CN110110210A - The method and apparatus that push shows information - Google Patents

The method and apparatus that push shows information Download PDF

Info

Publication number
CN110110210A
CN110110210A CN201810058428.XA CN201810058428A CN110110210A CN 110110210 A CN110110210 A CN 110110210A CN 201810058428 A CN201810058428 A CN 201810058428A CN 110110210 A CN110110210 A CN 110110210A
Authority
CN
China
Prior art keywords
information
displaying information
displaying
clicking rate
account
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201810058428.XA
Other languages
Chinese (zh)
Inventor
魏望
朱思宇
黄东波
谷俊青
黄浩
陈戈
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tencent Technology Beijing Co Ltd
Original Assignee
Tencent Technology Beijing Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tencent Technology Beijing Co Ltd filed Critical Tencent Technology Beijing Co Ltd
Priority to CN201810058428.XA priority Critical patent/CN110110210A/en
Publication of CN110110210A publication Critical patent/CN110110210A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9535Search customisation based on user profiles and personalisation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q30/00Commerce
    • G06Q30/02Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241Advertisements
    • G06Q30/0251Targeted advertisements
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q30/00Commerce
    • G06Q30/02Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241Advertisements
    • G06Q30/0277Online advertisement

Abstract

The present invention provides a kind of method and apparatus that push shows information, belong to field of computer technology.The described method includes: when receiving the acquisition request of displaying information of terminal transmission, obtain the account information of contextual information and the current logon account of the terminal, wherein, the account information include the essential information of account, before current point in time in preset duration the counting of each Starting mode and browsing displaying information, page iden-tity is carried in the acquisition request, the contextual information is other contents in the corresponding page of the page iden-tity in addition to showing information;According to the page iden-tity, the contextual information and the account information, in the displaying information aggregate of storage, displaying information to be presented is determined;The displaying information to be presented is sent to the terminal.Using the present invention, the validity for showing information can be improved.

Description

The method and apparatus that push shows information
Technical field
The present invention relates to field of computer technology, in particular to a kind of method and apparatus that push shows information.
Background technique
With the development of computer technology and network technology, Internet news business has obtained extensive popularization, and user can be with The application program for installing news category in the terminal, the news information of various channels, such as sports news are browsed by the application program Information, entertainment news information etc..The application program of news category can also show some exhibitions when showing news information in usual terminal Show information, such as advertising information, is browsed for user.
In the related technology, after general terminal detects that the application program of news category is opened, terminal can be to the clothes for showing information The acquisition request for device transmission displaying information of being engaged in, the server can randomly choose some to be presented from showing in information aggregate Show information, be pushed to terminal, terminal can display server push in the application displaying information.
In the implementation of the present invention, inventor find the relevant technologies the prior art has at least the following problems:
Since server is to push to show information at random to terminal, in this way, showing that information is likely to not meet the reality of user Border demand, user will not go to click, and show that the clicking rate of information is relatively low to will lead to.
Summary of the invention
The embodiment of the invention provides a kind of method and apparatus that push shows information, in order to solve asking for the relevant technologies Topic, technical scheme is as follows:
On the one hand, a kind of method that push shows information is provided, which comprises
When receiving the acquisition request of displaying information of terminal transmission, obtains contextual information and the terminal is currently stepped on Record the account information of account, wherein the account information includes the essential information of account, the preset duration before current point in time The counting of interior each Starting mode and the displaying information of browsing carry page iden-tity, the context letter in the acquisition request Breath is other contents in the corresponding page of the page iden-tity in addition to showing information;
According to the page iden-tity, the contextual information and the account information, in the displaying information aggregate of storage, Determine displaying information to be presented;
The displaying information to be presented is sent to the terminal.
On the one hand, a kind of method of determining clicking rate is provided, which comprises
It obtains page iden-tity, contextual information, account information and shows information aggregate, the account information includes account Essential information, before current point in time in preset duration the counting of each Starting mode and browsing displaying information, it is described up and down Literary information is other contents in the corresponding page of the page iden-tity in addition to showing information;
The account information, the contextual information and the displaying information aggregate are inputted into preset clicking rate and estimate mould Type obtains each clicking rate for showing information in the displaying information aggregate.
On the one hand, a kind of device of push displaying information is provided, described device includes:
Obtain module, for when receive terminal transmission displayings information acquisition request when, obtain contextual information with The account information of the current logon account of terminal, wherein the account information includes the essential information of account, in current time The counting of each Starting mode and the displaying information of browsing in preset duration before point carry page mark in the acquisition request Know, the contextual information is other contents in the corresponding page of the page iden-tity in addition to showing information;
Determining module is used for according to the page iden-tity, the contextual information and the account information, in the exhibition of storage Show in information aggregate, determines displaying information to be presented;
Sending module, for sending the displaying information to be presented to the terminal.
On the one hand, a kind of device of determining clicking rate is provided, described device includes:
Module is obtained, for obtaining page iden-tity, contextual information, account information and showing information aggregate, the account Information include the essential information of account, before current point in time in preset duration the counting of each Starting mode and browsing displaying Information, the contextual information are other contents in the corresponding page of the page iden-tity in addition to showing information;
Determining module, it is default for inputting the account information, the contextual information and the displaying information aggregate Clicking rate prediction model, obtain it is described displaying information aggregate in it is each show information clicking rate.
On the one hand, a kind of server is provided, the server includes processor and memory, is stored in the memory There is at least one instruction, described instruction is loaded by the processor and executed to realize and push the method for showing information and determining point The method for hitting rate.
On the one hand, a kind of computer readable storage medium is provided, at least one instruction is stored in the storage medium, Described instruction is loaded by processor and is executed to realize the method for pushing the method for showing information and determining clicking rate.
It is the account information and context in conjunction with account when determining displaying information to be presented in the embodiment of the present invention What information determined, the opposite demand that meet user improves user and clicks the probability for showing information, to can make to show information The validity comparison of displaying is high.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for For those of ordinary skill in the art, without creative efforts, it can also be obtained according to these attached drawings other Attached drawing.
Fig. 1 a is the schematic diagram for the application scenarios that a kind of push provided in an embodiment of the present invention shows information;
Fig. 1 b is the flow diagram for the method that a kind of push provided in an embodiment of the present invention shows information;
Fig. 2 is a kind of display schematic diagram for showing information provided in an embodiment of the present invention;
Fig. 3 is a kind of schematic diagram of trained clicking rate prediction model provided in an embodiment of the present invention;
Fig. 4 is a kind of method schematic diagram of data processing provided in an embodiment of the present invention;
Fig. 5 is a kind of minimum description principle cutting schematic diagram provided in an embodiment of the present invention;
Fig. 6 is the flow diagram that a kind of clicking rate provided in an embodiment of the present invention is estimated;
Fig. 7 is the structural schematic diagram for the device that a kind of push provided in an embodiment of the present invention shows information;
Fig. 8 is the structural schematic diagram for the device that a kind of push provided in an embodiment of the present invention shows information;
Fig. 9 is a kind of structural schematic diagram of the device of determining clicking rate provided in an embodiment of the present invention;
Figure 10 is a kind of structural schematic diagram of server provided in an embodiment of the present invention.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention Formula is described in further detail.
In the related technology, after the application program for opening news category, terminal can be sent out user to the server of the news information The acquisition request of news information is sent, and at the same time to the acquisition request for the server transmission displaying information for showing information.News letter The server of breath is after receiving the acquisition request of news information, available news information, then transmits news letter to terminal Breath.After showing the acquisition request that the server of information receives the displaying information of terminal transmission, show that the server of information can be with Pre-stored displaying information aggregate is obtained, from the displaying information aggregate, determines displaying information to be presented at random, then Displaying information to be presented is sent to terminal.After terminal receives displaying information to be presented, displaying to be presented can be shown Information.In this way, displaying information to be presented is to show that the server of information determines at random, it is possible to not meet the need of user It asks, user is also less likely to click, low so as to cause the validity comparison for showing that information is shown.
In scheme provided in an embodiment of the present invention, show that the server of information is receiving the displaying information of terminal transmission After acquisition request, the account information of the current logon account of available contextual information, terminal and preset displaying information aggregate, Account information includes the essential information of account, the counting of each Starting mode and the displaying information of browsing in preset duration, then According to the page iden-tity carried in acquisition request, contextual information, account information, show information aggregate and preset clicking rate Prediction model determines displaying information to be presented, displaying information to be presented is sent to terminal in showing information aggregate.This Sample, displaying information to be presented are that the account information of account and contextual information is combined to determine, the opposite need that meet user It asks, improves user and click the probability for showing information, to can make to show that the validity comparison that information is shown is high.
Before being implemented, the scene in the embodiment of the present invention is introduced first, and as shown in Figure 1a, terminal can be to exhibition Show that the server of information sends the acquisition request for showing information, and transmits news the acquisition of information to the server of news information Request, the server of news information can determine news information to be presented, news information to be presented sent to terminal, show The server of information can determine displaying information to be presented, and displaying information to be presented is sent to terminal, and terminal can be treated The news information of displaying and displaying information to be presented are shown.
The embodiment of the invention provides a kind of method that push shows information, the executing subject of this method is server.Clothes Business device can be the corresponding server for showing information of application program of news category, and processor can be set in server, connect Device and memory are sent out, processor can be used for handling the relevant treatment in push displaying information process, and receiving-transmitting device can be used for connecing Message is received and sends, memory can be used for storing the data that push shows the data and generation that need in information process.
As shown in Figure 1 b, the process flow of this method can be such that
Step 101, it when receiving the acquisition request of displaying information of terminal transmission, obtains contextual information and terminal is worked as The account information of preceding logon account.
In an implementation, when user wants to see news in the terminal, the application program of news category can be installed at the terminal, so The application program is opened in operation afterwards, and terminal can then show the login interface of the application program, oneself application can be used in user Account and password are logged in, and after logining successfully, terminal can determine the page to be presented, then determine the page to be presented Page iden-tity, the acquisition request for the information that then can transmit news to the server of news information, and at the same time to information is shown Server send the acquisition request for showing information, carry page iden-tity in two acquisition requests, be used to indicate displaying letter Cease the page to display.
The server of news information can be based on page iden-tity, and news information to be presented is sent to terminal.Show information Server receive terminal transmission displaying information acquisition request when, available current contextual information and terminal are worked as The account information of preceding logon account.Wherein, contextual information refers to the content for including in the page that terminal request is shown, the page institute The channel etc. of category, channel have military news channel, social news channel etc., in addition, understanding for convenience, contextual information can also To be interpreted as showing other contents in the information page to be shown in addition to showing information.Show that the server of information can be with Based on the page iden-tity that carries in the acquisition request for showing information, the content for including in the page that terminal request is shown is determined.
In account information including account essential information, before current point in time in preset duration each Starting mode meter The displaying information of number and browsing, the essential information of account includes age, the gender, interest that user inputs when registering the account Deng, the model of geographical location information, terminal, the network of connection etc..Preset duration can be preset by technical staff, and store to In server, such as 1 month, 1 week.
Starting mode includes active Starting mode and passive Starting mode two major classes, and active Starting mode refers to new by triggering The icon for hearing the application program of class, opens the mode of application program, and passive Starting mode is to open to apply journey by triggering link The mode of sequence.By different types of link, passive Starting mode is specific can be divided into following a few classes again: by triggering news category The PUSH message of application program open the mode of application program, taken by triggering the chain for the news information that other good friends share Open the mode of application program, in such a way that application program is opened in the link for the displaying information that other good friends share etc..Certain is opened The counting of flowing mode namely refers to the number for starting application program using this kind of Starting mode.
Step 102, it is determined to be presented in showing information aggregate according to page iden-tity, contextual information and account information Displaying information.
In an implementation, after showing that the server of information gets the account information and contextual information of current logon account, Then the set of the displaying information of available storage obtains preset clicking rate prediction model, contextual information, account are believed Breath shows that information aggregate is input to clicking rate prediction model, obtains showing the corresponding click of each displaying information in information aggregate Rate parses the page iden-tity wherein carried from acquisition request, the page iden-tity and each exhibition then obtained using parsing Show the corresponding clicking rate of information, from showing in information aggregate, determines displaying information to be presented.Wherein, clicking rate refers to displaying letter Breath is clicked the ratio of number Yu shown number, can be used for reflecting the concerned degree of the displaying information.
It should be noted that showing in information aggregate for above-mentioned storage includes showing information operator in the clothes for showing information The displaying information stored in business device shows that information can be advertising information etc., shows that the data of information generally comprise displaying information Picture, link and text etc..
In addition, since the displaying collection credit union stored in the server of displaying information is as the dispensing of information updates always, institute It can also be periodically variable (period can be one minute, two minutes etc.) with the displaying information aggregate of storage, then showing The displaying information aggregate that the server of information obtains is the displaying information aggregate stored when receiving the acquisition request for showing information.
Optionally, show that the server of information can determine the exhibition of browsing according to the displaying information of the browsing got Show the corresponding classification information of information, for example, the displaying information of browsing is that electronic product shows information, classification information is then electronics Product.Show that the server of information can be by the meter of essential information, various Starting mode in contextual information, account information Number, the classification information of the displaying information of browsing, displaying information aggregate are input to clicking rate prediction model, obtain showing information aggregate In the corresponding clicking rate of each displaying information.
Optionally, show that the server of information can be determined to be presented according to the page iden-tity carried in acquisition request Displaying information number, and then determine displaying information to be presented, the processing of corresponding step 102 can be such that
Contextual information, shows information aggregate and preset clicking rate prediction model at account information, believes showing In breath set, the first number is selected to show information, the first number displaying information is determined as displaying information to be presented.
In an implementation, show that the server of information can be according to pair of page iden-tity and the number of displaying information to be presented It should be related to (corresponding relationship can be stored in advance into the server for showing information by technical staff), that determines to receive obtains Take the first number of the corresponding displaying information to be presented of page iden-tity carried in request.Then by contextual information, account Information shows that information aggregate is input to clicking rate prediction model, obtains showing in information aggregate, the corresponding point of each displaying information Rate is hit, then selects the first number to show letter from showing in information aggregate using the corresponding clicking rate of each displaying information First number is shown information, is determined as displaying information to be presented by breath.
Optionally, it shows that the server of information can be used and shows the corresponding click unit price of information, determine exhibition to be presented Show information, the processing of corresponding step 102 can be such that
Contextual information and account information determine the corresponding point of each displaying information in the displaying information aggregate of storage Hit rate;According to the corresponding click unit price of the corresponding clicking rate of each displaying information, each displaying information and page iden-tity, determine to The displaying information of displaying.
In an implementation, show that the server of information gets the account information of current logon account and the displaying information of storage After set, available preset clicking rate prediction model is then defeated by contextual information, account information and displaying information aggregate Enter to clicking rate prediction model, obtains showing in information aggregate, the corresponding clicking rate of each displaying information.
Show that the operator of information when storage displaying information, marks in the server for showing information for each displaying information Corresponding click unit price is determined.When showing the click unit price of information indicates to show that information is clicked one time, the dispensing of information is shown The amount of money that person is paid by operator.
The available each click unit price for showing information of the server of information is shown, according to each correspondence for showing information Clicking rate, each displaying information corresponding click unit price and page iden-tity, determine displaying information to be presented.
Optionally, the corresponding clicking rate of each displaying information can be used and click the product of unit price, determine to be presented Show information, corresponding processing can be such that
For each displaying information, determine that showing that the corresponding clicking rate of information is corresponding with information is shown clicks monovalent the One product;According to page iden-tity and corresponding first product of each displaying information, displaying information to be presented is determined.
In an implementation, for each displaying information, show that the server of information can be by the clicking rate determined and click Unit price is multiplied, and obtains the first product, in this way, being all corresponding with first product for each displaying information.Show information Server the acquisition received can be determined according to the corresponding relationship of page iden-tity and the number of displaying information to be presented (corresponding relationship can be stored in advance to displaying information corresponding second number of the page iden-tity carried in request by technical staff Server in), then will show corresponding first product of each displaying information in information aggregate, be ranked up from big to small, The first product of the second number is chosen since maximum product, and the corresponding displaying of the first product of this second number is believed Breath, is determined as displaying information to be presented.In this way, displaying information to be presented can be determined in conjunction with unit price is clicked, so determining Displaying information to be presented it is relatively more accurate.
It optionally, can also be corresponding to information is shown first in order to keep the displaying information to be presented determined more acurrate Clicking rate corrected, corresponding processing can be such that
According to preset correction model, the corresponding clicking rate of each displaying information is corrected respectively, obtains each exhibition Clicking rate after showing the correction of information;For each displaying information, clicking rate and displaying after determining the correction for showing information are believed Cease the second product of corresponding unit price;According to page iden-tity and corresponding second product of each displaying information, determine to be presented Show information.
In an implementation, show that the server of information gets the account information of current logon account and the displaying information of storage After set, available preset clicking rate prediction model is then defeated by contextual information, account information and displaying information aggregate Enter to clicking rate prediction model, obtains showing in information aggregate, the corresponding clicking rate of each displaying information.
The available preset correction model of server for showing information makees the corresponding clicking rate of each displaying information Clicking rate for the input of correction model, after obtaining each correction for showing information.Show that the server of information is available every A click unit price for showing information shows that the server of information can will be after the correction that determined for each displaying information Clicking rate is multiplied with unit price is clicked, and the second product is obtained, in this way, being all corresponding with second product for each displaying information. Show that the server of information can be determined to receive according to the corresponding relationship of page iden-tity and the number of displaying information to be presented To acquisition request in carry corresponding second number of page iden-tity (corresponding relationship can be stored in advance by technical staff to In the server for showing information), then will show corresponding second product of each displaying information in information aggregate, from big to small into The second product of the second number is chosen in row sequence since maximum product, and the second product of this second number is respectively corresponded Displaying information, be determined as displaying information to be presented.In this way, since clicking rate have passed through correction, so the click determined Rate is more acurrate, and then the displaying information determined is more acurrate.
Optionally, in order to keep the displaying information determined more acurrate, it is also contemplated that show the preset quality coefficient of information, Corresponding processing can be such that
For each displaying information, determines the second product for showing information and show the third of the preset quality coefficient of information Product;According to page iden-tity and the corresponding third product of each displaying information, displaying information to be presented is determined.
In an implementation, the operator for showing information is each exhibition when storage displaying information in the server for showing information Showing that information has all demarcated corresponding preset quality coefficient, preset quality coefficient represents the quality high and low level of the displaying information, (0,1) is typically belonged to, shows that the preset quality coefficient of information is bigger, indicates that quality is higher, conversely, showing the preset quality of information Coefficient is smaller, indicate quality it is lower, if the unsound advertising presupposition quality coefficient of WORDS AND PHRASES IN ADVERTISEMENT is smaller, WORDS AND PHRASES IN ADVERTISEMENT health it is wide The preset quality coefficient of announcement is higher.
For each displaying information, show that the server of information can multiply the preset quality coefficient for showing information with second Product is multiplied, and obtains third product, displaying information each in this way corresponds to a third product.Show that the server of information can incite somebody to action It shows the corresponding third product of each displaying information in information aggregate, is ranked up, is selected since maximum product from big to small The second number third product is taken, this corresponding displaying information of the second number third product is determined as to be presented Show information.In this way, due to considering quality coefficient, so the displaying information determined is more acurrate.
Optionally, it can be such that using the method that correction model is corrected
According to the corresponding relationship of clicking rate range and correction value in preset correction model, determine that each displaying information is corresponding Clicking rate belonging to the corresponding correction value of clicking rate range, according to the corresponding correction value of each displaying information, respectively to each Show the clicking rate that the corresponding clicking rate of information is corrected respectively, after obtaining each correction for showing information.
In an implementation, the corresponding relationship of clicking rate range and correction value is provided in preset correction model, which closes The determination process of system is discussed in detail below.
After showing that the server of information determines the corresponding clicking rate of each displaying information, each displaying information can be determined Clicking rate belonging to clicking rate range, then use the corresponding relationship of clicking rate range and correction value, determine that each displayings is believed Cease the corresponding correction value of clicking rate range belonging to corresponding clicking rate.Correction value can form can be numerical value (deviation Value), it is also possible to coefficient, for any displaying information, if correction value is a deviation, shows that information server can incite somebody to action The corresponding clicking rate of displaying information is added with correction value, the clicking rate after obtaining the corresponding correction of displaying information.If rectified Positive value is a compensation coefficient, for any displaying information, shows that information server can be by the corresponding click of displaying information Rate is multiplied with correction value, the clicking rate after obtaining the corresponding correction of displaying information, in this way, available each displaying information pair Clicking rate after the correction answered.
Step 103, displaying information to be presented is sent to terminal.
It in an implementation, can be to terminal as shown in Fig. 2, after showing that the server of information determines displaying information to be presented Send displaying information to be presented.For each displaying information to be presented, putting person is both provided with corresponding display position, eventually After termination receives displaying information to be presented, the display position of displaying information to be presented can be determined, on the boundary currently shown Displaying information to be presented is shown at the display position in face.
Optionally, as shown in figure 3, in the embodiment of the present invention, the method for giving determining clicking rate prediction model, accordingly Treatment process can be such that
Step 301, sample set is obtained.
In an implementation, the click logs and exposure log, terminal for showing the available record of the server of information are examined every time It measures click and shows information, all can send click logs to the server of information is shown, show that the server of information will record a little Log is hit, and then can recorde the account information of each displaying information being clicked, if some shows that information is not clicked on, The corresponding account information of displaying information is 0.In addition, showing that also record has exposure log on the server of information, exposes log It is shown number for record display information, is shown time etc..
Show that the server of information can get a sample set, sample set by click logs and exposure log In conjunction include a period of time in multiple displaying information content and the corresponding account information of each displaying information, with And the corresponding contextual information of each displaying information.It include the displaying letter at age, gender, interest, historical viewings in account information Breath and browsing show the time (namely clicking the time for showing information) of information, the counting of various Starting mode, position letter Breath etc..Contextual information refers to channel belonging to other contents and the page in the page where showing information, and mentioned above is every A kind of information (age, gender etc.) can be known as showing the corresponding every one-dimensional characteristic of information.In addition, for each displaying Information, it is all corresponding to calculate actual click rate, that is, using being clicked number divided by display number.In this way, for each displaying Information, all calibration have the content for showing information, click the account information and contextual information, reality for showing the user of information Clicking rate (actual click rate is the label for showing information, is not feature).
In addition, can also be filtered to the data got, before continuing step 302 after execution if same displaying The duration that information is clicked twice by the same user is less than preset duration (such as 2 seconds), then it is assumed that and it is to click maliciously, it can be by this Click record deletion twice.
Each feature and label (clicking rate) for showing information is indicated with the form of table, it can be as shown in Table 1:
Table one
Step 302, according to sample set, repetitive exercise obtains clicking rate prediction model.
In an implementation, the server of information is shown after getting sample set, due to determining that clicking rate belongs to two classification Problem (i.e. there are two types of outputs only, respectively represents two classifications, click (classification 1) and do not click (classification 0)), and logic can be used Regression algorithm is trained, and particular by Sigmoid function, the rate of will click on is converted into 0~1, and formula expression can be such that
Assuming that the case where corresponding feature of each displaying information meets linear barrier in sample set, for linear barrier's Situation, boundary regime are as follows:
X in formula (2)iIndicate every one-dimensional characteristic, θiIndicating the corresponding parameter to be trained of every one-dimensional characteristic, T indicates transposition, Above-mentioned θ, x is vector, θ=[θ012,......,θn]T
Structure forecast function are as follows:
For the sample set of input, classification results (namely show that information is clicked for the probability of classification 1 and classification 0 Probability and the probability that is not clicked) be respectively as follows:
P (y=1 | x, θ)=hθ(x), P (y=0 | x, θ)=1-hθ(x) (4)
Probabilistic Synthesis is got up to obtain based on formula (4):
P (y | x, θ)=(hθ(x))y(1-hθ(x))1-y (5)
Likelihood function is taken to formula (5) are as follows:
In formula (6), m indicates that the number that information is shown in sample set, each displaying information have n feature.
It is that logarithm is taken to obtain to L (θ) in formula (6):
Then maximal possibility estimation is carried out to formula (7) and obtains θ so that l (θ) is maximized, θ is optimum value.
Wherein, when carrying out maximal possibility estimation, the embodiment of the invention also provides the iteration renewal processes of θ:
Assuming that loss function is J (θ), loss function J (θ) is taken as:
Because having multiplied a negative coefficient (- 1/m), when taking the minimum value of J (θ), the maximum value of as l (θ), also The optimum value of θ is obtained.
The subsequent minimum value that J (θ) is sought using gradient descent method.
Following present the renewal processes of θ:
Wherein, in formula (9), j indicates the number of iteration.
In this way, convolution (10), formula (9) can be indicated are as follows:
In this way, obtaining the minimum value of J (θ) based on above-mentioned successive ignition, the optimum value of θ is also just obtained, by the best of θ Value is updated to anticipation functionAlso clicking rate prediction model has just been obtained, it is subsequent to can be used for The clicking rate for showing information is predicted.
Optionally, since in the embodiment of the present invention, the format of the data of certain features is possible to not be instruction in sample set Practice the format of data, so can also be first in sample set, the data of various formats are handled, obtain model training Available format, corresponding processing can be such that
Data in sample set are handled, obtain can be used for trained data;According to can be used for trained data, Repetitive exercise obtains clicking rate prediction model.
In an implementation, show that the server of information can be handled multiple data for showing information, and to each exhibition Show that the data of the corresponding account information of information are handled, and to the corresponding contextual information of each displaying information Data handled, the data that obtain that treated.
Then using treated data, it is iterated training, clicking rate prediction model is obtained, how to train and clicked Rate prediction model has been described in detail in front, and details are not described herein again.
Optionally, as follows to the processing mode of the data of various features:
According to the corresponding processing mode of each feature, the data of each feature are handled.
In an implementation, as shown in figure 4, the data of general features can be divided into unordered type, continuous type and text-type etc., for Unordered type, have following several processing modes: property field (numberical range for limiting data), outlier processing (is demarcated exceptional value For null value or it is demarcated as undetected feature), vectorization (can carry out classification processing, clustering processing etc.), and Hash strategy (is breathed out Uncommon strategy can have the selection of hash space and hash function, and there are many selections of hash function, as MurMuHash32, MD5SUM etc., hash space have global Hash and local Hash), feature come into force frequency (by the frequency of occurrences of controlling feature into Row initial option, that is to say, that certain number occur and just counted).
As shown in figure 4, have following several processing modes: exceptional value (is demarcated as null value by outlier processing for continuous type Or it is demarcated as undetected feature), characteristic value bound (upper bound and the lower bound that determine characteristic value), feature discretization (waits frequency Discretization, equidistant discretization, division based on entropy etc.), feature normalization (extreme value offset and relative width).
Multiple data for showing information are usually picture and text etc., so belonging to the data of unordered type and the number of text-type According to can be handled according to any in several processing modes of unordered type.Contextual information usually includes showing in the page The content shown is typically also picture and text, and belongs to the data of unordered type and the data of text-type.Each displaying is believed The data for ceasing corresponding account information, since account information includes age, gender, interest, geographical location information, terminal Model, the counting of various Starting mode, time, the displaying information of historical viewings of clicking displaying information etc..So account information Data in existing unordered type data, also have the data of continuous type, carried out so processing corresponding to itself is all respectively adopted Processing.It is illustrated below with the processing mode of the counting of each Starting mode:
According to minimum description length principle, the counting of each Starting mode is handled.
The counting of general each Starting mode is one-hot coding, and one-hot coding is an efficient coding, for example, one shares three Kind color, yellow, green and red, yellow 0, green are 1, and red is 2, and after one-hot coding, yellow can be expressed as (1,0,0), green can be expressed as (0,1,0), and red can be expressed as (0,0,1).Start once due to every, it is corresponding Vector is increased by one-dimensional, in this way, vector can be made tediously long, so needing to handle the characteristic value of this category feature, adopts herein It is handled with minimum description length principle, carries out cutting processing.As shown in figure 5, characteristic value to be slit is added to be cut Unit's column first look at whether queue to be slit is sky, then terminate if it is sky, if non-empty, therefrom obtain characteristic value, into Row cutting carries out first time cutting, the entropy of the characteristic value before calculating cutting, E=SUM (pi*log pi), it calculates and cuts after cutting Entropy after point, SplitE=(| S1 | * E1+|S2|*E2)/| S |, after S1 indicates that the length of the first segment after cutting, S2 indicate cutting Second segment length, S indicate cutting before total length, E1The entropy of first segment after indicating cutting, E2Second after indicating cutting The entropy of section.Then information gain E-SplitE is calculated, (default value can be by technology people if information gain is greater than default value Member is default, and stores into the server for showing information), then the characteristic value after cutting is put into queue to be slit, continue into Row cutting, the information gain after n-th cutting are less than default value, the N-1 times cutting are determined as best cutting, so After rejudge queue to be slit whether be it is empty, if information gain is less than default value, rejudging queue to be slit is No is empty.
For example, if actively start it is primary be counted as 1, the summation of active Starting mode is 100, is compiled then carrying out solely heat Actively start the vector that primary vector is 100 dimensions after code, if after carrying out cutting, the segmentation result of active Starting mode is 0, 2,6,12,34,100, that is, be divided into 5 sections for 100, i.e., (0,2], (2,6], (6,12], (12,34], (34,100], this Sample, actively starting primary vector is (1,0,0,0,0), and the dimension for actively starting 5 times is (0,1,0,0,0), such vector Dimension can reduce.
In addition, complete training and process for using figure are additionally provided in the embodiment of the present invention, as shown in fig. 6, can be divided into Data processing platform (DPP), machine learning platform show information push platform, close in data processing platform (DPP) for collecting sample set, machine Device learning platform is used for training data, obtains the prediction model of clicking rate, shows that information push platform is used for based on clicking rate Prediction model push shows information.It is specific:
Show the server of information by obtaining click logs, exposure log and multiple displaying information in data processing platform (DPP) And the corresponding account information of each displaying information, the data that then will acquire are labeled, that is, mark out multiple exhibitions Show the corresponding account information of each displaying information, clicking rate and contextual information in information.Show that the server of information passes through number According to processing processing platform, above-mentioned sample set is transmitted to machine learning platform, it can be to getting by machine learning platform Sample set in data handled, obtain can be used for trained data, that is, the training log to be standardized, so After be trained to obtain clicking rate prediction model.The rate prediction model of will click on is sent to displaying information push platform, passes through displaying Clicking rate prediction model in information push platform can be pushed to information is shown.
In addition, being additionally provided with correction model in machine learning platform, the method for determination of correction model be can be such that
Machine learning platform is estimated by clicking rate of the clicking rate prediction model to the displaying information in sample set, The clicking rate estimated is compared with actual clicking rate, is then divided into actual click rate according to the method for cluster Multiple clicking rate ranges make the difference the actual click rate for including in some clicking rate range with the corresponding clicking rate estimated respectively Value, obtained multiple differences are averaged, and obtained average value is the corresponding correction value of clicking rate range, based on identical The corresponding correction value of the available each clicking rate range of method.In this way, just having obtained clicking rate model in correction model Enclose the corresponding relationship with correction value.
In addition, correction model can be sent to displaying letter by machine learning platform after determining correction model Breath push platform.
It should be noted that the data got in sample set are all remembered using distribution in the embodiment of the present invention Record.
It should also be noted that, in the embodiment of the present invention, the server for the news information mentioned and the service for showing information Device can also be same server.
In addition, being added to Starting mode, it is possible to make the click trained due in training clicking rate prediction model The corresponding AUC of rate prediction model (Area under curve, area under a curve) increases.The numerical value of general AUC is bigger, classification Device is better, so clicking rate prediction model can be made more accurate in conjunction with Starting mode.
It is the account information and context in conjunction with account when determining displaying information to be presented in the embodiment of the present invention What information determined, the opposite demand that meet user improves user and clicks the probability for showing information, to can make to show information The validity comparison of displaying is high.
In addition, additionally providing one kind in the embodiment of the present invention based on clicking rate prediction model, determining the method tool of clicking rate Body method can be such that
It obtains page iden-tity, contextual information, account information and shows information aggregate, account information includes the basic of account Information, before current point in time in preset duration the counting of each Starting mode and browsing displaying information, contextual information is Other contents in the corresponding page of page iden-tity in addition to showing information;By account information, contextual information and show information Set inputs preset clicking rate prediction model, obtains showing each clicking rate for showing information in information aggregate.
Concrete processing procedure may refer to the process that clicking rate is determined described in preceding embodiment, and details are not described herein again.
In the embodiment of the present invention, clicking rate is calculated due to combining Starting mode, it is described to make calculated click Rate is more acurrate.
Based on the same technical idea, the embodiment of the invention also provides the devices that a kind of push shows information, such as Fig. 7 institute Show, which includes:
Module 710 is obtained, for obtaining contextual information when receiving the acquisition request of displaying information of terminal transmission With the account information of the current logon account of the terminal, wherein the account information includes the essential information of account, when current Between put before the counting of each Starting mode and the displaying information of browsing in preset duration, carry page mark in the acquisition request Know, the contextual information is other contents in the corresponding page of the page iden-tity in addition to showing information;
Determining module 720, for storing according to the page iden-tity, the contextual information and the account information Displaying information aggregate in, determine displaying information to be presented;
Sending module 730, for sending the displaying information to be presented to the terminal.
Optionally, the determining module 720, is used for:
According to the contextual information and the account information, each displaying information in the displaying information aggregate of storage is determined Corresponding clicking rate;
According to the corresponding clicking rate of each displaying information, the corresponding click unit price of each displaying information and described Page iden-tity determines displaying information to be presented.
Optionally, the determining module 720, is used for:
For each displaying information, determine that the corresponding clicking rate of the displaying information is corresponding with the displaying information Click the first product of unit price;
According to the page iden-tity and corresponding first product of each displaying information, displaying letter to be presented is determined Breath.
Optionally, the determining module 720, is used for:
According to preset correction model, the corresponding clicking rate of each displaying information is corrected respectively, obtains institute Clicking rate after stating each correction for showing information;
For each displaying information, clicking rate and the displaying information pair after determining the correction for showing information Second product of the unit price answered;
According to the page iden-tity and corresponding second product of each displaying information, displaying letter to be presented is determined Breath.
Optionally, the determining module 720, is used for:
For each displaying information, second product for showing information and the default matter for showing information are determined The third product of coefficient of discharge;
According to the page iden-tity and the corresponding third product of each displaying information, displaying letter to be presented is determined Breath.
Optionally, the determining module 720, is used for:
According to the corresponding relationship of clicking rate range and correction value in preset correction model, each displaying information is determined The corresponding correction value of clicking rate range belonging to corresponding clicking rate;
According to the corresponding correction value of each displaying information, to the corresponding clicking rate of each displaying information respectively into Row correction, the clicking rate after obtaining each correction for showing information.
Optionally, the determining module 720, is used for:
According to the page iden-tity, the first number of displaying information to be presented is determined;
According to the contextual information and the account information, in the displaying information aggregate, first number is selected Mesh shows information, and first number displaying information is determined as displaying information to be presented.
Optionally, the determining module 720, is used for:
The page iden-tity, the contextual information, the account information and the displaying information aggregate of storage are input to Preset clicking rate prediction model obtains displaying information to be presented.
The acquisition module 710, is also used to:
Obtain sample set, wherein include multiple displaying information and the multiple displaying information in the sample set In the corresponding account information of each displaying information, contextual information, shown number;
As shown in figure 8, described device further include:
Training module 740, for according to the sample set, repetitive exercise to obtain the clicking rate prediction model.
Optionally, the training module 740, is used for:
Data in the sample set are handled, obtain can be used for trained data;
It can be used for trained data according to described, repetitive exercise obtains the clicking rate prediction model.
It is the account information and context in conjunction with account when determining displaying information to be presented in the embodiment of the present invention What information determined, the opposite demand that meet user improves user and clicks the probability for showing information, to can make to show information The validity comparison of displaying is high.
It should be understood that it is provided by the above embodiment push show information device push show information when, only with The division progress of above-mentioned each functional module can according to need and for example, in practical application by above-mentioned function distribution by not Same functional module is completed, i.e., the internal structure of device is divided into different functional modules, to complete whole described above Or partial function.In addition, the method that push provided by the above embodiment shows that the device of information and push show information is implemented Example belongs to same design, and specific implementation process is detailed in embodiment of the method, and which is not described herein again.
Based on the same technical idea, the embodiment of the invention also provides a kind of devices of determining clicking rate, such as Fig. 9 institute Show, which includes:
Module 910 is obtained, for obtaining page iden-tity, contextual information, account information and showing information aggregate, the account Family information include the essential information of account, before current point in time in preset duration the counting of each Starting mode and browsing exhibition Show information, the contextual information is other contents in the corresponding page of the page iden-tity in addition to showing information;
Determining module 920, it is pre- for inputting the account information, the contextual information and the displaying information aggregate If clicking rate prediction model, obtain it is described displaying information aggregate in it is each show information clicking rate.
In the embodiment of the present invention, clicking rate is calculated due to combining Starting mode, it is described to make calculated click Rate is more acurrate.
It should be understood that the device of determining clicking rate provided by the above embodiment is when determining clicking rate, only with above-mentioned The division progress of each functional module can according to need and for example, in practical application by above-mentioned function distribution by different Functional module is completed, i.e., the internal structure of device is divided into different functional modules, with complete it is described above whole or Partial function.In addition, the embodiment of the method for the device of determining clicking rate provided by the above embodiment and determining clicking rate belongs to together One design, specific implementation process are detailed in embodiment of the method, and which is not described herein again.
Referring to FIG. 10, the server can it illustrates the structural schematic diagram of server involved in the embodiment of the present invention The method for showing information with the push for implementing to provide in above-described embodiment.Specifically:
The server 1000 can generate bigger difference because configuration or performance are different, may include one or one with It upper central processing unit (central processing units, CPU) 1022 (for example, one or more processors) and deposits Storage medium 1030 (such as one or one of reservoir 1032, one or more storage application programs 1042 or data 1044 A above mass memory unit).Wherein, memory 1032 and storage medium 1030 can be of short duration storage or persistent storage.It deposits Storage may include one or more modules (diagram does not mark) in the program of storage medium 1030, and each module may include To the series of instructions operation in server.Further, central processing unit 1022 can be set to and storage medium 1030 Communication executes the series of instructions operation in storage medium 1030 on server 1000.
Server 1000 can also include one or more power supplys 1026, one or more wired or wireless nets Network interface 1050, one or more input/output interfaces 1058, one or more keyboards 1056, and/or, one or More than one operating system 1041, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM Etc..
Server 1000 may include have memory and one perhaps more than one program one of them or one A procedure above is stored in memory, and is configured to execute one or one by one or more than one processor A procedure above includes the method for showing information for carrying out above-mentioned push and determining clicking rate.
The embodiment of the invention also provides a kind of computer readable storage medium, stored in the computer readable storage medium There is at least one instruction, described instruction is loaded by processor and executed to realize that above-mentioned push shows the method and determination of information The method of clicking rate.
Those of ordinary skill in the art will appreciate that realizing that all or part of the steps of above-described embodiment can pass through hardware It completes, relevant hardware can also be instructed to complete by program, the program can store in a kind of computer-readable In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all in spirit of the invention and Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.

Claims (15)

1. a kind of method that push shows information, which is characterized in that the described method includes:
When receiving the acquisition request of displaying information of terminal transmission, obtains contextual information and the terminal currently logs in account The account information at family, wherein the account information includes the essential information of account, each in preset duration before current point in time The counting of Starting mode and the displaying information of browsing carry page iden-tity in the acquisition request, and the contextual information is Other contents in the corresponding page of the page iden-tity in addition to showing information;
According to the page iden-tity, the contextual information and the account information, in the displaying information aggregate of storage, determine Displaying information to be presented;
The displaying information to be presented is sent to the terminal.
2. the method according to claim 1, wherein described according to the page iden-tity, the contextual information Displaying information to be presented is determined in the displaying information aggregate of storage with the account information, comprising:
According to the contextual information and the account information, determine that each displaying information is corresponding in the displaying information aggregate of storage Clicking rate;
According to the corresponding clicking rate of each displaying information, each corresponding click unit price of displaying information and the page Mark, determines displaying information to be presented.
3. according to the method described in claim 2, it is characterized in that, described according to the corresponding click of each displaying information Rate, each corresponding click unit price of displaying information and the page iden-tity, determine displaying information to be presented, comprising:
For each displaying information, the click corresponding with the displaying information of the corresponding clicking rate of the displaying information is determined First product of unit price;
According to the page iden-tity and corresponding first product of each displaying information, displaying information to be presented is determined.
4. according to the method described in claim 2, it is characterized in that, described according to the corresponding click of each displaying information Rate, each corresponding click unit price of displaying information and the page iden-tity, determine displaying information to be presented, comprising:
According to preset correction model, the corresponding clicking rate of each displaying information is corrected respectively, is obtained described every Clicking rate after a correction for showing information;
For each displaying information, the clicking rate after determining the correction for showing information is corresponding with the displaying information Second product of unit price;
According to the page iden-tity and corresponding second product of each displaying information, displaying information to be presented is determined.
5. according to the method described in claim 4, it is characterized in that, described believe according to the page iden-tity with each displaying Corresponding second product is ceased, determines displaying information to be presented, comprising:
For each displaying information, second product for showing information and the preset quality system for showing information are determined Several third products;
According to the page iden-tity and the corresponding third product of each displaying information, displaying information to be presented is determined.
6. according to the method described in claim 3, it is characterized in that, described according to preset correction model, to each exhibition Show that the corresponding clicking rate of information is corrected respectively, the clicking rate after obtaining each correction for showing information, comprising:
According to the corresponding relationship of clicking rate range and correction value in preset correction model, determine that each displaying information is corresponding Clicking rate belonging to the corresponding correction value of clicking rate range;
According to the corresponding correction value of each displaying information, the corresponding clicking rate of each displaying information is rectified respectively Just, the clicking rate after obtaining each correction for showing information.
7. the method according to claim 1, wherein described according to the page iden-tity, the contextual information Displaying information to be presented is determined in the displaying information aggregate of storage with the account information, comprising:
According to the page iden-tity, the first number of displaying information to be presented is determined;
According to the contextual information and the account information, in the displaying information aggregate, first number is selected It shows information, first number displaying information is determined as displaying information to be presented.
8. the method according to claim 1, wherein described according to the page iden-tity, the contextual information Displaying information to be presented is determined in the displaying information aggregate of storage with the account information, comprising:
The page iden-tity, the contextual information, the account information and the displaying information aggregate of storage are input to default Clicking rate prediction model, obtain displaying information to be presented.
9. according to the method described in claim 8, it is characterized in that, the method also includes:
Obtain sample set, wherein include every in multiple displaying information and the multiple displaying information in the sample set The corresponding account information of a displaying information, contextual information, shown number;
According to the sample set, repetitive exercise obtains the clicking rate prediction model.
10. according to the method described in claim 9, repetitive exercise obtains institute it is characterized in that, described according to the sample set State clicking rate prediction model, comprising:
Data in the sample set are handled, obtain can be used for trained data;
It can be used for trained data according to described, repetitive exercise obtains the clicking rate prediction model.
11. a kind of method of determining clicking rate, which is characterized in that the described method includes:
It obtains page iden-tity, contextual information, account information and shows information aggregate, the account information includes the basic of account Information, before current point in time in preset duration the counting of each Starting mode and browsing displaying information, context letter Breath is other contents in the corresponding page of the page iden-tity in addition to showing information;
The account information, the contextual information and the displaying information aggregate are inputted into preset clicking rate prediction model, Obtain each clicking rate for showing information in the displaying information aggregate.
12. the device that a kind of push shows information, which is characterized in that described device includes:
Module is obtained, for when receiving the acquisition request of displayings information of terminal transmission, acquisition contextual information and described The account information of the current logon account of terminal, wherein the account information include account essential information, current point in time it The displaying information of the counting of each Starting mode and browsing in preceding preset duration carries page iden-tity in the acquisition request, institute Stating contextual information is other contents in the corresponding page of the page iden-tity in addition to showing information;
Determining module, for being believed in the displaying of storage according to the page iden-tity, the contextual information and the account information In breath set, displaying information to be presented is determined;
Sending module, for sending the displaying information to be presented to the terminal.
13. a kind of device of determining clicking rate, which is characterized in that the described method includes:
Module is obtained, for obtaining page iden-tity, contextual information, account information and showing information aggregate, the account information Essential information including account, the counting of each Starting mode and the displaying of browsing letter in preset duration before current point in time Breath, the contextual information are other contents in the corresponding page of the page iden-tity in addition to showing information;
Determining module, for the account information, the contextual information and the displaying information aggregate to be inputted preset point Rate prediction model is hit, each clicking rate for showing information in the displaying information aggregate is obtained.
14. a kind of server, which is characterized in that the server includes processor and memory, is stored in the memory At least one instruction, described instruction are loaded by the processor and are executed to realize pushing away as described in claims 1 to 10 is any The method and the method as claimed in claim 11 for determining clicking rate of information are shown in putting on display.
15. a kind of computer readable storage medium, which is characterized in that be stored at least one instruction, institute in the storage medium State instruction load by processor and is executed with realize as described in claims 1 to 10 is any push displaying information method and The method as claimed in claim 11 for determining clicking rate.
CN201810058428.XA 2018-01-22 2018-01-22 The method and apparatus that push shows information Pending CN110110210A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810058428.XA CN110110210A (en) 2018-01-22 2018-01-22 The method and apparatus that push shows information

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810058428.XA CN110110210A (en) 2018-01-22 2018-01-22 The method and apparatus that push shows information

Publications (1)

Publication Number Publication Date
CN110110210A true CN110110210A (en) 2019-08-09

Family

ID=67483467

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810058428.XA Pending CN110110210A (en) 2018-01-22 2018-01-22 The method and apparatus that push shows information

Country Status (1)

Country Link
CN (1) CN110110210A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111144974A (en) * 2019-12-04 2020-05-12 北京三快在线科技有限公司 Information display method and device
CN112765510A (en) * 2021-01-28 2021-05-07 维沃移动通信有限公司 Information display method and device, electronic equipment and readable storage medium

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105208113A (en) * 2015-08-31 2015-12-30 北京百度网讯科技有限公司 Information pushing method and device
CN105450586A (en) * 2014-07-24 2016-03-30 阿里巴巴集团控股有限公司 Information pushing method and system, server and client
CN105528408A (en) * 2015-12-03 2016-04-27 百度在线网络技术(北京)有限公司 Page display method and apparatus
CN105678335A (en) * 2016-01-08 2016-06-15 车智互联(北京)科技有限公司 Click rate pre-estimation method, device and calculating equipment
CN105678587A (en) * 2016-01-12 2016-06-15 腾讯科技(深圳)有限公司 Recommendation feature determination method and information recommendation method and device
CN106156878A (en) * 2015-04-21 2016-11-23 深圳市腾讯计算机系统有限公司 Ad click rate antidote and device
CN107015980A (en) * 2016-01-27 2017-08-04 腾讯科技(北京)有限公司 It is a kind of enter row information displaying method and apparatus

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105450586A (en) * 2014-07-24 2016-03-30 阿里巴巴集团控股有限公司 Information pushing method and system, server and client
CN106156878A (en) * 2015-04-21 2016-11-23 深圳市腾讯计算机系统有限公司 Ad click rate antidote and device
CN105208113A (en) * 2015-08-31 2015-12-30 北京百度网讯科技有限公司 Information pushing method and device
CN105528408A (en) * 2015-12-03 2016-04-27 百度在线网络技术(北京)有限公司 Page display method and apparatus
CN105678335A (en) * 2016-01-08 2016-06-15 车智互联(北京)科技有限公司 Click rate pre-estimation method, device and calculating equipment
CN105678587A (en) * 2016-01-12 2016-06-15 腾讯科技(深圳)有限公司 Recommendation feature determination method and information recommendation method and device
CN107015980A (en) * 2016-01-27 2017-08-04 腾讯科技(北京)有限公司 It is a kind of enter row information displaying method and apparatus

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111144974A (en) * 2019-12-04 2020-05-12 北京三快在线科技有限公司 Information display method and device
CN112765510A (en) * 2021-01-28 2021-05-07 维沃移动通信有限公司 Information display method and device, electronic equipment and readable storage medium
CN112765510B (en) * 2021-01-28 2024-02-27 维沃移动通信有限公司 Information display method, information display device, electronic equipment and readable storage medium

Similar Documents

Publication Publication Date Title
US10339220B2 (en) Monitoring conversations to identify topics of interest
US9455891B2 (en) Methods, apparatus, and articles of manufacture to determine a network efficacy
US20170351772A1 (en) Modeling actions for entity-centric search
US8417654B1 (en) Decision tree refinement
US11429988B2 (en) Method and system for increasing use of mobile devices to provide answer content in a question and answer based customer support system
US8542888B2 (en) Detecting facial similarity based on human perception of facial similarity
EP3115957A1 (en) System and method for identifying reviewers with incentives
US20120266191A1 (en) System and method to provide messages adaptive to a crowd profile
US20180300767A1 (en) Computer-based systems, apparatuses and methods for a social media platform for processing internet traffic through advertising revenue
US20110258256A1 (en) Predicting future outcomes
US20210357528A1 (en) Secured use of private user data by third party data consumers
US20190012591A1 (en) System and method for detecting the occurrence of an event and determining a response to the event
CN107809370B (en) User recommendation method and device
CN109191207A (en) Determine the method and device of advertisement Recommended Price
CN110782286A (en) Advertisement pushing method and device, server and computer readable storage medium
CN109493138A (en) Information recommendation method, device, server and storage medium
CN110110210A (en) The method and apparatus that push shows information
JP2010079683A (en) Program and advertisement distribution system
KR101811751B1 (en) Advertisement providing server using chatbot
CN108401005B (en) Expression recommendation method and device
US20130006705A1 (en) Small business intelligence tool
US20170365014A1 (en) Systems, methods and non-transitory computer readable storage media for tracking and evaluating predictions regarding relationships
CN114637927B (en) Content recommendation method and device, electronic equipment and storage medium
CN113362095A (en) Information delivery method and device
WO2023082864A1 (en) Training method and apparatus for content recommendation model, device, and storage medium

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