CN110473042B - Method and device for acquiring information - Google Patents
Method and device for acquiring information Download PDFInfo
- Publication number
- CN110473042B CN110473042B CN201810446322.7A CN201810446322A CN110473042B CN 110473042 B CN110473042 B CN 110473042B CN 201810446322 A CN201810446322 A CN 201810446322A CN 110473042 B CN110473042 B CN 110473042B
- Authority
- CN
- China
- Prior art keywords
- information
- sample
- process information
- webpage
- recommendation model
- 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.)
- Active
Links
Images
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 application discloses a method and a device for acquiring information. One embodiment of the method comprises: acquiring process information of a webpage browsed by a user, wherein the process information is used for representing information corresponding to the webpage browsed by the user; and importing the process information into a pre-trained article recommendation model to obtain predicted article information corresponding to the process information, wherein the article recommendation model is used for determining the predicted article information according to the process information. This embodiment improves the accuracy of obtaining the predicted item information.
Description
Technical Field
The embodiment of the application relates to the technical field of computers, in particular to a method and a device for acquiring information.
Background
With the development of network technology, more and more items are sold through a network. Generally, a user can log in a related website through various electronic devices and browse a webpage of related article information; then, the user puts the needed articles into a website shopping cart; finally, the user settles the account for the items in the shopping cart to complete the purchase of the items. The method and the device have the advantages that the articles are purchased through the network, so that a user can obtain a large amount of article information without going out, and the article obtaining efficiency of the user is improved. Correspondingly, the website technical personnel can also provide related article information for the user according to the purchase record of the user, and the efficiency of obtaining articles by the user is further improved.
Disclosure of Invention
The embodiment of the application provides a method and a device for acquiring information.
In a first aspect, an embodiment of the present application provides a method for acquiring information, where the method includes: acquiring process information of a webpage browsed by a user, wherein the process information is used for representing information corresponding to the webpage browsed by the user; and importing the process information into a pre-trained article recommendation model to obtain predicted article information corresponding to the process information, wherein the article recommendation model is used for determining the predicted article information according to the process information.
In some embodiments, the process information includes browsing characteristic information of at least one web page and a corresponding web page, and the browsing characteristic information includes at least one of: the user browses the information content of the webpage and the browsing time of the corresponding information content.
In some embodiments, the importing the process information into a pre-trained item recommendation model to obtain predicted item information corresponding to the process information includes: inputting the process information into the convolutional neural network to obtain a webpage feature vector corresponding to the process information, wherein the convolutional neural network is used for representing the corresponding relationship between the process information and the webpage feature vector; inputting the webpage feature vectors into the recurrent neural network to obtain webpage content feature vectors, wherein the recurrent neural network is used for representing the corresponding relation between the webpage feature vectors and the webpage content feature vectors, and the webpage content feature vectors are used for representing the incidence relation between the webpage feature vectors; and inputting the webpage content feature vector into the full-link layer to obtain the predicted article information corresponding to the process information, wherein the full-link layer is used for representing the corresponding relation between the webpage content feature vector and the predicted article information.
In some embodiments, the item recommendation model is trained by: acquiring sample target item information of a plurality of sample target items selected by a user through a webpage and sample process information corresponding to each sample target item in the plurality of sample target items; and training to obtain the item recommendation model by taking the sample process information corresponding to each sample target item in the plurality of sample target items as input and taking the sample item information of the sample target item corresponding to the sample process information as output.
In some embodiments, the training of the item recommendation model using the sample process information corresponding to each of the plurality of sample target items as an input and the sample item information of the sample target item corresponding to the sample process information as an output includes: the following training steps are performed: the method comprises the steps of sequentially inputting sample process information corresponding to each sample target object in a plurality of sample target objects into an initialized object recommendation model to obtain predicted target object information corresponding to the sample process information, comparing the predicted target object information corresponding to each sample process information with the sample target object information corresponding to the sample process information to obtain the predicted accuracy of the initialized object recommendation model, determining whether the predicted accuracy is greater than a preset accuracy threshold, and taking the initialized object recommendation model as the trained object recommendation model if the predicted accuracy is greater than the preset accuracy threshold.
In some embodiments, the training to obtain the item recommendation model by taking the sample process information corresponding to each of the plurality of sample target items as an input and taking the sample item information of the sample target item corresponding to the sample process information as an output further includes: and responding to the condition that the accuracy is not greater than the preset accuracy threshold, adjusting the parameters of the initialized item recommendation model, and continuing to execute the training step.
In a second aspect, an embodiment of the present application provides an apparatus for acquiring information, where the apparatus includes: the information receiving unit is configured to acquire process information of a webpage browsed by a user, wherein the process information is used for representing corresponding information when the user browses the webpage; and the information acquisition unit is configured to import the process information into a pre-trained article recommendation model to obtain the predicted article information corresponding to the process information, and the article recommendation model is used for determining the predicted article information through the process information.
In some embodiments, the process information includes browsing characteristic information of at least one web page and a corresponding web page, and the browsing characteristic information includes at least one of: the user browses the information content of the webpage and the browsing time of the corresponding information content.
In some embodiments, the information acquiring unit includes: the webpage feature vector acquisition subunit is configured to input the process information to the convolutional neural network to obtain a webpage feature vector corresponding to the process information, wherein the convolutional neural network is used for representing a corresponding relationship between the process information and the webpage feature vector; the webpage content feature vector acquisition subunit is configured to input the webpage feature vectors into the recurrent neural network to obtain webpage content feature vectors, wherein the recurrent neural network is used for representing the corresponding relation between the webpage feature vectors and the webpage content feature vectors, and the webpage content feature vectors are used for representing the incidence relation between the webpage feature vectors; and the predicted article information acquisition subunit is configured to input the webpage content feature vector to the fully-connected layer to obtain predicted article information corresponding to the process information, wherein the fully-connected layer is used for representing a corresponding relation between the webpage content feature vector and the predicted article information.
In some embodiments, the apparatus includes an item recommendation model training unit, the item recommendation model training unit including: the system comprises a sample acquisition subunit, a data processing unit and a data processing unit, wherein the sample acquisition subunit is configured to acquire sample target item information of a plurality of sample target items selected by a user through a webpage and sample process information corresponding to each sample target item in the plurality of sample target items; and the article recommendation model training subunit is configured to take the sample process information corresponding to each of the plurality of sample target articles as input, take the sample article information of the sample target article corresponding to the sample process information as output, and train to obtain the article recommendation model.
In some embodiments, the item recommendation model training subunit includes: the article recommendation model training module is configured to sequentially input sample process information corresponding to each sample target article in the plurality of sample target articles into the initialized article recommendation model to obtain predicted target article information corresponding to the sample process information, compare the predicted target article information corresponding to each sample process information with the sample target article information corresponding to the sample process information to obtain the prediction accuracy of the initialized article recommendation model, determine whether the prediction accuracy is greater than a preset accuracy threshold, and if the prediction accuracy is greater than the preset accuracy threshold, use the initialized article recommendation model as the trained article recommendation model.
In some embodiments, the item recommendation model training subunit includes: and the parameter adjusting module is configured to respond to the condition that the accuracy is not greater than the preset accuracy threshold value, adjust the parameters of the initialized item recommendation model and continue to execute the training step.
In a third aspect, an embodiment of the present application provides an electronic device, including: one or more processors; a memory having one or more programs stored thereon, which when executed by the one or more processors, cause the one or more processors to perform the method for obtaining information of the first aspect.
In a fourth aspect, the present application provides a computer-readable medium, on which a computer program is stored, where the computer program is executed by a processor to implement the method for acquiring information of the first aspect.
According to the method and the device for acquiring the information, firstly, the process information of a user browsing a webpage is acquired; and then, the process information is imported into the article recommendation model to obtain the predicted article information, so that the accuracy of obtaining the predicted article information is improved.
Drawings
Other features, objects and advantages of the present application will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, made with reference to the accompanying drawings in which:
FIG. 1 is an exemplary system architecture diagram in which the present application may be applied;
FIG. 2 is a flow diagram of one embodiment of a method for obtaining information according to the present application;
FIG. 3 is a flow diagram of one embodiment of an item recommendation model training method according to the present application;
FIG. 4 is a schematic illustration of an application scenario of a method for obtaining information according to the present application;
FIG. 5 is a schematic block diagram illustrating one embodiment of an apparatus for obtaining information according to the present application;
FIG. 6 is a schematic block diagram of a computer system suitable for use in implementing a server according to embodiments of the present application.
Detailed Description
The present application will be described in further detail with reference to the following drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the relevant invention and not restrictive of the invention. It should be noted that, for convenience of description, only the portions related to the related invention are shown in the drawings.
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.
Fig. 1 illustrates an exemplary system architecture 100 to which the method for acquiring information or the apparatus for acquiring information of the embodiments of the present application may be applied.
As shown in fig. 1, the system architecture 100 may include terminal devices 101, 102, 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the terminal devices 101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.
The user may use the terminal devices 101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The terminal devices 101, 102, 103 may have installed thereon various communication client applications, such as a web browser application, a shopping-type application, a search-type application, an instant messaging tool, and the like.
The terminal apparatuses 101, 102, and 103 may be hardware or software. When the terminal devices 101, 102, 103 are hardware, they may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like. When the terminal apparatuses 101, 102, 103 are software, they can be installed in the electronic apparatuses listed above. It may be implemented as multiple pieces of software or software modules (e.g., to provide distributed services) or as a single piece of software or software module. And is not particularly limited herein.
The server 105 may be a server that provides various services, for example, a server that performs data processing on process information corresponding to web pages transmitted from the terminal apparatuses 101, 102, 103. The server may analyze and otherwise process the received data, such as the process information, and feed back the processing result (e.g., the predicted item information) to the terminal device.
It should be noted that the method for acquiring information provided in the embodiment of the present application is generally performed by the server 105, and accordingly, the apparatus for acquiring information is generally disposed in the server 105.
The server may be hardware or software. When the server is hardware, it may be implemented as a distributed server cluster formed by multiple servers, or may be implemented as a single server. When the server is software, it may be implemented as multiple pieces of software or software modules (e.g., to provide distributed services), or as a single piece of software or software module. And is not particularly limited herein.
It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
With continued reference to FIG. 2, a flow 200 of one embodiment of a method for obtaining information in accordance with the present application is shown. The method for acquiring information comprises the following steps:
In the present embodiment, the execution subject of the method for acquiring information (e.g., the terminal devices 101, 102, 103 shown in fig. 1) may receive the process information from the terminal with which the user browses the web page through a wired connection manner or a wireless connection manner. The process information is used for representing information corresponding to the webpage browsed by the user. It should be noted that the wireless connection means may include, but is not limited to, a 3G/4G connection, a WiFi connection, a bluetooth connection, a WiMAX connection, a Zigbee connection, a uwb (ultra wideband) connection, and other wireless connection means now known or developed in the future.
In the prior art, when a website technician provides item information to a user, the related item information is generally provided to the user directly according to a purchase record of a user history. In practice, the user usually determines the needed item after a series of related operations such as browsing the image of the item, viewing some specific parameters of the item, and the like. These series of operations can reflect the point of interest of the user in selecting the item, directly affecting the item finally selected by the user. The existing method for providing the item information cannot reflect the details in the process of selecting the item by the user, and is not easy to determine the item really interested by the user according to the process of browsing the webpage by the user (the series of related operations). Therefore, the correlation between the article information provided by the conventional method and the article finally selected by the user is not large, and the accuracy of the article information is not high.
Therefore, when the webpage browsing of the user is monitored, the webpage browsing process information of the user can be obtained. Wherein the process information is used to characterize information viewed by the user from browsing the web page to finally selecting the item. That is, the process information of the present application may be used to represent information corresponding to a user browsing a web page.
In some optional implementation manners of this embodiment, the process information includes browsing characteristic information of at least one web page and a corresponding web page, where the browsing characteristic information includes at least one of: the user browses the information content of the webpage and the browsing time of the corresponding information content.
The browsing feature information of the present application may include at least one of: the information content of the webpage browsed by the user, the browsing time of the corresponding information content and the like. For example, a user logs on to a shopping site and starts searching for desired item information. The user opens a certain article webpage, first browses the parameters of the article for X minutes, and then checks the user evaluation for X minutes. And then, opening other at least one article webpage of the same type of article, checking the user evaluation for X minutes, and then checking the article parameter for X minutes and the article price for X minutes. In the process of querying the item information by the user, the process information described in the application can be composed of the browsed web pages, the item parameters, the user evaluation, the item price of each browsed web page, and the browsing time of the corresponding item parameters, the user evaluation, the item price, and the like. The information such as the item parameter, the user evaluation, the item price, etc. may be the information content of the web page. The browsing time corresponding to the item parameter, the user rating, the item price may be the browsing time of the information content. Wherein the execution subject may monitor the display content of the screen to determine the information content of the web page being browsed by the user; and the time of displaying the corresponding information content on the screen is taken as the browsing time of the information content. The information content of each web page and the browsing time of the corresponding information content may be browsing characteristic information corresponding to the web page. The browsing characteristic information can represent which characteristics of the item the user is interested in, and the attention points of the item selected by the user can be determined through the characteristics, and then the item finally selected by the user can be predicted based on the attention points.
In this embodiment, after acquiring the process information of the browsed web page, the execution subject may import the process information into a pre-trained article recommendation model to obtain the predicted article information corresponding to the process information. When browsing the web pages of the related article information, the information such as browsing sequence, browsing information content of the web pages, and browsing time of the information content among the web pages corresponding to different articles may affect the article finally selected by the user. Therefore, different weights can be given to the browsing order among web pages, the order of browsing the information contents of the web pages, and the browsing time of browsing the information contents when the user browses a plurality of web pages. All weights belonging to a web page (corresponding to an item) are added up as the total weight of the web page. The greater the weight, the higher the likelihood that the user will select the item corresponding to the weighted web page. For example, when browsing the related web pages of a mobile phone, the user first browses the web pages of a certain type of mobile phone of brand a. In the process of browsing the webpage, the user browses the product introduction part of the mobile phone firstly and then browses the user evaluation. After browsing the user evaluation, the user browses the user evaluation and then browses the product introduction and the price by looking up another mobile phone of the brand B. The execution subject can count the total weight of the web pages of the brand A mobile phone and the total weight of the web pages of the brand B mobile phone, and the mobile phone information corresponding to the web pages with large weights is used as the predicted article information.
In this embodiment, the item recommendation model may be an artificial neural network, which abstracts a human brain neuron network from an information processing perspective, establishes a simple model, and forms different networks according to different connection modes. An artificial neural network is generally composed of a large number of nodes (or neurons) interconnected with each other, each node representing a specific output function, called an excitation function. The connection between each two nodes represents a weighted value, called weight (also called parameter), for the signal passing through the connection, and the output of the network varies according to the connection mode, the weight value and the excitation function of the network. The item recommendation model generally includes a plurality of layers, each layer includes a plurality of nodes, and generally, the weight of the node of the same layer may be the same, and the weight of the node of different layers may be different, so the parameters of the plurality of layers of the item recommendation model may also be different. Here, the execution agent may input the process information from the input side of the item recommendation model, sequentially perform processing (for example, multiplication, convolution, or the like) of parameters of each layer in the item recommendation model, and output the process information from the output side of the item recommendation model, where the information output from the output side is predicted item information corresponding to the process information.
In this embodiment, the item recommendation model may be used to determine predicted item information from process information. The executing body may train an item recommendation model for determining predicted item information from the process information in various ways.
As an example, the execution subject may generate a correspondence table in which a plurality of correspondences between sample process information and sample article information corresponding to the sample process information are recorded, based on the statistics of a large amount of sample process information and sample article information of an article actually selected by the user corresponding to the sample process information, and use the correspondence table as the article recommendation model. In this way, the execution subject may sequentially compare the process information with the plurality of sample process information in the correspondence table. And if one sample process information in the corresponding relation table is the same as or similar to the process information, taking the sample article information corresponding to the process information in the corresponding relation table as the predicted article information corresponding to the process information.
As another example, the executing entity may first obtain a plurality of sample process information and sample article information corresponding to each of the plurality of sample process information; and then, taking each sample process information of the plurality of sample process information as input, taking the sample article information corresponding to each sample process information of the plurality of sample process information as output, and training to obtain an article recommendation model. Here, the executing entity may obtain a plurality of sample process information, and show to a person skilled in the art that the person skilled in the art may label, according to experience, each sample process information in the plurality of sample process information with corresponding sample article information. It may be that subject training is performed to initialize the item recommendation model. The initialized item recommendation model may be an untrained item recommendation model or an untrained completed item recommendation model. Each layer of the initialized item recommendation model may be provided with initial parameters, which may be continuously adjusted during the training of the item recommendation model. The initialized item recommendation model may be various types of untrained or untrained artificial neural networks or a model obtained by combining a plurality of untrained or untrained artificial neural networks. For example, the initialized item recommendation model may be an untrained convolutional neural network, an untrained cyclic neural network, or a model obtained by combining an untrained convolutional neural network, an untrained cyclic neural network, and an untrained full-connectivity layer. In this way, the execution body can input the process information from the input side of the item recommendation model, sequentially process the parameters of each layer in the item recommendation model, and output the process information from the output side of the item recommendation model, wherein the information output by the output side is the predicted item information corresponding to the process information.
In some optional implementations of the present embodiment, the item recommendation model may include a convolutional neural network, a cyclic neural network, and a fully-connected layer. Importing the process information into a pre-trained item recommendation model to obtain predicted item information corresponding to the process information, which may include the following steps:
firstly, inputting the process information into the convolutional neural network to obtain a webpage feature vector corresponding to the process information.
After the process information is obtained, the execution main body can lead the process information into the convolutional neural network of the article recommendation model, so that the webpage feature vector corresponding to the process information is obtained. The webpage feature vectors can be used for representing the incidence relation between webpages. For example, the web page feature vector may be used to characterize the browsing order between web pages when a user browses the web pages. The webpage feature vector can also represent the relationship between the articles corresponding to the webpage and the like, which is determined according to the actual situation.
In this embodiment, the convolutional neural network may be a feedforward neural network whose artificial neurons may respond to a portion of the coverage of surrounding cells, which may perform well for large image processing. In general, the basic structure of a convolutional neural network includes two layers, one of which is a feature extraction layer, and the input of each neuron is connected to the local acceptance domain of the previous layer and extracts the features of the local acceptance domain. Once the local feature is extracted, the position relation between the local feature and other features is determined; the other is a feature mapping layer, each calculation layer of the network is composed of a plurality of feature mappings, each feature mapping is a plane, and the weights of all neurons on the plane are equal. Here, the execution subject may input the process information from an input side of the convolutional neural network, sequentially perform processing of parameters of each layer in the convolutional neural network, and output the process information from an output side of the convolutional neural network, where the information output by the output side is the web page feature vector.
In this embodiment, the convolutional neural network may be used to represent a corresponding relationship between the process information and the webpage feature vector, and the execution main body may train the convolutional neural network that can represent a corresponding relationship between the process information and the webpage feature vector in a variety of ways.
As an example, the execution subject may generate a correspondence table storing correspondences of a plurality of sample process information and sample web page feature vectors of the sample process information based on counting a large amount of the sample process information and the sample web page feature vectors, and treat the correspondence table as a convolutional neural network. In this manner, the execution principal may compare the process information with a plurality of sample process information in the correspondence table. And if one sample process information in the corresponding relation table is the same as or similar to the process information, taking the sample webpage feature vector of the sample process information in the corresponding relation table as the webpage feature vector of the process information.
As another example, the executing entity may first obtain sample process information and a sample web page feature vector of the sample process information; and then, taking the sample process information as input, taking the sample webpage feature vector of the sample process information as output, and training to obtain the convolutional neural network capable of representing the corresponding relation between the process information and the webpage feature vector of the process information. In this way, the execution main body can input the process information from the input side of the convolutional neural network, sequentially process the parameters of each layer in the convolutional neural network, and output the process information from the output side of the convolutional neural network, wherein the information output by the output side is the webpage feature vector of the process information.
And secondly, inputting the webpage feature vector into the recurrent neural network to obtain a webpage content feature vector.
The execution subject can input the webpage feature vector into a recurrent neural network of the item recommendation model, so as to obtain the webpage content feature vector. The webpage content feature vectors can be used for representing the association relation between the webpage feature vectors. For example, the webpage content feature vector can be used for characterizing the attributes of the category, the parameter, the purpose and the like of the item. According to actual needs, the webpage content feature vector can also represent other attributes of the article, which is determined according to the actual needs.
In this embodiment, the recurrent neural network is an artificial neural network with nodes directionally connected into a ring. The essential feature of such a network is that there is both an internal feedback and a feed-forward connection between the processing units, the internal state of which may exhibit dynamic timing behavior.
In this embodiment, the recurrent neural network may be used to characterize the correspondence between the webpage feature vector and the webpage content feature vector. The execution subject can train out a recurrent neural network which can represent the corresponding relation between the webpage feature vector and the webpage content feature vector in various ways.
As an example, the execution subject may generate a correspondence table storing correspondences of a plurality of sample web page feature vectors and sample web page content feature vectors of the sample web page feature vectors based on counting a large number of sample web page feature vectors and sample web page content feature vectors of the sample web page feature vectors, and treat the correspondence table as a recurrent neural network. In this way, the execution subject may calculate the euclidean distance between the web page feature vector and the plurality of sample web page feature vectors in the correspondence table. And if the Euclidean distance between one sample webpage feature vector in the corresponding relation table and the webpage feature vector is smaller than a preset distance threshold, taking the sample webpage content feature vector of the sample webpage feature vector in the corresponding relation table as the webpage content feature vector of the webpage feature vector.
As another example, the executing entity may first obtain a sample web page feature vector and a sample web page content feature vector of the sample web page feature vector; and then taking the sample webpage feature vector as input, taking the sample webpage content feature vector of the sample webpage feature vector as output, and training to obtain the recurrent neural network capable of representing the corresponding relation between the webpage feature vector and the webpage content feature vector. In this way, the execution main body can input the webpage feature vector from the input side of the recurrent neural network, sequentially process the parameters of each layer in the recurrent neural network, and output the webpage feature vector from the output side of the recurrent neural network, wherein the information output by the output side is the webpage content feature vector of the webpage feature vector.
And thirdly, inputting the webpage content feature vector to the full-connection layer to obtain the predicted article information corresponding to the process information.
The execution subject can input the webpage content feature vector to a full connection layer of the item recommendation model, so that the predicted item information corresponding to the process information is obtained. For example, a web page content feature vector characterizes a number of attributes of a class of items. The plurality of attributes have information such as an association relationship. The fully-connected layer may use at least one item information satisfying these attribute information as the predicted item information.
In this embodiment, each node of the fully-connected layer is connected to all nodes of the output layer of the recurrent neural network, and is used to integrate the feature vectors output by the output layer of the recurrent neural network. The parameters of a fully connected layer are also typically the most due to its fully connected nature. Meanwhile, after the parameters of the full connection layer are utilized to carry out linear transformation on the webpage content feature vectors, a nonlinear excitation function can be added to convert the result of the linear transformation, so that nonlinear factors are introduced to enhance the expression capability of the item recommendation model. The excitation function may be a softmax function, which is a common excitation function in an artificial neural network and is not described in detail herein.
In this embodiment, the full link layer may be used to represent the correspondence between the webpage content feature vector and the predicted item information. The execution subject can train out a full connection layer which can represent the corresponding relation between the webpage content feature vector and the predicted article information in various ways.
As an example, the execution subject may generate a correspondence table storing correspondence between a plurality of sample web content feature vectors and sample target item information based on counting a large number of sample web content feature vectors and sample target item information corresponding to the sample web content feature vectors, and take the correspondence table as a full connection layer. In this way, the executing entity may calculate the euclidean distance between the web content feature vector and the sample web content feature vector. And if the Euclidean distance between a sample webpage content feature vector in the corresponding relation table and the webpage content feature vector is smaller than a preset distance threshold, taking the sample target article information of the sample webpage content feature vector in the corresponding relation table as the predicted article information of the webpage content feature vector.
As another example, the execution subject may first obtain a sample webpage content feature vector and sample target item information corresponding to the sample webpage content feature vector; and then taking the sample webpage content feature vector as input, taking the sample target article information corresponding to the sample webpage content feature vector as output, and training to obtain a full connection layer capable of representing the corresponding relation between the webpage content feature vector and the predicted article information. In this way, the execution main body can input the webpage content feature vector from the input side of the full connection layer, and output the webpage content feature vector from the output side of the full connection layer after the processing of the parameters and the excitation function of the full connection layer, wherein the information output by the output side is the predicted article information of the webpage content feature vector. In practice, the predicted item information may be one or more web pages that may contain information about other items that are relevant to the item being viewed by the user.
It should be noted that the convolutional neural network, the cyclic neural network, and the full connection layer in the item recommendation model may be trained separately, or may be trained simultaneously as a whole, which is not limited in this embodiment.
The method and the device lead the acquired process information of the user browsing webpage into the article recommendation model to obtain the predicted article information corresponding to the process information. The item recommendation model can determine the attention point of the user when the user selects the item through the process information, and obtains the predicted item information according to the attention point, so that the accuracy of obtaining the predicted item information is improved.
After the predicted article information is obtained through the article recommendation model, the execution main body can also display the predicted article information at the specified position of the current webpage browsed by the user, so that the user can conveniently inquire, and the information obtaining efficiency of the user is improved.
With further reference to FIG. 3, a flow 300 of one embodiment of an item recommendation model training method according to the present application is illustrated. The process 300 of the item recommendation model training method includes the following steps:
In this embodiment, the item recommendation model training method operating with an execution subject (e.g., the server in fig. 1) thereon may obtain sample target item information of a plurality of sample target items and sample process information corresponding to each of the plurality of sample target items.
In this embodiment, the executing entity may obtain sample process information of a plurality of sample target items, and present the sample process information to those skilled in the art. One skilled in the art may empirically tag sample process information corresponding to each of a plurality of sample target items.
In this embodiment, based on the sample target item information of the plurality of sample target items obtained in step 301, the executing entity may sequentially input the sample process information of each of the plurality of sample target items to the initialized item recommendation model, so as to obtain the prediction target item information corresponding to the sample process information of each of the plurality of sample target items. Here, the execution agent may input each sample process information from an input side of the initialized item recommendation model, sequentially pass through the processing of the parameters of the respective layers of the initialized item recommendation model, and output from an output side of the initialized item recommendation model. The output information is the predicted target article information corresponding to the sample process information. The initialized item recommendation model can be an untrained item recommendation model or an untrained item recommendation model, and each layer of the initialized item recommendation model is provided with initialization parameters which can be continuously adjusted in the training process of the item recommendation model.
The execution subject may compare the predicted target item information corresponding to each of the plurality of sample process information with the sample target item information corresponding to the sample process information, thereby obtaining the prediction accuracy of the initialized item recommendation model. Specifically, if the predicted target item information corresponding to one sample process information is the same as or similar to the sample target item information corresponding to the sample process information, the initialized item recommendation model predicts correctly; if the predicted target item information corresponding to one sample process information is different from or not similar to the sample target item information corresponding to the sample process information, the initialized item recommendation model is predicted incorrectly. Here, the execution subject may calculate a ratio of the number of prediction correctness to the total number of samples, and take the ratio as a prediction accuracy rate of the initialized item recommendation model.
The executive may compare the prediction accuracy of the initialized item recommendation model to a preset accuracy threshold. If the prediction accuracy is greater than the predetermined accuracy threshold, go to step 305; if not, go to step 306.
And 305, taking the initialized item recommendation model as a trained item recommendation model.
In this embodiment, when the prediction accuracy of the initialized item recommendation model is greater than the preset accuracy threshold, it indicates that the item recommendation model is trained completely. At this point, the executive may take the initialized item recommendation model as the trained item recommendation model.
In this embodiment, in the case that the prediction accuracy of the initialized item recommendation model is not greater than the preset accuracy threshold, the execution subject may adjust the parameters of the initialized item recommendation model, and return to execute step 302 until an item recommendation model capable of determining the predicted item information through the process information is trained.
The item recommendation model may be stored in the server 105, or may be stored in the terminal devices 101, 102, and 103.
With continued reference to fig. 4, fig. 4 is a schematic diagram of an application scenario of the method for acquiring information according to the present embodiment. In the application scenario of fig. 4, the user browses the web page of item a through the terminal device 103. The terminal device 103 may acquire the process information of the webpage of the article a being browsed by the user and send the process information to the server 105. The server 105 imports the acquired process information into the item recommendation model to obtain the predicted item information B corresponding to the item a. After that, the server 105 may transmit the predicted item information B to the terminal device 103. The terminal device 103 may display the predicted item information B at a specified position of the page currently viewed by the user.
The method provided by the embodiment of the application firstly acquires the process information of the webpage browsed by the user; then, the process information is imported into an article recommendation model to obtain the predicted article information, so that the accuracy of obtaining the predicted article information is improved; and finally, displaying the predicted article information on a webpage, so that a user can conveniently check the predicted article information.
With further reference to fig. 5, as an implementation of the methods shown in the above-mentioned figures, the present application provides an embodiment of an apparatus for acquiring information, which corresponds to the method embodiment shown in fig. 2, and which is particularly applicable to various electronic devices.
As shown in fig. 5, the apparatus 500 for acquiring information of the present embodiment may include: an information receiving unit 501 and an information acquiring unit 502. The information receiving unit 501 is configured to acquire process information of a user browsing a webpage, where the process information is used to represent information corresponding to the user browsing the webpage; the information obtaining unit 502 is configured to import the process information into a pre-trained item recommendation model for determining the predicted item information according to the process information, and obtain the predicted item information corresponding to the process information.
In some optional implementations of this embodiment, the process information includes browsing characteristic information of at least one web page and a corresponding web page, and the browsing characteristic information includes at least one of: the user browses the information content of the webpage and the browsing time of the corresponding information content.
In some optional implementation manners of this embodiment, the information obtaining unit 502 may include: a web page feature vector obtaining sub-unit (not shown), a web page content feature vector obtaining sub-unit (not shown), and a predicted item information obtaining sub-unit (not shown). The webpage feature vector acquisition subunit is configured to input the process information to the convolutional neural network to obtain a webpage feature vector corresponding to the process information, wherein the convolutional neural network is used for representing a corresponding relationship between the process information and the webpage feature vector; the webpage content feature vector acquisition subunit is configured to input the webpage feature vectors into the recurrent neural network to obtain webpage content feature vectors, wherein the recurrent neural network is used for representing the corresponding relation between the webpage feature vectors and the webpage content feature vectors, and the webpage content feature vectors are used for representing the incidence relation between the webpage feature vectors; the predicted article information acquisition subunit is configured to input the web content feature vector to the fully-connected layer, and obtain predicted article information corresponding to the process information, where the fully-connected layer is used to represent a corresponding relationship between the web content feature vector and the predicted article information.
In some optional implementations of this embodiment, the apparatus 500 for obtaining information may include an item recommendation model training unit (not shown in the figure), and the item recommendation model training unit may include: a sample acquisition subunit (not shown in the figure) and an item recommendation model training subunit (not shown in the figure). The system comprises a sample acquisition subunit, a data processing unit and a data processing unit, wherein the sample acquisition subunit is configured to acquire sample target item information of a plurality of sample target items selected by a user through a webpage and sample process information corresponding to each sample target item in the plurality of sample target items; the article recommendation model training subunit is configured to train the article recommendation model by taking, as an input, sample process information corresponding to each of the plurality of sample target articles and taking, as an output, sample article information of the sample target article corresponding to the sample process information.
In some optional implementations of this embodiment, the item recommendation model training subunit may include: an article recommendation model training module (not shown in the figure) is configured to sequentially input sample process information corresponding to each sample target article in the plurality of sample target articles into an initialized article recommendation model, obtain predicted target article information corresponding to the sample process information, compare the predicted target article information corresponding to each sample process information with the sample target article information corresponding to the sample process information, obtain a prediction accuracy of the initialized article recommendation model, determine whether the prediction accuracy is greater than a preset accuracy threshold, and if the prediction accuracy is greater than the preset accuracy threshold, use the initialized article recommendation model as a trained article recommendation model.
In some optional implementations of this embodiment, the item recommendation model training subunit may include: a parameter adjusting module (not shown in the figures) configured to adjust the parameters of the initialized item recommendation model in response to the accuracy not being greater than the preset accuracy threshold, and continue to perform the training step.
The present embodiment also provides an electronic device, including: one or more processors; a memory having one or more programs stored thereon, which when executed by the one or more processors, cause the one or more processors to perform the above-described method for obtaining information.
The present embodiment also provides a computer-readable medium, on which a computer program is stored, which program, when being executed by a processor, carries out the above-mentioned method for acquiring information.
Referring now to FIG. 6, a block diagram of a computer system 600 suitable for use in implementing a server (e.g., server 105 of FIG. 1) of an embodiment of the present application is shown. The server shown in fig. 6 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present application.
As shown in fig. 6, the computer system 600 includes a Central Processing Unit (CPU)601 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)602 or a program loaded from a storage section 608 into a Random Access Memory (RAM) 603. In the RAM 603, various programs and data necessary for the operation of the system 600 are also stored. The CPU 601, ROM 602, and RAM 603 are connected to each other via a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.
The following components are connected to the I/O interface 605: an input portion 606 including a keyboard, a mouse, and the like; an output portion 607 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section 608 including a hard disk and the like; and a communication section 609 including a network interface card such as a LAN card, a modem, or the like. The communication section 609 performs communication processing via a network such as the internet. The driver 610 is also connected to the I/O interface 605 as needed. A removable medium 611 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 610 as necessary, so that a computer program read out therefrom is mounted in the storage section 608 as necessary.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 609, and/or installed from the removable medium 611. The computer program performs the above-described functions defined in the method of the present application when executed by a Central Processing Unit (CPU) 601.
It should be noted that the computer readable medium mentioned above in the present application may be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In this application, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units described in the embodiments of the present application may be implemented by software or hardware. The described units may also be provided in a processor, and may be described as: a processor includes an information receiving unit and an information acquiring unit. Here, the names of these units do not constitute a limitation of the unit itself in some cases, and for example, the information acquiring unit may also be described as "a unit for acquiring predicted item information by process information".
As another aspect, the present application also provides a computer-readable medium, which may be contained in the apparatus described in the above embodiments; or may be present separately and not assembled into the device. The computer readable medium carries one or more programs which, when executed by the apparatus, cause the apparatus to: acquiring process information of a webpage browsed by a user, wherein the process information is used for representing information corresponding to the webpage browsed by the user; and importing the process information into a pre-trained article recommendation model to obtain predicted article information corresponding to the process information, wherein the article recommendation model is used for determining the predicted article information according to the process information.
The above description is only a preferred embodiment of the application and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the invention herein disclosed is not limited to the particular combination of features described above, but also encompasses other arrangements formed by any combination of the above features or their equivalents without departing from the spirit of the invention. For example, the above features may be replaced with (but not limited to) features having similar functions disclosed in the present application.
Claims (12)
1. A method for obtaining information, the method comprising:
acquiring process information of a webpage browsed by a user, wherein the process information is used for representing information corresponding to the webpage browsed by the user;
importing the process information into a pre-trained article recommendation model to obtain predicted article information corresponding to the process information, wherein the article recommendation model is used for determining the predicted article information according to the process information and comprises the following steps: inputting the process information into a convolutional neural network to obtain a webpage feature vector corresponding to the process information, wherein the convolutional neural network is used for representing the corresponding relation between the process information and the webpage feature vector;
inputting the webpage feature vectors into a recurrent neural network to obtain webpage content feature vectors, wherein the recurrent neural network is used for representing the corresponding relation between the webpage feature vectors and the webpage content feature vectors, and the webpage content feature vectors are used for representing the incidence relation between the webpage feature vectors;
and inputting the webpage content feature vector to a full-link layer to obtain the predicted article information corresponding to the process information, wherein the full-link layer is used for representing the corresponding relation between the webpage content feature vector and the predicted article information.
2. The method of claim 1, wherein the process information includes browsing characteristic information of at least one web page and corresponding web pages, the browsing characteristic information including at least one of: the user browses the information content of the webpage and the browsing time of the corresponding information content.
3. The method of claim 1, wherein the item recommendation model is trained by:
acquiring sample target item information of a plurality of sample target items selected by a user through a webpage and sample process information corresponding to each sample target item in the plurality of sample target items;
and taking the sample process information corresponding to each sample target object in the plurality of sample target objects as input, taking the sample target object information of the sample target object corresponding to the sample process information as output, and training to obtain the object recommendation model.
4. The method according to claim 3, wherein the training of the item recommendation model using the sample process information corresponding to each of the plurality of sample target items as an input and the sample target item information of the sample target item corresponding to the sample process information as an output comprises:
the following training steps are performed: the method comprises the steps of sequentially inputting sample process information corresponding to each sample target object in a plurality of sample target objects into an initialized object recommendation model to obtain predicted target object information corresponding to the sample process information, comparing the predicted target object information corresponding to each sample process information with the sample target object information corresponding to the sample process information to obtain the predicted accuracy of the initialized object recommendation model, determining whether the predicted accuracy is greater than a preset accuracy threshold, and if the predicted accuracy is greater than the preset accuracy threshold, taking the initialized object recommendation model as the trained object recommendation model.
5. The method of claim 4, wherein the training the item recommendation model using the sample process information corresponding to each of the plurality of sample target items as input and the sample target item information corresponding to the sample target item as output further comprises:
and responding to the condition that the accuracy is not larger than the preset accuracy threshold, adjusting the parameters of the initialized item recommendation model, and continuing to execute the training step.
6. An apparatus for obtaining information, the apparatus comprising:
the information receiving unit is configured to acquire process information of a webpage browsed by a user, wherein the process information is used for representing corresponding information when the user browses the webpage;
an information obtaining unit configured to import the process information into a pre-trained item recommendation model for determining predicted item information through the process information, and obtain predicted item information corresponding to the process information, wherein the information obtaining unit includes: the webpage feature vector acquisition subunit is configured to input the process information to a convolutional neural network to obtain a webpage feature vector corresponding to the process information, wherein the convolutional neural network is used for representing a corresponding relationship between the process information and the webpage feature vector;
the webpage content feature vector acquisition subunit is configured to input the webpage feature vectors into a recurrent neural network to obtain the webpage content feature vectors, wherein the recurrent neural network is used for representing the corresponding relation between the webpage feature vectors and the webpage content feature vectors, and the webpage content feature vectors are used for representing the incidence relation between the webpage feature vectors;
and the predicted article information acquisition subunit is configured to input the webpage content feature vector to a full connection layer to obtain predicted article information corresponding to the process information, wherein the full connection layer is used for representing the corresponding relation between the webpage content feature vector and the predicted article information.
7. The apparatus of claim 6, wherein the process information comprises browsing characteristic information of at least one web page and corresponding web pages, the browsing characteristic information comprising at least one of: the user browses the information content of the webpage and the browsing time of the corresponding information content.
8. The apparatus of claim 6, wherein the apparatus comprises an item recommendation model training unit comprising:
a sample acquiring subunit configured to acquire sample target item information of a plurality of sample target items selected by a user through a web page and sample process information corresponding to each of the plurality of sample target items;
and the article recommendation model training subunit is configured to take the sample process information corresponding to each sample target article in the plurality of sample target articles as input, take the sample target article information of the sample target article corresponding to the sample process information as output, and train to obtain the article recommendation model.
9. The apparatus of claim 8, wherein the item recommendation model training subunit comprises:
an item recommendation model training module configured to perform the following training steps: the method comprises the steps of sequentially inputting sample process information corresponding to each sample target object in a plurality of sample target objects into an initialized object recommendation model to obtain predicted target object information corresponding to the sample process information, comparing the predicted target object information corresponding to each sample process information with the sample target object information corresponding to the sample process information to obtain the predicted accuracy of the initialized object recommendation model, determining whether the predicted accuracy is greater than a preset accuracy threshold, and if the predicted accuracy is greater than the preset accuracy threshold, taking the initialized object recommendation model as the trained object recommendation model.
10. The apparatus of claim 9, wherein the item recommendation model training subunit comprises:
a parameter adjustment module configured to adjust a parameter of the initialized item recommendation model in response to not being greater than the preset accuracy threshold, and to continue performing the training step.
11. An electronic device, comprising:
one or more processors;
a storage device having one or more programs stored thereon,
when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-5.
12. A computer-readable medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1 to 5.
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 CN110473042A (en) | 2019-11-19 |
| CN110473042B true 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) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110473042A (en) | 2019-11-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109460513B (en) | Method and apparatus for generating click rate prediction model | |
| CN110473042B (en) | Method and device for acquiring information | |
| WO2018192491A1 (en) | Information pushing method and device | |
| CN108805594B (en) | Information pushing method and device | |
| CN108197652B (en) | Method and apparatus for generating information | |
| CN108920717B (en) | Method and device for displaying information | |
| CN109976997B (en) | Test method and device | |
| CN107679211A (en) | Method and apparatus for pushed information | |
| CN109388548B (en) | Method and apparatus for generating information | |
| CN111538907B (en) | Object recommendation method, system and device | |
| CN108182472A (en) | For generating the method and apparatus of information | |
| CN111695041B (en) | Method and device for recommending information | |
| CN113743971A (en) | Data processing method and device | |
| CN111680213B (en) | Information recommendation method, data processing method and device | |
| CN109582854B (en) | Method and apparatus for generating information | |
| CN109978594B (en) | Order processing method, device and medium | |
| US20220224765A1 (en) | Method and apparatus for pushing information | |
| US20220198487A1 (en) | Method and device for processing user interaction information | |
| CN110555747A (en) | method and device for determining target user | |
| US20240062237A1 (en) | Information processing method and apparatus | |
| US20220027722A1 (en) | Deep Relational Factorization Machine Techniques for Content Usage Prediction via Multiple Interaction Types | |
| CN113792952A (en) | Method and apparatus for generating a model | |
| CN115392943A (en) | Method, device, equipment and storage medium for sending display information | |
| CN110110199B (en) | Information output method and device | |
| CN112348615A (en) | Method and device for auditing 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 |