CN106708829A - Data recommendation method and data recommendation system - Google Patents
Data recommendation method and data recommendation system Download PDFInfo
- Publication number
- CN106708829A CN106708829A CN201510465978.XA CN201510465978A CN106708829A CN 106708829 A CN106708829 A CN 106708829A CN 201510465978 A CN201510465978 A CN 201510465978A CN 106708829 A CN106708829 A CN 106708829A
- Authority
- CN
- China
- Prior art keywords
- data
- information
- recommended
- label
- attribute information
- 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
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Embodiments of the invention disclose a data recommendation method and a data recommendation system, which are applied to the technical field of information processing. According to the recommendation method, at least one piece of to-be-recommended data related to recommendation data of user terminal operation is found firstly when the recommendation system sends the to-be-recommended data; the correlation between the recommendation data and the at least one piece of to-be-recommended data is calculated according to information of the recommendation data and information of the at least one piece of to-be-recommended data; and then which the to-be-recommended data is sent to a user terminal is determined according to the correlation. Therefore, the recommendation system can sent the to-be-recommended data, which a user is interested in, to the user terminal by taking the recommendation data of the user terminal operation as information which the user is interested in, and the corresponding data is recommended according to the interests of the user.
Description
Technical field
The present invention relates to technical field of information processing, more particularly to a kind of data recommendation method and commending system.
Background technology
In existing commending system (such as ad system, news commending system, program of radio station commending system
Deng) in, businessman can by commending system directionally or it is nondirectional will need recommend data is activation arrive
Each user terminal, the purposes such as a certain product or some information are promoted so as to reach.Generally, push away
The system of recommending can be by the such as most popular program of radio station of newest recommending data or nearest news etc. be sent to
User terminal, the data indistinction between the ues so recommended, without personalization.
The content of the invention
The embodiment of the present invention provides a kind of data recommendation method and commending system, realizes according to the emerging of user
Interest recommends corresponding data.
The embodiment of the present invention provides a kind of data recommendation method, including：
First label information of user is obtained, first label information includes：The user is to application
At least one first attribute information of the recommending data of family terminal operation and respectively corresponding first weighted value；
Obtain at least one to be recommended data related at least one first attribute information；
According at least one first attribute information and corresponding first weighted value respectively, and it is described at least
One the second label information of data to be recommended, calculates the recommending data and is treated with described at least one respectively
The degree of correlation of recommending data；Wherein, second label information includes described at least one number to be recommended
According at least one second attribute information and corresponding second weighted value respectively；
User terminal is answered to send the data to be recommended to described according to the degree of correlation.
The embodiment of the present invention also provides a kind of commending system, including：
First label acquiring unit, the first label information for obtaining user, first label information
Include：At least one first attribute information of the recommending data of user correspondence user terminal operations and
Corresponding first weighted value of difference；
Related acquiring unit, for obtaining obtained with the first label acquiring unit at least one first
At least one related data to be recommended of attribute information；
Correlation calculation unit, for being obtained according to the first label acquiring unit at least one first belongs to
Property information and corresponding first weighted value respectively, and the related acquiring unit obtain at least one wait to push away
The second label information of data is recommended, the recommending data is calculated respectively with described at least one data to be recommended
The degree of correlation；Wherein, second label information includes described at least one data to be recommended at least
A kind of second attribute information and respectively corresponding second weighted value；
Data transmission unit, for the degree of correlation that is calculated according to the correlation calculation unit to the application family
Terminal sends the data to be recommended.
It can be seen that, in the recommendation method of the present embodiment, commending system is first looked into when data to be recommended are sent
At least one related to the recommending data of user terminal operations data to be recommended are found, and according to recommendation number
According to the information of information and at least one data to be recommended calculate recommending data respectively with least one band recommendation
The degree of correlation between data, then determines which data to be recommended sent to user terminal according to the degree of correlation.
So by by the recommending data of user terminal operations as user's information interested, it is possible to by recommending
System by user's data is activation to be recommended interested to user terminal, realize according to the interest of user come
Recommend corresponding data.
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 reality
The accompanying drawing to be used needed for example or description of the prior art is applied to be briefly described, it should be apparent that, below
Accompanying drawing in description is only some embodiments of the present invention, for those of ordinary skill in the art,
Without having to pay creative labor, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 is a kind of flow chart of data recommendation method provided in an embodiment of the present invention；
Fig. 2 is the flow chart of another data recommendation method provided in an embodiment of the present invention；
Fig. 3 is a kind of structural representation of commending system provided in an embodiment of the present invention；
Fig. 4 is the structural representation of another commending system provided in an embodiment of the present invention；
Fig. 5 is the structural representation of another commending system provided in an embodiment of the present invention；
Fig. 6 is the schematic diagram of the label information of bag sending station program in Application Example of the present invention.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out
Clearly and completely describe, it is clear that described embodiment is only a part of embodiment of the invention, and
It is not all, of embodiment.Based on the embodiment in the present invention, those of ordinary skill in the art are without work
Go out the every other embodiment obtained under the premise of creative work, belong to the scope of protection of the invention.
Term " first ", " second ", " in description and claims of this specification and above-mentioned accompanying drawing
Three " (if present) such as " 4th " is for distinguishing similar object, without specific suitable for describing
Sequence or precedence.It should be appreciated that the data for so using can be exchanged in the appropriate case, so as to here
The embodiments of the invention of description for example can be with the order in addition to those for illustrating herein or describing
Implement.Additionally, term " comprising " and " having " and their any deformation, it is intended that covering is not exclusively
Include, for example, containing process, method, system, product or the equipment of series of steps or unit
Those steps or the unit clearly listed are not necessarily limited to, but be may include not listing clearly or right
In these processes, method, product or other intrinsic steps of equipment or unit.
A kind of data recommendation method of embodiment of the present invention offer, mainly commending system (such as ad system,
News commending system, program of radio station commending system etc.) recommending data is sent to the method for user terminal.
The method of the present embodiment be the method performed by commending system, flow chart as shown in figure 1, including：
Step 101, obtains first label information of user, includes in the first label information：The user couple
Answer at least one first attribute information and corresponding first weight of difference of the recommending data of user terminal operations
Value.
It is appreciated that first label information of user is primarily used to describe the recommendation number interested to user
According to information, specifically, (ratio is operated when the corresponding user terminal of user to a certain bar recommending data
As collected, the operation such as click or download) when, commending system can store into the information of this recommending data
The form of the first label information.Wherein, because user terminal understands timing or sporadically report of user terminal
To the operation information of recommending data, then commending system can learn that user terminal is carried out to which recommending data
Operation.The method of step 101 to 104 is that commending system is needing to be sent out to a certain user terminal in the present embodiment
The flow triggered during data to be recommended is sent, when obtaining the first label information in this step 101, commending system
Can in reading system before the first label information for having stored.
Here the first attribute information can be the classification (category, abbreviation C) of recommending data；Recommend number
According to title (title, abbreviation T)；The description (description, abbreviation D) of recommending data；Or recommend number
According to ID, such as the main broadcaster (singer, abbreviation S) in radio station in program of radio station commending system, i.e.,
User profile belonging to radio station etc..It should be noted that can include in a certain attribute information a plurality of
The attribute information of recommending data and respectively corresponding first weighted value, such as in the classification of recommending data
In information can include recommending data 1 classification and corresponding first weighted value, and recommending data 2 classification
And corresponding first weighted value etc..
Step 102, obtains at least one to be recommended data related at least one first attribute information.
When commending system needs the data to be recommended for being sent to user terminal, such as the new radio station section released
The advertisement wait recommending data that mesh, nearest news, or needs are promoted, commending system will can be waited to push away in advance
The information for recommending data stores into the form of the 3rd label information, and when this step 102 are performed, commending system
Can be searched in the 3rd label information of system storage and related to above-mentioned at least one attribute information owned
The information of data to be recommended.
Wherein, can include in the 3rd label information：At least one dimension sort out under data to be recommended
Information, it can be a certain specific classification that dimension here is sorted out, a certain specific title, a certain specific descriptions,
Or a certain particular user mark, such as in the main broadcaster etc. in a certain radio station.It should be noted that in a certain kind
Data to be recommended under dimension is sorted out can include a plurality of data to be recommended and respectively corresponding 3rd weighted value,
Data to be recommended 1 and corresponding 3rd weighted value can be included such as in the data to be recommended of a certain classification,
And data 2 to be recommended and corresponding 3rd weighted value etc..
Step 103, corresponds to according at least one first attribute information obtained in above-mentioned steps 101 and respectively
The first weighted value, and in above-mentioned steps 102 obtain at least one data to be recommended the second label information,
It is to be recommended with least one that above-mentioned recommending data (recommending data of the first label information description) is calculated respectively
The degree of correlation of data.
Here the second label information is primarily used to description commending system to be needed to be sent to treating for user terminal
The information of recommending data, such as the new program of radio station released, nearest news, or need the advertisement promoted
Wait the information of recommending data.Include at least the one of at least one data to be recommended in the second label information
Plant the second attribute information and respectively corresponding second weighted value, second label information and above-mentioned first label
The form of information is similar to, unlike, the second label information describes data to be recommended, and above-mentioned the
One label information describes the recommending data that the commending system of user terminal operations has sent.
Commending system in the relatedness computation in performing this step, specifically can using cosine similarity or
The modes such as Euclidean distance are calculated, during calculating, mainly by least one first attribute information and
Corresponding first weighted value represents the vector information of recommending data respectively, and passes through at least one second and belong to
Property information and respectively corresponding second weighted value represent the vector information of data to be recommended, and then by each
From vector information calculate the degree of correlation.
Step 104, data to be recommended are sent according to the degree of correlation to user's correspondence user terminal, specifically, can
With by the data is activation a plurality of to be recommended user terminal higher of the degree of correlation between recommending data.
It should be noted that above-mentioned steps 101 to 104 are to send data to be recommended to certain in commending system
Data recommendation method during one user terminal.
It can be seen that, in the recommendation method of the present embodiment, commending system is first looked into when data to be recommended are sent
At least one related to the recommending data of user terminal operations data to be recommended are found, and according to recommendation number
According to the information of information and at least one data to be recommended calculate recommending data respectively with least one band recommendation
The degree of correlation between data, then determines which data to be recommended sent to user terminal according to the degree of correlation.
So by by the recommending data of user terminal operations as user's information interested, it is possible to by recommending
System by user's data is activation to be recommended interested to user terminal, realize according to the interest of user come
Recommend corresponding data.
In a specific embodiment, the method in order to realize the embodiment of the present invention, commending system needs
User's information interested is stored in advance, then before above-mentioned steps 101 are performed, commending system can also lead to
The storage that the following two kinds mode realizes the recommending data to user's operation is crossed, specifically：
(1) solid arrow with reference to shown in Fig. 2 is pointed to, in one case, before step 101 also
Including step A.
A, when user terminal is operated to recommending data, commending system determines original tag information, just
Beginning label information includes：At least one attribute information of recommending data, and at least one attribute information point
Not corresponding weighted value.
It should be noted that such case is when commending system have sent a recommending data first, and use
Family terminal-pair this recommending data is operated, storage of the commending system to the information of this recommending data,
When the transmission flow of above-mentioned data to be recommended is triggered, commending system is specifically performing above-mentioned steps 101, i.e.,
When obtaining first label information of user, the original tag information can be directly read and believed as the first label
Breath.
(2) dotted arrow is pointed to as shown in Figure 2, in another case, before step 101 also
Including step A and B.
A, when user terminal is operated to recommending data, commending system determines original tag information, just
Beginning label information includes：At least one attribute information of recommending data, and at least one attribute information point
Not corresponding weighted value.
B, original tag information and the existing label information that system is stored merge obtains final label letter
Breath, final label information includes the attribute information and corresponding weighted value after fusion, the attribute after fusion
Information includes at least one attribute information of recommending data.
Commending system will can specifically be wrapped in the mixing operation in performing this step in original tag information
At least one attribute information for including accordingly is stored with the corresponding attribute information that has, and reduces existing attribute
The corresponding existing weighted value of information.Wherein, existing label information includes existing attribute information.It can be seen that,
Commending system take into account the time AF of the user stored in system information interested, drop
The weight of the existing attribute information of low existing recommending data, the user stored in system of having given prominence to the key points is newest
Information interested.
It should be noted that such case is to have stored pushing away for user terminal operations in commending system
After recommending the information of data, commending system have sent a recommending data again, and user terminal is recommended this
Data are operated, storage of the commending system to the information of this recommending data, above-mentioned to be recommended when triggering
During the transmission flow of data, commending system is specifically performing above-mentioned steps 101, that is, obtain first mark of user
During label information, the final label information can be directly read as the first label information.
In another specific embodiment, in order to calculate the recommendation of data to be recommended and user terminal operations
The degree of correlation between data, commending system is needed data storage to be recommended into specific form in advance, then
Before above-mentioned steps 102 are performed, commending system can also determine the second label information of data to be recommended,
Second label information includes at least one second attribute information of data to be recommended and respectively corresponding the
Two weighted values；Then the second label information is converted into the 3rd label information, the 3rd label information includes
The information of the data to be recommended under the classification of at least one dimension and respectively corresponding 3rd weighted value, wherein,
Dimension classifies as some second attribute information.It can be seen that, the 3rd label information here is to treat each to push away
The information after data are sorted out according to any one attribute information is recommended, commending system is facilitated and is performed above-mentioned steps 102
In the related data to be recommended of acquisition.
Wherein, the second attribute information can be the classification of data to be recommended, and the title of data to be recommended is pushed away
Recommend the description of data, or data to be recommended ID etc., then for the classification and use of data to be recommended
Family identifies, and commending system is it is determined that during corresponding second weighted value of the second attribute information, directly carry out tax power
Value.
For the title of data to be recommended, commending system is it is determined that corresponding second weight of the second attribute information
During value, the title that can first treat recommending data carries out participle, and then each participle respectively to title is assigned
Weights obtain corresponding second weighted value of title of data to be recommended.Wherein, participle is carried out to title, it is main
If removing auxiliary word and stop words etc. from title, such as to " the daily luck tendency brief introduction of 12 constellations and point
Analysis " is obtained after carrying out participle：" 12, constellation, daily, luck tendency, brief introduction, analysis "；To participle
When carrying out tax weights, can be using word frequency-reverse document-frequency (term frequency-inverse document
Frequency, TF-IDF) method carry out to each participle assign weights.
For the description of data to be recommended, commending system is it is determined that corresponding second weight of the second attribute information
During value, the description that can first treat recommending data carries out participle, and the participle of description is clustered into multiple shallow words
Then multiple shallow meaning of a word are assigned corresponding second weighted value of description that weights obtain data to be recommended by justice respectively.
Wherein, commending system can be based on semantic model, such as be to hide border using a topic model algorithm
Distribution (Latent Dirichlet Allocation, LDA), or open source projects word2vec etc. is clustered,
So multiple participles can be clustered into a shallow meaning of a word, such as by " splendid magic, very Austria
Contend with, only I am ancient military most strong " decile term clustering, into the shallow meaning of a word such as " magic " or " magical ", reduces
The dimension of data to be recommended.
It should be noted that in above-mentioned first label information the determination method of the first weighted value with this second mark
The determination method of the second weighted value is similar in label information, unlike, the determination of the first label information is base
In recommending data, and the determination of the second label information is based on data to be recommended.
The embodiment of the present invention also provides a kind of commending system, and its structural representation is as shown in figure 3, specifically can be with
Including：
First label acquiring unit 10, the first label information for obtaining user, the first label letter
Breath includes：At least one first attribute information of the recommending data of user's correspondence user terminal operations
And corresponding first weighted value of difference.Here the first attribute information can be the classification of recommending data；Push away
Recommend the title of data；The description of recommending data；Or the ID of recommending data etc..
Related acquiring unit 11, for obtaining at least one obtained with the first label acquiring unit 10
At least one related data to be recommended of first attribute information.
Correlation calculation unit 12, for obtained according to the first label acquiring unit 10 at least one the
One attribute information and respectively corresponding first weighted value, and the related acquiring unit 11 obtain at least one
Second label information of bar data to be recommended, calculates the recommending data and waits to push away with described at least one respectively
Recommend the degree of correlation of data；Wherein, second label information includes described at least one data to be recommended
At least one second attribute information and corresponding second weighted value respectively.Correlation calculation unit 12 specifically may be used
Calculated with using modes such as cosine similarity or Euclidean distances.
Data transmission unit 13, the degree of correlation for being calculated according to the correlation calculation unit 12 is answered to described
User terminal sends the data to be recommended.Data transmission unit 13 specifically can be by between recommending data
The degree of correlation data is activation a plurality of to be recommended user terminal higher.
It can be seen that, when the commending system of the present embodiment is in transmission data to be recommended, related acquiring unit 11 is first
At least one related to the recommending data of user terminal operations data to be recommended are found, and by correlometer
Calculate unit 12 and recommendation number is calculated according to the information of recommending data and the information of at least one data to be recommended respectively
According at least one with the degree of correlation between recommending data, then data transmission unit 13 is true according to the degree of correlation
Which data to be recommended is directional user's terminal send.So work as by by the recommending data of user terminal operations
Make user's information interested, it is possible to user's data is activation to be recommended interested is given by commending system
User terminal, realizes the interest according to user to recommend corresponding data.
With reference to shown in Fig. 4, in a specific embodiment, commending system is except including as shown in Figure 3
Outside structure, can also include originally determined unit 14, the tag determination unit 16 of integrated unit 15 and second,
Wherein：
Originally determined unit 14, for when the user terminal is operated to the recommending data, really
Determine original tag information, the original tag information includes：At least one attribute of the recommending data
Information, and at least one attribute information distinguishes corresponding weighted value.
Integrated unit 15, original tag information and system for the originally determined unit 14 to be determined are stored up
The existing label information deposited merge and obtains final label information, and the final label information includes melting
Attribute information and corresponding weighted value after conjunction, the attribute information after the fusion include the recommendation number
According at least one attribute information.Specifically, integrated unit 15, specifically for by the originally determined list
At least one attribute information that the original tag information that unit 14 determines includes has attribute information with corresponding
Accordingly store, reduce the corresponding existing weighted value of existing attribute information；Wherein, the existing label letter
Breath includes the existing attribute information.
It is appreciated that in one case, commending system have sent recommending data first, and by user's end
When end operates the recommending data, originally determined unit 14 can determine the initial labels letter of the recommending data
Breath, then the first label acquiring unit 10 is specifically for reading the first of the determination of originally determined unit 14
Beginning label information is used as first label information.In another case, commending system have sent recommendation
Data, and by after the user terminal operations recommending datas, recommending data having been stored in commending system
Information, then the first label acquiring unit 10 obtained most specifically for reading the integrated unit 15
Whole label information is used as first label information.
Second tag determination unit 16, second label information for determining the data to be recommended；
Second label information is converted into the 3rd label information, the 3rd label information includes at least one
The information of the data to be recommended under individual dimension classification and respectively corresponding 3rd weighted value；Wherein, the dimension
Degree classifies as some described second attribute information.
For the classification and ID of data to be recommended, the second tag determination unit 16 is it is determined that the second category
During corresponding second weighted value of property information, tax weights are directly carried out.
If second attribute information includes the title of the data to be recommended, second label determines
Unit 16, carries out participle, respectively to each of the title specifically for the title to the data to be recommended
Individual participle assigns corresponding second weighted value of title that weights obtain the data to be recommended；
If second attribute information includes the description of the data to be recommended, second label determines
Unit 16, participle is carried out specifically for the description to the data to be recommended, and the participle of the description is gathered
Class assigns the description correspondence that weights obtain the data to be recommended to multiple shallow meaning of a word respectively into multiple shallow meaning of a word
The second weighted value.
The information of data to be recommended is so assured that by the second tag determination unit 16, has so been
Looked into the 3rd label information that related acquiring unit 11 in system can determine from the second tag determination unit 16
Look for all to be recommended data related to the first attribute information.
The embodiment of the present invention also provides a kind of commending system, structural representation as shown in figure 5, the commending system
Can include one or more centres because of configuration or performance is different and the larger difference of producing ratio
Reason device (central processing units, CPU) 20 (for example, one or more processors) and
The storage medium 22 of memory 21, one or more storage application programs 221 or data 222 is (for example
One or more mass memory units).Wherein, memory 21 and storage medium 22 can be of short duration depositing
Storage or persistently storage.The program stored in storage medium 22 can include one or more module (figures
Show and do not mark), each module can be included to the series of instructions operation in terminal.Further, in
Central processor 30 could be arranged to be communicated with storage medium 22, in performing storage medium 22 on commending system
Series of instructions operation.
Commending system can also include one or more power supplys 23, one or more wired or nothings
Wired network interface 24, one or more input/output interfaces 25, and/or, one or more behaviour
Make system 223, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM,
FreeBSDTM etc..
The step as performed by commending system described in above method embodiment can be based on shown in the Fig. 5
Commending system structure.
The method that the embodiment of the present invention is illustrated with a specific application example below, in the present embodiment
The commending system that method is mainly used in is program of radio station commending system, then program of radio station commending system sends
Recommending data is program of radio station, then with reference to shown in Fig. 6, program of radio station commending system is triggering program of radio station
Transmission flow before need to pre-process program of radio station, specifically：
(1) sending station program is treated to be pre-processed
First, the second label information of program of radio station to be sent is determined：Treat the main broadcaster of sending station program
(S), classification (C), title (T) and description (D) is respectively processed, wherein, to electricity to be sent
The main broadcaster of platform program and classification directly assign weights, such as the weighted value of the main broadcaster 1 in Fig. 6 is w1, classification 2
Weighted value for w2 etc.；The title for treating sending station program carries out participle, then using the side of TF-IDF
Method assigns weights to participle, such as the weighted value of the participle of project (term) 1 is w1, the participle of project 2
Weighted value is w2；Treating the description of sending station program carries out the label (tag) that participle obtains each participle,
It is then based on semantic model and the label of each participle is respectively mapped to multiple themes (topic), it is plurality of
The label of participle can be mapped as a theme, be thus a shallow meaning of a word, drop by multiple participles cluster
The low dimension of the description of program of radio station to be sent.
Second label information of program of radio station to be sent is converted into the 3rd label information：By the second label letter
The program of radio station all to be sent stored in breath is sorted out according to each attribute information, specifically, can be by the
Two label informations carry out upset and can obtain the 3rd label information, for any attribute in the second label information
Information, using the attribute information as index key (key), and by the second label information with the index
The corresponding program of radio station to be sent of keyword is attributed under the attribute information, and under the index key each
The weighted value of program of radio station to be sent, respectively with the second label information each program of radio station to be sent should
The weighted value of attribute information is identical.In such as Fig. 6, the weight of the classification 1 in radio station 1 in the second label information
It is w1 to be worth, then after the 3rd label information is converted into, radio station 1 is ranged under classification 1, then the radio station 1
Weighted value then be w1.
(2) program of radio station that program of radio station commending system has sent is pre-processed
When program of radio station commending system have sent a program of radio station first, and the user terminal operations electricity
Platform program, and after operation information is reported into program of radio station commending system, program of radio station commending system can be by
The program of radio station is saved as user terminal correspondence user's according to the method for determining above-mentioned second label information
Initial label information.
When program of radio station commending system have sent other program of radio station, and other radio station of user terminal operations
Program, and after operation information is reported into program of radio station commending system, program of radio station commending system can be according to
Determine that the method for above-mentioned second label information determines the label information of other program of radio station, then by other electricity
The label information of platform program with the initial label information for having stored merged after label information,
Specifically, it is the attribute information of other program of radio station is corresponding with corresponding attribute information in initial label information
Ground storage, and reduce the corresponding weighted value of the attribute information that initial label information includes.For example, most
Whole label information includes classification 1 and weighted value w1, and the attribute information of other program of radio station be classification 2 and
Weighted value is w2, then in fusion process, classification 1 and classification 2 are accordingly stored, and reduces the value of w1,
And represent w2 with larger value.
Thus can be using the information of the program of radio station of user terminal operations as correspondence user letter interested
Breath.
Program of radio station commending system can use as follows after the transmission flow for triggering program of radio station, specifically
Method sending station program to user terminal, specifically：
(1) user's information interested is obtained, that is, obtains program of radio station commending system by above-mentioned pretreatment
Label information after the fusion that process is obtained, or initially label information.
(2) in the 3rd label information obtained above, search with merge after label information (or initial mark
Label information) in program of radio station the related program of radio station all to be sent of attribute information.
(3) method according to cosine similarity calculates the program of radio station and above-mentioned step of user terminal operations respectively
Suddenly the degree of correlation between the program of radio station all to be sent searched in (2), in this process, will be pending
Second label information of the vectorial program of radio station to be sent obtained with above-mentioned pretreatment of program of radio station is sent come table
Show, label information after the fusion that the vector of the program of radio station of user terminal operations is obtained with above-mentioned pretreatment
(or initially label information) represents.
(4) degree of correlation calculated in above-mentioned steps (3) is ranked up, higher degree of relation is corresponding
Multiple program of radio station to be sent are sent to user terminal.
It can be seen that, in the present embodiment, program of radio station commending system using the program of radio station of user terminal operations as
Correspondence user's information interested, then by program of radio station that the information correlation interested with user is higher
It is sent to user terminal.Such as to preference magic user correspondence user terminal send pass through wait correlation electricity
Platform program.
One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment
Rapid to can be by program to instruct the hardware of correlation to complete, the program can be stored in a computer can
Read in storage medium, storage medium can include：Read-only storage (ROM), random access memory
RAM), disk or CD etc..
The data recommendation method and commending system for being provided the embodiment of the present invention above are described in detail,
Specific case used herein is set forth to principle of the invention and implementation method, above example
Explanation be only intended to help and understand the method for the present invention and its core concept；Simultaneously for this area
Those skilled in the art, according to thought of the invention, have change in specific embodiments and applications
Become part, in sum, this specification content should not be construed as limiting the invention.
Claims (12)
1. a kind of data recommendation method, it is characterised in that including：
First label information of user is obtained, first label information includes：The user is to application
At least one first attribute information of the recommending data of family terminal operation and respectively corresponding first weighted value；
Obtain at least one to be recommended data related at least one first attribute information；
According at least one first attribute information and corresponding first weighted value respectively, and it is described at least
One the second label information of data to be recommended, calculates the recommending data and is treated with described at least one respectively
The degree of correlation of recommending data；Wherein, second label information includes described at least one number to be recommended
According at least one second attribute information and corresponding second weighted value respectively；
User terminal is answered to send the data to be recommended to described according to the degree of correlation.
2. the method for claim 1, it is characterised in that the first label letter of the acquisition user
Before breath, methods described also includes：
When the user terminal is operated to the recommending data, original tag information is determined, it is described
Original tag information includes：At least one attribute information of the recommending data, and at least one
Attribute information distinguishes corresponding weighted value；
First label information for obtaining user, specifically includes：Read the original tag information conduct
First label information.
3. the method for claim 1, it is characterised in that the first label letter of the acquisition user
Before breath, methods described also includes：
When the user terminal is operated to the recommending data, original tag information is determined, it is described
Original tag information includes：At least one attribute information of the recommending data, and at least one
Attribute information distinguishes corresponding weighted value；
The original tag information with the existing label information that system is stored merge and obtains final label
Information, the final label information includes the attribute information and corresponding weighted value after fusion, described to melt
Attribute information after conjunction includes at least one attribute information of the recommending data；
First label information for obtaining user, specifically includes：Read the final label information conduct
First label information.
4. method as claimed in claim 3, it is characterised in that by the original tag information and system
The existing label information of storage merge and obtains final label information, is specifically included：
At least one attribute information that the original tag information is included has attribute information with corresponding
Accordingly store, reduce the corresponding existing weighted value of existing attribute information；Wherein, the existing label letter
Breath includes the existing attribute information.
5. the method as described in any one of Claims 1-4, it is characterised in that the acquisition with it is described extremely
Before at least one related data to be recommended of few a kind of first attribute information, also include：
Determine second label information of the data to be recommended；
Second label information is converted into the 3rd label information, the 3rd label information include to
The information of the data to be recommended under few dimension classification and respectively corresponding 3rd weighted value；
Wherein, the dimension classifies as some described second attribute information.
6. method as claimed in claim 5, it is characterised in that
If title of second attribute information including the data to be recommended, wait to push away described in the determination
The second weighted value that second label information of data includes is recommended, is specifically included：To described to be recommended
The title of data carries out participle, and each participle tax weights to the title obtain the number to be recommended respectively
According to corresponding second weighted value of title；
If description of second attribute information including the data to be recommended, wait to push away described in the determination
The second weighted value that second label information of data includes is recommended, is specifically included：To described to be recommended
The description of data carries out participle, and the participle of the description is clustered into multiple shallow meaning of a word, shallow to multiple respectively
The meaning of a word assigns corresponding second weighted value of description that weights obtain the data to be recommended.
7. a kind of commending system, it is characterised in that including：
First label acquiring unit, the first label information for obtaining user, first label information
Include：At least one first attribute information of the recommending data of user correspondence user terminal operations and
Corresponding first weighted value of difference；
Related acquiring unit, for obtaining obtained with the first label acquiring unit at least one first
At least one related data to be recommended of attribute information；
Correlation calculation unit, for being obtained according to the first label acquiring unit at least one first belongs to
Property information and corresponding first weighted value respectively, and the related acquiring unit obtain at least one wait to push away
The second label information of data is recommended, the recommending data is calculated respectively with described at least one data to be recommended
The degree of correlation；Wherein, second label information includes described at least one data to be recommended at least
A kind of second attribute information and respectively corresponding second weighted value；
Data transmission unit, for the degree of correlation that is calculated according to the correlation calculation unit to the application family
Terminal sends the data to be recommended.
8. system as claimed in claim 7, it is characterised in that the system also includes：
Originally determined unit, for when the user terminal is operated to the recommending data, it is determined that
Original tag information, the original tag information includes：At least one attribute letter of the recommending data
Breath, and at least one attribute information distinguishes corresponding weighted value；
Then the first label acquiring unit, it is initial specifically for read that the originally determined unit determines
Label information is used as first label information.
9. system as claimed in claim 7, it is characterised in that the system also includes：
Originally determined unit, for when the user terminal is operated to the recommending data, it is determined that
Original tag information, the original tag information includes：At least one attribute letter of the recommending data
Breath, and at least one attribute information distinguishes corresponding weighted value；
Integrated unit, for the original tag information and the system storage that determine the originally determined unit
Existing label information merge and obtains final label information, after the final label information includes fusion
Attribute information and corresponding weighted value, the attribute information after the fusion includes the recommending data
At least one attribute information；
Then the first label acquiring unit, specifically for reading the final label that the integrated unit is obtained
Information is used as first label information.
10. system as claimed in claim 9, it is characterised in that
The integrated unit, specifically for will be wrapped in the original tag information of the originally determined unit determination
At least one attribute information for including accordingly is stored with the corresponding attribute information that has, and reduces existing attribute letter
The corresponding existing weighted value of breath；Wherein, the existing label information includes the existing attribute information.
11. system as described in any one of claim 7 to 10, it is characterised in that the system also includes：
Second tag determination unit, second label information for determining the data to be recommended；Will
Second label information is converted into the 3rd label information, and the 3rd label information includes at least one
The information of the data to be recommended under dimension classification and respectively corresponding 3rd weighted value；
Wherein, the dimension classifies as some described second attribute information.
12. systems as claimed in claim 11, it is characterised in that
If second attribute information includes the title of the data to be recommended, second label determines
Unit, participle is carried out specifically for the title to the data to be recommended, respectively to the title each
Participle assigns corresponding second weighted value of title that weights obtain the data to be recommended；
If second attribute information includes the description of the data to be recommended, second label determines
Unit, participle is carried out specifically for the description to the data to be recommended, and the participle of the description is clustered
Into multiple shallow meaning of a word, the description for obtaining the data to be recommended to multiple shallow meaning of a word tax weights respectively is corresponding
Second weighted value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510465978.XA CN106708829B (en) | 2015-07-31 | 2015-07-31 | Data recommendation method and recommendation system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510465978.XA CN106708829B (en) | 2015-07-31 | 2015-07-31 | Data recommendation method and recommendation system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106708829A true CN106708829A (en) | 2017-05-24 |
| CN106708829B CN106708829B (en) | 2022-05-10 |
Family
ID=58923202
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510465978.XA Active CN106708829B (en) | 2015-07-31 | 2015-07-31 | Data recommendation method and recommendation system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106708829B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555131A (en) * | 2018-03-27 | 2019-12-10 | 优酷网络技术（北京）有限公司 | Content recommendation method, content recommendation device and electronic equipment |
| CN110555135A (en) * | 2018-03-27 | 2019-12-10 | 优酷网络技术（北京）有限公司 | Content recommendation method, content recommendation device and electronic equipment |
| CN110555157A (en) * | 2018-03-27 | 2019-12-10 | 优酷网络技术（北京）有限公司 | Content recommendation method, content recommendation device and electronic equipment |
| CN110837598A (en) * | 2019-11-11 | 2020-02-25 | 腾讯科技（深圳）有限公司 | Information recommendation method, device, equipment and storage medium |
| WO2020143351A1 (en) * | 2019-01-08 | 2020-07-16 | 北京三快在线科技有限公司 | Area comprehensive recommendation method, electronic device, and readable storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102654859A (en) * | 2011-03-01 | 2012-09-05 | 北京彩云在线技术开发有限公司 | Method and system for recommending songs |
| CN103023971A (en) * | 2012-11-15 | 2013-04-03 | 广州酷狗计算机科技有限公司 | Information pushing method and system of music sharing radio stations |
| CN103870448A (en) * | 2012-12-07 | 2014-06-18 | 盛乐信息技术（上海）有限公司 | Method and method for recommending data |
| CN104424210A (en) * | 2013-08-22 | 2015-03-18 | 腾讯科技（深圳）有限公司 | Information recommendation method, information recommendation system and server |
| CN104426934A (en) * | 2013-08-24 | 2015-03-18 | 上海莞东拿信息科技有限公司 | New information pushing method for tourism mobile terminal |
-
2015
- 2015-07-31 CN CN201510465978.XA patent/CN106708829B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102654859A (en) * | 2011-03-01 | 2012-09-05 | 北京彩云在线技术开发有限公司 | Method and system for recommending songs |
| CN103023971A (en) * | 2012-11-15 | 2013-04-03 | 广州酷狗计算机科技有限公司 | Information pushing method and system of music sharing radio stations |
| CN103870448A (en) * | 2012-12-07 | 2014-06-18 | 盛乐信息技术（上海）有限公司 | Method and method for recommending data |
| CN104424210A (en) * | 2013-08-22 | 2015-03-18 | 腾讯科技（深圳）有限公司 | Information recommendation method, information recommendation system and server |
| CN104426934A (en) * | 2013-08-24 | 2015-03-18 | 上海莞东拿信息科技有限公司 | New information pushing method for tourism mobile terminal |
Non-Patent Citations (1)
| Title |
|---|
| 薛福亮著: "《电子商务推荐相关技术及其改进机制》", 30 June 2014 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555131A (en) * | 2018-03-27 | 2019-12-10 | 优酷网络技术（北京）有限公司 | Content recommendation method, content recommendation device and electronic equipment |
| CN110555135A (en) * | 2018-03-27 | 2019-12-10 | 优酷网络技术（北京）有限公司 | Content recommendation method, content recommendation device and electronic equipment |
| CN110555157A (en) * | 2018-03-27 | 2019-12-10 | 优酷网络技术（北京）有限公司 | Content recommendation method, content recommendation device and electronic equipment |
| WO2020143351A1 (en) * | 2019-01-08 | 2020-07-16 | 北京三快在线科技有限公司 | Area comprehensive recommendation method, electronic device, and readable storage medium |
| CN110837598A (en) * | 2019-11-11 | 2020-02-25 | 腾讯科技（深圳）有限公司 | Information recommendation method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106708829B (en) | 2022-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105224699B (en) | News recommendation method and device | |
| CN110209808B (en) | Event generation method based on text information and related device | |
| CN108629665A (en) | A kind of individual commodity recommendation method and system | |
| CN104111941B (en) | The method and apparatus that information is shown | |
| CN106708829A (en) | Data recommendation method and data recommendation system | |
| CN104751354B (en) | A kind of advertisement crowd screening technique | |
| CN108090208A (en) | Fused data processing method and processing device | |
| CN103106262B (en) | The method and apparatus that document classification, supporting vector machine model generate | |
| CN104537115A (en) | Method and device for exploring user interests | |
| CN108446964B (en) | User recommendation method based on mobile traffic DPI data | |
| CN103914492A (en) | Method for query term fusion, method for commodity information publish and method and system for searching | |
| CN104199822A (en) | Method and system for identifying demand classification corresponding to searching | |
| CN109885674B (en) | Method and device for determining and recommending information of subject label | |
| CN104391999A (en) | Information recommendation method and device | |
| CN109074366A (en) | Gain adjustment component for computer network routed infrastructure | |
| CN102004772A (en) | Method and equipment for sequencing search results according to terms | |
| CN106897359A (en) | Internet information is collected and correlating method | |
| CN109558533A (en) | A kind of personalization content recommendation method and device based on multiple cluster | |
| Zubiaga et al. | Content-based clustering for tag cloud visualization | |
| CN109885651A (en) | A kind of question pushing method and device | |
| CN108629047A (en) | A kind of song list generation method and terminal device | |
| CN112417121A (en) | Client intention recognition method and device, computer equipment and storage medium | |
| CN114265927A (en) | Data query method and device, storage medium and electronic device | |
| CN115098777A (en) | User personalized recommendation method and system based on data analysis | |
| CN110717089A (en) | User behavior analysis system and method based on weblog |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |