CN109190046A - Content recommendation method, device and content recommendation service device - Google Patents
Content recommendation method, device and content recommendation service device Download PDFInfo
- Publication number
- CN109190046A CN109190046A CN201811087626.5A CN201811087626A CN109190046A CN 109190046 A CN109190046 A CN 109190046A CN 201811087626 A CN201811087626 A CN 201811087626A CN 109190046 A CN109190046 A CN 109190046A
- Authority
- CN
- China
- Prior art keywords
- vector collection
- keyword
- network model
- neural network
- internet resources
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention provides a kind of content recommendation method, device and content recommendation service device, is related to computer data processing technology field.This method clicks at least one the corresponding keyword of first network resource checked by user terminal in the first preset period of time by collecting user, and obtain text information corresponding to a plurality of second Internet resources, then keyword is converted into primary vector collection using default neural network model, and text information is converted into secondary vector collection；Then it determines the similarity of primary vector collection and secondary vector collection, and similarity in a plurality of second Internet resources is greater than or equal to the corresponding Internet resources of preset threshold and pushes to user terminal.It based on this, avoids because pushing the single problem of content caused by according to tag match, so that the content for recommending user is more abundant.
Description
Technical field
The present invention relates to computer data processing technology field, in particular to a kind of content recommendation method, device and
Content recommendation service device.
Background technique
The mode of information flow Products Show article at present, mainly according to the label of article and user interest tag match
Degree creates the correlation between article and user, to recommend to meet the article of user interest label.The recommended method, depends on
The calculating of user interest label and the calculating of story label, the article that user sees enrich degree and depend on own interests label feelings
Condition.In the prior art, the accuracy of user and document match depends on the accuracy of story label itself, and numerous dimensions
Label also complicate matching primitives.Meanwhile the recommendation of article is carried out according to user tag, the article collection for recommending to come out can be made
In content corresponding to label, be easy to cause the unification of content.
Summary of the invention
In order to overcome the deficiencies in the prior art described above, the present invention provides a kind of content recommendation method, device and content and pushes away
Recommend server.
To achieve the goals above, technical solution provided by the embodiment of the present invention is as follows:
In a first aspect, the embodiment of the present invention provides a kind of content recommendation method, it is applied to content recommendation service device, it is described interior
Hold recommendation server to connect with user terminal communication, which comprises
It is corresponding to obtain the first network resource that user is checked in the first preset period of time by user terminal click
At least one keyword, and obtain text information corresponding to a plurality of second Internet resources；
At least one described keyword is inputted into default neural network model, is obtained corresponding at least one described keyword
Primary vector collection, the primary vector collection include M dimension primary vector；And the text information is inputted into the default nerve
Network model obtains secondary vector collection corresponding with the text information, and the secondary vector collection includes M dimension secondary vector,
In, M is the integer greater than 0；
The similarity of the primary vector collection Yu the secondary vector collection is determined according to the default neural network model；
When the similarity is greater than or equal to preset threshold, the similarity is greater than or equal to the preset threshold
Second Internet resources corresponding to secondary vector collection push to the user terminal.
Optionally, above-mentioned that at least one described keyword is inputted into default neural network model, it obtains and described at least one
The corresponding primary vector collection of a keyword, comprising:
At least one described keyword is inputted into default neural network model, the default neural network model is based on described
The determining same or similar word of meaning at least one keyword of at least one keyword, and based on it is described at least one
Keyword and the primary vector collection is determined with the same or similar word of meaning of at least one keyword.
Optionally, above-mentioned that the text information is inputted into the default neural network model, it obtains and the text information
Corresponding secondary vector collection, comprising:
The text information is inputted into the default neural network model, by the default neural network model by the text
This information is split to obtain multiple keywords, determines the secondary vector collection based on obtained multiple keywords are split.
Optionally, above-mentioned acquisition user clicks the first network checked by the user terminal in the first preset period of time
Before at least one corresponding keyword of resource, the method also includes:
Learning training, the nerve net that will be obtained after training are carried out to neural network model based on the sample dictionary constructed in advance
Network model is as the default neural network model.
Optionally, above-mentioned Internet resources include at least one of text information, pictorial information, video information.
Optionally, above-mentioned that the similarity is greater than or equal to the second network corresponding to the secondary vector collection of preset threshold
Resource supplying is to the user terminal, comprising:
The URL information of the second Internet resources corresponding with the secondary vector collection is pushed into the user terminal.
Optionally, above-mentioned default neural network model includes word incorporation model.
Optionally, above-mentioned that the similarity is greater than or equal to second corresponding to the secondary vector collection of the preset threshold
Internet resources push to the user terminal, comprising:
It is chosen from the Internet resources that the first platform for issuing the first network resource is issued and meets the similarity
Internet resources corresponding to secondary vector collection more than or equal to the preset threshold, using as second Internet resources, and
Second Internet resources are pushed into the user terminal.
Second aspect, the embodiment of the present invention provide a kind of content recommendation device, are applied to content recommendation service device, described interior
Hold recommendation server to connect with user terminal communication, described device includes:
Acquiring unit clicks the first net checked by the user terminal in the first preset period of time for obtaining user
At least one corresponding keyword of network resource, and obtain text information corresponding to a plurality of second Internet resources；
Input unit, at least one described keyword to be inputted default neural network model, obtain with it is described at least
The corresponding primary vector collection of one keyword, the primary vector collection include M dimension primary vector；And it is the text information is defeated
Enter the default neural network model, obtains secondary vector collection corresponding with the text information, the secondary vector collection includes M
Tie up secondary vector, wherein M is the integer greater than 0；
Similarity determining unit, for determining the primary vector collection and described the according to the default neural network model
The similarity of two vector sets；
Push unit, for when the similarity is greater than or equal to preset threshold, the similarity to be greater than or equal to
Second Internet resources corresponding to the secondary vector collection of the preset threshold push to the user terminal.
The third aspect, the embodiment of the present invention provide a kind of content recommendation service device, the content recommendation service device and user
Terminal communication connection, the content recommendation service device include:
Storage unit；
Processing unit；And
Content recommendation device is stored in the storage unit including one or more and is executed by the processing unit
Software function module, the content recommendation device include:
Acquiring unit clicks the first net checked by the user terminal in the first preset period of time for obtaining user
At least one corresponding keyword of network resource, and obtain text information corresponding to a plurality of second Internet resources；
Input unit, at least one described keyword to be inputted default neural network model, obtain with it is described at least
The corresponding primary vector collection of one keyword, the primary vector collection include M dimension primary vector；And it is the text information is defeated
Enter the default neural network model, obtains secondary vector collection corresponding with the text information, the secondary vector collection includes M
Tie up secondary vector, wherein M is the integer greater than 0；
Similarity determining unit, for determining the primary vector collection and described the according to the default neural network model
The similarity of two vector sets；
Push unit, for when the similarity is greater than or equal to preset threshold, the similarity to be greater than or equal to
Second Internet resources corresponding to the secondary vector collection of the preset threshold push to the user terminal.
In terms of existing technologies, content recommendation method provided by the invention, device and content recommendation service device be at least
Have the advantages that this method was checked in the first preset period of time by user terminal click by collecting user
At least one corresponding keyword of first network resource, and text information corresponding to a plurality of second Internet resources is obtained, so
Keyword is converted into primary vector collection using default neural network model afterwards, and text information is converted into secondary vector
Collection；Then determine the similarity of primary vector collection and secondary vector collection, and similarity in a plurality of second Internet resources is greater than or
Internet resources corresponding equal to preset threshold push to user terminal.Based on this, avoid because pushing caused by according to tag match
The single problem of content, so that the content for recommending user is more abundant.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, the embodiment of the present invention is cited below particularly, and match
Appended attached drawing is closed, is described in detail below.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described.It should be appreciated that the following drawings illustrates only certain embodiments of the present invention, therefore it is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 is that the interaction of content recommendation service device provided in an embodiment of the present invention, user terminal and content source server is shown
It is intended to.
Fig. 2 is the block diagram of content recommendation service device provided in an embodiment of the present invention.
Fig. 3 is the flow diagram of content recommendation method provided in an embodiment of the present invention.
Fig. 4 is the block diagram of content recommendation device provided in an embodiment of the present invention.
Icon: 10- content recommendation service device；11- processing unit；12- communication unit；13- storage unit；20- user is whole
End；30- content source server；100- content recommendation device；110- acquiring unit；120- input unit；130- similarity determines
Unit；140- push unit.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description.Obviously, described embodiment is only a part of the embodiments of the present invention, instead of all the embodiments.It is logical
The component for the embodiment of the present invention being often described and illustrated herein in the accompanying drawings can be arranged and be designed with a variety of different configurations.
Therefore, the detailed description of the embodiment of the present invention provided in the accompanying drawings is not intended to limit below claimed
The scope of the present invention, but be merely representative of selected embodiment of the invention.Based on the embodiment of the present invention, those skilled in the art
Member's every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
It should also be noted that similar label and letter indicate similar terms in following attached drawing, therefore, once a certain Xiang Yi
It is defined in a attached drawing, does not then need that it is further defined and explained in subsequent attached drawing.In addition, term " the
One ", " second " etc. is only used for distinguishing description, is not understood to indicate or imply relative importance.
With reference to the accompanying drawing, it elaborates to some embodiments of the present invention.In the absence of conflict, following
Feature in embodiment and embodiment can be combined with each other.
Fig. 1 is please referred to, is content recommendation service device 10 provided in an embodiment of the present invention, user terminal 20 and content source service
The interaction schematic diagram of device 30.Content recommendation service device 10 provided by the invention can be communicated with user terminal 20 by network foundation
Connection, to carry out data interaction, content recommendation service device 10 can also be established by network with data source server and be communicated to connect,
To carry out data interaction.Wherein, content recommendation service device 10 can be liked according to user, obtain phase from content source server 30
The Internet resources answered simultaneously push to user terminal 20.The form of push can be Web page push link, allow user from
Family terminal 20 views the shorthand information of Internet resources, if user clicks the link of push, will get detailed network money
Source.It is (such as all kinds of new that its Internet resources includes, but are not limited to text, picture, video, application program, the platform of delivery network resource
Hear website) etc..
It is worth noting that content recommendation service device 10 can be previously stored with Internet resources, it is also possible to from content source
Server 30 gets Internet resources.That is, in other embodiments, content recommendation service device 10 also can have content source clothes
Then the function of business device 30 carries out commending contents further according to the Internet resources of storage for storing Internet resources.
Further, user terminal 20 may be, but not limited to, smart phone, PC (personal
Computer, PC), tablet computer, personal digital assistant (personal digital assistant, PDA), mobile Internet access set
Standby (mobile Internet device, MID) etc..Content source server 30 may be, but not limited to, Cloud Server, cluster clothes
Business device, distributed server etc..Network may be, but not limited to, cable network or wireless network.
It referring to figure 2., is the block diagram of content recommendation service device 10 provided in an embodiment of the present invention.In the present embodiment
In, content recommendation service device 10 may include processing unit 11, communication unit 12, storage unit 13 and content recommendation device
100, between processing unit 11, communication unit 12, storage unit 13 and each element of content recommendation device 100 directly or indirectly
Ground is electrically connected, to realize the transmission or interaction of data.For example, these elements between each other can be total by one or more communication
Line or signal wire, which are realized, to be electrically connected.
Processing unit 11 may be a kind of IC chip, the processing capacity with signal.Above-mentioned processing unit 11 can
To be general processor.For example, the processor can be central processing unit (Central Processing Unit, CPU), figure
Shape processor (Graphics Processing Unit, GPU), network processing unit (Network Processor, NP) etc.；Also
Can be digital signal processor (DSP), specific integrated circuit (ASIC), field programmable gate array (FPGA) or other can
Programmed logic device, discrete gate or transistor logic, discrete hardware components.It may be implemented or execute present invention implementation
Disclosed each method, step and logic diagram in example.
Communication unit 12 is used to establish content recommendation service device 10 and user terminal 20 and content source server by network
30 communication connection, and pass through network sending and receiving data.
Storage unit 13 may be, but not limited to, random access memory, read-only memory, programmable read only memory,
Erasable Programmable Read Only Memory EPROM, electrically erasable programmable read-only memory etc..In the present embodiment, storage unit 13 can be with
For storing default neural network model.Certainly, storage unit 13 can be also used for storage program, and processing unit 11 is receiving
After executing instruction, the program is executed.
Further, content recommendation device 100 includes that at least one can be deposited in the form of software or firmware (firmware)
The software for being stored in storage unit 13 or being solidificated in 10 operating system of content recommendation service device (operating system, OS)
Functional module.Processing unit 11 is for executing the executable module stored in storage unit 13, such as 100 institute of content recommendation device
Including software function module and computer program etc..
It is understood that structure shown in Fig. 2 is only a kind of structural schematic diagram of content recommendation service device 10, content is pushed away
Recommending server 10 can also include than more or fewer components shown in Fig. 2.Each component shown in Fig. 2 can using hardware,
Software or combinations thereof is realized.
It referring to figure 3., is the flow diagram of content recommendation method provided in an embodiment of the present invention.In provided by the invention
Holding recommended method can be applied to above-mentioned content recommendation service device 10, execute content recommendation method by content recommendation service device 10
Each step, can be avoided in the prior art because pushing the single problem of content caused by according to tag match, so that recommending use
The content at family is more abundant.
Each step of content recommendation method shown in Fig. 3 will be described in detail below, in the present embodiment, content pushes away
The method of recommending may comprise steps of:
Step S210 obtains user and clicks the first network resource checked by user terminal 20 in the first preset period of time
At least one corresponding keyword, and obtain text information corresponding to a plurality of second Internet resources.Wherein, first it is default when
Section can be configured according to the actual situation, be not especially limited here.
Wherein, the process for obtaining text information corresponding to a plurality of second Internet resources, may include: content recommendation service
Device 10 is previously stored with a plurality of second Internet resources, and content recommendation service device 10 can get more from the storage unit 13 of itself
The second Internet resources of item；Alternatively, content recommendation service device 10 can obtain 30 institute of content source server from content source server 30
Second Internet resources of storage.
Before step S210, method can also include:
Learning training, the nerve net that will be obtained after training are carried out to neural network model based on the sample dictionary constructed in advance
Network model is as default neural network model.
The default neural network model may include word2vec word incorporation model, Recognition with Recurrent Neural Network model etc..
At least one keyword is inputted default neural network model, obtained and at least one keyword pair by step S220
The primary vector collection answered, primary vector collection include M dimension primary vector；And text information is inputted into default neural network model,
Secondary vector collection corresponding with text information is obtained, secondary vector collection includes M dimension secondary vector, wherein M is the integer greater than 0.
Optionally, at least one keyword is inputted into default neural network model, obtained corresponding at least one keyword
Primary vector collection the step of, may include: that at least one keyword is inputted into default neural network model, preset neural network
Model is based at least one based on the determining same or similar word of meaning at least one keyword of at least one keyword
Keyword and primary vector collection is determined with the same or similar word of meaning of at least one keyword.Wherein, true based on keyword
The process of fixed and the keyword the same or similar word of meaning, can be the meaning vector in advance by word all in dictionary
Change, then match the same or similar word of vector corresponding with the keyword, the same or similar word being matched to is and this
The same or like word of the meaning of keyword.Based on this, it can avoid the occurrence of and in the prior art be made according to tag match
At the single problem of content.
The step of text information is inputted default neural network model, obtains secondary vector collection corresponding with text information,
May include:
Text information is inputted into default neural network model, by default neural network model by text information split with
Multiple keywords are obtained, determine secondary vector collection based on obtained multiple keywords are split.
Step S230 determines the similarity of primary vector collection Yu secondary vector collection according to default neural network model.
Similarity is greater than or equal to the of preset threshold when similarity is greater than or equal to preset threshold by step S240
Second Internet resources corresponding to two vector sets push to user terminal 20.Wherein, preset threshold can according to the actual situation into
Row setting, is not especially limited here.
Step S240 can also include: to choose from the Internet resources that the first platform of publication first network resource is issued
Meet Internet resources corresponding to secondary vector collection of the similarity more than or equal to preset threshold, using as the second Internet resources,
And the second Internet resources are pushed into user terminal 20.Based on this, the content resource of identical platform can be made to have more
Chance for exposure, while the Internet resources for being pushed to user terminal 20 can be enriched.Understandably, which can be virtual flat
Platform, such as news website or other websites.
Step S240 may include: that the URL information of the second Internet resources corresponding with secondary vector collection is pushed to user
Terminal 20.Wherein, URL is the abbreviation (Uniform Resource Locator, URL) of uniform resource locator.
Based on above-mentioned design, the present embodiment can be by the text feature direct vectorization of article (Internet resources), Yong Hute
The integrity degree of article of seeking peace feature is retained, eliminate it is characteristic labeling in the prior art after carry out causing in matching process again
Information lose, increase accuracy in computation.In addition, by the article characteristic aggregation of same content source to content source itself, later
The correlation between user vector and content source vector is carried out again, relative to article and End-user relevance is calculated, can substantially be mentioned
Rise computational efficiency.Furthermore using the content source set and article calculated when recommending to user, make to recommend the interior of user
Hold and more enrich, avoids the homogeneity of the content according to caused by tag match merely.By to user's recommendation source set and
Other articles that these platforms are issued make content source relative to before, there has also been more chances for exposure, peomote interior
The dispatch enthusiasm of appearance source publisher.
It referring to figure 4., is the block diagram of content recommendation device 100 provided in an embodiment of the present invention.The present invention is implemented
The content recommendation device 100 that example provides can be applied to above-mentioned content recommendation service device 10, can be used for executing commending contents
Each step of method, to avoid in the prior art because pushing the single problem of content caused by according to tag match, so that recommending
The content of user is more abundant.Wherein, the content recommendation device 100 may include acquiring unit 110, it is input unit 120, similar
Spend determination unit 130 and push unit 140.
Acquiring unit 110 clicks check first by user terminal 20 in the first preset period of time for obtaining user
At least one corresponding keyword of Internet resources, and obtain text information corresponding to a plurality of second Internet resources.
Acquiring unit 110 clicks the first network checked by user terminal 20 in the first preset period of time in acquisition user
Before at least one corresponding keyword of resource, content recommendation device 100 can also include training unit, be used for: based on preparatory
The sample dictionary of building carries out learning training to neural network model, using the neural network model obtained after training as default mind
Through network model.
Input unit 120 obtains closing at least one at least one keyword to be inputted default neural network model
The corresponding primary vector collection of keyword, primary vector collection include M dimension primary vector；And text information is inputted into default neural network
Model, obtains secondary vector collection corresponding with text information, and secondary vector collection includes M dimension secondary vector, wherein M is greater than 0
Integer.
Input unit 120 can be also used for: at least one keyword being inputted default neural network model, presets nerve net
Network model is based at least one based on the determining same or similar word of meaning at least one keyword of at least one keyword
A keyword and primary vector collection is determined with the same or similar word of meaning of at least one keyword.
Input unit 120 can be also used for, and text information be inputted default neural network model, by presetting neural network mould
Type splits text information to obtain multiple keywords, determines secondary vector collection based on obtained multiple keywords are split.
Similarity determining unit 130, for determining primary vector collection and secondary vector collection according to default neural network model
Similarity.
Push unit 140, for when similarity is greater than or equal to preset threshold, similarity to be greater than or equal to default threshold
Second Internet resources corresponding to the secondary vector collection of value push to user terminal 20.
Push unit 140 can be also used for: the URL information of the second Internet resources corresponding with secondary vector collection is pushed to
User terminal 20.
Push unit 140 can be also used for: from the Internet resources that the first platform of publication first network resource is issued
It chooses and meets Internet resources corresponding to secondary vector collection of the similarity more than or equal to preset threshold, to be provided as the second network
Source, and the second Internet resources are pushed into user terminal 20.
The embodiment of the present invention also provides a kind of computer readable storage medium.Calculating is stored in the readable storage medium storing program for executing
Machine program, when the computer program is run on computers, so that the computer executes described in above-described embodiment
Content recommendation method.
Through the above description of the embodiments, those skilled in the art can be understood that the present invention can lead to
Hardware realization is crossed, the mode of necessary general hardware platform can also be added to realize by software, based on this understanding, this hair
Bright technical solution can be embodied in the form of software products, which can store in a non-volatile memories
In medium (can be CD-ROM, USB flash disk, mobile hard disk etc.), including some instructions are used so that a computer equipment (can be
Personal computer, server or network equipment etc.) execute method described in each implement scene of the present invention.
As stated above, the present invention provides a kind of content recommendation method, device and content recommendation service device.This method passes through receipts
Collect user and at least one the corresponding keyword of first network resource checked clicked by user terminal in the first preset period of time,
And text information corresponding to a plurality of second Internet resources is obtained, then keyword is converted using default neural network model
For primary vector collection, and text information is converted into secondary vector collection；Then primary vector collection and secondary vector collection are determined
Similarity, and similarity in a plurality of second Internet resources is greater than or equal to the corresponding Internet resources of preset threshold and pushes to user
Terminal.Based on this, avoid because pushing the single problem of content caused by according to tag match, so that recommending the content of user more
Add abundant.
In embodiment provided by the present invention, it should be understood that disclosed devices, systems, and methods can also lead to
Other modes are crossed to realize.Devices, systems, and methods embodiment described above is only schematical, for example, in attached drawing
Flow chart and block diagram show that the system of multiple embodiments according to the present invention, the possibility of method and computer program product are real
Existing architecture, function and operation.In this regard, each box in flowchart or block diagram can represent module, a journey
A part of sequence section or code, a part of the module, section or code include one or more for realizing defined
The executable instruction of logic function.It should also be noted that in some implementations as replacement, function marked in the box
It can also occur in a different order than that indicated in the drawings.For example, two continuous boxes can actually be substantially in parallel
It executes, they can also be executed in the opposite order sometimes, and this depends on the function involved.It is also noted that block diagram and/
Or the combination of each box in flow chart and the box in block diagram and or flow chart, can with execute as defined in function or
The dedicated hardware based system of movement is realized, or can be realized using a combination of dedicated hardware and computer instructions.
In addition, each functional module in each embodiment of the present invention can integrate one independent part of formation together, it can also be with
It is modules individualism, an independent part can also be integrated to form with two or more modules.
It can replace, can be realized wholly or partly by software, hardware, firmware or any combination thereof.When
When using software realization, can entirely or partly it realize in the form of a computer program product.The computer program product
Including one or more computer instructions.It is all or part of when loading on computers and executing the computer program instructions
Ground is generated according to process or function described in the embodiment of the present invention.The computer can be general purpose computer, special purpose computer,
Computer network or other programmable devices.The computer instruction may be stored in a computer readable storage medium, or
Person is transmitted from a computer readable storage medium to another computer readable storage medium, for example, the computer instruction
Wired (such as coaxial cable, optical fiber, digital subscriber can be passed through from a web-site, computer, server or data center
Line (DSL)) or wireless (such as infrared, wireless, microwave etc.) mode to another web-site, computer, server or data
It is transmitted at center.The computer readable storage medium can be any usable medium that computer can access and either wrap
The data storage devices such as server, the data center integrated containing one or more usable mediums.The usable medium can be magnetic
Property medium, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or semiconductor medium (such as solid state hard disk
Solid State Disk (SSD)) etc..
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of content recommendation method, which is characterized in that be applied to content recommendation service device, the content recommendation service device and use
The communication connection of family terminal, which comprises
It is corresponding at least to obtain the first network resource that user is checked in the first preset period of time by user terminal click
One keyword, and obtain text information corresponding to a plurality of second Internet resources；
At least one described keyword is inputted into default neural network model, obtains corresponding at least one described keyword the
One vector set, the primary vector collection include M dimension primary vector；And the text information is inputted into the default neural network
Model, obtains secondary vector collection corresponding with the text information, and the secondary vector collection includes M dimension secondary vector, wherein M
For the integer greater than 0；
The similarity of the primary vector collection Yu the secondary vector collection is determined according to the default neural network model；
When the similarity is greater than or equal to preset threshold, the similarity is greater than or equal to the second of the preset threshold
Second Internet resources corresponding to vector set push to the user terminal.
2. the method according to claim 1, wherein described will the default nerve of at least one keyword input
Network model obtains primary vector collection corresponding at least one described keyword, comprising:
At least one described keyword is inputted into default neural network model, the default neural network model be based on it is described at least
The determining same or similar word of meaning at least one keyword of one keyword, and based at least one described key
Word and the primary vector collection is determined with the same or similar word of meaning of at least one keyword.
3. the method according to claim 1, wherein described input the default nerve net for the text information
Network model obtains secondary vector collection corresponding with the text information, comprising:
The text information is inputted into the default neural network model, by the default neural network model by the text envelope
Breath is split to obtain multiple keywords, determines the secondary vector collection based on obtained multiple keywords are split.
4. the method according to claim 1, wherein the acquisition user is in the first preset period of time by described
Before user terminal clicks at least one the corresponding keyword of first network resource checked, the method also includes:
Learning training, the neural network mould that will be obtained after training are carried out to neural network model based on the sample dictionary constructed in advance
Type is as the default neural network model.
5. the method according to claim 1, wherein the Internet resources include text information, pictorial information, view
At least one of frequency information.
6. the method according to claim 1, wherein described be greater than or equal to preset threshold for the similarity
Second Internet resources corresponding to secondary vector collection push to the user terminal, comprising:
The URL information of the second Internet resources corresponding with the secondary vector collection is pushed into the user terminal.
7. the method according to claim 1, wherein the default neural network model includes word incorporation model.
8. the method according to claim 1, wherein described be greater than or equal to the default threshold for the similarity
Second Internet resources corresponding to the secondary vector collection of value push to the user terminal, comprising:
Selection meets the similarity and is greater than from the Internet resources that the first platform for issuing the first network resource is issued
Or Internet resources corresponding to the secondary vector collection equal to the preset threshold, using as second Internet resources, and by institute
It states the second Internet resources and pushes to the user terminal.
9. a kind of content recommendation device, which is characterized in that be applied to content recommendation service device, the content recommendation service device and use
The communication connection of family terminal, described device include:
Acquiring unit clicks the first network checked money by the user terminal in the first preset period of time for obtaining user
At least one corresponding keyword of source, and obtain text information corresponding to a plurality of second Internet resources；
Input unit, at least one described keyword to be inputted default neural network model, obtain with it is described at least one
The corresponding primary vector collection of keyword, the primary vector collection include M dimension primary vector；And the text information is inputted into institute
Default neural network model is stated, obtains secondary vector collection corresponding with the text information, the secondary vector collection includes M dimension the
Two vectors, wherein M is the integer greater than 0；
Similarity determining unit, for according to the default neural network model determine the primary vector collection with described second to
The similarity of quantity set；
Push unit, for the similarity being greater than or equal to described when the similarity is greater than or equal to preset threshold
Second Internet resources corresponding to the secondary vector collection of preset threshold push to the user terminal.
10. a kind of content recommendation service device, which is characterized in that the content recommendation service device is connect with user terminal communication, institute
Stating content recommendation service device includes:
Storage unit；
Processing unit；And
Content recommendation device, the software for being stored in the storage unit and being executed by the processing unit including one or more
Functional module, the content recommendation device include:
Acquiring unit clicks the first network checked money by the user terminal in the first preset period of time for obtaining user
At least one corresponding keyword of source, and obtain text information corresponding to a plurality of second Internet resources；
Input unit, at least one described keyword to be inputted default neural network model, obtain with it is described at least one
The corresponding primary vector collection of keyword, the primary vector collection include M dimension primary vector；And the text information is inputted into institute
Default neural network model is stated, obtains secondary vector collection corresponding with the text information, the secondary vector collection includes M dimension the
Two vectors, wherein M is the integer greater than 0；
Similarity determining unit, for according to the default neural network model determine the primary vector collection with described second to
The similarity of quantity set；
Push unit, for the similarity being greater than or equal to described when the similarity is greater than or equal to preset threshold
Second Internet resources corresponding to the secondary vector collection of preset threshold push to the user terminal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811087626.5A CN109190046A (en) | 2018-09-18 | 2018-09-18 | Content recommendation method, device and content recommendation service device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811087626.5A CN109190046A (en) | 2018-09-18 | 2018-09-18 | Content recommendation method, device and content recommendation service device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109190046A true CN109190046A (en) | 2019-01-11 |
Family
ID=64911981
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811087626.5A Pending CN109190046A (en) | 2018-09-18 | 2018-09-18 | Content recommendation method, device and content recommendation service device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109190046A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110162701A (en) * | 2019-05-10 | 2019-08-23 | 腾讯科技（深圳）有限公司 | Content delivery method, device, computer equipment and storage medium |
| CN111221936A (en) * | 2020-01-02 | 2020-06-02 | 中科鼎富（北京）科技发展有限公司 | Information matching method and device, electronic equipment and storage medium |
| CN111538907A (en) * | 2020-06-05 | 2020-08-14 | 支付宝(杭州)信息技术有限公司 | Object recommendation method, system and device |
| CN111652664A (en) * | 2019-03-04 | 2020-09-11 | 富士通株式会社 | Apparatus and method for training mixed element learning network |
| CN111914079A (en) * | 2020-08-07 | 2020-11-10 | 上海梅斯医药科技有限公司 | Topic recommendation method and system based on user tags |
| CN112182460A (en) * | 2019-07-04 | 2021-01-05 | 腾讯科技（北京）有限公司 | Resource pushing method and device, storage medium and electronic device |
| CN112861001A (en) * | 2021-02-18 | 2021-05-28 | 咪咕数字传媒有限公司 | Method and device for generating recommendation value of digital content, electronic equipment and storage medium |
| CN113051480A (en) * | 2021-04-22 | 2021-06-29 | 深圳壹账通智能科技有限公司 | Resource pushing method and device, electronic equipment and storage medium |
| CN113360745A (en) * | 2020-03-04 | 2021-09-07 | 阿里巴巴集团控股有限公司 | Data object recommendation method and device and electronic equipment |
| CN113377928A (en) * | 2021-08-11 | 2021-09-10 | 明品云(北京)数据科技有限公司 | Text recommendation method, system, device and medium |
| CN113538108A (en) * | 2021-07-27 | 2021-10-22 | 北京沃东天骏信息技术有限公司 | Resource information determination method and device, electronic equipment and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007105909A1 (en) * | 2006-03-16 | 2007-09-20 | Nhn Corporation | Method for targeting web advertisement clickers based on click pattern by using a collaborative filtering system with neural networks and system thereof |
| CN102831234A (en) * | 2012-08-31 | 2012-12-19 | 北京邮电大学 | Personalized news recommendation device and method based on news content and theme feature |
| US20120323725A1 (en) * | 2010-12-15 | 2012-12-20 | Fourthwall Media | Systems and methods for supplementing content-based attributes with collaborative rating attributes for recommending or filtering items |
| CN104239512A (en) * | 2014-09-16 | 2014-12-24 | 电子科技大学 | Text recommendation method |
| CN104899273A (en) * | 2015-05-27 | 2015-09-09 | 东南大学 | Personalized webpage recommendation method based on topic and relative entropy |
| CN105279288A (en) * | 2015-12-04 | 2016-01-27 | 深圳大学 | Online content recommending method based on deep neural network |
| CN105631009A (en) * | 2015-12-25 | 2016-06-01 | 广州视源电子科技股份有限公司 | Word vector similarity based retrieval method and system |
-
2018
- 2018-09-18 CN CN201811087626.5A patent/CN109190046A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007105909A1 (en) * | 2006-03-16 | 2007-09-20 | Nhn Corporation | Method for targeting web advertisement clickers based on click pattern by using a collaborative filtering system with neural networks and system thereof |
| US20120323725A1 (en) * | 2010-12-15 | 2012-12-20 | Fourthwall Media | Systems and methods for supplementing content-based attributes with collaborative rating attributes for recommending or filtering items |
| CN102831234A (en) * | 2012-08-31 | 2012-12-19 | 北京邮电大学 | Personalized news recommendation device and method based on news content and theme feature |
| CN104239512A (en) * | 2014-09-16 | 2014-12-24 | 电子科技大学 | Text recommendation method |
| CN104899273A (en) * | 2015-05-27 | 2015-09-09 | 东南大学 | Personalized webpage recommendation method based on topic and relative entropy |
| CN105279288A (en) * | 2015-12-04 | 2016-01-27 | 深圳大学 | Online content recommending method based on deep neural network |
| CN105631009A (en) * | 2015-12-25 | 2016-06-01 | 广州视源电子科技股份有限公司 | Word vector similarity based retrieval method and system |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111652664A (en) * | 2019-03-04 | 2020-09-11 | 富士通株式会社 | Apparatus and method for training mixed element learning network |
| CN110162701B (en) * | 2019-05-10 | 2023-12-19 | 深圳市雅阅科技有限公司 | Content pushing method, device, computer equipment and storage medium |
| CN110162701A (en) * | 2019-05-10 | 2019-08-23 | 腾讯科技（深圳）有限公司 | Content delivery method, device, computer equipment and storage medium |
| CN112182460B (en) * | 2019-07-04 | 2024-03-15 | 深圳市雅阅科技有限公司 | Resource pushing method and device, storage medium and electronic device |
| CN112182460A (en) * | 2019-07-04 | 2021-01-05 | 腾讯科技（北京）有限公司 | Resource pushing method and device, storage medium and electronic device |
| CN111221936A (en) * | 2020-01-02 | 2020-06-02 | 中科鼎富（北京）科技发展有限公司 | Information matching method and device, electronic equipment and storage medium |
| CN111221936B (en) * | 2020-01-02 | 2023-11-07 | 鼎富智能科技有限公司 | Information matching method and device, electronic equipment and storage medium |
| CN113360745A (en) * | 2020-03-04 | 2021-09-07 | 阿里巴巴集团控股有限公司 | Data object recommendation method and device and electronic equipment |
| CN111538907B (en) * | 2020-06-05 | 2023-04-18 | 支付宝(杭州)信息技术有限公司 | Object recommendation method, system and device |
| CN111538907A (en) * | 2020-06-05 | 2020-08-14 | 支付宝(杭州)信息技术有限公司 | Object recommendation method, system and device |
| CN111914079A (en) * | 2020-08-07 | 2020-11-10 | 上海梅斯医药科技有限公司 | Topic recommendation method and system based on user tags |
| CN112861001A (en) * | 2021-02-18 | 2021-05-28 | 咪咕数字传媒有限公司 | Method and device for generating recommendation value of digital content, electronic equipment and storage medium |
| CN113051480A (en) * | 2021-04-22 | 2021-06-29 | 深圳壹账通智能科技有限公司 | Resource pushing method and device, electronic equipment and storage medium |
| CN113538108A (en) * | 2021-07-27 | 2021-10-22 | 北京沃东天骏信息技术有限公司 | Resource information determination method and device, electronic equipment and storage medium |
| CN113377928A (en) * | 2021-08-11 | 2021-09-10 | 明品云(北京)数据科技有限公司 | Text recommendation method, system, device and medium |
| CN113377928B (en) * | 2021-08-11 | 2022-05-27 | 明品云(北京)数据科技有限公司 | Text recommendation method, system, device and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109190046A (en) | Content recommendation method, device and content recommendation service device | |
| US11669579B2 (en) | Method and apparatus for providing search results | |
| US9141906B2 (en) | Scoring concept terms using a deep network | |
| WO2018192491A1 (en) | Information pushing method and device | |
| US9449271B2 (en) | Classifying resources using a deep network | |
| US11172040B2 (en) | Method and apparatus for pushing information | |
| CN107025216A (en) | Sentence extracting method and system | |
| CN113204691B (en) | Information display method, device, equipment and medium | |
| CN106330846A (en) | Cross-platform object recommendation method and device | |
| CN110363206B (en) | Clustering of data objects, data processing and data identification method | |
| CN109819002B (en) | Data pushing method and device, storage medium and electronic device | |
| CN113688310A (en) | Content recommendation method, device, equipment and storage medium | |
| CN110059172B (en) | Method and device for recommending answers based on natural language understanding | |
| CN110058992B (en) | Text template effect feedback method and device and electronic equipment | |
| CN110060075B (en) | Method, apparatus, system, and medium for predicting sales | |
| CN112650942A (en) | Product recommendation method, device, computer system and computer-readable storage medium | |
| CN112818213A (en) | Multimedia service data pushing method, device, equipment and storage medium | |
| CN112446214A (en) | Method, device and equipment for generating advertisement keywords and storage medium | |
| CN117216393A (en) | Information recommendation method, training method and device of information recommendation model and equipment | |
| CN114139052B (en) | Ranking model training method for intelligent recommendation, intelligent recommendation method and device | |
| CN110750708A (en) | Keyword recommendation method and device and electronic equipment | |
| WO2021204039A1 (en) | Method and apparatus for pushing information | |
| KR20210084641A (en) | Method and apparatus for transmitting information | |
| CN107483595A (en) | Information-pushing method and device | |
| CN110110199B (en) | Information output method and device |
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 |