention adopt following technical scheme:
First aspect, provides a kind of recommend method of multi-medium data, comprising:
Obtain the attribute information of multi-medium data, described attribute information comprises the mark of the multi-medium data of multimedia data type belonging to the mark of user, described multi-medium data and user's viewing;
According to the attribute information of described multi-medium data, generator matrix R and matrix S corresponding to each user, the row and column of described matrix R represents the mark of user and the mark of multi-medium data respectively, the element R of described matrix R ijrepresent whether user i watches multi-medium data j, the row and column of described matrix S is respectively the mark of multimedia data type and the viewed multi-medium data of described user, the element S of described matrix S vurepresent whether the multi-medium data v that described user watches belongs to multimedia data type u;
The first similarity between the described first user calculated according to the multi-medium data set of the first user obtained from described matrix R and the multi-medium data set of the second user and described second user, and the second similarity between the described first user to calculate according to matrix S 1 corresponding to described first user and matrix S 2 corresponding to described second user and described second user, obtain the similarity between described first user and described second user, described second user is other the arbitrary users except described first user, the multi-medium data set of user comprises the viewed all multi-medium datas of described user,
Similarity between each user in described first user and other users is sorted, and determines the similar users of described first user according to the number of default similar users;
According to the multi-medium data number that the similarity between the similar users of the mark of the similar users of described first user, described matrix S, described first user and described first user and needing is recommended for described first user, determine the multi-medium data that described first user is recommended;
Wherein, described i ∈ 1,2 ..., n; J, v ∈ 1,2 ..., m; U ∈ 1,2 ..., k; Described n is user's number, and described m is multi-medium data number, and described k is multimedia data type number.
Second aspect, provides a kind of recommendation apparatus of multi-medium data, comprising:
Acquisition module, for obtaining the attribute information of multi-medium data, described attribute information comprises the mark of the multi-medium data of multimedia data type belonging to the mark of user, described multi-medium data and user's viewing;
Generation module, for the attribute information of described multi-medium data obtained according to described acquisition module, generator matrix R and matrix S corresponding to each user, the row and column of described matrix R represents the mark of user and the mark of multi-medium data respectively, the element R of described matrix R ijrepresent whether user i watches multi-medium data j, the row and column of described matrix S is respectively the mark of multimedia data type and the viewed multi-medium data of described user, the element S of described matrix S vurepresent whether the multi-medium data v that described user watches belongs to multimedia data type u;
Computing module, the first similarity between the described first user that the multi-medium data set of first user obtained from the described matrix R that described generation module generates for basis and the multi-medium data set of the second user calculate and described second user, and the second similarity between the described first user that calculates of matrix S 1 corresponding to the described first user to generate according to described generation module and matrix S 2 corresponding to described second user and described second user, obtain the similarity between described first user and described second user, described second user is other the arbitrary users except described first user, the multi-medium data set of user comprises the viewed all multi-medium datas of described user,
First determination module, sorts for the similarity between each user in the described first user that calculated by described computing module and other users, and determines the similar users of described first user according to the number of default similar users;
Second determination module, the multi-medium data number that similarity between the similar users of the described first user that the described matrix S generated for the mark of the similar users of described first user determined according to described first determination module, described generation module and described first user and described computing module calculate and needing is recommended for described first user, determines the multi-medium data recommended described first user;
Wherein, described i ∈ 1,2 ..., n; J, v ∈ 1,2 ..., m; U ∈ 1,2 ..., k; Described n is user's number, and described m is multi-medium data number, and described k is multimedia data type number.
The recommend method of the multi-medium data that embodiments of the invention provide and device, according to the attribute information of multi-medium data, generator matrix R and each user generate corresponding matrix S, and the row and column of this matrix R represents the mark of user and the mark of multi-medium data respectively, the element R of this matrix R ijrepresent whether user i watches multi-medium data j, the row and column of this matrix S is respectively the mark of multimedia data type and the viewed multi-medium data of described user, the element S of this matrix S vurepresent whether the multi-medium data v that described user watches belongs to multimedia data type u, the matrix S 1 corresponding according to this matrix R, first user and matrix S 2 corresponding to other users, calculate the similarity between first user and other users, then according to the similarity between this first user and other users, preset similar users number and need the multi-medium data number of recommending for first user, determine the multi-medium data to first user recommendation.The proportion degree of the multimedia data type between the multi-medium data watched by each user embodied in the user embodied in the matrix R matrix S 1 corresponding with the relevance between multi-medium data and first user and matrix S 2 corresponding to other users like this, to the differentiation that the preference of multimedia data type of the multi-medium data of user's viewing more becomes more meticulous, thus improve the accuracy that terminal recommends user and multi-medium data.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, be briefly described to the accompanying drawing used required in embodiment or description of the prior art below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.
The schematic flow sheet of the recommend method of a kind of multi-medium data that Fig. 1 provides for embodiments of the invention;
The structural representation of the recommendation apparatus of a kind of multi-medium data that Fig. 2 provides for embodiments of the invention;
The structural representation of the recommendation apparatus of the another kind of multi-medium data that Fig. 3 provides for embodiments of the invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, be clearly and completely described the technical scheme in the embodiment of the present invention, obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art, not making the every other embodiment obtained under creative work prerequisite, belong to the scope of protection of the invention.
Embodiments of the invention provide a kind of recommend method of multi-medium data, and as shown in Figure 1, the method specifically comprises the steps:
101, the recommendation apparatus of multi-medium data obtains the attribute information of multi-medium data.
Exemplary, the multi-medium data in the present invention is the multimedia file data such as video, music, text document.The attribute information of above-mentioned multi-medium data comprise user mark and, multimedia data type belonging to the mark of the multi-medium data of user's viewing and this multi-medium data.Such as, if when this multi-medium data is film, this multimedia data type comprises science fiction, animation, the story of a play or opera, war, ancient costume, comedy etc.The kind of the multimedia data type in the present embodiment can be set in advance by technician, and determine the multimedia data type belonging to each multi-medium data, it should be noted that, each multi-medium data can belong to a multimedia data type also can belong to multiple multimedia data type simultaneously, such as, not only a certain film belongs to ancient costume type but also belong to comedy type.Wherein, the parameter information that this attribute information also comprises multi-medium data comprises file attribute information, and such as, when this multi-medium data is video, the parameter information of this multi-medium data comprises: video performer title, director title, video type etc.
Wherein, the mark of above-mentioned user can for the login account of this user or other uniquely can represent and the mark of this user in the present embodiment, adopt U1, U2, U3 ... Un form represents the mark of different user; The mark of multi-medium data can for the title of this multi-medium data or other uniquely can represent the mark of this multi-medium data, adopt in the present embodiment B1, B2, B3 ..., Bm represents the mark of different multimedia data; The record that user watches multi-medium data represents the relation between the multi-medium data of user and viewing.
Preferably, in a step 101, a update cycle can be set, the length of update cycle can set according to multimedia data storehouse update status, such as, can one month be set to, one week or one day, the present invention does not limit this, obtains the attribute information of the multi-medium data in described each update cycle and upgrade within each update cycle.The following each step of the present embodiment is all described for current period.
102, the recommendation apparatus of multi-medium data is according to the attribute information of multi-medium data, and generator matrix R and each user generate corresponding matrix S.
Wherein, the row and column of above-mentioned matrix R represents the mark of user and the mark of multi-medium data respectively, the element R of this matrix R ijrepresent whether user i watches multi-medium data j; The row and column of above-mentioned matrix S is respectively the mark of multimedia data type and the viewed multi-medium data of respective user, the element S of this matrix S vurepresent whether the multi-medium data v that described user watches belongs to multimedia data type u.
Above-mentioned i ∈ 1,2 ..., n; J, v ∈ 1,2 ..., m; U ∈ 1,2 ..., k; Above-mentioned n is user's number, the unduplicated multi-medium data summation that above-mentioned m watches for the user of n in record, and k is multimedia data type number, and namely in the present embodiment, multimedia data type comprises K kind.
Exemplary, if being video with multi-medium data is example, suppose that the set of user and video is respectively U={U 1, U 2..., U nand B={B 1, B 2..., B m, if using user ID as row matrix, using multi-medium data mark as rectangular array, then the relational matrix defined between user and video is matrix r ijrepresent user u iwhether viewed video B jinformation.If R on intelligent television ijrepresent user u iviewed video B j, then R ij=1, otherwise be R ij=0.Here with table 1, the implication to matrix R is described, it should be noted that, real matrix R has the dimension of 1,000,000 grades, and following table 1 is only be described the implication of matrix R, is only a kind of example.
Table 1
Exemplary, if when predetermined multimedia data type K is 4, suppose user U 1viewed video set is combined into B={B 1, B 2, B 4, B 6, user U 2viewed video set is combined into B={B 1, B 2, B 6, B 7, the implication of user's matrix S can be described with table 2 table 3 here, and concrete, table 2 is for representing user U 1matrix S 1, table 2 is for representing user U 2matrix S 2, be noted that in practical application, user can watch tens or video up to a hundred, is only described for 4 videos here, is only a kind of example.
Table 2
Table 3
103, the first similarity between the first user that calculates according to the multi-medium data set of the first user obtained from matrix R and the multi-medium data set of the second user of the recommendation apparatus of multi-medium data and the second user, and the second similarity between the first user to calculate according to matrix S 1 corresponding to first user and matrix S 2 corresponding to the second user and the second user, obtain the similarity between first user and the second user.
Wherein, the second above-mentioned user is other the arbitrary users except first user, and the multi-medium data set of above-mentioned user comprises the viewed all multi-medium datas of this user.Concrete, this multi-medium data recommendation apparatus can calculate the similarity between each user in this first user and other users.Preferably, the recommendation apparatus of multi-medium data is when calculating the similarity between first user and the second user, the first similarity between the first user calculated with the second user and the second similarity between first user with the second user can be multiplied or be added, thus obtain the similarity between first user and the second user.It should be noted that, above-mentioned to be multiplied or addition form is only a kind of preferred form, similarity can be calculated by other form of calculation in practical application, not limit here.
Concrete, the computation process of the first similarity between the first user that in step 103, the multi-medium data set of the first user that the recommendation apparatus of multi-medium data obtains from matrix R and the multi-medium data set of the second user calculate and the second user is as shown in following step:
The recommendation apparatus of a1, multi-medium data obtains multi-medium data set corresponding to each user from matrix R.
The recommendation apparatus of a2, multi-medium data is according to the multi-medium data set I of the first calculating formula of similarity, first user 1with the multi-medium data set I of the second user 2, calculate the first similarity between first user and the second user.
Wherein, the second above-mentioned user is the arbitrary user in other users except first user, and the first above-mentioned calculating formula of similarity is:
Concrete, the computation process of the second similarity between the first user that in step 103, the recommendation apparatus of multi-medium data calculates according to matrix S 1 corresponding to first user and matrix S 2 corresponding to the second user and the second user is as shown in following step:
The recommendation apparatus of b1, multi-medium data, according to matrix S 1 corresponding to the second calculating formula of similarity, first user and matrix S 2 corresponding to the second user, calculates the second similarity between first user and the second user.
Wherein, the second above-mentioned user is the arbitrary user in other users except first user; The second above-mentioned calculating formula of similarity this x, y ∈ 1,2 ..., k, p arefer to all elements in each multimedia data type column in described matrix S 1 add up after the vector of numerical value composition, p bto refer in matrix S 2 all elements in each multimedia data type column add up after the vector of numerical value composition.
Concrete, known based on the content described in step 103, the calculating formula of similarity between first user and the second user is: R = R 1 × R 2 = | I 1 ∩ I 2 | | I 1 ∪ I 2 | × Σ p x ∈ P a , p y ∈ P b k p x p y | p a | | p b | .
Exemplary, the recommendation apparatus of multi-medium data is when calculating the similarity between first user and the second user, the matrix S 1 corresponding according to first user is needed to obtain the multi-medium data preference vector of matrix S 1, the matrix S 2 corresponding according to the second user obtains the multi-medium data preference vector of matrix S 2, and the multi-medium data preference vector form of this matrix S 1 is p a(P 1, P 1..., P k) in like manner, the multi-medium data preference vector form of this matrix S 2 is p b(P 1, P 1..., P k).
Exemplary, if the matrix S of user U1 is as shown in table 2, the matrix S 2 of user U2 is as shown in table 3, the multi-medium data set I of this user U1 1(B 1, B 2, B 4, B 6), the multi-medium data set I of this user U2 2(B 1, B 2, B 6, B 7), the multi-medium data preference vector form of the matrix S 1 of this user U1 is p (4,3,1,2), and the multi-medium data preference vector form of the matrix S 2 of this user U2 is p (3,4,2,1).
Based on the first calculating formula of similarity, can show that the first similarity R1 between user U1 and user U2 is:
R 1 = | I 1 ∩ I 2 | | I 1 ∪ I 2 | = B 1 , B 2 , B 6 B 1 , B 2 , B 4 , B 6 , B 7 = 3 5 = 60 % ;
Further, based on the second calculating formula of similarity, can show that the second similarity R2 between user U1 and user U2 is:
R 2 = Σ p x ∈ P a , p y ∈ P b k p x p y | p a | | p b | = 4 × 3 + 3 × 4 + 1 × 2 + 2 × 1 4 2 + 3 2 + 1 2 + 2 2 × 3 2 + 4 2 + 2 2 + 1 2 = 93.3 % ;
Finally, according to formula R=R1 × R2, the similarity R between user U1 and user U2 is: R=60%*93.3%=56%.
104, the similarity between each user in first user and other users sorts by the recommendation apparatus of multi-medium data, and determines the similar users of first user according to default similar users number.
Exemplary, the similarity between each user in first user and other users sorts by the recommendation apparatus of multi-medium data, and descending is inserted in default chained list, and the greatest member number that this default chained list comprises is identical with default similar users number.
105, the recommendation apparatus of the multi-medium data multi-medium data number of recommending for first user according to the similarity between the similar users of the mark of the similar users of=first user, matrix S, first user and this first user and needing, determines the multi-medium data recommended first user.
Optionally, step 104 specifically comprises the steps:
The mark of the multi-medium data that the recommendation apparatus of 105a, multi-medium data is not watched according to the mark of matrix R, similar users and first user, generator matrix Y.
Wherein, the row and column of this matrix Y represents the mark of the multi-medium data that the mark of similar users and the viewed and first user of similar users are not watched respectively.
The recommendation apparatus of 105b, multi-medium data, according to the similarity between first user and similar users and matrix Y, calculates the relating value of first user to each multi-medium data in matrix Y.
Exemplary, the recommendation apparatus of multi-medium data can pass through following process implementation when calculating the relating value of first user to the arbitrary multi-medium data in the multi-medium data in matrix Y: the recommendation apparatus of multi-medium data selects arbitrary multi-medium data from the multi-medium data matrix Y, then the similarity between the similar users of first user and each viewed arbitrary multi-medium data is added up, obtain the relating value between first user and arbitrary multi-medium data.
Relating value between each multi-medium data in first user and described matrix Y sorts by the recommendation apparatus of 105c, multi-medium data, and the multi-medium data number of recommending for first user is as required determined as the multi-medium data that described first user is recommended.
Exemplary, relating value between each multi-medium data in first user and described matrix Y sorts by the recommendation apparatus of multi-medium data, descending is inserted in default chained list, and the greatest member number that this default chained list comprises is identical with the multi-medium data number needing to recommend for first user.
The recommend method of the multi-medium data that embodiments of the invention provide, according to the attribute information of multi-medium data, generator matrix R and each user generate corresponding matrix S, and the row and column of this matrix R represents the mark of user and the mark of multi-medium data respectively, the element R of this matrix R ijrepresent whether user i watches multi-medium data j, the row and column of this matrix S is respectively the mark of multimedia data type and the viewed multi-medium data of described user, the element S of matrix S vurepresent whether the multi-medium data v of user's viewing belongs to multimedia data type u, the matrix S 1 corresponding according to this matrix R, first user and matrix S 2 corresponding to other users, calculate the similarity between first user and other users, then according to the similarity between this first user and other users, preset similar users number and need the multi-medium data number of recommending for first user, determine the multi-medium data to first user recommendation.The proportion degree of the multimedia data type between the multi-medium data watched by each user embodied in the user embodied in the matrix R matrix S 1 corresponding with the relevance between multi-medium data and first user and matrix S 2 corresponding to other users like this, to the differentiation that the preference of multimedia data type of the multi-medium data of user's viewing more becomes more meticulous, thus improve the accuracy that terminal recommends user and multi-medium data.
Embodiments of the invention provide a kind of recommendation apparatus of multi-medium data, this device is for realizing above-mentioned recommend method, as shown in Figure 2, this device comprises acquisition module 21, generation module 22, computing module 23, first determination module 24 and the second determination module 25, wherein:
Acquisition module 21, for obtaining the attribute information of multi-medium data, this attribute information comprises the mark of the multi-medium data of multimedia data type belonging to the mark of user, multi-medium data and user's viewing.
Generation module 22, for the attribute information of multi-medium data obtained according to acquisition module 21, generator matrix R and matrix S corresponding to each user, the row and column of matrix R represents the mark of user and the mark of multi-medium data respectively, the element R of matrix R ijrepresent whether user i watches multi-medium data j, the row and column of matrix S is respectively the mark of the multimedia data type multi-medium data viewed with this user, the element S of matrix S vurepresent whether the multi-medium data v of user's viewing belongs to multimedia data type u.
Computing module 23, the first similarity between the first user that the multi-medium data set of first user obtained from the matrix R that generation module 22 generates for basis and the multi-medium data set of the second user calculate and the second user, and the second similarity between the first user that calculates of matrix S 1 corresponding to first user to generate according to generation module 22 and matrix S 2 corresponding to the second user and the second user, obtain the similarity between first user and the second user, second user is other the arbitrary users except first user, the multi-medium data set of user comprises the viewed all multi-medium datas of this user.
First determination module 24, sorts for the similarity between each user in the first user that calculated by computing module 23 and other users, and determines the similar users of first user according to the number of default similar users.
Second determination module 25, the multi-medium data number that similarity between the similar users of the first user that the matrix S generated for the mark of the similar users of first user determined according to the first determination module 24, generation module 22 and first user and computing module 23 calculate and needing is recommended for first user, determines the multi-medium data to first user recommendation.
Wherein, above-mentioned i ∈ 1,2 ..., n; J, v ∈ 1,2 ..., m; U ∈ 1,2 ..., k; N is user's number, and m is multi-medium data number, and k is multimedia data type number.
Optionally, computing module 23 is specifically for the first similarity between the first user calculate the multi-medium data set of the first user obtained in the matrix R generated from generation module 22 and the multi-medium data set of the second user and the second user, and the first user that the matrix S 2 corresponding with the second user according to the matrix S 1 that the first user of generation module 22 generation is corresponding calculates is multiplied with the second similarity between the second user, obtains the similarity between first user and the second user.
Optionally, specifically comprise during first similarity of computing module 23 between the first user calculated according to the multi-medium data set of first user that obtains and the multi-medium data set of the second user the matrix R generated from generation module 22 and the second user:
Multi-medium data set corresponding to each user is obtained from the matrix R that generation module 22 generates;
According to the multi-medium data set I of the first calculating formula of similarity, first user 1with the multi-medium data set I of the second user 2, calculate the first similarity between first user and the second user, the second user is the arbitrary user in other users except first user;
Wherein, above-mentioned first calculating formula of similarity is:
Optionally, specifically comprise during second similarity of computing module 23 between the first user that matrix S 1 corresponding to the first user generated according to generation module 22 and matrix S 2 corresponding to the second user calculate and the second user:
Matrix S 1 corresponding to first user generated according to the second calculating formula of similarity, generation module 22 and matrix S 2 corresponding to the second user, calculate the second similarity between first user and the second user, the second user is the arbitrary user in other users except first user;
Wherein, the second calculating formula of similarity x, y ∈ 1,2 ..., k, p arefer to all elements in each multimedia data type column in described matrix S 1 add up after the vector of numerical value composition, p bto refer in matrix S 2 all elements in each multimedia data type column add up after the vector of numerical value composition.
Optionally, the second determination module 25 specifically for:
The mark of the multi-medium data that the mark of the similar users that the matrix R, the first determination module 24 that generate according to generation module 22 are determined and first user are not watched, the row and column of generator matrix Y, matrix Y represents the mark of the multi-medium data that the mark of similar users and the viewed and first user of similar users are not watched respectively;
Similarity between the first user calculated according to computing module 23 and similar users and matrix Y, calculate the relating value of first user to each multi-medium data in matrix Y;
Relating value between each multi-medium data in first user and matrix Y is sorted, and the multi-medium data number of recommending for first user is as required determined as the multi-medium data that first user is recommended.
Further alternative, the similarity of the second determination module 25 between the first user calculated according to computing module 23 and similar users and matrix Y, specifically comprise when calculating the relating value of first user to each multi-medium data in matrix Y:
Arbitrary multi-medium data is selected from the multi-medium data matrix Y;
Similarity between the similar users of first user and each viewed arbitrary multi-medium data is added up, obtains the relating value between first user and arbitrary multi-medium data.
Optionally, described first determination module 24 specifically for:
Similarity between each user in the first user calculate computing module 23 and other users sorts, and descending is inserted in default chained list, and wherein, the greatest member number that default chained list comprises is identical with default similar users number.
Optionally, as shown in Figure 3, this device also comprises, and arranges module 26, for arranging the update cycle;
Acquisition module 21 specifically for: according to arrange update cycle, within each update cycle, obtain the attribute information of the multi-medium data in each update cycle.
The recommendation apparatus of the multi-medium data that embodiments of the invention provide, according to the attribute information of multi-medium data, generator matrix R and each user generate corresponding matrix S, and the row and column of this matrix R represents the mark of user and the mark of multi-medium data respectively, the element R of this matrix R ijrepresent whether user i watches multi-medium data j, the row and column of this matrix S is respectively the mark of multimedia data type and the viewed multi-medium data of described user, the element S of matrix S vurepresent whether the multi-medium data v of user's viewing belongs to multimedia data type u, the matrix S 1 corresponding according to this matrix R, first user and matrix S 2 corresponding to other users, calculate the similarity between first user and other users, then according to the similarity between this first user and other users, preset similar users number and need the multi-medium data number of recommending for first user, determine the multi-medium data to first user recommendation.The proportion degree of the multimedia data type between the multi-medium data watched by each user embodied in the user embodied in the matrix R matrix S 1 corresponding with the relevance between multi-medium data and first user and matrix S 2 corresponding to other users like this, to the differentiation that the preference of multimedia data type of the multi-medium data of user's viewing more becomes more meticulous, thus improve the accuracy that terminal recommends user and multi-medium data.
Those skilled in the art can be well understood to, for convenience and simplicity of description, only be illustrated with the division of above-mentioned each functional module, in practical application, can distribute as required and by above-mentioned functions and be completed by different functional modules, inner structure by device is divided into different functional modules, to complete all or part of function described above.The system of foregoing description, the specific works process of device and unit, with reference to the corresponding process in preceding method embodiment, can not repeat them here.
In several embodiments that the application provides, should be understood that disclosed apparatus and method can realize by another way.Such as, device embodiment described above is only schematic, such as, the division of described module or unit, be only a kind of logic function to divide, actual can have other dividing mode when realizing, such as multiple unit or assembly can in conjunction with or another system can be integrated into, or some features can be ignored, or do not perform.Another point.In addition, each functional unit in each embodiment of the application can be integrated in a processing unit, also can be that the independent physics of unit exists, also can two or more unit in a unit integrated.Above-mentioned integrated unit both can adopt the form of hardware to realize, and the form of SFU software functional unit also can be adopted to realize.
The above, above embodiment only in order to the technical scheme of the application to be described, is not intended to limit; Although with reference to previous embodiment to present application has been detailed description, those of ordinary skill in the art is to be understood that: it still can be modified to the technical scheme described in foregoing embodiments, or carries out equivalent replacement to wherein portion of techniques feature; And these amendments or replacement, do not make the essence of appropriate technical solution depart from the spirit and scope of each embodiment technical scheme of the application.

Claims (11)

1. a recommend method for multi-medium data, is characterized in that, comprising:
Obtain the attribute information of multi-medium data, described attribute information comprises the mark of the multi-medium data of multimedia data type belonging to the mark of user, described multi-medium data and user's viewing;
According to the attribute information of described multi-medium data, generator matrix R and matrix S corresponding to each user, the row and column of described matrix R represents the mark of user and the mark of multi-medium data respectively, the element R of described matrix R ijrepresent whether user i watches multi-medium data j, the row and column of described matrix S is respectively the mark of multimedia data type and the viewed multi-medium data of described user, the element S of described matrix S vurepresent whether the multi-medium data v that described user watches belongs to multimedia data type u;
The first similarity between the described first user calculated according to the multi-medium data set of the first user obtained from described matrix R and the multi-medium data set of the second user and described second user, and the second similarity between the described first user to calculate according to matrix S 1 corresponding to described first user and matrix S 2 corresponding to described second user and described second user, obtain the similarity between described first user and described second user, described second user is other the arbitrary users except described first user, the multi-medium data set of user comprises the viewed all multi-medium datas of described user,
Similarity between each user in described first user and other users is sorted, and determines the similar users of described first user according to the number of default similar users;
According to the multi-medium data number that the similarity between the similar users of the mark of the similar users of described first user, described matrix S, described first user and described first user and needing is recommended for described first user, determine the multi-medium data that described first user is recommended;
Wherein, described i ∈ 1,2 ..., n; J, v ∈ 1,2 ..., m; U ∈ 1,2 ..., k; Described n is user's number, and described m is multi-medium data number, and described k is multimedia data type number.
2. method according to claim 1, it is characterized in that, the first similarity between the described first user that the multi-medium data set of the first user that described basis obtains from described matrix R and the multi-medium data set of the second user calculate and described second user, and the second similarity between the described first user to calculate according to matrix S 1 corresponding to described first user and matrix S 2 corresponding to described second user and described second user, the similarity obtained between described first user and described second user specifically comprises:
The first similarity between the described first user that the multi-medium data set of the first user obtained from described matrix R and the multi-medium data set of the second user are calculated and described second user, and the described first user that the matrix S 2 corresponding with described second user according to the matrix S 1 that described first user is corresponding calculates is multiplied with the second similarity between described second user, obtains the similarity between described first user and described second user.
3. method according to claim 1 and 2, it is characterized in that, the first similarity between the described first user that the multi-medium data set of the first user that described basis obtains from described matrix R and the multi-medium data set of the second user calculate and described second user specifically comprises:
The multi-medium data set that each user is corresponding is obtained from described matrix R;
According to the multi-medium data set I of the first calculating formula of similarity, described first user 1with the multi-medium data set I of the second user 2, calculate the first similarity between described first user and described second user, described second user is the arbitrary user in other users except described first user;
Wherein, described first calculating formula of similarity is:
4. method according to claim 1 and 2, is characterized in that, the second similarity between the described first user that the described matrix S 1 corresponding according to described first user and matrix S 2 corresponding to described second user calculate and described second user specifically comprises:
The matrix S 1 corresponding according to the second calculating formula of similarity, described first user and matrix S 2 corresponding to described second user, calculate the second similarity between described first user and described second user, described second user is the arbitrary user in other users except described first user;
Wherein, described second calculating formula of similarity described x, y ∈ 1,2 ..., k, described p arefer to all elements in each multimedia data type column in described matrix S 1 add up after the vector of numerical value composition, described p bto refer in matrix S 2 all elements in each multimedia data type column add up after the vector of numerical value composition.
5. method according to claim 1, it is characterized in that, the multi-medium data number that similarity between the similar users of the mark of the described similar users according to described first user, described matrix S, described first user and described first user and needing is recommended for described first user, determine that the multi-medium data to described first user is recommended specifically comprises:
According to the mark of the multi-medium data that mark and the described first user of matrix R, described similar users are not watched, the row and column of generator matrix Y, described matrix Y represents the mark of the multi-medium data that the mark of similar users and the viewed and described first user of described similar users are not watched respectively;
According to the similarity between described first user and described similar users and matrix Y, calculate the relating value of described first user to each multi-medium data in described matrix Y;
Relating value between each multi-medium data in described first user and described matrix Y is sorted, and the multi-medium data number of recommending for described first user is as required determined as the multi-medium data that described first user is recommended.
6. method according to claim 5, is characterized in that, described according to the similarity between described first user and described similar users and matrix Y, calculates the relating value of described first user to each multi-medium data in described matrix Y and specifically comprises:
Arbitrary multi-medium data is selected from the multi-medium data described matrix Y;
Similarity between the similar users of described first user and each viewed described arbitrary multi-medium data is added up, obtains the relating value between described first user and described arbitrary multi-medium data.
7. a recommendation apparatus for multi-medium data, is characterized in that, comprising:
Acquisition module, for obtaining the attribute information of multi-medium data, described attribute information comprises the mark of the multi-medium data of multimedia data type belonging to the mark of user, described multi-medium data and user's viewing;
Generation module, for the attribute information of described multi-medium data obtained according to described acquisition module, generator matrix R and matrix S corresponding to each user, the row and column of described matrix R represents the mark of user and the mark of multi-medium data respectively, the element R of described matrix R ijrepresent whether user i watches multi-medium data j, the row and column of described matrix S is respectively the mark of multimedia data type and the viewed multi-medium data of described user, the element S of described matrix S vurepresent whether the multi-medium data v that described user watches belongs to multimedia data type u;
Computing module, the first similarity between the described first user that the multi-medium data set of first user obtained from the described matrix R that described generation module generates for basis and the multi-medium data set of the second user calculate and described second user, and the second similarity between the described first user that calculates of matrix S 1 corresponding to the described first user to generate according to described generation module and matrix S 2 corresponding to described second user and described second user, obtain the similarity between described first user and described second user, described second user is other the arbitrary users except described first user, the multi-medium data set of user comprises the viewed all multi-medium datas of described user,
First determination module, sorts for the similarity between each user in the described first user that calculated by described computing module and other users, and determines the similar users of described first user according to the number of default similar users;
Second determination module, the multi-medium data number that similarity between the similar users of the described first user that the described matrix S generated for the mark of the similar users of described first user determined according to described first determination module, described generation module and described first user and described computing module calculate and needing is recommended for described first user, determines the multi-medium data recommended described first user;
Wherein, described i ∈ 1,2 ..., n; J, v ∈ 1,2 ..., m; U ∈ 1,2 ..., k; Described n is user's number, and described m is multi-medium data number, and described k is multimedia data type number.
8. device according to claim 7, it is characterized in that, specifically comprise when the multi-medium data set of first user that described computing module obtains among the described matrix R that basis generates from described generation module and the first similarity between the described first user that the multi-medium data set of the second user calculates and described second user:
Multi-medium data set corresponding to each user is obtained from the described matrix R that described generation module generates;
According to the multi-medium data set I of the first calculating formula of similarity, described first user 1with the multi-medium data set I of the second user 2, calculate the first similarity between described first user and described second user, described second user is the arbitrary user in other users except described first user;
Wherein, described first calculating formula of similarity is:
9. device according to claim 7, it is characterized in that, specifically comprise during second similarity of described computing module between the described first user that matrix S 1 corresponding to the described first user generated according to described generation module and matrix S 2 corresponding to described second user calculate and described second user:
Matrix S 1 corresponding to described first user generated according to the second calculating formula of similarity, described generation module and matrix S 2 corresponding to described second user, calculate the second similarity between described first user and described second user, described second user is the arbitrary user in other users except described first user;
Wherein, described second calculating formula of similarity described x, y ∈ 1,2 ..., k, described p arefer to all elements in each multimedia data type column in described matrix S 1 add up after the vector of numerical value composition, described p bto refer in matrix S 2 all elements in each multimedia data type column add up after the vector of numerical value composition.
10. device according to claim 7, is characterized in that, described second determination module specifically for:
The mark of the multi-medium data that the matrix R generated according to described generation module, the mark of described similar users and described first user are not watched, the row and column of generator matrix Y, described matrix Y represents the mark of the multi-medium data that the mark of similar users and the viewed and described first user of described similar users are not watched respectively;
Similarity between the described first user calculated according to described computing module and described similar users and matrix Y, calculate the relating value of described first user to each multi-medium data in described matrix Y;
Relating value between each multi-medium data in described first user and described matrix Y is sorted, and the multi-medium data number of recommending for described first user is as required determined as the multi-medium data that described first user is recommended.
11. devices according to claim 10, it is characterized in that, the similarity of described second determination module between the described first user calculated according to described computing module and described similar users and matrix Y, specifically comprise when calculating the relating value of described first user to each multi-medium data in described matrix Y:
Arbitrary multi-medium data is selected from the multi-medium data described matrix Y;
Similarity between the similar users of described first user and each viewed described arbitrary multi-medium data is added up, obtains the relating value between described first user and described arbitrary multi-medium data.
CN201510438746.5A 2015-07-23 2015-07-23 A kind of recommendation method and device of multi-medium data Active CN105095442B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201810034582.3A CN108108491B (en) 2015-07-23 2015-07-23 Multimedia data recommendation method and device
CN201510438746.5A CN105095442B (en) 2015-07-23 2015-07-23 A kind of recommendation method and device of multi-medium data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510438746.5A CN105095442B (en) 2015-07-23 2015-07-23 A kind of recommendation method and device of multi-medium data

Related Child Applications (1)

Application Number Title Priority Date Filing Date
CN201810034582.3A Division CN108108491B (en) 2015-07-23 2015-07-23 Multimedia data recommendation method and device

Publications (2)

Publication Number Publication Date
CN105095442A true CN105095442A (en) 2015-11-25
CN105095442B CN105095442B (en) 2018-04-24

Family

ID=54575879

Family Applications (2)

Application Number Title Priority Date Filing Date
CN201510438746.5A Active CN105095442B (en) 2015-07-23 2015-07-23 A kind of recommendation method and device of multi-medium data
CN201810034582.3A Active CN108108491B (en) 2015-07-23 2015-07-23 Multimedia data recommendation method and device

Family Applications After (1)

Application Number Title Priority Date Filing Date
CN201810034582.3A Active CN108108491B (en) 2015-07-23 2015-07-23 Multimedia data recommendation method and device

Country Status (1)

Country Link
CN (2) CN105095442B (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105404700A (en) * 2015-12-30 2016-03-16 山东大学 Collaborative filtering-based video program recommendation system and recommendation method
CN105512252A (en) * 2015-12-01 2016-04-20 海信集团有限公司 Method and device obtaining multimedia data correlation
CN105574198A (en) * 2015-12-28 2016-05-11 海信集团有限公司 Column recommendation method and device
CN105589916A (en) * 2016-01-11 2016-05-18 西华大学 Extraction method for explicit and implicit interest knowledge
CN105868317A (en) * 2016-03-25 2016-08-17 华中师范大学 Digital education resource recommendation method and system
CN105956061A (en) * 2016-04-26 2016-09-21 海信集团有限公司 Method and device for determining similarity between users
CN106294800A (en) * 2016-08-16 2017-01-04 武汉斗鱼网络科技有限公司 Method and system recommended by direct broadcasting room based on weighting k neighbour scoring
CN106534984A (en) * 2016-12-23 2017-03-22 深圳Tcl数字技术有限公司 TV program pushing method and device
CN106604068A (en) * 2016-12-30 2017-04-26 中广热点云科技有限公司 Method and system for updating media program
CN109063080A (en) * 2018-07-25 2018-12-21 北京小度互娱科技有限公司 A kind of video recommendation method and device
CN111711838A (en) * 2020-06-23 2020-09-25 广州酷狗计算机科技有限公司 Video switching method, device, terminal, server and storage medium

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112559856B (en) * 2020-12-04 2023-06-02 中国联合网络通信集团有限公司 Application ordering method and device

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104035934A (en) * 2013-03-06 2014-09-10 腾讯科技（深圳）有限公司 Multimedia information recommending method and device
CN104123315A (en) * 2013-04-28 2014-10-29 百度在线网络技术（北京）有限公司 Multi-media file recommendation method and recommendation server

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103327045B (en) * 2012-03-21 2017-03-22 腾讯科技（深圳）有限公司 User recommendation method and system in social network
CN103327111A (en) * 2013-07-01 2013-09-25 百度在线网络技术（北京）有限公司 multimedia file recommendation method, system thereof and server
CN103942712A (en) * 2014-05-09 2014-07-23 北京联时空网络通信设备有限公司 Product similarity based e-commerce recommendation system and method thereof

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104035934A (en) * 2013-03-06 2014-09-10 腾讯科技（深圳）有限公司 Multimedia information recommending method and device
US20140289334A1 (en) * 2013-03-06 2014-09-25 Tencent Technology (Shenzhen) Company Limited System and method for recommending multimedia information
CN104123315A (en) * 2013-04-28 2014-10-29 百度在线网络技术（北京）有限公司 Multi-media file recommendation method and recommendation server

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
夏培勇: "个性化推荐技术中的协同过滤算法研究", 《中国优秀博士论文全文数据库》 *
郭昆等: "基于局部近邻传播及用户特征的社区识别算法", 《通信学报》 *

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105512252B (en) * 2015-12-01 2019-03-05 海信集团有限公司 The method and device of correlation between a kind of acquisition multi-medium data
CN105512252A (en) * 2015-12-01 2016-04-20 海信集团有限公司 Method and device obtaining multimedia data correlation
CN105574198A (en) * 2015-12-28 2016-05-11 海信集团有限公司 Column recommendation method and device
CN105574198B (en) * 2015-12-28 2019-12-06 海信集团有限公司 column recommendation method and device
CN105404700A (en) * 2015-12-30 2016-03-16 山东大学 Collaborative filtering-based video program recommendation system and recommendation method
CN105404700B (en) * 2015-12-30 2019-04-16 山东大学 A kind of video column recommendation system and recommended method based on collaborative filtering
CN105589916A (en) * 2016-01-11 2016-05-18 西华大学 Extraction method for explicit and implicit interest knowledge
CN105589916B (en) * 2016-01-11 2020-05-08 西华大学 Method for extracting explicit and implicit interest knowledge
CN105868317A (en) * 2016-03-25 2016-08-17 华中师范大学 Digital education resource recommendation method and system
CN105868317B (en) * 2016-03-25 2017-04-12 华中师范大学 Digital education resource recommendation method and system
CN105956061A (en) * 2016-04-26 2016-09-21 海信集团有限公司 Method and device for determining similarity between users
CN105956061B (en) * 2016-04-26 2020-01-03 海信集团有限公司 Method and device for determining similarity between users
WO2018032790A1 (en) * 2016-08-16 2018-02-22 武汉斗鱼网络科技有限公司 Weighted k-nearest-neighbor scoring-based live broadcast room recommendation method and system
CN106294800A (en) * 2016-08-16 2017-01-04 武汉斗鱼网络科技有限公司 Method and system recommended by direct broadcasting room based on weighting k neighbour scoring
CN106534984A (en) * 2016-12-23 2017-03-22 深圳Tcl数字技术有限公司 TV program pushing method and device
CN106604068A (en) * 2016-12-30 2017-04-26 中广热点云科技有限公司 Method and system for updating media program
CN106604068B (en) * 2016-12-30 2019-11-05 中广热点云科技有限公司 A kind of method and its system of more new media program
CN109063080A (en) * 2018-07-25 2018-12-21 北京小度互娱科技有限公司 A kind of video recommendation method and device
CN109063080B (en) * 2018-07-25 2022-01-21 北京小度互娱科技有限公司 Video recommendation method and device
CN111711838A (en) * 2020-06-23 2020-09-25 广州酷狗计算机科技有限公司 Video switching method, device, terminal, server and storage medium

Also Published As

Publication number Publication date
CN108108491B (en) 2021-07-30
CN105095442B (en) 2018-04-24
CN108108491A (en) 2018-06-01

Similar Documents

Publication Publication Date Title
CN105095442A (en) Multimedia data recommendation method and device
CN106547767B (en) Method and device for determining video cover picture
US11601703B2 (en) Video recommendation based on video co-occurrence statistics
CN105404700B (en) A kind of video column recommendation system and recommended method based on collaborative filtering
CN105677715A (en) Multiuser-based video recommendation method and apparatus
CN105426548A (en) Video recommendation method and device based on multiple users
CN104933135A (en) Method and device for clustering multimedia data
US20150205580A1 (en) Method and System for Sorting Online Videos of a Search
CN103440335A (en) Video recommendation method and device
CN104182449A (en) System and method for personalized video recommendation based on user interests modeling
CN105404680A (en) Searching recommendation method and apparatus
CN105593841A (en) Automatically determining a size for a content item for a web page
CN106131703A (en) A kind of method of video recommendations and terminal
CN105139021B (en) A kind of method and system for realizing TV user Fast Classification based on rough set theory
CN106407446B (en) Network questionnaire construction method and device
CN105744370A (en) Radio and television system based on group viewing behaviors and personalized program recommendation method thereof
CN109840732B (en) Checking system for integrity of whole vehicle configurable bill of materials under free selection and matching condition
CN105681910A (en) Video recommending method and device based on multiple users
CN106599291B (en) Data grouping method and device
US9251113B1 (en) System for enabling participants to discuss, debate, connect and compare media and information
CN109063080B (en) Video recommendation method and device
CN115048571A (en) Online education recommendation management system based on cloud platform
CN114461871A (en) Recommendation model training method, object recommendation device and storage medium
CN109829110A (en) A kind of personalized recommendation method of learning materials
CN106777276B (en) Method and device for index storage of user information and native information display

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant