CN107766379A - A kind of recommendation method and apparatus of Web content - Google Patents
A kind of recommendation method and apparatus of Web content Download PDFInfo
- Publication number
- CN107766379A CN107766379A CN201610701119.0A CN201610701119A CN107766379A CN 107766379 A CN107766379 A CN 107766379A CN 201610701119 A CN201610701119 A CN 201610701119A CN 107766379 A CN107766379 A CN 107766379A
- Authority
- CN
- China
- Prior art keywords
- web content
- user
- mrow
- content
- similarity
- 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.)
- Pending
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
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The invention discloses a kind of recommendation method and apparatus of Web content, are related to Internet technical field, and method therein includes:The historical information that user downloads or accesses Web content is obtained, establishes user and the corresponding relation for the Web content downloaded or accessed;The corresponding relation of Web content based on user and download or access determines the similarity between Web content;Interest value of the user to Web content is determined according to the similarity between Web content and historical information;The Web content recommended based on the true directional user of interest value.Methods and apparatus of the present invention, the accuracy rate and recall rate of content recommendation can be lifted, and by reducing the weight of any active ues, lift the coverage and diversity of Web content, enhance user's stickiness by reducing the weight of hot content.
Description
Technical field
The present invention relates to the recommendation method and apparatus of Internet technical field, more particularly to a kind of Web content.
Background technology
Personalized recommendation is to recommend its content interested to user according to the Characteristic of Interest and behavior of user, is mainly solved
How user's information interested is found in magnanimity information.Conventional recommendation method includes information filtering and collaborative filtering, association
It is divided into the UserCF algorithms based on user and the ItemCF algorithms based on article with filtering.At present, each mobile interchange of China Telecom
Net content platform is such as liked to play, and typically more focuses on consumer consumption behavior rather than Social behaviors, more real using ItemCF algorithms
With.Similarity measurement is the core content of collaborative filtering, for example, cosine similarity refers to two vector angles in vector space
Cosine value as weigh two interindividual variations size, closer to 1 show it is more similar.In existing personalized recommendation method
In, two be present when moving the personalized recommendation of internet platform by similarity:1. hot content (hot topic trip
Play, books) can be similar with a lot of other contents;2. the behavior of Showed Very Brisk user (being partly malicious user), can make similarity
Matrix is excessively dense, influences the accuracy of personalized recommendation.
The content of the invention
In view of this, the invention solves a technical problem be to provide the recommendation method and dress of a kind of Web content
Put.
According to an aspect of the present invention, there is provided a kind of recommendation method of Web content, including:User is obtained to download or visit
The historical information of Web content is asked, establishes user and the corresponding relation for the Web content downloaded or accessed;Based on the user with
The corresponding relation for the Web content downloaded or accessed determines the similarity between Web content;According between the Web content
Similarity and the historical information determine interest value of the user to Web content;Based on the interest value, true directional user recommends
Web content.
Alternatively, it is described to obtain user's download or access the historical information of Web content, establish user with downloading or accessing
The corresponding relation of Web content include:The historical record downloaded or accessed by accessing user obtains user and downloads or access net
The set of network content;Collection based on the download or access Web content is combined into each user and establishes user and content relation square
Battle array;Wherein, the quantity for the Web content species that whole users download or accessed is n, and the user and content relation matrix are n rows
The matrix of × n row, the user and the nonzero element U in content relation matrixijTo download or accessing Web content i and net simultaneously
Network content j user.
Alternatively, it is described based on the user with download or access Web content corresponding relation determine Web content it
Between similarity include:It will be added with the user corresponding to each user with content relation matrix, obtain Web content similarity
Matrix;Wherein, the Web content similarity matrix is the matrix of n rows × n row, non-in the Web content similarity matrix
Neutral element AijTo download or accessing Web content i and Web content j user's set simultaneously;According to the Web content similarity
Similarity between matrix computations Web content.
Alternatively, the corresponding relation of the Web content based on the user and download or access is determined between Web content
Similarity is:
Wherein, WijFor the similarity between Web content i and Web content j, N (i) is download or accesses Web content i
User gathers, and N (j) is download or accesses Web content j user's set, and u is the user during N (i) and N (j) occurs simultaneously, | N (u) |
It is the quantity that user u downloaded or accessed Web content, | N (i) | it is download or access Web content i number of users, | N (j) |
It is download or access Web content j number of users.
Alternatively, the similarity according between the Web content and the historical information determine user to network
The interest value of content is:
Wherein, PujInterest value for user u to Web content j, D (u) are the collection for the Web content that user u is downloaded or accessed
Close, the set for the k game that S (i, k) is default and Web content i similarities are high, WjiIt is Web content j and Web content i
Similarity, ruiIt is interest values of the default user u to Web content i.
Alternatively, the Web content recommended based on the true directional user of the interest value is included:Height based on the interest value
It is low that Web content is ranked up;Web content is extracted according to default recommended amount and recommended to user;Wherein, Web content
Including:Game, books, film.
According to another aspect of the present invention, there is provided a kind of recommendation apparatus of Web content, including:Network-content acquisition mould
Block, the historical information of Web content is downloaded or accessed for obtaining user, establish user and the Web content of download or access
Corresponding relation;Content similarities determining module, for based on the user and the corresponding relation for the Web content downloaded or accessed
Determine the similarity between Web content;Content interest value determining module, for according to the similarity between the Web content
And the historical information determines interest value of the user to Web content;Content recommendation determining module, for based on the interest
It is worth the Web content that true directional user recommends.
Alternatively, the network-content acquisition module, the historical record for downloading or accessing by accessing user obtain
User downloads or accessed the set of Web content;The content similarities determining module, for based on the download or access net
The collection of network content is combined into each user and establishes user and content relation matrix;Wherein, the network that whole users download or accessed
The quantity of content type is n, the user and the matrix that content relation matrix is n rows × n row, the user and content relation square
Nonzero element U in battle arrayijTo download or accessing Web content i and Web content j user simultaneously.
Alternatively, the content similarities determining module, it is additionally operable to close with the user corresponding to each user and content
It is that matrix is added, Web content similarity matrix is obtained, according between the Web content similarity matrix calculating network content
Similarity;Wherein, the Web content similarity matrix is the matrix of n rows × n row, in the Web content similarity matrix
Nonzero element AijTo download or accessing Web content i and Web content j user's set simultaneously.
Alternatively, the content similarities determining module, it is additionally operable to based in the user and the network downloaded or accessed
The corresponding relation of appearance determines that the similarity between Web content is:
Wherein, WijFor the similarity between Web content i and Web content j, N (i) is download or accesses Web content i
User gathers, and N (j) is download or accesses Web content j user's set, and u is the user during N (i) and N (j) occurs simultaneously, | N (u) |
It is the quantity that user u downloaded or accessed Web content, | N (i) | it is download or access Web content i number of users, | N (j) |
It is download or access Web content j number of users.
Alternatively, the content interest value determining module, for according to the similarity between the Web content and institute
State historical information and determine that user is to the interest value of Web content:
Wherein, PujInterest value for user u to Web content j, D (u) are the collection for the Web content that user u is downloaded or accessed
Close, the set for the k game that S (i, k) is default and Web content i similarities are high, WjiIt is Web content j and Web content i
Similarity, ruiIt is interest values of the default user u to Web content i.
Alternatively, the Web content recommended based on the true directional user of the interest value is included:The content recommendation determines mould
Block, Web content is ranked up for the height based on the interest value, Web content is extracted according to default recommended amount
And recommend to user;Wherein, Web content includes:Game, books, film.
The recommendation method and apparatus of the Web content of the present invention, it is corresponding with the Web content of download or access based on user
Similarity between relation calculating network content, and user is calculated to the interested of Web content according to similarity and historical information
Degree, recommended based on interest level, by reducing the weight of hot content, accuracy rate and recall rate can be lifted, and
And by reducing the weight of any active ues, lift the coverage and diversity of Web content.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are only
Some embodiments of the present invention, for those of ordinary skill in the art, without having to pay creative labor, also
Other accompanying drawings can be obtained according to these accompanying drawings.
Fig. 1 is the schematic flow sheet according to one embodiment of the recommendation method of the Web content of the present invention;
Fig. 2 is the schematic flow sheet according to another embodiment of the recommendation method of the Web content of the present invention;
Fig. 3, which is that user is corresponding with content in another embodiment according to the recommendation method of the Web content of the present invention, to close
System establishes schematic diagram;
Fig. 4 is the module diagram according to one embodiment of the recommendation apparatus of the Web content of the present invention.
Embodiment
The present invention is described more fully with reference to the accompanying drawings, wherein illustrating the exemplary embodiment of the present invention.Under
The accompanying drawing that face will be combined in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, and shows
So, described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.Based on the reality in the present invention
Example is applied, the every other embodiment that those of ordinary skill in the art are obtained under the premise of creative work is not made, is all belonged to
In the scope of protection of the invention.Many descriptions are carried out to technical scheme with reference to each figure and embodiment.
Fig. 1 is according to the schematic flow sheet of one embodiment of the recommendation method of the Web content of the present invention, such as Fig. 1 institutes
Show:
Step 101, the historical information that user downloads or accesses Web content is obtained, establishes user and the net downloaded or accessed
The corresponding relation of network content, Web content include user's download or the game accessed, song, video etc..
Step 102, the corresponding relation based on user and download or the Web content of access determines the phase between Web content
Like degree, the similarity between many algorithms calculating network content can be used, for example with cosine similarity algorithm etc..
Step 103, the historical information of Web content is downloaded or accessed according to the similarity between Web content and user,
Determine interest value of the user to Web content.
Step 104, the Web content recommended based on the true directional user of interest value.
The recommendation method of Web content in above-described embodiment, the corresponding relation calculating network based on user and Web content
Similarity between content, and interest level of the user to Web content is calculated according to similarity and historical information, based on sense
Level of interest carries out commending contents, can lift accuracy, covering and the diversity of recommendation.
Fig. 2 is according to the schematic flow sheet of another embodiment of the recommendation method of the Web content of the present invention, such as Fig. 2 institutes
Show:
Step 201, user accesses the platform for providing Web content, such as gaming platform by client, clicks to enter and appoints
Meaning game details page, client send the access request of user to background program.
Step 201, background program obtains the game ID of user's current accessed, and with obtaining the history of user according to ID
Record.
Step 203, the similarity matrix of the game of user's download is established.Personalized recommendation emphasis to user is similarity
The calculating of matrix and personalized recommendation list is generated according to similarity matrix and the historical behavior of user.
As shown in figure 3, Web content is game, Far Left is the set of user's download games, and user downloaded the trip of certain money
Play, then it is assumed that the user is interested in the game.User and content relation matrix are established, it is interested to represent a user per a line
Game set, for each article set, the article of the inside is combined two-by-two, obtains a new matrix, these squares
Battle array, which is added, obtains the matrix on the right, the user list of element representation while download games i and game j in matrix.
The historical record downloaded or accessed by accessing user obtains the set that user downloads or accesses Web content, is based on
Download or the collection of access Web content is combined into each user and establishes user and content relation matrix.As shown in figure 3, all 5
The quantity for the Web content species that user downloads or accessed is that 5,5 users and content relation matrix are all the square that 5 rows × 5 arrange
Battle array, user and the nonzero element U in content relation matrixijTo download or accessing Web content i and Web content j use simultaneously
Family.
It will be added with the user corresponding to 5 users with content relation matrix, obtain Web content similarity matrix, network
Content similarity matrix is the matrix of the row of 5 rows × 5, and the nonzero element in Web content similarity matrix is downloads or accessed simultaneously
Web content i and Web content j user gather, according to similar between Web content similarity matrix calculating network content
Degree.
Step 204,205, the similarity between calculating network content.
The corresponding relation of Web content based on user and download or access determines the meter of the similarity between Web content
Calculating formula is:
Wherein, WijFor the similarity between Web content i and Web content j, N (i) is download or accesses Web content i
User gathers, and N (j) is download or accesses Web content j user's set, and u is the user during N (i) and N (j) occurs simultaneously, | N (u) |
It is the quantity that user u downloaded or accessed Web content, | N (i) | it is download or access Web content i number of users, | N (j) |
It is download or access Web content j number of users.
Denominator in above formula reduces game (Web content) j weight, can mitigate hot game and much play similar
Possibility, so as to lift the quality of recommendation.
For gaming platform, part malicious downloading user be present, in order to ensure the reliability of similarity between playing,
Need to correct contribution of the Showed Very Brisk user to similarity of playing, i.e., 100 sections of game have been downloaded for a game
User contribution degree be less than only downloaded 10 sections game users.
Determine that user is to the interest value of Web content according to the similarity between Web content:
Wherein, PujInterest value for user u to Web content j, D (u) are the collection for the Web content that user u is downloaded or accessed
Close, the set for the k game that S (i, k) is default and Web content i similarities are high, WjiIt is Web content j and Web content i
Similarity, ruiIt is interest values of the default user u to Web content i.
Pass through the calculating of above-mentioned interest value so that the game more similar to game interested in user's history, then more have
Higher ranking may be obtained in the recommendation list of user.Can is recommended user after the completion of calculating.
Step 206, the list of games that user may like is ranked according to similarity, TopN game before taking, and returned
As a result client is given.
The recommendation method of Web content in above-described embodiment, it is corresponding with the Web content of download or access based on user
Similarity between relation calculating network content, and user is calculated to the interested of Web content according to similarity and historical information
Degree, recommended based on interest level, by reducing the weight of hot content, accuracy rate and recall rate can be lifted, and
And by reducing the weight of any active ues, lift the coverage and diversity of Web content.
In one embodiment, as shown in figure 4, the present invention provides a kind of recommendation apparatus 40 of Web content, including:Network
Content obtaining module 41, content similarities determining module 42, content interest value determining module 43 and content recommendation determining module 44.
Network-content acquisition module 41 obtains the historical information that user downloads or accesses Web content, establishes user with downloading
Or the corresponding relation of the Web content accessed.Content similarities determining module 42 is based on user and download or the Web content accessed
Corresponding relation determine similarity between Web content.Content interest value determining module 43 is according to similar between Web content
Degree and historical information determine interest value of the user to Web content.Content recommendation determining module 44 based on interest value determine to
The Web content that family is recommended.
The historical record that network-content acquisition module 41 is downloaded or accessed by accessing user obtains user and downloads or access
The set of Web content.Content similarities determining module 42 is combined into each user based on the collection for downloading or accessing Web content and built
Vertical user and content relation matrix;Wherein, the quantity for the Web content species that whole users download or accessed is n, user with it is interior
Appearance relational matrix is the matrix of n rows × n row, user and the nonzero element U in content relation matrixijTo download or accessing net simultaneously
Network content i and Web content j user.
Content similarities determining module 42 will be added with the user corresponding to each user with content relation matrix, obtain net
Network content similarity matrix, according to the similarity between Web content similarity matrix calculating network content.Web content is similar
Degree matrix is the matrix of n rows × n row, the nonzero element A in Web content similarity matrixijTo download or accessing in network simultaneously
Hold i and Web content j user's set.
Corresponding relation of the content similarities determining module 43 based on Web content of the user with downloading or accessing determines network
Similarity between content is:
Wherein, WijFor the similarity between Web content i and Web content j, N (i) is download or accesses Web content i
User gathers, and N (j) is download or accesses Web content j user's set, and u is the user during N (i) and N (j) occurs simultaneously, | N (u) |
It is the quantity that user u downloaded or accessed Web content, | N (i) | it is download or access Web content i number of users, | N (j) |
It is download or access Web content j number of users.
Content interest value determining module 43 determines interest of the user to Web content according to the similarity between Web content
It is worth and is:
Wherein, PujInterest value for user u to Web content j, D (u) are the collection for the Web content that user u is downloaded or accessed
Close, the set for the k game that S (i, k) is default and Web content i similarities are high, WjiIt is Web content j and Web content i
Similarity, ruiIt is interest values of the default user u to Web content i.
Height of the content recommendation determining module 44 based on interest value is ranked up to Web content, according to default recommendation number
Amount extraction Web content is simultaneously recommended to user;Wherein, Web content includes:Game, books, film etc..
The recommendation method and apparatus of Web content in above-described embodiment, based on user and download or the Web content accessed
Corresponding relation calculating network content between similarity, and user is calculated to Web content according to similarity and historical information
Interest level, recommended based on interest level, by reducing the weight of hot content, accuracy rate can be lifted and recalled
Rate, and by reducing the weight of any active ues, the coverage and diversity of Web content are lifted, so as to strengthen commending system hair
Dig long-tail ability so that be available for calculate content it is more, then corresponding precision and coverage rate are higher.
The method and system of the present invention may be achieved in many ways.For example, can by software, hardware, firmware or
Software, hardware, firmware any combinations come realize the present invention method and system.The said sequence of the step of for method is only
Order described in detail above is not limited in order to illustrate, the step of method of the invention, is especially said unless otherwise
It is bright.In addition, in certain embodiments, the present invention can be also embodied as recording program in the recording medium, these programs include
For realizing the machine readable instructions of the method according to the invention.Thus, the present invention also covering storage is used to perform according to this hair
The recording medium of the program of bright method.
Description of the invention provides for the sake of example and description, and is not exhaustively or by the present invention
It is limited to disclosed form.Many modifications and variations are obvious for the ordinary skill in the art.Select and retouch
State embodiment and be to more preferably illustrate the principle and practical application of the present invention, and one of ordinary skill in the art is managed
The present invention is solved so as to design the various embodiments with various modifications suitable for special-purpose.
Claims (12)
1. a kind of recommendation method of Web content, it is characterised in that including:
The historical information that user downloads or accesses Web content is obtained, it is corresponding with the Web content of download or access to establish user
Relation;
The corresponding relation of Web content based on the user and download or access determines the similarity between Web content;
Interest value of the user to Web content is determined according to the similarity between the Web content and the historical information;
The Web content recommended based on the true directional user of the interest value.
2. the method as described in claim 1, it is characterised in that the history letter for obtaining user and downloading or access Web content
Ceasing, establish the corresponding relation of Web content of the user with downloading or accessing includes:
The historical record downloaded or accessed by accessing user obtains the set that user downloads or accesses Web content;
Collection based on the download or access Web content is combined into each user and establishes user and content relation matrix;
Wherein, the quantity for the Web content species that whole users download or accessed is n, and the user and content relation matrix are n
The matrix of row × n row, the user and the nonzero element U in content relation matrixijFor simultaneously download or access Web content i and
Web content j user, i, j are the natural number less than n.
3. method as claimed in claim 2, it is characterised in that described based on the user and download or the Web content accessed
Corresponding relation determine that the similarity between Web content includes:
It will be added with the user corresponding to each user with content relation matrix, obtain Web content similarity matrix;Wherein, institute
State matrix of the Web content similarity matrix for n rows × n row, the nonzero element A in the Web content similarity matrixijTo be same
When download or access Web content i and Web content j user set;
According to the similarity between the Web content similarity matrix calculating network content.
4. the method as described in any one of claims 1 to 3, it is characterised in that based on the user and download or the net accessed
The corresponding relation of network content determines that the similarity between Web content is:
<mrow>
<msub>
<mi>w</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>&Sigma;</mi>
<mrow>
<mi>u</mi>
<mo>&Element;</mo>
<mi>N</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>&cap;</mo>
<mi>N</mi>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</mrow>
</msub>
<mfrac>
<mn>1</mn>
<mrow>
<mi>log</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>+</mo>
<mo>|</mo>
<mi>N</mi>
<mo>(</mo>
<mi>u</mi>
<mo>)</mo>
<mo>|</mo>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
<msqrt>
<mrow>
<mo>|</mo>
<mi>N</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
<mo>|</mo>
<mi>N</mi>
<mo>(</mo>
<mi>j</mi>
<mo>)</mo>
<mo>|</mo>
</mrow>
</msqrt>
</mfrac>
<mo>;</mo>
</mrow>
Wherein, WijFor the similarity between Web content i and Web content j, N (i) is download or the user for accessing Web content i
Set, N (j) is download or accesses Web content j user's set, and u is the user during N (i) and N (j) occurs simultaneously, | N (u) | it is to use
Family u downloads or accessed the quantity of Web content, | N (i) | it is download or access Web content i number of users, | N (j) | under being
Carry or access Web content j number of users.
5. method as claimed in claim 4, it is characterised in that the similarity and institute according between the Web content
State historical information and determine that user is to the interest value of Web content:
<mrow>
<msub>
<mi>p</mi>
<mrow>
<mi>u</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>D</mi>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>)</mo>
</mrow>
<mo>&cap;</mo>
<mi>S</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
<mo>)</mo>
</mrow>
</mrow>
</munder>
<msub>
<mi>w</mi>
<mrow>
<mi>j</mi>
<mi>i</mi>
</mrow>
</msub>
<msub>
<mi>r</mi>
<mrow>
<mi>u</mi>
<mi>i</mi>
</mrow>
</msub>
<mo>;</mo>
</mrow>
Wherein, PujInterest value for user u to Web content j, D (u) be user u download or access Web content set, S
The set for the k game that (i, k) is default and Web content i similarities are high, WjiIt is Web content j and Web content i phase
Like degree, ruiIt is interest values of the default user u to Web content i.
6. method as claimed in claim 5, it is characterised in that the Web content recommended based on the true directional user of the interest value
Including:
Height based on the interest value is ranked up to the Web content;
The Web content is extracted according to default recommended amount and recommended to user;
Wherein, the Web content includes:Game, books, film.
A kind of 7. recommendation apparatus of Web content, it is characterised in that including:
Network-content acquisition module, the historical information of Web content is downloaded or accessed for obtaining user, establish user with downloading
Or the corresponding relation of the Web content accessed;
Content similarities determining module, net is determined for the corresponding relation based on Web content of the user with downloading or accessing
Similarity between network content;
Content interest value determining module, for determining to use according to the similarity between the Web content and the historical information
Interest value of the family to Web content;
Content recommendation determining module, for the Web content recommended based on the true directional user of the interest value.
8. device as claimed in claim 7, it is characterised in that:
The network-content acquisition module, the historical record for downloading or accessing by accessing user obtain user and download or visit
Ask the set of Web content;
The content similarities determining module, it is combined into each user for the collection based on the download or access Web content and builds
Vertical user and content relation matrix;Wherein, the quantity for the Web content species that whole users download or accessed is n, the user
With the matrix that content relation matrix is n rows × n row, the user and the nonzero element U in content relation matrixijTo download simultaneously
Or access Web content i and Web content j user.
9. device as claimed in claim 8, it is characterised in that:
The content similarities determining module, it is additionally operable to be added with content relation matrix with the user corresponding to each user,
Web content similarity matrix is obtained, according to the similarity between the Web content similarity matrix calculating network content;Its
In, the Web content similarity matrix is the matrix of n rows × n row, the nonzero element in the Web content similarity matrix
AijTo download or accessing Web content i and Web content j user's set simultaneously.
10. the device as described in any one of claim 7 to 9, it is characterised in that:
The content similarities determining module, it is additionally operable to based on the user and the corresponding relation for the Web content downloaded or accessed
Determine that the similarity between Web content is:
<mrow>
<msub>
<mi>w</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>&Sigma;</mi>
<mrow>
<mi>u</mi>
<mo>&Element;</mo>
<mi>N</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>&cap;</mo>
<mi>N</mi>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</mrow>
</msub>
<mfrac>
<mn>1</mn>
<mrow>
<mi>log</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>+</mo>
<mo>|</mo>
<mi>N</mi>
<mo>(</mo>
<mi>u</mi>
<mo>)</mo>
<mo>|</mo>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
<msqrt>
<mrow>
<mo>|</mo>
<mi>N</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
<mo>|</mo>
<mi>N</mi>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
</msqrt>
</mfrac>
<mo>;</mo>
</mrow>
Wherein, WijFor the similarity between Web content i and Web content j, N (i) is download or the user for accessing Web content i
Set, N (j) is download or accesses Web content j user's set, and u is the user during N (i) and N (j) occurs simultaneously, | N (u) | it is to use
Family u downloads or accessed the quantity of Web content, | N (i) | it is download or access Web content i number of users, | N (j) | under being
Carry or access Web content j number of users.
11. device as claimed in claim 10, it is characterised in that:
The content interest value determining module, for true according to the similarity between the Web content and the historical information
Determine user is to the interest value of Web content:
<mrow>
<msub>
<mi>p</mi>
<mrow>
<mi>u</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>D</mi>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>)</mo>
</mrow>
<mo>&cap;</mo>
<mi>S</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
<mo>)</mo>
</mrow>
</mrow>
</munder>
<msub>
<mi>w</mi>
<mrow>
<mi>j</mi>
<mi>i</mi>
</mrow>
</msub>
<msub>
<mi>r</mi>
<mrow>
<mi>u</mi>
<mi>i</mi>
</mrow>
</msub>
<mo>;</mo>
</mrow>
Wherein, PujInterest value for user u to Web content j, D (u) be user u download or access Web content set, S
The set for the k game that (i, k) is default and Web content i similarities are high, WjiIt is Web content j and Web content i phase
Like degree, ruiIt is interest values of the default user u to Web content i.
12. device as claimed in claim 11, it is characterised in that in the network recommended based on the true directional user of the interest value
Appearance includes:
The content recommendation determining module, is ranked up for the height based on the interest value to Web content, according to default
Recommended amount extraction Web content and to user recommend;
Wherein, the Web content includes:Game, books, film.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610701119.0A CN107766379A (en) | 2016-08-22 | 2016-08-22 | A kind of recommendation method and apparatus of Web content |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610701119.0A CN107766379A (en) | 2016-08-22 | 2016-08-22 | A kind of recommendation method and apparatus of Web content |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107766379A true CN107766379A (en) | 2018-03-06 |
Family
ID=61263570
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610701119.0A Pending CN107766379A (en) | 2016-08-22 | 2016-08-22 | A kind of recommendation method and apparatus of Web content |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107766379A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109299351A (en) * | 2018-09-21 | 2019-02-01 | 北京百度网讯科技有限公司 | Content recommendation method and device, electronic equipment and computer-readable medium |
CN111737558A (en) * | 2020-05-21 | 2020-10-02 | 苏宁金融科技(南京)有限公司 | Information recommendation method and device and computer readable storage medium |
CN115631746A (en) * | 2022-12-20 | 2023-01-20 | 深圳元象信息科技有限公司 | Hot word recognition method and device, computer equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101105795A (en) * | 2006-10-27 | 2008-01-16 | 北京搜神网络技术有限责任公司 | Network behavior based personalized recommendation method and system |
US20100211576A1 (en) * | 2009-02-18 | 2010-08-19 | Johnson J R | Method And System For Similarity Matching |
CN102184199A (en) * | 2011-04-22 | 2011-09-14 | 北京志腾新诺科技有限公司 | Network information recommending method and system |
CN104598643A (en) * | 2015-02-13 | 2015-05-06 | 成都品果科技有限公司 | Article similarity contribution factor, similarity acquiring method, as well as article recommendation method and system thereof |
-
2016
- 2016-08-22 CN CN201610701119.0A patent/CN107766379A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101105795A (en) * | 2006-10-27 | 2008-01-16 | 北京搜神网络技术有限责任公司 | Network behavior based personalized recommendation method and system |
US20100211576A1 (en) * | 2009-02-18 | 2010-08-19 | Johnson J R | Method And System For Similarity Matching |
CN102184199A (en) * | 2011-04-22 | 2011-09-14 | 北京志腾新诺科技有限公司 | Network information recommending method and system |
CN104598643A (en) * | 2015-02-13 | 2015-05-06 | 成都品果科技有限公司 | Article similarity contribution factor, similarity acquiring method, as well as article recommendation method and system thereof |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109299351A (en) * | 2018-09-21 | 2019-02-01 | 北京百度网讯科技有限公司 | Content recommendation method and device, electronic equipment and computer-readable medium |
CN109299351B (en) * | 2018-09-21 | 2022-04-12 | 北京百度网讯科技有限公司 | Content recommendation method and device, electronic equipment and computer readable medium |
CN111737558A (en) * | 2020-05-21 | 2020-10-02 | 苏宁金融科技(南京)有限公司 | Information recommendation method and device and computer readable storage medium |
CN115631746A (en) * | 2022-12-20 | 2023-01-20 | 深圳元象信息科技有限公司 | Hot word recognition method and device, computer equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103995907B (en) | Determining method of access users | |
CN104102648B (en) | Interest based on user behavior data recommends method and device | |
CN102214169B (en) | The offer method and device of key word information and target information | |
CN103916436B (en) | Information-pushing method, device, terminal and server | |
CN103259805B (en) | The domain name access control method evaluated based on user and system | |
CN103353920A (en) | Method and device for recommending games based on SNS | |
CN107766379A (en) | A kind of recommendation method and apparatus of Web content | |
CN103780625B (en) | User interest finds method and apparatus | |
CN102902717A (en) | Method, system and device for organizing a plurality of applications in app store | |
CN103366009B (en) | A kind of book recommendation method based on self-adaption cluster | |
CN107402961A (en) | One kind recommends method and device, electronic equipment | |
CN109492153A (en) | A kind of Products Show method and apparatus | |
CN108926843A (en) | The control method and system won the game in a kind of mahjong class game | |
CN103888466B (en) | user interest discovery method and device | |
CN108280560A (en) | A kind of anti-brush method and device of subject evaluation | |
CN104679791A (en) | Processing method and device for acquiring data packets | |
CN110309410A (en) | A kind of information recommended method, platform and computer readable storage medium | |
CN104035978B (en) | Combo discovering method and system | |
JP5450712B2 (en) | GAME PROPOSITION DEVICE, GAME PROPOSITION METHOD, AND PROGRAM | |
CN114470791A (en) | Game item recommendation method and device, computer equipment, storage medium and product | |
CN107122805A (en) | A kind of user clustering method and apparatus | |
CN105574183A (en) | App (application) recommendation method based on collaborative filtering recommendation algorithm-KNN (K-nearest neighbor) classification algorithm | |
CN103577541B (en) | The ranking fraud detection method and ranking fraud detection system of application program | |
US20170039627A1 (en) | Cigar Band Recognition Interface | |
CN107967637A (en) | The recommendation method, apparatus and electronic equipment of a kind of merchandise items model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180306 |