CN103150663A - Method and device for placing network placement data - Google Patents
Method and device for placing network placement data Download PDFInfo
- Publication number
- CN103150663A CN103150663A CN2013100529133A CN201310052913A CN103150663A CN 103150663 A CN103150663 A CN 103150663A CN 2013100529133 A CN2013100529133 A CN 2013100529133A CN 201310052913 A CN201310052913 A CN 201310052913A CN 103150663 A CN103150663 A CN 103150663A
- Authority
- CN
- China
- Prior art keywords
- data
- user
- active user
- click probability
- neighbour
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Information Transfer Between Computers (AREA)
Abstract
The invention provides a method and a device for placing network placement data. The method comprises the following steps of: counting a click probability of a current user aiming at the network placement data in different categories; acquiring a neighbor user related to the current user; extracting the click probability of the neighbor user aiming at the network placement data in different categories; by the adoption of the click probability of the neighbor user aiming at the network placement data in different categories, adjusting the click probability of the current user aiming at the network placement data in relevant categories; and extracting the network placement data with the click probability larger than a preset threshold value after adjustment, and placing the network placement data to the current user. According to the method and the device, the click probability of the current user aiming at the network placement data in different categories can be precisely forecast, and the accuracy on placing the network placement data is effectively improved.
Description
Technical field
The present invention relates to the technical field that internet and ad data are thrown in, particularly relate to a kind of network and throw in the method that data are thrown in, and a kind of network is thrown in the device that data are thrown in.
Background technology
In the evolution of internet, it is very important business model that the network such as ad data is thrown in data always, and under the continuous promotion of Internet advertising data, internet industry just is able to prosperity like this just.Simultaneously, along with the fast development of Internet technology, people can see different ad datas when browsing same page, realize that the personalization of ad data is showed.In this way, can show that more accurately it may interested ad data to the user.
In correlation technique, ad data is precisely thrown in strategy and is mainly comprised following three classes: search trigger (sponsored search), content matching (content match), user behavior directed (behavioral targeting).
Wherein, the search trigger strategy is mainly to carry out the ad data retrieval according to the keyword that the active user submits to search engine, has directly reflected active user's interest due to keyword, therefore can push the ad data relevant to search content to the active user; The content matching strategy is mainly that the content of the webpage of active user's browsing is carried out modeling analysis, and then shows the ad data close with web page contents to the active user.
But, search trigger strategy and content matching strategy are not considered active user's personalized interest, when different active users searches for same keyword or browses same webpage, the ad data of showing is identical often, but actual conditions are, when two active users searched for same keyword or browse same webpage, its focus may be not identical.
The user behavior directional strategy is mainly the historical behavior record according to the active user, and its interest is carried out modeling analysis, then shows to the active user ad data that meets its characteristics interest.According to the user behavior directional strategy, when different active users searched for same keyword or browses same webpage, the ad data of seeing was not identical; But, the user behavior directional strategy the prediction active user during to the click probability of ad data often not precisely really, thereby can not throw in accurately the interested ad data of active user.
Therefore, need at present the urgent technical matters that solves of those skilled in the art to be exactly: to propose a kind of network and throw in the method that data are thrown in, in order to predict that accurately the active user throws in the click probability of data for different classes of network, effectively improve network and throw in the accuracy that data are thrown in.
Summary of the invention
The invention provides a kind of network and throw in the method that data are thrown in, can predict accurately that the active user throws in the click probability of data for different classes of network, effectively improve network and throw in the accuracy that data are thrown in.
Accordingly, the present invention also provides the device of a kind of network input data input in order to guarantee the implementation and application of said method.
In order to address the above problem, the invention discloses a kind of network and throw in the method that data are thrown in, comprising:
The statistics active user throws in the click probability of data for different classes of network;
Obtain the neighbour user related with described active user;
Extract described neighbour user and throw in the click probability of data for different classes of network;
Adopt described neighbour user to throw in the click probability of data for different classes of network, adjust the active user and throw in the click probability of data for the respective classes network;
The click probability that extracts after described adjustment is thrown in data greater than the network of predetermined threshold value, and it is thrown in to the active user.
Preferably, described network is thrown in data and is comprised ad data, and described statistics active user comprises for the step of the click probability of different classes of ad data:
Obtain active user's historical behavior information;
Extract the active user for the number of clicks of different classes of ad data from described active user's historical behavior information, and the active user is for the displaying number of times of different classes of ad data;
According to the number of clicks of described active user for different classes of ad data, and the active user calculates the active user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
Preferably, described active user's historical behavior information comprises active user's Webpage search information and active user's web page access information.
Preferably, the described neighbour user related with the active user comprises the user who has mutual network behavior with the active user, and/or, with the user of active user's behavior similarity greater than predetermined threshold value.
Preferably, described extraction neighbour user comprises for the step of the click probability of different classes of ad data:
Obtain neighbour user's historical behavior information;
Extract neighbour user for the number of clicks of different classes of ad data from described neighbour user's historical behavior information, and neighbour user is for the displaying number of times of different classes of ad data;
According to the number of clicks of described neighbour user for different classes of ad data, and neighbour user calculates neighbour user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
Preferably, described employing neighbour user is for the click probability of different classes of ad data, and the adjustment active user comprises for the step of the click probability of respective classes ad data:
Adopt following formula to adjust described active user for the click probability of respective classes ad data,
Wherein, u is described active user, Ctr
uBe the click probability of described active user for the respective classes ad data, g is the neighbour user related with described active user, Ctr
gBe the click probability of the neighbour user related with described active user for the respective classes ad data, outdeg (g) is out-degree, and α is ratio of damping.
Preferably, described employing neighbour user is for the click probability of different classes of ad data, and the adjustment active user also comprises for the step of the click probability of respective classes ad data:
Calculate it for the mean value of the click probability of respective classes ad data to described active user and with active user's behavior similarity greater than the user of predetermined threshold value.
The invention also discloses a kind of network and throw in the device that data are thrown in, comprising:
Click the probability statistics module, be used for the statistics active user and throw in the click probability of data for different classes of network;
Neighbour user's acquisition module is used for obtaining the neighbour user related with described active user;
Click the probability extraction module, be used for extracting described neighbour user and throw in the click probability of data for different classes of network;
Click the probability adjusting module, be used for adopting described neighbour user to throw in the click probability of data for different classes of network, adjust the active user and throw in the click probability of data for the respective classes network;
Putting module, the click probability that is used for extracting after described adjustment is thrown in data greater than the network of predetermined threshold value, and it is thrown in to the active user.
Preferably, described network is thrown in data and is comprised ad data, and described click probability statistics module comprises:
The first historical behavior acquisition of information submodule is for the historical behavior information of obtaining the active user;
First extracts submodule, be used for extracting the active user for the number of clicks of different classes of ad data from described active user's historical behavior information, and the active user is for the displaying number of times of different classes of ad data;
First clicks the probability calculation submodule, be used for according to the number of clicks of described active user for different classes of ad data, and the active user calculates the active user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
Preferably, described active user's historical behavior information comprises active user's Webpage search information and active user's web page access information.
Preferably, the described neighbour user related with the active user comprises the user who has mutual network behavior with the active user, and/or, with the user of active user's behavior similarity greater than predetermined threshold value.
Preferably, described click probability extraction module comprises:
The second historical behavior acquisition of information submodule is for the historical behavior information of obtaining neighbour user;
Second extracts submodule, be used for extracting neighbour user for the number of clicks of different classes of ad data from described neighbour user's historical behavior information, and neighbour user is for the displaying number of times of different classes of ad data;
Second clicks the probability calculation submodule, be used for according to the number of clicks of described neighbour user for different classes of ad data, and neighbour user calculates neighbour user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
Preferably, described click probability adjusting module comprises:
Adjust submodule, be used for adopting following formula to adjust described active user for the click probability of respective classes ad data,
Wherein, u is described active user, Ctr
uBe the click probability of described active user for the respective classes ad data, g is the neighbour user related with described active user, Ctr
gBe the click probability of the neighbour user related with described active user for the respective classes ad data, outdeg (g) is out-degree, and α is ratio of damping.
Preferably, described click probability adjusting module also comprises:
The mean value calculation submodule is used for calculating it for the mean value of the click probability of respective classes ad data to described active user and with active user's behavior similarity greater than the user of predetermined threshold value.
Compare with background technology, the present invention has the following advantages:
The present invention is by adopting neighbour user to throw in the click probability of data for different classes of network, adjust the active user and throw in the click probability of data for the respective classes network, can predict accurately that the active user throws in the click probability of data for different classes of network, effectively improve network and throw in the accuracy that data are thrown in.
Description of drawings
Fig. 1 shows the flow chart of steps that a kind of network provided by the invention is thrown in the embodiment of the method 1 of data input;
Fig. 2 shows the flow chart of steps that a kind of network provided by the invention is thrown in the embodiment of the method 2 of data input; What Fig. 3 showed that the inventive method embodiment provides represents the neighbour user's related with the active user schematic diagram with graph model;
What Fig. 4 showed that the inventive method embodiment provides represents the neighbour user's related with the active user schematic diagram with adjacency matrix;
Fig. 5 shows the structured flowchart that a kind of network provided by the invention is thrown in the device embodiment of data input.
Embodiment
For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, the present invention is further detailed explanation below in conjunction with the drawings and specific embodiments.
One of core idea of the embodiment of the present invention is, by adopting neighbour user to throw in the click probability of data for different classes of network, adjust the active user and throw in the click probability of data for the respective classes network, can predict accurately that the active user throws in the click probability of data for the respective classes network, effectively improve network and throw in the accuracy that data are thrown in.
With reference to Fig. 1, show the flow chart of steps that a kind of network provided by the invention is thrown in the embodiment of the method 1 of data input, specifically can comprise the following steps:
Step 102 is obtained the neighbour user related with described active user;
Step 103 is extracted described neighbour user and is thrown in the click probability of data for different classes of network;
Step 104 adopts described neighbour user to throw in the click probability of data for different classes of network, adjusts the active user and throws in the click probability of data for the respective classes network;
As the concrete a kind of example used of the present invention, described network is thrown in data can be ad data.Use the embodiment of the present invention, can throw in the click probability of data for different classes of network by adopting neighbour user, adjust the active user and throw in the click probability of data for the respective classes network, can predict accurately that the active user throws in the click probability of data for the respective classes network, effectively improve network and throw in the accuracy that data are thrown in, in addition, the method also can reduce the unnecessary computing resource waste of server.
With reference to Fig. 2, show the flow chart of steps that a kind of network provided by the invention is thrown in the embodiment of the method 2 of data input, specifically can comprise the following steps: step 201, the statistics active user is for the click probability of different classes of ad data;
In a preferred embodiment of the present invention, described step 201 can comprise following substep:
Substep S211 obtains active user's historical behavior information;
As the concrete a kind of example used of the present invention, described active user's historical behavior information can comprise active user's Webpage search information and active user's web page access information.
In specific implementation, can obtain according to active user's historical search message active user's Webpage search information, according to browse the web page access information that the advertisement message message and active user before clicking or showed obtain active user of active user to webpage.
Particularly, active user's historical search message and active user to webpage browse active user's communications access message that message can obtain according to operator's interface and the User action log of server record obtains, the information such as the search engine that wherein active user's historical search message can comprise that the active user indicates, use active user IP, active user region of living in, search time, active user, search content; The active user to webpage browse that message can comprise that the active user indicates, the information such as active user IP, active user region of living in, access time, access URL, refer (the source page), UA; The active user clicked in the past or the advertisement message showed can extract from the advertisement putting daily record that release platform is collected, and can comprise usually that the active user indicates, the information such as whether the advertising display page, ad spot information, adline, ad related information, displaying time, user are showed, click time, click location.
particularly, active user's Webpage search information can be according to extract the query word information of using from active user's historical search message, then query word is carried out participle, again the result after participle is described as active user's query word and generated, and, the content of active user's Webpage search information can comprise the correlated characteristic the when active user utilized search engine to retrieve in the past, as carry out relevant TF or the TF*IDF feature of query word after participle, the statistical natures such as the average length of the query word that uses, number etc., active user's web page access information can be according to from the active user, browsing of webpage being extracted the page that the active user accesses message, then can grasp the page, analyze the content of pages feature, extract relevant content, keyword, webpage classification etc., and click before the active user or the advertisement message showed generates, and, the content of active user's web page access information can comprise that the active user is to access times and the residence time of different web sites, the active user is the represented characteristic set of content of the webpage of browsing in the past, TF as relevant in content, the TF*IDF feature, classification and the residence time of institute's browsing page, the active user is for displaying number of times and the number of clicks of different classes of ad data, the active user clicks or showed the content characteristic of landing page itself of bid word and content description correlated characteristic and the ad data of ad data.
Substep S212 extract the active user for the number of clicks of different classes of ad data from described active user's historical behavior information, and the active user is for the displaying number of times of different classes of ad data;
Substep S213, according to the number of clicks of described active user for different classes of ad data, and the active user calculates the active user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
As the concrete a kind of example used of the embodiment of the present invention, the computing method of described click probability can be for described active user for the number of clicks of different classes of ad data divided by the displaying number of times of active user for different classes of ad data.For example, the active user is 3 for the number of clicks of game class ad data, and the active user is 10 for the displaying number of times of game class ad data, and the active user is 3 divided by 10 for the click probability of game class ad data so, namely 0.3.
In another kind of preferred embodiment of the present invention, described active user can also calculate by active user's behavioral targeting model for the click probability of different classes of ad data.
particularly, can first obtain active user's historical behavior information, described active user's historical behavior information can comprise active user's Webpage search information and active user's web page access information, then can analyze active user's behavior according to active user's Webpage search information and active user's web page access information, as analyze the interested classification of active user etc., for example, the active user often searches for the relevant keyword of NBA, or often access NBA match related pages, usually show that the active user may be interested in " basketball " or " NBA ", thereby obtain active user's category of interest and be " NBA " or " basketball ", if the user often searches for the relevant content of cosmetology, show that the user is likely a women user, it may be marked as " cosmetology " relevant interest group classification, in addition, can also analyze active user's the information such as sex, age, educational background, the level of consumption, concern theme.
On the basis of active user's behavioural analysis, can make up from the historical behavior information of different dimensions to the active user, form active user's information Multidimensional numerical, this array can be described the active user from different dimensions, in this case, can utilize this array to be proper vector, the active user is desired value to the number of clicks of different classes of ad data, training active user behavioral targeting model.
As the concrete a kind of example used of the embodiment of the present invention, described proper vector is the concrete quantification of each feature on the active user, such as sex is a feature, " man ", " female " are exactly the value of sex, also can represent the man with 0,1 expression female is to certain active user, its value on sex character is exactly 0 or 1, characteristic value jointly consist of active user's proper vector.
After training active user behavioral targeting model, for the new behavior of browsing of active user, can utilize this model prediction active user for the click probability of different classes of ad data.
Step 202 is obtained the neighbour user related with described active user;
As the concrete a kind of example used of the embodiment of the present invention, the described neighbour user related with the active user can comprise the user who has mutual network behavior with the active user, and/or, with the user of active user's behavior similarity greater than predetermined threshold value.
Particularly, the user that described and active user have mutual network behavior can be included in the user who is connected with the active user in social networks, the follow/follower relation on the twitter/ microblogging for example, good friend's relation on facebook etc., can also comprise the user in active user's the Internet, the forwarding relation on twitter for example, relation, the mail interactive relation in mailing system etc. are replied in the comment in blog; Described and active user's behavior similarity can comprise the user in the interest relational network between the user who builds according to search or access behavior greater than the user of predetermined threshold value, for example used the link that can build between the user of same keyword based on common keyword, the user who accessed same web site/webpage also can build linking relationship based on webpage etc., need to prove, may be directed networks based on above-mentioned network, may be also undirected network.
In specific implementation, the described neighbour user related with the active user can represent with graph model, also can be expressed as adjacency matrix, with reference to Fig. 3, show a kind of schematic diagram that represents the neighbour user related with the active user with graph model, with reference to Fig. 4, show a kind of schematic diagram that represents the neighbour user related with the active user with adjacency matrix.
Step 203 is extracted described neighbour user for the click probability of different classes of ad data;
In a preferred embodiment of the present invention, described step 203 can comprise following substep:
Substep S231 obtains neighbour user's historical behavior information;
Substep S232 extract neighbour user for the number of clicks of different classes of ad data from described neighbour user's historical behavior information, and neighbour user is for the displaying number of times of different classes of ad data;
Substep S233, according to the number of clicks of described neighbour user for different classes of ad data, and neighbour user calculates neighbour user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
In the embodiment of the present invention in step 203 substep S231 substep S211 to substep S233 and the step 201 similar to substep S213, be not described in detail in this.
In another kind of preferred embodiment of the present invention, described neighbour user can also calculate by neighbour's user behavior recursive model for the click probability of different classes of ad data.
In the generation of described neighbour's user behavior recursive model and step 201 generation of active user's behavioral targeting model is similar, is not described in detail in this.
Step 204 adopts described neighbour user for the click probability of different classes of ad data, adjusts the active user for the click probability of respective classes ad data;
In a preferred embodiment of the present invention, described employing neighbour user is for the click probability of different classes of ad data, and the adjustment active user can comprise for the step of the click probability of respective classes ad data:
Adopt following formula to adjust described active user for the click probability of respective classes ad data,
Wherein, u is described active user, Ctr
uBe the click probability of described active user for the respective classes ad data, g is the neighbour user related with described active user, Ctr
gBe the click probability of the neighbour user related with described active user for the respective classes ad data, outdeg (g) is out-degree, and α is ratio of damping.
particularly, can only consider in neighbour user and the impact of active user's neighboring user, adopt above-mentioned formula to carry out influence power and propagate, perhaps can copy the PageRank method, adopting above-mentioned formula to carry out iteration to neighbour user upgrades, in this case, because active user and neighbour user thereof can influence each other to each other, so active user and neighbour user's thereof click probability can just can reach basicly stable through a dynamic process, at this moment with regard to the default iterations of needs or until iteration convergence, for example, in theory, need adopting above-mentioned formula to carry out 10000 iteration to active user and neighbour user thereof clicks probability and just can reach stable, but in practice, when iteration to 100 time, active user and neighbour user's thereof click probability is basicly stable, in this case, iterations can be set as 100, perhaps, can adopt above-mentioned formula with clicking rate with the user of active user's behavior similarity greater than predetermined threshold value in expand.
In another preferred embodiment of the present invention, described employing neighbour user is for the click probability of different classes of ad data, and the adjustment active user can also comprise for the step of the click probability of respective classes ad data:
Calculate it for the mean value of the click probability of respective classes ad data to described active user and with active user's behavior similarity greater than the user of predetermined threshold value.
In specific implementation, can carry out arithmetic mean greater than the user of predetermined threshold value for the click probability of respective classes ad data to described active user and with active user's behavior similarity; Perhaps, can be weighted on average for the click probability of respective classes ad data greater than the user of predetermined threshold value to described active user and with active user's behavior similarity; For example, the active user is 0.3 for the click probability of amusement series advertisements data, active user's neighbour user A is 0.2 for amusement series advertisements data, active user's neighbour user B is 0.6 for amusement series advertisements data, active user's neighbour user C is 0.4 for amusement series advertisements data, so described active user and with active user's behavior similarity be (0.3+0.2+0.6+0.4)/4 greater than the user of predetermined threshold value for the arithmetic mean of the click probability of respective classes ad data, namely 0.375; If in above-mentioned example, active user's weight is 0.5, the weight of active user's neighbour user A is 0.1, the weight of active user's neighbour user B is 0.25, the weight of active user's neighbour user C is 0.15, so described active user and with active user's behavior similarity be 0.3*0.5+0.2*0.1+0.6*0.25+0.4*0.15 greater than the user of predetermined threshold value for the weighted mean value of the click probability of respective classes ad data, namely 0.38.
Step 205 is extracted click probability after described adjustment greater than the ad data of predetermined threshold value, and it is thrown in to the active user.
Need to prove, for above-mentioned embodiment of the method, for simple description, therefore it all is expressed as a series of combination of actions, but those skilled in the art should know, the present invention is not subjected to the restriction of described sequence of movement, because according to the present invention, some step can adopt other orders or carry out simultaneously.Secondly, those skilled in the art also should know, the embodiment described in instructions all belongs to preferred embodiment, and related action and module might not be that the present invention is necessary.
With reference to Fig. 5, show the structured flowchart that a kind of network provided by the invention is thrown in the device embodiment of data input, specifically can comprise with lower module:
Click probability statistics module 501, be used for the statistics active user and throw in the click probability of data for different classes of network;
In specific implementation, described network is thrown in data can be ad data.
In a kind of preferred embodiment of the embodiment of the present invention, described network is thrown in data can comprise ad data, and described click probability statistics module 501 can comprise:
The first historical behavior acquisition of information submodule S511 is for the historical behavior information of obtaining the active user;
As the concrete a kind of example used of the embodiment of the present invention, described active user's historical behavior information can comprise active user's Webpage search information and active user's web page access information.
First extracts submodule S512, be used for extracting the active user for the number of clicks of different classes of ad data from described active user's historical behavior information, and the active user is for the displaying number of times of different classes of ad data;
First clicks probability calculation submodule S513, be used for according to the number of clicks of described active user for different classes of ad data, and the active user calculates the active user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
Neighbour user's acquisition module 502 is used for obtaining the neighbour user related with described active user;
As the concrete a kind of example used of the embodiment of the present invention, the described neighbour user related with the active user can comprise the user who has mutual network behavior with the active user, and/or, with the user of active user's behavior similarity greater than predetermined threshold value.
Click probability extraction module 503, be used for extracting described neighbour user and throw in the click probability of data for different classes of network;
In a kind of preferred embodiment of the embodiment of the present invention, described click probability extraction module 503 can comprise:
The second historical behavior acquisition of information submodule S531 is for the historical behavior information of obtaining neighbour user;
Second extracts submodule S532, be used for extracting neighbour user for the number of clicks of different classes of ad data from described neighbour user's historical behavior information, and neighbour user is for the displaying number of times of different classes of ad data;
Second clicks probability calculation submodule S533, be used for according to the number of clicks of described neighbour user for different classes of ad data, and neighbour user calculates neighbour user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
Click probability adjusting module 504, be used for adopting described neighbour user to throw in the click probability of data for different classes of network, adjust the active user and throw in the click probability of data for the respective classes network;
In a kind of preferred embodiment of the embodiment of the present invention, described click probability adjusting module 504 can comprise:
Adjust submodule, be used for adopting following formula to adjust described active user for the click probability of respective classes ad data,
Wherein, u is described active user, Ctr
uBe the click probability of described active user for the respective classes ad data, g is the neighbour user related with described active user, Ctr
gBe the click probability of the neighbour user related with described active user for the respective classes ad data, outdeg (g) is out-degree, and α is ratio of damping.
In the another kind of preferred embodiment of the embodiment of the present invention, described click probability adjusting module 504 can also comprise:
The mean value calculation submodule is used for calculating it for the mean value of the click probability of respective classes ad data to described active user and with active user's behavior similarity greater than the user of predetermined threshold value.
Putting module 505, the click probability that is used for extracting after described adjustment is thrown in data greater than the network of predetermined threshold value, and it is thrown in to the active user.
For device embodiment, because it is substantially similar to embodiment of the method, so description is fairly simple, relevant part gets final product referring to the part explanation of embodiment of the method.
Those skilled in the art should understand, embodiments of the invention can be provided as method, system or computer program.Therefore, the present invention can adopt complete hardware implementation example, implement software example or in conjunction with the form of the embodiment of software and hardware aspect fully.And the present invention can adopt the form that wherein includes the upper computer program of implementing of computer-usable storage medium (including but not limited to magnetic disk memory, CD-ROM, optical memory etc.) of computer usable program code one or more.
The present invention describes with reference to process flow diagram and/or the block scheme of the method according to this invention, terminal device (system) and computer program.Should understand can be by the flow process in each flow process in computer program instructions realization flow figure and/or block scheme and/or square frame and process flow diagram and/or block scheme and/or the combination of square frame.Can provide these computer program instructions to the processor of multi-purpose computer, special purpose computer, Embedded Processor or other programmable data processing terminal equipment to produce a machine, make the instruction of carrying out by the processor of computing machine or other programmable data processing terminal equipment produce to be used for the device of realizing in the function of flow process of process flow diagram or a plurality of flow process and/or square frame of block scheme or a plurality of square frame appointments.
These computer program instructions also can be stored in energy vectoring computer or the computer-readable memory of other programmable data processing terminal equipment with ad hoc fashion work, make the instruction that is stored in this computer-readable memory produce the manufacture that comprises command device, this command device is realized the function of appointment in flow process of process flow diagram or a plurality of flow process and/or square frame of block scheme or a plurality of square frame.
These computer program instructions also can be loaded on computing machine or other programmable data processing terminal equipment, make on computing machine or other programmable terminal equipment and to carry out the sequence of operations step producing computer implemented processing, thereby be provided for realizing the step of the function of appointment in flow process of process flow diagram or a plurality of flow process and/or square frame of block scheme or a plurality of square frame in the instruction of carrying out on computing machine or other programmable terminal equipment.
Although described the preferred embodiments of the present invention, in a single day those skilled in the art get the basic creative concept of cicada, can make other change and modification to these embodiment.So claims are intended to all changes and the modification that are interpreted as comprising preferred embodiment and fall into the scope of the invention.
At last, also need to prove, in this article, relational terms such as the first and second grades only is used for an entity or operation are separated with another entity or operational zone, and not necessarily requires or hint and have the relation of any this reality or sequentially between these entities or operation.And, term " comprises ", " comprising " or its any other variant are intended to contain comprising of nonexcludability, thereby make the process, method, article or the terminal device that comprise a series of key elements not only comprise those key elements, but also comprise other key elements of clearly not listing, or also be included as the intrinsic key element of this process, method, article or terminal device.In the situation that not more restrictions, the key element that is limited by statement " comprising ... ", and be not precluded within process, method, article or the terminal device that comprises described key element and also have other identical element.
Above method of a kind of network provided by the present invention being thrown in the data input, and, a kind of network is thrown in the device that data are thrown in, be described in detail, used specific case herein principle of the present invention and embodiment are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, all will change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (14)
1. a network is thrown in the method that data are thrown in, and it is characterized in that, comprising:
The statistics active user throws in the click probability of data for different classes of network;
Obtain the neighbour user related with described active user;
Extract described neighbour user and throw in the click probability of data for different classes of network;
Adopt described neighbour user to throw in the click probability of data for different classes of network, adjust the active user and throw in the click probability of data for the respective classes network;
The click probability that extracts after described adjustment is thrown in data greater than the network of predetermined threshold value, and it is thrown in to the active user.
2. method according to claim 1, is characterized in that, described network is thrown in data and comprised ad data, and described statistics active user comprises for the step of the click probability of different classes of ad data:
Obtain active user's historical behavior information;
Extract the active user for the number of clicks of different classes of ad data from described active user's historical behavior information, and the active user is for the displaying number of times of different classes of ad data;
According to the number of clicks of described active user for different classes of ad data, and the active user calculates the active user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
3. method according to claim 2, is characterized in that, described active user's historical behavior information comprises active user's Webpage search information and active user's web page access information.
4. method according to claim 1 and 2, is characterized in that, the described neighbour user related with the active user comprises the user who has mutual network behavior with the active user, and/or, with the user of active user's behavior similarity greater than predetermined threshold value.
5. method according to claim 1 and 2, is characterized in that, described extraction neighbour user comprises for the step of the click probability of different classes of ad data:
Obtain neighbour user's historical behavior information;
Extract neighbour user for the number of clicks of different classes of ad data from described neighbour user's historical behavior information, and neighbour user is for the displaying number of times of different classes of ad data;
According to the number of clicks of described neighbour user for different classes of ad data, and neighbour user calculates neighbour user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
6. 1 or 2 described methods as requested, is characterized in that, described employing neighbour user is for the click probability of different classes of ad data, adjusts the active user and comprise for the step of the click probability of respective classes ad data:
Adopt following formula to adjust described active user for the click probability of respective classes ad data,
Wherein, u is described active user, Ctr
uBe the click probability of described active user for the respective classes ad data, g is the neighbour user related with described active user, Ctr
gBe the click probability of the neighbour user related with described active user for the respective classes ad data, outdeg (g) is out-degree, and α is ratio of damping.
7. 1 or 2 described methods as requested, is characterized in that, described employing neighbour user is for the click probability of different classes of ad data, adjusts the active user and also comprise for the step of the click probability of respective classes ad data:
Calculate it for the mean value of the click probability of respective classes ad data to described active user and with active user's behavior similarity greater than the user of predetermined threshold value.
8. a network is thrown in the device that data are thrown in, and it is characterized in that, comprising:
Click the probability statistics module, be used for the statistics active user and throw in the click probability of data for different classes of network;
Neighbour user's acquisition module is used for obtaining the neighbour user related with described active user;
Click the probability extraction module, be used for extracting described neighbour user and throw in the click probability of data for different classes of network;
Click the probability adjusting module, be used for adopting described neighbour user to throw in the click probability of data for different classes of network, adjust the active user and throw in the click probability of data for the respective classes network;
Putting module, the click probability that is used for extracting after described adjustment is thrown in data greater than the network of predetermined threshold value, and it is thrown in to the active user.
9. device according to claim 8, is characterized in that, described network is thrown in data and comprised ad data, and described click probability statistics module comprises:
The first historical behavior acquisition of information submodule is for the historical behavior information of obtaining the active user;
First extracts submodule, be used for extracting the active user for the number of clicks of different classes of ad data from described active user's historical behavior information, and the active user is for the displaying number of times of different classes of ad data;
First clicks the probability calculation submodule, be used for according to the number of clicks of described active user for different classes of ad data, and the active user calculates the active user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
10. device according to claim 9, is characterized in that, described active user's historical behavior information comprises active user's Webpage search information and active user's web page access information.
11. according to claim 8 or 9 described devices is characterized in that, the described neighbour user related with the active user comprises the user who has mutual network behavior with the active user, and/or, with the user of active user's behavior similarity greater than predetermined threshold value.
12. according to claim 8 or 9 described devices is characterized in that, described click probability extraction module comprises:
The second historical behavior acquisition of information submodule is for the historical behavior information of obtaining neighbour user;
Second extracts submodule, be used for extracting neighbour user for the number of clicks of different classes of ad data from described neighbour user's historical behavior information, and neighbour user is for the displaying number of times of different classes of ad data;
Second clicks the probability calculation submodule, be used for according to the number of clicks of described neighbour user for different classes of ad data, and neighbour user calculates neighbour user for the click probability of different classes of ad data for the displaying number of times of different classes of ad data.
13. 8 or 9 described devices, is characterized in that as requested, described click probability adjusting module comprises:
Adjust submodule, be used for adopting following formula to adjust described active user for the click probability of respective classes ad data,
Wherein, u is described active user, Ctr
uBe the click probability of described active user for the respective classes ad data, g is the neighbour user related with described active user, Ctr
gBe the click probability of the neighbour user related with described active user for the respective classes ad data, outdeg (g) is out-degree, and α is ratio of damping.
14. 8 or 9 described devices, is characterized in that as requested, described click probability adjusting module also comprises:
The mean value calculation submodule is used for calculating it for the mean value of the click probability of respective classes ad data to described active user and with active user's behavior similarity greater than the user of predetermined threshold value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100529133A CN103150663A (en) | 2013-02-18 | 2013-02-18 | Method and device for placing network placement data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100529133A CN103150663A (en) | 2013-02-18 | 2013-02-18 | Method and device for placing network placement data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103150663A true CN103150663A (en) | 2013-06-12 |
Family
ID=48548719
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013100529133A Pending CN103150663A (en) | 2013-02-18 | 2013-02-18 | Method and device for placing network placement data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103150663A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103390067A (en) * | 2013-08-15 | 2013-11-13 | 北京国双科技有限公司 | Data processing method and device for internet entity analysis |
| CN104598511A (en) * | 2014-10-23 | 2015-05-06 | 腾讯科技（深圳）有限公司 | Method, device and system for recommending search results |
| CN104834647A (en) * | 2014-02-12 | 2015-08-12 | 腾讯科技（深圳）有限公司 | Method and device for obtaining informative abstract |
| CN105634924A (en) * | 2015-12-30 | 2016-06-01 | 腾讯科技（深圳）有限公司 | Display method of media information, server and client end |
| WO2016169427A1 (en) * | 2015-04-21 | 2016-10-27 | 腾讯科技（深圳）有限公司 | Correction method for advertisement click-through rate and advertisement delivery server |
| CN107590691A (en) * | 2017-09-06 | 2018-01-16 | 晶赞广告（上海）有限公司 | A kind of information issuing method and device, storage medium, terminal |
| WO2018027361A1 (en) * | 2016-08-06 | 2018-02-15 | 胡明祥 | Method for adaptive adjustment of moments information-based advertisement pushing and pushing system |
| CN107967341A (en) * | 2017-12-08 | 2018-04-27 | 广东欧珀移动通信有限公司 | A kind of image display method, terminal device and storage medium |
| WO2019019396A1 (en) * | 2017-07-24 | 2019-01-31 | 上海壹账通金融科技有限公司 | Push result prediction method and apparatus, computer device and storage medium |
| CN109801100A (en) * | 2018-12-26 | 2019-05-24 | 北京达佳互联信息技术有限公司 | Advertisement placement method, device and computer readable storage medium |
| CN110324418A (en) * | 2019-07-01 | 2019-10-11 | 阿里巴巴集团控股有限公司 | Method and apparatus based on customer relationship transmission service |
| CN110378714A (en) * | 2018-04-13 | 2019-10-25 | 北京京东尚科信息技术有限公司 | A kind of method and apparatus of processing access data |
| CN110968802A (en) * | 2019-12-04 | 2020-04-07 | 上海风秩科技有限公司 | User characteristic analysis method, analysis device and readable storage medium |
| CN111861542A (en) * | 2020-06-15 | 2020-10-30 | 北京雷石天地电子技术有限公司 | Method, device, terminal and non-transitory computer readable storage medium for delivering advertisement |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080147635A1 (en) * | 2006-12-13 | 2008-06-19 | Il Im | System, apparatus and method for providing weight to information gathering engine according to situation of user and computer readable medium processing the method |
| CN101520878A (en) * | 2009-04-03 | 2009-09-02 | 华为技术有限公司 | Method, device and system for pushing advertisements to users |
| CN101685521A (en) * | 2008-09-23 | 2010-03-31 | 北京搜狗科技发展有限公司 | Method for showing advertisements in webpage and system |
| CN102521767A (en) * | 2011-12-13 | 2012-06-27 | 亿赞普(北京)科技有限公司 | Method and system for publishing network advertising information |
-
2013
- 2013-02-18 CN CN2013100529133A patent/CN103150663A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080147635A1 (en) * | 2006-12-13 | 2008-06-19 | Il Im | System, apparatus and method for providing weight to information gathering engine according to situation of user and computer readable medium processing the method |
| CN101685521A (en) * | 2008-09-23 | 2010-03-31 | 北京搜狗科技发展有限公司 | Method for showing advertisements in webpage and system |
| CN101520878A (en) * | 2009-04-03 | 2009-09-02 | 华为技术有限公司 | Method, device and system for pushing advertisements to users |
| CN102521767A (en) * | 2011-12-13 | 2012-06-27 | 亿赞普(北京)科技有限公司 | Method and system for publishing network advertising information |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015021868A1 (en) * | 2013-08-15 | 2015-02-19 | 北京国双科技有限公司 | Data processing method and apparatus for internet entity analysis |
| US10282464B2 (en) | 2013-08-15 | 2019-05-07 | Beijing Gridsum Technology Co., Ltd. | Data processing method and device for internet entity analysis |
| CN103390067B (en) * | 2013-08-15 | 2016-10-05 | 北京国双科技有限公司 | The data processing method analyzed for internet entity and device |
| CN103390067A (en) * | 2013-08-15 | 2013-11-13 | 北京国双科技有限公司 | Data processing method and device for internet entity analysis |
| CN104834647B (en) * | 2014-02-12 | 2019-04-23 | 腾讯科技（深圳）有限公司 | A kind of method and device obtaining informative abstract |
| CN104834647A (en) * | 2014-02-12 | 2015-08-12 | 腾讯科技（深圳）有限公司 | Method and device for obtaining informative abstract |
| CN104598511A (en) * | 2014-10-23 | 2015-05-06 | 腾讯科技（深圳）有限公司 | Method, device and system for recommending search results |
| WO2016169427A1 (en) * | 2015-04-21 | 2016-10-27 | 腾讯科技（深圳）有限公司 | Correction method for advertisement click-through rate and advertisement delivery server |
| US10650076B2 (en) | 2015-12-30 | 2020-05-12 | Tencent Technology (Shenzhen) Company Limited | System for determining placement parameters for a media content |
| CN105634924B (en) * | 2015-12-30 | 2020-07-07 | 腾讯科技（深圳）有限公司 | Media information display method, server and client |
| CN105634924A (en) * | 2015-12-30 | 2016-06-01 | 腾讯科技（深圳）有限公司 | Display method of media information, server and client end |
| WO2018027361A1 (en) * | 2016-08-06 | 2018-02-15 | 胡明祥 | Method for adaptive adjustment of moments information-based advertisement pushing and pushing system |
| WO2019019396A1 (en) * | 2017-07-24 | 2019-01-31 | 上海壹账通金融科技有限公司 | Push result prediction method and apparatus, computer device and storage medium |
| CN107590691B (en) * | 2017-09-06 | 2021-01-15 | 晶赞广告（上海）有限公司 | Information publishing method and device, storage medium and terminal |
| CN107590691A (en) * | 2017-09-06 | 2018-01-16 | 晶赞广告（上海）有限公司 | A kind of information issuing method and device, storage medium, terminal |
| CN107967341A (en) * | 2017-12-08 | 2018-04-27 | 广东欧珀移动通信有限公司 | A kind of image display method, terminal device and storage medium |
| CN110378714A (en) * | 2018-04-13 | 2019-10-25 | 北京京东尚科信息技术有限公司 | A kind of method and apparatus of processing access data |
| CN110378714B (en) * | 2018-04-13 | 2024-02-13 | 北京京东尚科信息技术有限公司 | Method and device for processing access data |
| CN109801100A (en) * | 2018-12-26 | 2019-05-24 | 北京达佳互联信息技术有限公司 | Advertisement placement method, device and computer readable storage medium |
| CN110324418A (en) * | 2019-07-01 | 2019-10-11 | 阿里巴巴集团控股有限公司 | Method and apparatus based on customer relationship transmission service |
| CN110324418B (en) * | 2019-07-01 | 2022-09-20 | 创新先进技术有限公司 | Method and device for pushing service based on user relationship |
| CN110968802A (en) * | 2019-12-04 | 2020-04-07 | 上海风秩科技有限公司 | User characteristic analysis method, analysis device and readable storage medium |
| CN110968802B (en) * | 2019-12-04 | 2023-05-05 | 上海风秩科技有限公司 | Analysis method and analysis device for user characteristics and readable storage medium |
| CN111861542A (en) * | 2020-06-15 | 2020-10-30 | 北京雷石天地电子技术有限公司 | Method, device, terminal and non-transitory computer readable storage medium for delivering advertisement |
| CN111861542B (en) * | 2020-06-15 | 2024-02-02 | 北京雷石天地电子技术有限公司 | Method, device, terminal and non-transitory computer readable storage medium for advertising |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103150663A (en) | Method and device for placing network placement data | |
| US9767182B1 (en) | Classification of search queries | |
| CN101957834B (en) | Content recommending method and device based on user characteristics | |
| CN104008109B (en) | Web information Push Service system based on user interest | |
| CN103064945B (en) | Based on the Situational searching method of body | |
| CN102426610B (en) | Microblog rank searching method and microblog searching engine | |
| CN103914468A (en) | Method and device for searching for released information | |
| CN103530416B (en) | Project data forecasting grading library generating and project data pushing method and project data forecasting grading library generating and project data pushing system | |
| CN104252456B (en) | A kind of weight method of estimation, apparatus and system | |
| US10296535B2 (en) | Method and system to randomize image matching to find best images to be matched with content items | |
| CN103514191A (en) | Method and device for determining keyword matching mode of target popularization information | |
| US20160125028A1 (en) | Systems and methods for query rewriting | |
| CN103838756A (en) | Method and device for determining pushed information | |
| Yang | OntoCrawler: A focused crawler with ontology-supported website models for information agents | |
| CN105677857B (en) | method and device for accurately matching keywords with marketing landing pages | |
| US9330135B2 (en) | Method, apparatus and computer readable recording medium for a search using extension keywords | |
| CN105389329A (en) | Open source software recommendation method based on group comments | |
| CN103309960A (en) | Method and device for extracting multidimensional information of network public sentiment event | |
| CN105868267A (en) | Modeling method for mobile social network user interests | |
| CN108959641A (en) | A kind of content information recommended method and system based on artificial intelligence | |
| CN103745380A (en) | Advertisement delivery method and apparatus | |
| CN104268230A (en) | Method for detecting objective points of Chinese micro-blogs based on heterogeneous graph random walk | |
| CN103412958A (en) | Display method and device for searching result | |
| CN104503988A (en) | Searching method and device | |
| CN103365876A (en) | Method and device for generating network operation auxiliary information based on relation maps |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20130612 |
|
| RJ01 | Rejection of invention patent application after publication |