CN110473042A - For obtaining the method and device of information - Google Patents
For obtaining the method and device of information Download PDFInfo
- Publication number
- CN110473042A CN110473042A CN201810446322.7A CN201810446322A CN110473042A CN 110473042 A CN110473042 A CN 110473042A CN 201810446322 A CN201810446322 A CN 201810446322A CN 110473042 A CN110473042 A CN 110473042A
- Authority
- CN
- China
- Prior art keywords
- information
- article
- sample
- web page
- recommended models
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the present application discloses the method and device for obtaining information.One specific embodiment of this method includes: the procedural information for obtaining user and browsing webpage, wherein above process information is used to characterize user and browses information corresponding when webpage;Above process information is imported to article recommended models trained in advance, obtains the prediction Item Information of corresponding process information, above-mentioned article recommended models are used to determine prediction Item Information by procedural information.This embodiment improves the accuracys for obtaining prediction Item Information.
Description
Technical field
The invention relates to field of computer technology, and in particular to for obtaining the method and device of information.
Background technique
With the development of network technology, more and more articles are sold by network.In general, user can be by each
Kind electronic equipment logs in relevant website, and browses the webpage of relative article information;Then, the article of needs is put into net by user
It stands shopping cart;Finally, user settles accounts the article in shopping cart, the purchase of article is completed.It, can by Online Shopping article
So that user obtains a large amount of Item Information with staying indoors, the efficiency that user obtains article is improved.Correspondingly, website skill
Art personnel can also provide relevant Item Information according to the purchaser record of user for user, further improve user and obtain object
The efficiency of product.
Summary of the invention
The embodiment of the present application proposes the method and device for obtaining information.
In a first aspect, the embodiment of the present application provides a kind of method for obtaining information, this method comprises: obtaining user
Browse the procedural information of webpage, wherein above process information is used to characterize user and browses information corresponding when webpage;It will be above-mentioned
Procedural information imports article recommended models trained in advance, obtains the prediction Item Information of corresponding process information, above-mentioned article pushes away
Model is recommended for determining prediction Item Information by procedural information.
In some embodiments, above process information includes the browsing characteristic information of at least one webpage and corresponding webpage,
Above-mentioned browsing characteristic information include at least one of the following: user browse webpage the information content and corresponding informance content browsing when
Between.
In some embodiments, above-mentioned that above process information is imported to article recommended models trained in advance, it is corresponded to
The prediction Item Information of procedural information, comprising: by above process information input to above-mentioned convolutional neural networks, obtain the above process
The corresponding web page characteristics vector of information, wherein above-mentioned convolutional neural networks for characterize procedural information and web page characteristics vector it
Between corresponding relationship;Above-mentioned web page characteristics vector is input to above-mentioned Recognition with Recurrent Neural Network, obtains web page contents feature vector,
In, above-mentioned Recognition with Recurrent Neural Network is used to characterize the corresponding relationship between web page characteristics vector and web page contents feature vector, webpage
Content feature vector is used to characterize the incidence relation between web page characteristics vector;Above-mentioned web page contents feature vector is input to
Full articulamentum is stated, the corresponding prediction Item Information of above process information is obtained, wherein above-mentioned full articulamentum is for characterizing in webpage
Hold feature vector and predicts the corresponding relationship between Item Information.
In some embodiments, training obtains above-mentioned article recommended models as follows: obtaining user and passes through webpage
Each sample mesh in the sample object Item Information and above-mentioned multiple sample object articles of multiple sample object articles of selection
Mark the corresponding sample processes information of article;By the corresponding sample of each sample object article in above-mentioned multiple sample object articles
Procedural information is as input, using the sample Item Information of the corresponding sample object article of the sample processes information as output, instruction
Get above-mentioned article recommended models.
In some embodiments, the above-mentioned corresponding sample of each sample object article by above-mentioned multiple sample object articles
This procedural information as input, using the sample Item Information of the corresponding sample object article of the sample processes information as export,
Training obtains above-mentioned article recommended models, comprising: executes following training step: by each sample in multiple sample object articles
The corresponding sample processes information of target item is sequentially input to initialization article recommended models, is obtained corresponding to sample processes information
Prediction target item information, by prediction target item information corresponding to each sample processes information and the sample processes information
Corresponding sample object Item Information is compared, and obtains the predictablity rate of above-mentioned initialization article recommended models, is determined
Whether above-mentioned predictablity rate is greater than default accuracy rate threshold value, if more than above-mentioned default accuracy rate threshold value, then by above-mentioned initialization
The article recommended models that article recommended models are completed as training.
In some embodiments, the above-mentioned corresponding sample of each sample object article by above-mentioned multiple sample object articles
This procedural information as input, using the sample Item Information of the corresponding sample object article of the sample processes information as export,
Training obtains above-mentioned article recommended models, further includes: in response to being not more than above-mentioned default accuracy rate threshold value, adjusts above-mentioned initialization
The parameter of article recommended models, and continue to execute above-mentioned training step.
Second aspect, the embodiment of the present application provide it is a kind of for obtaining the device of information, the device include: information receive
Unit is configured to obtain the procedural information that user browses webpage, wherein above process information browses webpage for characterizing user
When corresponding information;Information acquisition unit is configured to importing above process information into article recommended models trained in advance,
The prediction Item Information of corresponding process information is obtained, above-mentioned article recommended models are used to determine prediction article letter by procedural information
Breath.
In some embodiments, above process information includes the browsing characteristic information of at least one webpage and corresponding webpage,
Above-mentioned browsing characteristic information include at least one of the following: user browse webpage the information content and corresponding informance content browsing when
Between.
In some embodiments, above- mentioned information acquiring unit include: web page characteristics vector obtain subelement, be configured to by
Above process information input obtains the corresponding web page characteristics vector of above process information to above-mentioned convolutional neural networks, wherein on
Convolutional neural networks are stated for characterizing the corresponding relationship between procedural information and web page characteristics vector;Web page contents feature vector obtains
Take subelement, be configured to above-mentioned web page characteristics vector being input to above-mentioned Recognition with Recurrent Neural Network, obtain web page contents feature to
Amount, wherein above-mentioned Recognition with Recurrent Neural Network is used to characterize the corresponding relationship between web page characteristics vector and web page contents feature vector,
Web page contents feature vector is used to characterize the incidence relation between web page characteristics vector;Predict that Item Information obtains subelement, quilt
It is configured to above-mentioned web page contents feature vector being input to above-mentioned full articulamentum, obtains the corresponding prediction article of above process information
Information, wherein above-mentioned full articulamentum is used to characterize web page contents feature vector and predicts the corresponding relationship between Item Information.
In some embodiments, above-mentioned apparatus includes article recommended models training unit, above-mentioned article recommended models training
Unit includes: sample acquisition subelement, is configured to obtain the sample for multiple sample object articles that user is selected by webpage
The corresponding sample processes information of each sample object article in target item information and above-mentioned multiple sample object articles;Article
Recommended models train subelement, are configured to the corresponding sample of each sample object article in above-mentioned multiple sample object articles
This procedural information as input, using the sample Item Information of the corresponding sample object article of the sample processes information as export,
Training obtains above-mentioned article recommended models.
In some embodiments, above-mentioned article recommended models training subelement includes: article recommended models training module, quilt
It is configured to sequentially input the corresponding sample processes information of each sample object article in multiple sample object articles to initial
Compound product recommended models obtain prediction target item information corresponding to sample processes information, by each sample processes information institute
Corresponding prediction target item information is compared with sample object Item Information corresponding to the sample processes information, is obtained
The predictablity rate for stating initialization article recommended models, determines whether above-mentioned predictablity rate is greater than default accuracy rate threshold value, if
Greater than above-mentioned default accuracy rate threshold value, then using above-mentioned initialization article recommended models as the article recommended models of training completion.
In some embodiments, above-mentioned article recommended models training subelement includes: parameter adjustment module, is configured to ring
Ying Yu is not more than above-mentioned default accuracy rate threshold value, adjusts the parameter of above-mentioned initialization article recommended models, and continue to execute above-mentioned
Training step.
The third aspect, the embodiment of the present application provide a kind of electronic equipment, comprising: one or more processors;Memory,
One or more programs are stored thereon with, when said one or multiple programs are executed by said one or multiple processors, are made
It obtains said one or multiple processors executes the method for obtaining information of above-mentioned first aspect.
Fourth aspect, the embodiment of the present application provide a kind of computer-readable medium, are stored thereon with computer program,
It is characterized in that, which realizes the method for obtaining information of above-mentioned first aspect when being executed by processor.
The process that the method and device provided by the embodiments of the present application for being used to obtain information, first acquisition user browse webpage
Information;Then procedural information is imported into article recommended models, obtains prediction Item Information, improved and obtain prediction Item Information
Accuracy.
Detailed description of the invention
By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, the application's is other
Feature, objects and advantages will become more apparent upon:
Fig. 1 is that this application can be applied to exemplary system architecture figures therein;
Fig. 2 is the flow chart according to one embodiment of the method for obtaining information of the application;
Fig. 3 is the flow chart according to one embodiment of the article recommended models training method of the application;
Fig. 4 is the schematic diagram according to an application scenarios of the method for obtaining information of the application;
Fig. 5 is the structural schematic diagram according to one embodiment of the device for obtaining information of the application;
Fig. 6 is adapted for the structural schematic diagram for the computer system for realizing the server of the embodiment of the present application.
Specific embodiment
The application is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to
Convenient for description, part relevant to related invention is illustrated only in attached drawing.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
Fig. 1 is shown can the method for obtaining information using the embodiment of the present application or the device for obtaining information
Exemplary system architecture 100.
As shown in Figure 1, system architecture 100 may include terminal device 101,102,103, network 104 and server 105.
Network 104 between terminal device 101,102,103 and server 105 to provide the medium of communication link.Network 104 can be with
Including various connection types, such as wired, wireless communication link or fiber optic cables etc..
User can be used terminal device 101,102,103 and be interacted by network 104 with server 105, to receive or send out
Send message etc..Various telecommunication customer end applications can be installed, such as web browser is answered on terminal device 101,102,103
With, shopping class application, searching class application, instant messaging tools etc..
Terminal device 101,102,103 can be hardware, be also possible to software.When terminal device 101,102,103 is hard
When part, it can be the various electronic equipments with display screen and supported web page browsing, including but not limited to smart phone, plate
Computer, pocket computer on knee and desktop computer etc..When terminal device 101,102,103 is software, can install
In above-mentioned cited electronic equipment.Multiple softwares or software module may be implemented into (such as providing distributed clothes in it
Business), single software or software module also may be implemented into.It is not specifically limited herein.
Server 105 can be to provide the server of various services, for example, terminal device 101,102,103 is sent with
The corresponding procedural information of webpage carries out the server of data processing.Server can carry out the data such as the procedural information received
The processing such as analysis, and processing result (such as prediction Item Information) is fed back into terminal device.
It should be noted that the method provided by the embodiment of the present application for obtaining information is generally held by server 105
Row, correspondingly, the device for obtaining information is generally positioned in server 105.
It should be noted that server can be hardware, it is also possible to software.When server is hardware, may be implemented
At the distributed server cluster that multiple servers form, individual server also may be implemented into.It, can when server is software
To be implemented as multiple softwares or software module (such as providing Distributed Services), single software or software also may be implemented into
Module.It is not specifically limited herein.
It should be understood that the number of terminal device, network and server in Fig. 1 is only schematical.According to realization need
It wants, can have any number of terminal device, network and server.
With continued reference to Fig. 2, the process of one embodiment of the method for obtaining information according to the application is shown
200.This be used for obtain information method the following steps are included:
Step 201, the procedural information that user browses webpage is obtained.
In the present embodiment, for obtain the method for information executing subject (such as terminal device shown in FIG. 1 101,
102,103) it by wired connection mode or radio connection from user using its terminal for carrying out web page browsing can be connect
Receive procedural information.Wherein, above process information is used to characterize user and browses information corresponding when webpage.It should be pointed out that
Above-mentioned radio connection can include but is not limited to 3G/4G connection, WiFi connection, bluetooth connection, WiMAX connection, Zigbee
Connection, UWB (ultra wideband) connection and other currently known or exploitation in the future radio connections.
In the prior art, when web technology personnel provide a user Item Information, usually directly with the purchase of user's history
It buys and is recorded as according to providing a user relevant Item Information.And in practice, user is usually in browsing images of items, checks
It is just final after a series of relevant operations such as certain specific parameters of article to determine the article needed.These a series of operations
It is able to reflect the focus that user selects article, directly influences the article of user's final choice.Existing offer Item Information
Method can not embody user select article during details, be not easy according to user browsing webpage process (above-mentioned one
Series relevant operation) determine user's really interested article.Therefore, existing method provide Item Information and user most
The article correlation selected eventually is little, and the accuracy of Item Information is not high.
For this purpose, the application, when monitoring that user browses webpage, available user browses the procedural information of webpage.Its
In, procedural information is used to characterize user's information for being checked during the browsing webpage to final choice article.That is, the application
Procedural information can be used for characterizing corresponding information when user browses webpage.
In some optional implementations of the present embodiment, above process information includes at least one webpage and corresponding net
The browsing characteristic information of page, the browsing characteristic information include at least one of the following: that user browses the information content of webpage and right
Answer the browsing time of the information content.
The browsing characteristic information of the application may include at least one: the information content and corresponding informance of user's browsing webpage
The browsing time etc. of content.For example, user logs in a certain shopping website, start the Item Information for searching for a certain needs.User beats
A certain article webpage is opened, has browsed the parameter of article first X minutes, has then checked user's evaluation X minutes again.Later, it opens
Other at least one article webpages of same type article, first check user's evaluation X minutes, then check item parameter X minutes
And item price X minutes.Then, during user inquires Item Information herein, the webpage that browsed, each webpage checked
Item parameter, user's evaluation, item price, and the information such as browsing time of corresponding item parameter, user's evaluation, item price
Procedural information described herein can be formed.Wherein, the information such as item parameter, user's evaluation, item price can be net
The information content of page.Corresponding item parameter, user's evaluation, item price browsing time when can be the browsing of the information content
Between.Wherein, executing subject can monitor the display content of screen to determine the information content of webpage that user is browsing;And it will
Screen shows browsing time of the time of corresponding informance content as the information content.The information content of each webpage and corresponding letter
The browsing time of breath content can be the browsing characteristic information of the corresponding webpage.Browsing characteristic information can characterize user's concern
Which feature of article can determine that user selects the focus of article by these features, subsequent to be paid close attention to based on these
Point predicts the article of user's final choice.
Step 202, above process information is imported to article recommended models trained in advance, obtains the pre- of corresponding process information
Survey Item Information.
In the present embodiment, after the procedural information for getting browsing webpage, executing subject can import procedural information pre-
First trained article recommended models, obtain the prediction Item Information of corresponding process information.Wherein, user is in browsing relative article letter
When the webpage of breath, the information content and browsing information of browsing sequence, browsing webpage between the corresponding webpage of different articles
The information such as the time of content may all influence the article of user's final choice.Therefore, when can browse multiple webpages for user
The sequence of the information content of browsing sequence, browsing webpage between webpage, and the browsing time of the browsing information content assign not
Same weight.All weights for belonging to a certain webpage (corresponding a certain article) are added up to total weight as the webpage.Weight
It is bigger, illustrate that a possibility that user selects the webpage of the weight corresponding article is higher.For example, user browses the associated nets of mobile phone
When page, what is browsed first is the webpage of certain Mobile phone of A brand.During browsing webpage, user has browsed hand first
The product introduction part of machine has browsed user's evaluation again later.After user has browsed user's evaluation, and the another of B brand is checked
Mobile phone has browsed user's evaluation first, has browsed product introduction and price later.Executing subject can count A brand
Total weight of the webpage of the mobile phone of total weight and B brand of the webpage of mobile phone makees the corresponding cellphone information of the big webpage of weight
To predict Item Information.
In the present embodiment, article recommended models can be artificial neural network, it is from information processing angle to human brain mind
It is abstracted through metanetwork, establishes certain naive model, different networks is formed by different connection types.Artificial neural network
It is usually constituted by being coupled to each other between a large amount of node (or neuron), a kind of each specific output function of node on behalf,
Referred to as excitation function.Connection between every two node all represents a weighted value for passing through the connection signal, referred to as weighs
The output of weight (be called and do parameter), network is then different according to the difference of the connection type of network, weighted value and excitation function.Article
Recommended models generally include multiple layers, and each layer includes multiple nodes, in general, the weight of the node of same layer can be identical, no
The weight of the node of same layer can be different, therefore multiple layers of article recommended models of parameter can also be different.Here, executing subject
Procedural information can be inputted from the input side of article recommended models, successively by the parameter of each layer in article recommended models
It handles (such as product, convolution etc.), and is exported from the outlet side of article recommended models, the information of outlet side output is to correspond to
The prediction Item Information of journey information.
In the present embodiment, article recommended models can be used for determining prediction Item Information by procedural information.Execute master
Body can train the article recommended models for determining prediction Item Information by procedural information in several ways.
As an example, executing subject can be based on the use to great amount of samples procedural information and corresponding sample processes information
The sample Item Information of the article of family actual selection counted and generate be stored with it is multiple record have sample processes information and sample
The mapping table of corresponding relationship between the corresponding sample Item Information of this procedural information, and using the mapping table as object
Product recommended models.In this way, executing subject can be by multiple sample processes information in procedural information and the mapping table successively
It is compared.If the sample processes information and the procedural information in the mapping table are same or similar, by the correspondence
Prediction Item Information of the corresponding sample Item Information of the procedural information as the corresponding procedural information in relation table.
As another example, executing subject can obtain multiple sample processes information and multiple sample processes information first
In each sample processes information corresponding to sample Item Information;Then by each sample processes of multiple sample processes information
Information is as input, using sample Item Information corresponding to each sample processes information in multiple sample processes information as defeated
Out, training obtains article recommended models.Here, the available multiple sample processes information of executing subject, and be art technology
Personnel show that those skilled in the art can be rule of thumb to each sample processes information labeling in multiple sample processes information
Corresponding sample Item Information.Executing subject training can be initialization article recommended models.Initialize article recommended models
It can be unbred article recommended models or the article recommended models that training is not completed.The article recommended models of initialization
Initial parameter has can be set in each layer, and parameter can be continuously adjusted in the training process of article recommended models.Initialization
Article recommended models can be various types of indisciplines or not training complete artificial neural network or to it is a variety of without
Training or the artificial neural network that training is not completed are combined obtained model.For example, initialization article recommended models can
To be unbred convolutional neural networks, it is also possible to unbred Recognition with Recurrent Neural Network, can also be to indiscipline
Convolutional neural networks, unbred Recognition with Recurrent Neural Network and unbred full articulamentum be combined obtained mould
Type.In this way, executing subject can input procedural information from the input side of article recommended models, successively pass through article recommended models
In each layer parameter processing, and from the outlet side of article recommended models export, outlet side output information be corresponded to
The prediction Item Information of journey information.
In some optional implementations of the present embodiment, article recommended models may include convolutional neural networks, follow
Ring neural network and full articulamentum.Above process information is imported to article recommended models trained in advance, obtains corresponding process letter
The prediction Item Information of breath, may comprise steps of:
Above process information input to above-mentioned convolutional neural networks is obtained the corresponding net of above process information by the first step
Page feature vector.
After obtaining procedural information, procedural information can be imported the convolutional neural networks of article recommended models by executing subject,
To obtain the corresponding web page characteristics vector of the procedural information.Wherein, web page characteristics vector can be used for characterizing between webpage
Incidence relation.For example, the web page characteristics vector can be used for characterizing browsing sequence when user browses webpage between webpage.The net
Page feature vector can also characterize the relationship etc. between the corresponding article of webpage, specifically depending on actual conditions.
In the present embodiment, convolutional neural networks can be a kind of feedforward neural network, its artificial neuron can ring
The surrounding cells in a part of coverage area are answered, have outstanding performance for large-scale image procossing.In general, the base of convolutional neural networks
This structure includes two layers, and one is characterized extract layer, and the input of each neuron is connected with the local acceptance region of preceding layer, and mentions
Take the feature of the part.After the local feature is extracted, its positional relationship between other feature is also decided therewith;
The second is Feature Mapping layer, each computation layer of network is made of multiple Feature Mappings, and each Feature Mapping is a plane, is put down
The weight of all neurons is equal on face.Here, executing subject can be defeated by the input side of procedural information from convolutional neural networks
Enter, successively by the processing of the parameter of each layer in convolutional neural networks, and is exported from the outlet side of convolutional neural networks, output
The information of side output is web page characteristics vector.
In the present embodiment, convolutional neural networks can be used for characterizing corresponding between procedural information and web page characteristics vector
Relationship, executing subject can train the corresponding pass that can be characterized between procedural information and web page characteristics vector in several ways
The convolutional neural networks of system.
As an example, executing subject can be based on to great amount of samples procedural information and the progress of sample web page feature vector
Count and generate the corresponding relationship for the sample web page feature vector for being stored with multiple sample processes information and sample processes information
Mapping table, and using the mapping table as convolutional neural networks.In this way, executing subject can be right with this by procedural information
Multiple sample processes information in relation table are answered to be compared.If a sample processes information and process in the mapping table
Information is same or similar, then using the sample web page feature vector of the sample processes information in the mapping table as the process
The web page characteristics vector of information.
As another example, executing subject can obtain the sample net of sample processes information and sample processes information first
Page feature vector;Then using sample processes information as input, using the sample web page feature vector of sample processes information as defeated
Out, training obtains to characterize the convolutional Neural net of the corresponding relationship between procedural information and the web page characteristics vector of procedural information
Network.In this way, executing subject can input procedural information from the input side of convolutional neural networks, successively pass through convolutional neural networks
In each layer parameter processing, and from the outlet side of convolutional neural networks export, outlet side output information be process letter
The web page characteristics vector of breath.
Above-mentioned web page characteristics vector is input to above-mentioned Recognition with Recurrent Neural Network, obtains web page contents feature vector by second step.
Web page characteristics vector can be input to the Recognition with Recurrent Neural Network of article recommended models by executing subject, to obtain net
Page content feature vector.Wherein, web page contents feature vector can be used for characterizing the incidence relation between web page characteristics vector.Example
Such as, web page contents feature vector can be used for characterizing the attributes such as the classification, parameter, purposes of article.According to actual needs, in webpage
Appearance feature vector can also characterize other attributes of article, specific depending on actual needs.
In the present embodiment, Recognition with Recurrent Neural Network is a kind of artificial neural network of node orientation connection cyclization.This net
The substantive characteristics of network is that the feedback link of the existing inside between processing unit has feedforward to connect again, and internal state can be shown
Dynamic time sequence behavior.
In the present embodiment, Recognition with Recurrent Neural Network can be used for characterizing web page characteristics vector and web page contents feature vector it
Between corresponding relationship.Executing subject can train in several ways can characterize web page characteristics vector and web page contents feature
The Recognition with Recurrent Neural Network of corresponding relationship between vector.
As an example, executing subject can based on to a large amount of sample web page feature vector and sample web page feature to
The sample web page content feature vector of amount is counted and generates and be stored with multiple sample web page feature vectors and sample web page spy
The mapping table of the corresponding relationship of the sample web page content feature vector of vector is levied, and using the mapping table as circulation mind
Through network.In this way, executing subject can calculate multiple sample web page features in web page characteristics vector and the mapping table to
Euclidean distance between amount.If a sample web page feature vector in the mapping table and the Europe between web page characteristics vector
Family name's distance is less than preset distance threshold, then the sample web page content of the sample web page feature vector in the mapping table is special
Levy web page contents feature vector of the vector as the web page characteristics vector.
As another example, executing subject sample web page feature vector available first and sample web page feature vector
Sample web page content feature vector;Then using sample web page feature vector as input, by the sample of sample web page feature vector
This webpage content feature vector obtains to characterize between web page characteristics vector and web page contents feature vector as output, training
Corresponding relationship Recognition with Recurrent Neural Network.In this way, executing subject can be by web page characteristics vector from the input of Recognition with Recurrent Neural Network
Side input successively by the processing of the parameter of each layer in Recognition with Recurrent Neural Network, and is exported from the outlet side of Recognition with Recurrent Neural Network,
The information of outlet side output is the web page contents feature vector of web page characteristics vector.
Above-mentioned web page contents feature vector is input to above-mentioned full articulamentum by third step, and it is corresponding to obtain above process information
Prediction Item Information.
Web page contents feature vector can be input to the full articulamentum of article recommended models by executing subject, to be obtained
The corresponding prediction Item Information of journey information.For example, web page contents feature vector characterizes multiple attributes of certain a kind of article.And
And there are the information such as incidence relation between multiple attributes.Then full articulamentum can by meet these attribute informations at least one
Item Information is as prediction Item Information.
In the present embodiment, all node phases of each node of full articulamentum and the output layer of Recognition with Recurrent Neural Network
Even, the feature vector for Recognition with Recurrent Neural Network output layer is exported integrates.Due to the characteristic that it is connected entirely, generally connect entirely
It is also most for connecing the parameter of layer.Meanwhile linear transformation is carried out to web page contents feature vector in the parameter using full articulamentum
Afterwards, the result of linear transformation can be converted plus a nonlinear activation function, so that non-linear factor is introduced, to increase
The ability to express of strong article recommended models.Wherein, excitation function can be softmax function, and softmax function is artificial neuron
Common a kind of excitation function in network, in this not go into detail.
In the present embodiment, full articulamentum can be used for characterizing between web page contents feature vector and prediction Item Information
Corresponding relationship.Executing subject can train in several ways can characterize web page contents feature vector and prediction Item Information
Between corresponding relationship full articulamentum.
As an example, executing subject can be based on to great amount of samples web page contents feature vector and sample web page content
The corresponding sample object Item Information of feature vector counted and generate be stored with multiple sample web page content feature vectors with
The mapping table of the corresponding relationship of sample object Item Information, and using the mapping table as full articulamentum.In this way, executing
Main body can calculate the Euclidean distance between web page contents feature vector and sample web page content feature vector.If the corresponding relationship
A sample web page content feature vector in table and the Euclidean distance between web page contents feature vector are less than preset distance
Threshold value, then using the sample object Item Information of the sample web page content feature vector in the mapping table as web page contents
The prediction Item Information of feature vector.
As another example, executing subject can obtain sample web page content feature vector and sample web page content first
The corresponding sample object Item Information of feature vector;Then using sample web page content feature vector as input, by sample web page
The corresponding sample object Item Information of content feature vector as output, training obtain capable of characterizing web page contents feature vector with
Predict the full articulamentum of the corresponding relationship between Item Information.In this way, executing subject can be by web page contents feature vector from complete
The input side of articulamentum inputs, the processing of parameter and excitation function by full articulamentum, and defeated from the outlet side of full articulamentum
Out, the information of outlet side output is the prediction Item Information of web page contents feature vector.In practice, prediction Item Information can be with
It is one or more webpages, which may include the letter of other articles relevant to the article of user's browsing
Breath.
It should be noted that the convolutional neural networks, Recognition with Recurrent Neural Network and full articulamentum in article recommended models can divide
Training is opened, can also be used as an entirety while training, the present embodiment is to this without limiting.
The procedural information that the user that the application will acquire browses webpage imports article recommended models, obtains corresponding process information
Prediction Item Information.Article recommended models can be determined by procedural information user select article when focus, and according to
Focus obtains prediction Item Information, improves the accuracy for obtaining prediction Item Information.
After obtaining prediction Item Information by article recommended models, executing subject can also show prediction Item Information
The designated position of the current web page of user's browsing improves the efficiency that user obtains information so as to user query.
With further reference to Fig. 3, it illustrates according to one embodiment of the article recommended models training method of the application
Process 300.The process 300 of the article recommended models training method, comprising the following steps:
Step 301, the sample object Item Information of multiple sample object articles that user is selected by webpage and upper is obtained
State the corresponding sample processes information of each sample object article in multiple sample object articles.
In the present embodiment, the operation of article recommended models training method and the executing subject (service of example as shown in figure 1 thereon
Device) available multiple sample object articles sample object Item Information and multiple sample object articles in each sample mesh
Mark the corresponding sample processes information of article.
In the present embodiment, the sample processes information of the available multiple sample object articles of executing subject, and be ability
Field technique personnel show.Those skilled in the art can be rule of thumb to each sample object object in multiple sample object articles
The corresponding sample processes information of product is marked.
Step 302, successively by the corresponding sample processes information of each sample object article in multiple sample object articles
Initialization article recommended models are input to, prediction target item information corresponding to sample processes information is obtained.
In the present embodiment, the sample object Item Information based on step 301 multiple sample object articles obtained, holds
Row main body can sequentially input the sample processes information of each sample object article in multiple sample object articles to initial
Compound product recommended models, so that the sample processes information institute for obtaining each sample object article in multiple sample object articles is right
The prediction target item information answered.Herein, executing subject can by each sample processes information from initialization article recommended models
Input side input, successively by initialization article recommended models each layer parameter processing, and from initialization article recommend
The outlet side of model exports.The information of output is prediction target item information corresponding to the sample processes information.Wherein, just
The article recommended models that beginning compound product recommended models can be unbred article recommended models or training is not completed, each layer
It is provided with initiation parameter, initiation parameter can be continuously adjusted in the training process of article recommended models.
Step 303, by prediction target item information corresponding to each sample processes information and the sample processes information institute
Corresponding sample object Item Information is compared, and obtains the predictablity rate of above-mentioned initialization article recommended models.
Executing subject can be by prediction object corresponding to each sample processes information in multiple sample processes information
Product information is compared with sample object Item Information corresponding to the sample processes information, is recommended to obtain initialization article
The predictablity rate of model.Specifically, if prediction target item information corresponding to a sample processes information and the sample mistake
Sample object Item Information corresponding to journey information is same or similar, then it is correct to initialize the prediction of article recommended models;If one
Sample object Item Information corresponding to prediction target item information and the sample processes information corresponding to sample processes information
It is different or not close, then initialize article recommended models prediction error.Here, executing subject can calculate the correct number of prediction
With the ratio of total sample number, and using the ratio as initialization article recommended models predictablity rate.
Step 304, determine whether above-mentioned predictablity rate is greater than default accuracy rate threshold value.
The predictablity rate for initializing article recommended models can be compared by executing subject with default accuracy rate threshold value.
If predictablity rate is greater than default accuracy threshold value, 305 are thened follow the steps;If executing step no more than default accuracy threshold value
Rapid 306.
Step 305, the article recommended models above-mentioned initialization article recommended models completed as training.
In the present embodiment, the case where the prediction accuracy for initializing article recommended models is greater than default accuracy threshold value
Under, illustrate that article recommended models training is completed.At this point, executing subject can will initialization article recommended models as having trained
At article recommended models.
Step 306, the parameter of above-mentioned initialization article recommended models is adjusted.
In the present embodiment, the feelings of default accuracy threshold value are not more than in the prediction accuracy of initialization article recommended models
Under condition, the parameter of the adjustable initialization article recommended models of executing subject, and 302 are returned to step, until training energy
Until the article recommended models for enough determining prediction Item Information by procedural information.
It should be noted that article recommended models can store on server 105, also can store in terminal device
101, on 102,103.
With continued reference to the signal that Fig. 4, Fig. 4 are according to the application scenarios of the method for obtaining information of the present embodiment
Figure.In the application scenarios of Fig. 4, user browses the webpage of article A by terminal device 103.Terminal device 103 is available to be arrived
Procedural information and is sent to server 105 in the procedural information of the webpage of browsing article A by user.What server 105 will acquire
Procedural information imports article recommended models, obtains the prediction Item Information B of corresponding article A.Later, server 105 can will be pre-
It surveys Item Information B and is sent to terminal device 103.Prediction Item Information B can be shown and currently be browsed in user by terminal device 103
The page designated position.
The method provided by the above embodiment of the application obtains the procedural information that user browses webpage first;Then by process
Information imports article recommended models, obtains prediction Item Information, improves the accuracy for obtaining prediction Item Information;It finally will be pre-
It surveys Item Information to show on webpage, be checked convenient for user.
With further reference to Fig. 5, as the realization to method shown in above-mentioned each figure, this application provides one kind for obtaining letter
One embodiment of the device of breath, the Installation practice is corresponding with embodiment of the method shown in Fig. 2, which can specifically answer
For in various electronic equipments.
As shown in figure 5, the device 500 for obtaining information of the present embodiment may include: information receiving unit 501 and letter
Cease acquiring unit 502.Wherein, information receiving unit 501 is configured to obtain the procedural information that user browses webpage, wherein on
Corresponding information when stating procedural information for characterizing user's browsing webpage;Information acquisition unit 502 is configured to above-mentioned mistake
Journey information imports article recommended models trained in advance, obtains the prediction Item Information of corresponding process information, and above-mentioned article is recommended
Model is used to determine prediction Item Information by procedural information.
In some optional implementations of the present embodiment, above process information includes at least one webpage and corresponding net
The browsing characteristic information of page, above-mentioned browsing characteristic information include at least one of the following: that user browses the information content of webpage and right
Answer the browsing time of the information content.
In some optional implementations of the present embodiment, above- mentioned information acquiring unit 502 may include: web page characteristics
Vector obtains subelement (not shown), web page contents feature vector obtains subelement (not shown) and prediction article letter
Breath obtains subelement (not shown).Wherein, web page characteristics vector acquisition subelement is configured to above process information is defeated
Enter to above-mentioned convolutional neural networks, obtains the corresponding web page characteristics vector of above process information, wherein above-mentioned convolutional neural networks
For characterizing the corresponding relationship between procedural information and web page characteristics vector;Web page contents feature vector obtains subelement and is configured
It is input to above-mentioned Recognition with Recurrent Neural Network at by above-mentioned web page characteristics vector, obtains web page contents feature vector, wherein above-mentioned circulation
Neural network is used to characterize the corresponding relationship between web page characteristics vector and web page contents feature vector, web page contents feature vector
For characterizing the incidence relation between web page characteristics vector;Prediction Item Information acquisition subelement is configured to will be in above-mentioned webpage
Hold feature vector and be input to above-mentioned full articulamentum, obtains the corresponding prediction Item Information of above process information, wherein above-mentioned to connect entirely
Layer is connect for characterizing web page contents feature vector and predicting the corresponding relationship between Item Information.
It is above-mentioned for obtain the device 500 of information to may include object in some optional implementations of the present embodiment
Product recommended models training unit (not shown), above-mentioned article recommended models training unit may include: that sample acquisition is single
First (not shown) and article recommended models training subelement (not shown).Wherein, sample acquisition subelement is configured
At the sample object Item Information and above-mentioned multiple sample objects for obtaining multiple sample object articles that user is selected by webpage
The corresponding sample processes information of each sample object article in article;Article recommended models training subelement is configured to will be upper
The corresponding sample processes information of each sample object article in multiple sample object articles is stated as input, by the sample processes
The sample Item Information of the corresponding sample object article of information obtains above-mentioned article recommended models as output, training.
In some optional implementations of the present embodiment, above-mentioned article recommended models training subelement may include:
Article recommended models training module (not shown) is configured to each sample object object in multiple sample object articles
The corresponding sample processes information of product is sequentially input to initialization article recommended models, obtains prediction corresponding to sample processes information
Target item information, will be corresponding to prediction target item information corresponding to each sample processes information and the sample processes information
Sample object Item Information be compared, obtain the predictablity rate of above-mentioned initialization article recommended models, determine above-mentioned pre-
It surveys whether accuracy rate is greater than default accuracy rate threshold value, if more than above-mentioned default accuracy rate threshold value, then pushes away above-mentioned initialization article
Recommend the article recommended models that model is completed as training.
In some optional implementations of the present embodiment, above-mentioned article recommended models training subelement may include:
Parameter adjustment module (not shown) is configured in response to adjust above-mentioned initial no more than above-mentioned default accuracy rate threshold value
The parameter of compound product recommended models, and continue to execute above-mentioned training step.
The present embodiment additionally provides a kind of electronic equipment, comprising: one or more processors;Memory is stored thereon with
One or more programs, when said one or multiple programs are executed by said one or multiple processors, so that said one
Or multiple processors execute the above-mentioned method for obtaining information.
The present embodiment additionally provides a kind of computer-readable medium, is stored thereon with computer program, and the program is processed
Device realizes the above-mentioned method for obtaining information when executing.
Below with reference to Fig. 6, it illustrates the servers for being suitable for being used to realize the embodiment of the present application (for example, the service in Fig. 1
Device 105) computer system 600 structural schematic diagram.Server shown in Fig. 6 is only an example, should not be to the application
The function and use scope of embodiment bring any restrictions.
As shown in fig. 6, computer system 600 includes central processing unit (CPU) 601, it can be read-only according to being stored in
Program in memory (ROM) 602 or be loaded into the program in random access storage device (RAM) 603 from storage section 608 and
Execute various movements appropriate and processing.In RAM 603, also it is stored with system 600 and operates required various programs and data.
CPU 601, ROM 602 and RAM 603 are connected with each other by bus 604.Input/output (I/O) interface 605 is also connected to always
Line 604.
I/O interface 605 is connected to lower component: the importation 606 including keyboard, mouse etc.;It is penetrated including such as cathode
The output par, c 607 of spool (CRT), liquid crystal display (LCD) etc. and loudspeaker etc.;Storage section 608 including hard disk etc.;
And the communications portion 609 of the network interface card including LAN card, modem etc..Communications portion 609 via such as because
The network of spy's net executes communication process.Driver 610 is also connected to I/O interface 605 as needed.Detachable media 611, such as
Disk, CD, magneto-optic disk, semiconductor memory etc. are mounted on as needed on driver 610, in order to read from thereon
Computer program be mounted into storage section 608 as needed.
Particularly, in accordance with an embodiment of the present disclosure, it may be implemented as computer above with reference to the process of flow chart description
Software program.For example, embodiment of the disclosure includes a kind of computer program product comprising be carried on computer-readable medium
On computer program, which includes the program code for method shown in execution flow chart.In such reality
It applies in example, which can be downloaded and installed from network by communications portion 609, and/or from detachable media
611 are mounted.When the computer program is executed by central processing unit (CPU) 601, limited in execution the present processes
Above-mentioned function.
It should be noted that the above-mentioned computer-readable medium of the application can be computer-readable signal media or meter
Calculation machine readable storage medium storing program for executing either the two any combination.Computer readable storage medium for example can be --- but not
Be limited to --- electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor system, device or device, or any above combination.Meter
The more specific example of calculation machine readable storage medium storing program for executing can include but is not limited to: have the electrical connection, just of one or more conducting wires
Taking formula computer disk, hard disk, random access storage device (RAM), read-only memory (ROM), erasable type may be programmed read-only storage
Device (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory device,
Or above-mentioned any appropriate combination.In this application, computer readable storage medium can be it is any include or storage journey
The tangible medium of sequence, the program can be commanded execution system, device or device use or in connection.And at this
In application, computer-readable signal media may include in a base band or as carrier wave a part propagate data-signal,
Wherein carry computer-readable program code.The data-signal of this propagation can take various forms, including but unlimited
In electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be that computer can
Any computer-readable medium other than storage medium is read, which can send, propagates or transmit and be used for
By the use of instruction execution system, device or device or program in connection.Include on computer-readable medium
Program code can transmit with any suitable medium, including but not limited to: wireless, electric wire, optical cable, RF etc. are above-mentioned
Any appropriate combination.
Flow chart and block diagram in attached drawing are illustrated according to the system of the various embodiments of the application, method and computer journey
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part of one module, program segment or code of table, a part of the module, program segment or code include one or more use
The executable instruction of the logic function as defined in realizing.It should also be noted that in some implementations as replacements, being marked in box
The function of note can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are actually
It can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it to infuse
Meaning, the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart can be with holding
The dedicated hardware based system of functions or operations as defined in row is realized, or can use specialized hardware and computer instruction
Combination realize.
Being described in unit involved in the embodiment of the present application can be realized by way of software, can also be by hard
The mode of part is realized.Described unit also can be set in the processor, for example, can be described as: a kind of processor packet
Include information receiving unit and information acquisition unit.Wherein, the title of these units is not constituted under certain conditions to the unit
The restriction of itself, for example, information acquisition unit is also described as " for obtaining prediction Item Information by procedural information
Unit ".
As on the other hand, present invention also provides a kind of computer-readable medium, which be can be
Included in device described in above-described embodiment;It is also possible to individualism, and without in the supplying device.Above-mentioned calculating
Machine readable medium carries one or more program, when said one or multiple programs are executed by the device, so that should
Device: the procedural information that user browses webpage is obtained, wherein corresponding when above process information is for characterizing user's browsing webpage
Information;Above process information is imported to article recommended models trained in advance, obtains the prediction article letter of corresponding process information
Breath, above-mentioned article recommended models are used to determine prediction Item Information by procedural information.
Above description is only the preferred embodiment of the application and the explanation to institute's application technology principle.Those skilled in the art
Member is it should be appreciated that invention scope involved in the application, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from foregoing invention design, it is carried out by above-mentioned technical characteristic or its equivalent feature
Any combination and the other technical solutions formed.Such as features described above has similar function with (but being not limited to) disclosed herein
Can technical characteristic replaced mutually and the technical solution that is formed.
Claims (14)
1. a kind of method for obtaining information, which comprises
Obtain the procedural information that user browses webpage, wherein corresponding when the procedural information is for characterizing user's browsing webpage
Information;
The procedural information is imported to article recommended models trained in advance, obtains the prediction Item Information of corresponding process information,
The article recommended models are used to determine prediction Item Information by procedural information.
2. according to the method described in claim 1, wherein, the procedural information includes the clear of at least one webpage and corresponding webpage
Look at characteristic information, the browsing characteristic information includes at least one of the following: that user browses the information content and corresponding informance of webpage
The browsing time of content.
3. according to the method described in claim 1, wherein, the article trained in advance that the procedural information is imported recommends mould
Type obtains the prediction Item Information of corresponding process information, comprising:
The procedural information is input to the convolutional neural networks, obtains the corresponding web page characteristics vector of the procedural information,
Wherein, the convolutional neural networks are used to characterize the corresponding relationship between procedural information and web page characteristics vector;
The web page characteristics vector is input to the Recognition with Recurrent Neural Network, obtains web page contents feature vector, wherein described to follow
Ring neural network is used to characterize corresponding relationship between web page characteristics vector and web page contents feature vector, web page contents feature to
Amount is for characterizing the incidence relation between web page characteristics vector;
The web page contents feature vector is input to the full articulamentum, obtains the corresponding prediction article letter of the procedural information
Breath, wherein the full articulamentum is used to characterize web page contents feature vector and predicts the corresponding relationship between Item Information.
4. according to the method described in claim 1, wherein, training obtains the article recommended models as follows:
Obtain the sample object Item Information and the multiple sample mesh of multiple sample object articles that user is selected by webpage
Mark the corresponding sample processes information of each sample object article in article;
Using the corresponding sample processes information of each sample object article in the multiple sample object article as input, by this
As output, training obtains the article and recommends mould the sample Item Information of the corresponding sample object article of sample processes information
Type.
5. according to the method described in claim 4, wherein, each sample object by the multiple sample object article
The corresponding sample processes information of article believes the sample article of the corresponding sample object article of the sample processes information as input
Breath obtains the article recommended models as output, training, comprising:
Execute following training step: by the corresponding sample processes information of each sample object article in multiple sample object articles
It sequentially inputs to initialization article recommended models, obtains prediction target item information corresponding to sample processes information, it will be each
Sample object Item Information corresponding to prediction target item information and the sample processes information corresponding to sample processes information
It is compared, obtains the predictablity rate of the initialization article recommended models, it is pre- to determine whether the predictablity rate is greater than
If accuracy rate threshold value, if more than the default accuracy rate threshold value, then completed the initialization article recommended models as training
Article recommended models.
6. according to the method described in claim 5, wherein, each sample object by the multiple sample object article
The corresponding sample processes information of article believes the sample article of the corresponding sample object article of the sample processes information as input
Breath obtains the article recommended models as output, training, further includes:
In response to being not more than the default accuracy rate threshold value, the parameter of the initialization article recommended models is adjusted, and continue to hold
The row training step.
7. a kind of for obtaining the device of information, described device includes:
Information receiving unit is configured to obtain the procedural information that user browses webpage, wherein the procedural information is for characterizing
User browses information corresponding when webpage;
Information acquisition unit is configured to importing the procedural information into article recommended models trained in advance, be corresponded to
The prediction Item Information of journey information, the article recommended models are used to determine prediction Item Information by procedural information.
8. device according to claim 7, wherein the procedural information includes the clear of at least one webpage and corresponding webpage
Look at characteristic information, the browsing characteristic information includes at least one of the following: that user browses the information content and corresponding informance of webpage
The browsing time of content.
9. device according to claim 7, wherein the information acquisition unit includes:
Web page characteristics vector obtains subelement, is configured to the procedural information being input to the convolutional neural networks, obtain
The corresponding web page characteristics vector of the procedural information, wherein the convolutional neural networks are special for characterizing procedural information and webpage
Levy the corresponding relationship between vector;
Web page contents feature vector obtains subelement, is configured to the web page characteristics vector being input to the circulation nerve net
Network obtains web page contents feature vector, wherein the Recognition with Recurrent Neural Network is special for characterizing web page characteristics vector and web page contents
The corresponding relationship between vector is levied, web page contents feature vector is used to characterize the incidence relation between web page characteristics vector;
It predicts that Item Information obtains subelement, is configured to the web page contents feature vector being input to the full articulamentum,
Obtain the corresponding prediction Item Information of the procedural information, wherein the full articulamentum is for characterizing web page contents feature vector
With the corresponding relationship between prediction Item Information.
10. device according to claim 7, wherein described device includes article recommended models training unit, the article
Recommended models training unit includes:
Sample acquisition subelement is configured to obtain the sample object object for multiple sample object articles that user is selected by webpage
The corresponding sample processes information of each sample object article in product information and the multiple sample object article;
Article recommended models train subelement, are configured to each sample object article in the multiple sample object article
Corresponding sample processes information makees the sample Item Information of the corresponding sample object article of the sample processes information as input
For output, training obtains the article recommended models.
11. device according to claim 10, wherein article recommended models training subelement includes:
Article recommended models training module is configured to each sample object article in multiple sample object articles is corresponding
Sample processes information is sequentially input to initialization article recommended models, obtains prediction target item corresponding to sample processes information
Information, by sample mesh corresponding to prediction target item information corresponding to each sample processes information and the sample processes information
Mark Item Information is compared, and is obtained the predictablity rate of the initialization article recommended models, is determined the predictablity rate
Whether it is greater than default accuracy rate threshold value, if more than the default accuracy rate threshold value, then makees the initialization article recommended models
The article recommended models completed for training.
12. device according to claim 11, wherein article recommended models training subelement includes:
Parameter adjustment module is configured in response to adjust the initialization article no more than the default accuracy rate threshold value and push away
The parameter of model is recommended, and continues to execute the training step.
13. a kind of electronic equipment, comprising:
One or more processors;
Storage device is stored thereon with one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processors are real
The now method as described in any in claim 1 to 6.
14. a kind of computer-readable medium, is stored thereon with computer program, such as right is realized when which is executed by processor
It is required that any method in 1 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810446322.7A CN110473042B (en) | 2018-05-11 | 2018-05-11 | Method and device for acquiring information |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810446322.7A CN110473042B (en) | 2018-05-11 | 2018-05-11 | Method and device for acquiring information |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110473042A true CN110473042A (en) | 2019-11-19 |
CN110473042B CN110473042B (en) | 2022-02-01 |
Family
ID=68504244
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810446322.7A Active CN110473042B (en) | 2018-05-11 | 2018-05-11 | Method and device for acquiring information |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110473042B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111738754A (en) * | 2020-06-01 | 2020-10-02 | 上海晶赞融宣科技有限公司 | Object recommendation method and device, storage medium and computer equipment |
CN111754062A (en) * | 2019-11-27 | 2020-10-09 | 北京沃东天骏信息技术有限公司 | Method and device for establishing article quality detection model |
CN113781150A (en) * | 2021-01-28 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Article recommendation method and device |
CN113781139A (en) * | 2020-10-19 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Item recommendation method, item recommendation device, equipment and medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103839169A (en) * | 2012-11-21 | 2014-06-04 | 大连灵动科技发展有限公司 | Personalized commodity recommendation method based on frequency matrix and text similarity |
CN105869024A (en) * | 2016-04-20 | 2016-08-17 | 北京小米移动软件有限公司 | Commodity recommending method and device |
CN107730343A (en) * | 2017-09-15 | 2018-02-23 | 广州唯品会研究院有限公司 | A kind of user's merchandise news method for pushing and equipment based on picture attribute extraction |
CN107895024A (en) * | 2017-09-13 | 2018-04-10 | 同济大学 | The user model construction method and recommendation method recommended for web page news classification |
CN107909037A (en) * | 2017-11-16 | 2018-04-13 | 百度在线网络技术(北京)有限公司 | Information output method and device |
-
2018
- 2018-05-11 CN CN201810446322.7A patent/CN110473042B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103839169A (en) * | 2012-11-21 | 2014-06-04 | 大连灵动科技发展有限公司 | Personalized commodity recommendation method based on frequency matrix and text similarity |
CN105869024A (en) * | 2016-04-20 | 2016-08-17 | 北京小米移动软件有限公司 | Commodity recommending method and device |
CN107895024A (en) * | 2017-09-13 | 2018-04-10 | 同济大学 | The user model construction method and recommendation method recommended for web page news classification |
CN107730343A (en) * | 2017-09-15 | 2018-02-23 | 广州唯品会研究院有限公司 | A kind of user's merchandise news method for pushing and equipment based on picture attribute extraction |
CN107909037A (en) * | 2017-11-16 | 2018-04-13 | 百度在线网络技术(北京)有限公司 | Information output method and device |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111754062A (en) * | 2019-11-27 | 2020-10-09 | 北京沃东天骏信息技术有限公司 | Method and device for establishing article quality detection model |
CN111754062B (en) * | 2019-11-27 | 2023-01-31 | 北京沃东天骏信息技术有限公司 | Method and device for establishing article quality detection model |
CN111738754A (en) * | 2020-06-01 | 2020-10-02 | 上海晶赞融宣科技有限公司 | Object recommendation method and device, storage medium and computer equipment |
CN113781139A (en) * | 2020-10-19 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Item recommendation method, item recommendation device, equipment and medium |
CN113781150A (en) * | 2021-01-28 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Article recommendation method and device |
Also Published As
Publication number | Publication date |
---|---|
CN110473042B (en) | 2022-02-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109460513A (en) | Method and apparatus for generating clicking rate prediction model | |
CN107832468B (en) | Demand recognition methods and device | |
CN107105031A (en) | Information-pushing method and device | |
CN109460514A (en) | Method and apparatus for pushed information | |
CN110473042A (en) | For obtaining the method and device of information | |
CN107391680A (en) | Content recommendation method, device and equipment | |
CN108805594A (en) | Information-pushing method and device | |
CN109976997A (en) | Test method and device | |
CN107908740A (en) | Information output method and device | |
CN108520324A (en) | Method and apparatus for generating information | |
CN106484766B (en) | Searching method and device based on artificial intelligence | |
CN109492160A (en) | Method and apparatus for pushed information | |
CN109598576A (en) | Service recommendation method, device and equipment | |
CN108960316A (en) | Method and apparatus for generating model | |
CN107944481A (en) | Method and apparatus for generating information | |
CN110298716A (en) | Information-pushing method and device | |
CN108388563A (en) | Information output method and device | |
CN109388548A (en) | Method and apparatus for generating information | |
CN108182472A (en) | For generating the method and apparatus of information | |
CN107944956A (en) | Method and apparatus for generating information | |
CN109447246A (en) | Method and apparatus for generating model | |
CN108920717A (en) | For showing the method and device of information | |
CN110377522A (en) | The test method for scene of trading, calculates equipment and medium at device | |
CN110413872A (en) | Method and apparatus for showing information | |
CN107451869A (en) | Method and apparatus for pushed information |
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 |