CN109981785A

CN109981785A - Method and apparatus for pushed information

Info

Publication number: CN109981785A
Application number: CN201910261259.4A
Authority: CN
Inventors: 李磊
Original assignee: Beijing ByteDance Network Technology Co Ltd
Current assignee: Douyin Vision Co Ltd; Douyin Vision Beijing Co Ltd
Priority date: 2019-04-02
Filing date: 2019-04-02
Publication date: 2019-07-05
Anticipated expiration: 2039-04-02
Also published as: CN109981785B; WO2020199662A1

Abstract

Embodiment of the disclosure discloses the method and apparatus for pushed information.One specific embodiment of this method includes: the first location information obtained for indicating the position of target user, obtains candidate pushed information collection；For the candidate pushed information that candidate pushed information is concentrated, the first match information of candidate's pushed information is determined；According to the first match information, corresponding first operation information of candidate's pushed information is determined, the first operation information is for characterizing the probability that target user operates candidate's pushed information performance objective；According to corresponding first operation information, is concentrated from candidate pushed information and choose candidate pushed information；The candidate pushed information selected to the corresponding terminal device push of target user.The embodiment helps to promote the accuracy of pushed information.

Description

Method and apparatus for pushed information

Technical field

Embodiment of the disclosure is related to field of computer technology, and in particular to the method and apparatus for pushed information.

Background technique

It can include some information relevant to region to the information that user pushes in the push class application of some information. For example, for describe generation in the information of event in certain city, the traffic condition for introducing some street etc..

For this kind of information relevant to region, the user for usually only corresponding to region, which may compare these information, feels emerging Interest.Based on this, some information push class applications are normally based on the position where user, where pushing some and user to user The relevant information in position.

Summary of the invention

Embodiment of the disclosure proposes the method and apparatus for pushed information.

In a first aspect, embodiment of the disclosure provides a kind of method for pushed information, used this method comprises: obtaining First location information in the position for indicating target user, and obtain candidate pushed information collection；For candidate pushed information collection In candidate pushed information, execute and following determine step: determining the first match information of candidate's pushed information, wherein first Match information is based on first location information, preference of the target user to candidate's pushed information for characterizing；According to first Match information determines corresponding first operation information of candidate's pushed information, wherein the first operation information is for characterizing based on the One match information, the probability that target user operates candidate's pushed information performance objective；According to corresponding first operation information, It is concentrated from candidate pushed information and chooses candidate pushed information；The candidate selected to the corresponding terminal device push of target user pushes away It delivers letters breath.

In some embodiments, the above method further include: according to first location information, determine at least one second confidence Breath；And determine step further include: for the second location information at least one second location information, determine that the candidate pushes Second match information of information, wherein the second match information is based on the second location information for characterizing, and target user is to the time Select the preference of pushed information；According to the second match information, corresponding second operation information of candidate's pushed information is determined, In, the second operation information is based on the second match information for characterizing, and target user operates candidate's pushed information performance objective Probability；And it according to corresponding first operation information, is concentrated from candidate pushed information and chooses candidate pushed information, comprising: root According to corresponding first operation information and at least one second operation information, is concentrated from candidate pushed information and choose candidate push letter Breath.

In some embodiments, according to first location information, at least one second location information is determined, comprising: according to One location information and at least one preset distance threshold, determine at least one second location information, wherein distance threshold is used for Characterize the position of first location information instruction and the distance between the position of second location information instruction.

In some embodiments, according to the first match information, corresponding first operation information of candidate's pushed information is determined, Include: that the first match information is input to information prediction model trained in advance, obtains candidate's pushed information corresponding first Operation information, wherein information prediction model is used to characterize the corresponding relationship between match information and operation information；And according to Two match informations determine corresponding second operation information of candidate's pushed information, comprising: the second match information is input to information Prediction model obtains corresponding second operation information of candidate's pushed information.

In some embodiments, the first match information is included at least one of the following: for characterizing target user couple and first Location information instruction the relevant information in position preference information, for characterizing candidate's pushed information and first position Information instruction position the degree of correlation information, for characterize first location information instruction position attention rate information.

In some embodiments, the second match information include at least one of the following: for characterize target user couple and this Two location informations instruction the relevant information in position preference information, for characterize candidate's pushed information and this second Location information instruction position the degree of correlation information, for characterize the second location information instruction position attention rate letter Breath.

In some embodiments, training obtains information prediction model as follows: obtaining training sample set, wherein The training sample that training sample is concentrated includes sample matches information and sample instruction information, wherein sample matches information includes mesh The match information of the corresponding history pushed information of user is marked, sample instruction information is used to indicate whether target user goes through to corresponding History pushed information executed object run；Obtain the initial model pre-established；Using the method for machine learning, based on training sample This collection and preset loss function, are trained initial model, and the initial model after training is determined as information prediction Model.

In some embodiments, it according to corresponding first operation information and at least one second operation information, is pushed away from candidate Breath of delivering letters, which is concentrated, chooses candidate pushed information, comprising: for the candidate pushed information that candidate pushed information is concentrated, in response to determination The corresponding destination probability of candidate's pushed information is greater than preset probability threshold value, chooses candidate's pushed information, wherein target is general Rate is used to indicate the average value of the maximum value in the corresponding probability set of candidate's pushed information or each probability in probability set, In, probability and corresponding at least one second operation of the probability set by the corresponding first operation information characterization of candidate's pushed information The probability composition that information characterizes respectively.

In some embodiments, candidate pushed information collection is obtained, comprising: according to first location information, obtain candidate push Information collection, wherein the candidate pushed information that candidate pushed information is concentrated is related to the position that first location information indicates.

In some embodiments, the candidate pushed information selected to the corresponding terminal device push of target user, comprising: In response to detecting the predetermined registration operation of target user, the candidate pushed information selected to terminal device push, and control exist Selected candidate pushed information is shown on terminal device, wherein predetermined registration operation is for requesting push to refer to first location information The relevant information in the position shown.

Second aspect, embodiment of the disclosure provide a kind of device for pushed information, which includes: to obtain list Member is configured to obtain the first location information for indicating the position of target user, and obtains candidate pushed information collection；Behaviour Make information determination unit, be configured to the candidate pushed information concentrated for candidate pushed information, executes following determining step: really First match information of fixed candidate's pushed information, wherein the first match information is based on first location information, target for characterizing Preference of the user to candidate's pushed information；According to the first match information, candidate's pushed information corresponding first is determined Operation information, wherein the first operation information is based on the first match information for characterizing, and target user holds candidate's pushed information The probability of row object run；Selection unit is configured to concentrate and select from candidate pushed information according to corresponding first operation information Take candidate pushed information；Push unit is configured to the candidate push selected to the corresponding terminal device push of target user Information.

In some embodiments, the above-mentioned device for pushed information further include: location information determination unit is configured to According to first location information, at least one second location information is determined；And aforesaid operations information determination unit, further matched It is set to for the second location information at least one second location information, determines the second matching letter of candidate's pushed information Breath, wherein the second match information is based on the second location information, preference of the target user to candidate's pushed information for characterizing Degree；According to the second match information, corresponding second operation information of candidate's pushed information is determined, wherein the second operation information The second match information, the probability that target user operates candidate's pushed information performance objective are based on for characterizing；And it is above-mentioned Selection unit is further configured to be pushed away according to corresponding first operation information and at least one second operation information from candidate Breath of delivering letters, which is concentrated, chooses candidate pushed information.

In some embodiments, above-mentioned location information determination unit is further configured to according to first location information, really At least one fixed second location information, comprising: according to first location information and at least one preset distance threshold, determine at least One second location information, wherein the position that distance threshold is used to characterize first location information instruction refers to second location information The distance between position shown.

In some embodiments, aforesaid operations information determination unit is further configured to for the first match information being input to Trained information prediction model in advance, obtains corresponding first operation information of candidate's pushed information, wherein information prediction model For characterizing the corresponding relationship between match information and operation information；And aforesaid operations information determination unit is further configured At: the second match information is input to information prediction model, obtains corresponding second operation information of candidate's pushed information.

In some embodiments, above-mentioned selection unit is further configured to push away the candidate that candidate pushed information is concentrated Breath of delivering letters is chosen the candidate and is pushed away in response to determining that the corresponding destination probability of candidate's pushed information is greater than preset probability threshold value It delivers letters breath, wherein destination probability is used to indicate in maximum value or the probability set in the corresponding probability set of candidate's pushed information The average value of each probability, wherein probability that probability set is characterized by corresponding first operation information of candidate's pushed information and right The probability composition that the second operation information of at least one answered characterizes respectively.

In some embodiments, above-mentioned acquiring unit is further configured to obtain candidate according to first location information and push away It delivers letters breath collection, wherein the candidate pushed information that candidate pushed information is concentrated is related to the position that first location information indicates.

In some embodiments, above-mentioned push unit is further configured to the default behaviour in response to detecting target user Make, the candidate pushed information selected to terminal device push, and control show that selected candidate pushes away on the terminal device It delivers letters breath, wherein predetermined registration operation pushes information relevant to the position that first location information indicates for request.

The third aspect, embodiment of the disclosure provide a kind of server, which includes: one or more processing Device；Storage device, for storing one or more programs；When one or more programs are executed by one or more processors, make Obtain method of the one or more processors realization as described in implementation any in first aspect.

Fourth aspect, embodiment of the disclosure provide a kind of computer-readable medium, are stored thereon with computer program, The method as described in implementation any in first aspect is realized when the computer program is executed by processor.

Information relevant to position for one, usually only lives in a certain range of user near corresponding position for a long time It may be interested in the information.Wherein, it might have part to live near corresponding position in a certain range but to this kind of and position Set the relevant uninterested user of information.It therefore, is opposite to the number of an interested user of information relevant to position Less.When pushing this kind of information relevant to position, it is only necessary to it is pushed to the interested user of this category information, to avoid Unnecessary information push belt carrys out additional resource consumption.

The method and apparatus for pushed information that embodiment of the disclosure provides, the position based on user where it Place, predicts the preference of candidate pushed information the probability that user operates candidate pushed information performance objective.Prediction Probability can reflect that user, in turn, can be according to the probability of prediction to the interest level of the relevant information in its position Selectively to user's pushed information, as much as possible to the relevant information in its position interested user push pair Information is answered, also avoids pushing corresponding informance to the uninterested user of the relevant information in its position, to reduce information During push due to to the uninterested user of the relevant information in its position push corresponding informance caused by need not The resource consumption wanted.

Detailed description of the invention

By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, the disclosure is other Feature, objects and advantages will become more apparent upon:

Fig. 1 is that one embodiment of the disclosure can be applied to exemplary system architecture figure therein；

Fig. 2 is the flow chart according to one embodiment of the method for pushed information of the disclosure；

Fig. 3 is the flow chart according to another embodiment of the method for pushed information of the disclosure；

Fig. 4 is the schematic diagram of an application scenarios of the method according to an embodiment of the present disclosure for pushed information；

Fig. 5 is the structural schematic diagram according to one embodiment of the device for pushed information of the disclosure；

Fig. 6 is adapted for the structural schematic diagram for realizing the electronic equipment of embodiment of the disclosure.

Specific embodiment

The disclosure is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to Convenient for description, part relevant to related invention is illustrated only in attached drawing.

It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the disclosure can phase Mutually combination.The disclosure is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

Fig. 1 is shown can be using the method for pushed information of the disclosure or the implementation of the device for pushed information The exemplary architecture 100 of example.

As shown in Figure 1, system architecture 100 may include terminal device 101,102,103, network 104 and server 105. Network 104 between terminal device 101,102,103 and server 105 to provide the medium of communication link.Network 104 can be with Including various connection types, such as wired, wireless communication link or fiber optic cables etc..

Terminal device 101,102,103 is interacted by network 104 with server 105, to receive or send message etc..Terminal Various client applications can be installed in equipment 101,102,103.For example, the application of browser class, searching class application, Instant Messenger Letter tool, the application of reading class, information share class application etc..

Terminal device 101,102,103 can be hardware, be also possible to software.When terminal device 101,102,103 is hard When part, it can be the various electronic equipments with display screen, including but not limited to smart phone, tablet computer, on knee portable Computer and desktop computer etc..When terminal device 101,102,103 is software, above-mentioned cited electricity may be mounted at In sub- equipment.Multiple softwares or software module may be implemented into (such as providing multiple softwares of Distributed Services or soft in it Part module), single software or software module also may be implemented into.It is not specifically limited herein.

Server 105 can be to provide the server of various services, for example, install on terminal device 101,102,103 Some applications (such as information shares class application) provide the back-end server supported.Server 105 can based on terminal device 101, 102, the position of 103 corresponding users and candidate pushed information collection, operation information of the prediction user to each candidate pushed information. And then it can be concentrated from candidate pushed information and choose candidate pushed information according to corresponding operation information, and by the time of selection Pushed information is selected to push to terminal device 101,102,103.

It should be noted that the method provided by embodiment of the disclosure for pushed information is generally by server 105 It executes, correspondingly, the device for pushed information is generally positioned in server 105.

It may also be noted that can also be equipped with information processing class application in terminal device 101,102,103, terminal is set Standby 101,102,103 class can also apply at the candidate pushed information concentrated to candidate pushed information based on information processing Reason.At this point, the method for pushed information can also be executed by terminal device 101,102,103, correspondingly, it to be used for pushed information Device also can be set in terminal device 101,102,103.At this point, exemplary system architecture 100 can there is no services Device 105 and network 104.

It should be noted that server 105 can be hardware, it is also possible to software.It, can when server 105 is hardware To be implemented as the distributed server cluster that multiple servers form, individual server also may be implemented into.When server 105 is When software, multiple softwares or software module may be implemented into (such as providing multiple softwares of Distributed Services or software mould Block), single software or software module also may be implemented into.It is not specifically limited herein.

It should be understood that the number of terminal device, network and server in Fig. 1 is only schematical.According to realization need It wants, can have any number of terminal device, network and server.

With continued reference to Fig. 2, it illustrates the processes according to one embodiment of the method for pushed information of the disclosure 200.This for pushed information method the following steps are included:

Step 201, the first location information for indicating the position of target user is obtained, and obtains candidate pushed information Collection.

In the present embodiment, target user can refer to (as shown in Figure 1 with the executing subject of the method for pushed information Server 105) communication connection the corresponding user of terminal device (terminal device 101,102,103 as shown in Figure 1).First Confidence breath can be being capable of locative various information.For example, first location information can be latitude and longitude information, address Information etc..

Above-mentioned executing subject can determine the position of target user based on existing various location technologies.For example, can be with base The position of target user is determined in GPS (Global Positioning System, global positioning system) technology.Optionally, may be used Can also be obtained from local or other storage equipment pre- directly by the position of determining target user as first location information If for indicate determining target user position information as first location information.

In the present embodiment, candidate pushed information collection can be the collection being made of the preassigned some information of technical staff It closes, is also possible to the set of some information determined according to preset screening conditions composition.The time that candidate pushed information is concentrated Pushed information is selected to can be various types of information.For example, candidate pushed information can be image, text, video, audio etc. Deng.Above-mentioned executing subject can be obtained from local or other storage equipment (database, the third party's data platform that such as connect) Candidate pushed information collection.

In some optional implementations of the present embodiment, candidate push letter can be obtained according to first location information Breath collection.Wherein, the candidate pushed information that candidate pushed information is concentrated is related to the position that first location information indicates.

Thus, it is possible to avoiding candidate pushed information from concentrating there is the information unrelated with the position that first location information indicates, To help to reduce the number of subsequent candidate pushed information to be treated, speed up processing.

Whether candidate pushed information and the position that first location information indicates are related can be according to pre-set related item Part judges.For example, correlated condition can be the position phase that the content of candidate pushed information description is indicated with first location information It closes.As an example, candidate pushed information is an article for describing certain food, if described food refers to for first location information The food of specialty in certain area where the position shown, then it is considered that candidate pushed information and first location information indicate Position it is related.

In another example correlated condition can be corresponding to the position for having first location information instruction in candidate pushed information Predetermined keyword concentrate predetermined keyword.Wherein, predetermined keyword collection may include the position of first location information instruction Various describing modes.Predetermined keyword collection also may include first location information instruction position where certain area in institute The corresponding various describing modes in all positions of covering.

It should be appreciated that generally, the position of target user can be a very specific position.For example, target The position of user can be as accurate as specific longitude and latitude.In this case, higher information related to the position of target user can It can be less.It therefore, can be extensive to where the position of target user by the position of target user according to actual application demand In certain area.At this point, the information to relevant in the certain area where the position of target user is construed as and target The relevant information in the position of user.

Above-mentioned executing subject can according to first location information, from local or other storage equipment (as connection database, Third party's data platform etc.) obtain some candidate pushed informations of information composition relevant to the position that first location information indicates Collection.

Step 202, the candidate pushed information concentrated for candidate pushed information executes following determining step:

Step 2021, the first match information of candidate's pushed information is determined.

In the present embodiment, the first match information can be used for characterizing based on first location information, and target user is to the time Select the preference of pushed information.Preference can weigh according to the various relevant informations of user and candidate's pushed information Amount.

It specifically, can different, one or more can be with application demand flexible choice according to actual application scenarios Characterization is based on first location information, and target user believes the information of the preference of candidate's pushed information as the first matching Breath.

Optionally, the first match information may include that there are user some history of higher preference to push away within a preset period of time It delivers letters the similarity of breath and candidate's pushed information.Wherein, user has some history of higher preference within a preset period of time Pushed information can be according within a preset period of time, and historical behavior data of the user at the position that first location information indicates are come It determines.

For example, can be pushed away within a preset period of time, and when user is at the position that first location information indicates to user The history pushed information that user clicks or commented on is chosen in each history pushed information sent has higher preference as user History pushed information.

Generally, there are some history pushed informations of higher preference and the candidate to push within a preset period of time and believe by user The similarity of breath is higher, user can be indicated at the position that first location information indicates, to the preference of candidate's pushed information Degree is higher.

Optionally, the first match information may include at least one of following: for characterizing target user couple and first position Information instruction the relevant information in position preference information, for characterizing candidate's pushed information and first location information The information of the degree of correlation of the position of instruction, for characterize first location information instruction position attention rate information.

Wherein, the preference of target user couple information relevant to the position that first location information indicates can be according to reality The application scenarios on border are adopted and are determined in various manners.

It is alternatively possible to according to target user, the position phase with first location information instruction that pushed for history The historical behavior data of the information of pass are inclined come the relevant information in the position of analyzing target user couple to first location information indicates Good degree.

Generally, user has some interactive operations to the pushed information of preference.For example, clicking operation, comment operation, receipts Hiding or the operation of link address etc. for sharing pushed information.User is for interactive operation performed by different pushed informations Some attribute informations would also vary from.Such as number, the browsing duration etc. of clicking operation.It therefore, can be according to user's Historical behavior data analyze user to the preference of the information of push.Under normal conditions, it is more to execute interactive operation by user Or it is more frequent, it can indicate that user is higher to the preference of corresponding pushed information.

It is being pushed to target user, refer to first location information as an example, can count in the target histories period The total number of the relevant information in the position shown is executed as the first number, and in these information of push by target user The number of the information of predetermined registration operation is crossed as the second number.Later, the quotient that can use the second number and the first number carrys out table Levy the preference of target user couple information relevant to the position that first location information indicates.

Wherein, the target histories period can be by technical staff's preassigned any time period.For example, the object time Section can be the period corresponding to three months using current time as starting point, before current time.Predetermined registration operation can be By the preassigned various user's operations of technical staff.

It is alternatively possible to using the model (such as clicking rate prediction model) for prediction of existing some open sources come pre- Survey the operation information (such as clicking rate) of target user couple information relevant to the position that first location information indicates.It, can based on this With the operation information using prediction come the preference of the relevant information in the position that characterizes target user couple to first location information indicates Degree.

Wherein, a large amount of user's history behavioral data training can also be advanced with by technical staff for the model of prediction It obtains.For example, information prediction model can be according to the feature vector for characterizing user and for characterizing candidate push letter The feature vector of breath predicts the user to the operation information (such as clicking rate) of candidate pushed information.

In the present embodiment, the degree of correlation for the position that candidate pushed information is indicated with first location information can refer to that candidate pushes away The correlation degree delivered letters between breath and the position of first location information instruction.For example, candidate pushed information " A " is one and retouches State first location information instruction position where region in sight spot article.Candidate pushed information " B " is one and has Indicate the title of the position of first location information instruction but the content article unrelated with the position that first location information indicates. So, the degree of correlation that can be generally considered as the position of candidate pushed information " A " and first location information instruction is higher than candidate push The degree of correlation of the position of information " B " and first location information instruction.

Wherein, the determination method of the degree of correlation can choose different methods according to different application scenarios.For example, can be first Obtain the corresponding comment information collection of candidate pushed information.Wherein, comment information collection may include that some of candidate pushed information comment By information.Later, the location information where the corresponding user of each comment information that available comment information is concentrated, obtains To position information set.Wherein, the corresponding user of comment information can refer to the corresponding user of terminal device of publication comment information.It Afterwards, the certain area where the position of first location information instruction can be determined as target area.It is then possible to count position Position that information is concentrated, instruction belongs to the number of the location information of target area as the first number, and determines position letter The total number for the location information that breath collection includes is as the second number.In turn, it can use the quotient of the first number Yu the second number To characterize the degree of correlation for the position that candidate pushed information and first location information indicate.Generally, the first number and the second number Quotient it is bigger, can indicate that the degree of correlation for the position that candidate pushed information and first location information indicate is higher.

Due to being usually for a long time first to the interested user of information relevant to the position that first location information indicates The user in certain area coverage where the position of location information instruction.So, the position for being indicated with first location information It sets for relevant information, the user that can be commented on these information is also for a long time in the position institute of first location information instruction Certain area coverage in user.It therefore, can be according to the corresponding user's of each comment information of candidate pushed information Location information come analyze candidate pushed information and first location information instruction position the degree of correlation.

It is alternatively possible to extract the keyword of candidate pushed information using existing various keyword extracting methods, obtain The corresponding keyword set of candidate pushed information.It is then possible to obtain preset corresponding with the position of first location information instruction Keyword set.Later, the position that can use the corresponding keyword set of candidate pushed information and first location information instruction is corresponding Keyword set similarity come characterize candidate pushed information and first location information instruction position the degree of correlation.Generally, The similarity of the corresponding keyword set of candidate pushed information and the corresponding keyword set in position of first location information instruction is higher, The degree of correlation that the position that candidate pushed information and first location information indicate can be characterized is higher.

Wherein it is determined that the algorithm of the similarity of two keyword sets can use existing various related algorithms.For example, base In the calculation of TF-IDF (Term Frequency-Inverse Document Frequency, word frequency-inverse document frequency) Method, is based on WMD (Word Mover ' s at the algorithm based on SimRank (a kind of model for measuring any two object similarity degree) Distance, a method of measuring the distance between two objects) algorithm etc..

It is alternatively possible to determine model using the degree of correlation trained in advance to analyze candidate pushed information and first position and believe Cease the degree of correlation of the position of instruction.Wherein, the degree of correlation determines that model can be used for believing according to candidate pushed information and first position Breath determines the degree of correlation between candidate pushed information and first location information.Specifically, the degree of correlation determines that model can be based on big The training sample training of amount obtains.Each training sample may include history pushed information, location information and history push letter The degree of correlation between breath and location information.Wherein, the history pushed information in training sample and the degree of correlation between location information It can use aforementioned some degrees of correlation and determine that method determines, can also be labeled in advance by technical staff.

In the present embodiment, the attention rate of the position of first location information instruction can be used for characterizing first location information and refer to The concerned degree of the position shown.For example, generally, the attention rate of sight spot position is generally higher than non-sight spot position Attention rate.

Since under normal circumstances, the attention rate of each position is all more stable.Therefore, each position pair can be preset The attention rate answered.Certainly, in some cases, the attention rate of certain positions needs dynamic to adjust.For example, some position location More well-known activity has been carried out in domain in the recent period, in this case, so that it may the attention rate of this position suitably be turned up.

It should be appreciated that being used to characterize target user couple the preference of information relevant to the position that a location information indicates The information of degree can be various forms of information.For example, being referred to using clicking rate characterization target user couple with a location information When the preference of the relevant information in the position shown, clicking rate can be that characterization target user couple indicates with a location information The information of the preference of the relevant information in position.At this point, the corresponding expression information of different clicking rates can also be preset, into And the corresponding expression information of clicking rate is obtained as characterization target user couple letter relevant to the position that a location information indicates The information of the preference of breath.

Similarly, for characterizing the information of the degree of correlation of the position of candidate's pushed information and first location information instruction It can be various forms of information.Information for characterizing the attention rate of the position of first location information instruction is also possible to various The information of form.Specific representation method can be by technical staff according to actual application demand flexible setting.

Step 2022, according to the first match information, corresponding first operation information of candidate's pushed information is determined.

In the present embodiment, the first operation information can be used for characterizing based on the first match information, and target user is to the time The probability for selecting pushed information performance objective to operate.Wherein, object run can be preassigned user's operation (as indicating The user's operation of clicking operation, for indicating the user's operation etc. of comment operation), it is also possible to be determined according to candidate pushed information User's operation (such as object run can be any user's operation that candidate pushed information is supported).

In different application scenarios, according to the first match information, can choosing different determining, candidate pushed information is corresponding The first operation information method.

It is alternatively possible to choose the maximum value for each information instruction that the first match information includes or each information indicates The average value work of numerical value is set to the first operation information.

In some optional implementations of the present embodiment, the first match information can be input to letter trained in advance Prediction model is ceased, corresponding first operation information of candidate pushed information is obtained.Wherein, information prediction model can be used for characterization With the corresponding relationship between information and operation information.

In some optional implementations of the present embodiment, information prediction model can be trained as follows It arrives:

Step 1 obtains training sample set.Wherein, the training sample that training sample is concentrated may include sample matches information Information is indicated with sample.Wherein, sample matches information may include the match information of the corresponding history pushed information of target user. Sample instruction information can serve to indicate that whether target user executed object run to corresponding history pushed information.

In this step, sample matches information may include at least one of following: for table according to different application demands Levy target user to the information of the preference of the relevant information in the corresponding position of target user, for characterizing history pushed information The information of the degree of correlation of position corresponding with target user, the information of attention rate for characterizing the corresponding position of target user. Wherein, the corresponding position of target user can be directed toward the position that user pushes identified target user when history pushed information.

Optionally, history pushed information can refer to and the relevant information in position corresponding with target user.

Step 2 obtains the initial model pre-established.

In this step, the various artificial neural networks that initial model can be indiscipline or training is completed, can also be with It is that various artificial neural networks are combined obtained model.Initial model can also be by technical staff according to actual application Demand (such as needs which network layer, the parameter of each network layer, the size of convolution kernel), utilizes some deep learning frames (such as TensorFlow, Caffe) is built.

It is alternatively possible to be based on existing some regression models (such as the model of logic-based recurrence, based on successive Regression Model etc.) build initial model.

Step 3 is based on training sample set and preset loss function using the method for machine learning, to initial model into Row training, and the initial model after training is determined as information prediction model.

In this step, loss function can be pre-designed by technical staff.Loss function is commonly used according to introductory die Type exports and the difference between corresponding desired output.It is generally desirable to which the value of loss function is small as far as possible.

The selection of loss function can also be determined according to initial model.Different initial models can be with corresponding selection not Same loss function.For example, the initial model that logic-based returns can choose and existing be pushed away by Maximum-likelihood estimation The loss function led.

Specifically, it can be concentrated every time from training sample and choose at least one training sample.Then by the training sample of selection Sample matches information in this is separately input into initial model, the corresponding output operation letter of the training sample chosen Breath.Later, output operation information characterization can be determined according to preset probability threshold value is that target user can be to the sample of input The corresponding history pushed information performance objective operation of match information or target user will not be corresponding to the sample matches information of input History pushed information performance objective operation.

For example, if probability threshold value is 0.5.It so, can be with if the probability of output operation information characterization is greater than 0.5 Think that target user can operate the corresponding history pushed information performance objective of the sample matches information of input.Accordingly, if it is defeated The probability of information representation is operated out less than 0.5, it may be considered that target user will not be corresponding to the sample matches information of input The operation of history pushed information performance objective.

It is then possible to indicate information according to output operation information and sample, the value of preset loss function is determined.Later, Determine whether initial model trains completion according to the value of loss function.If determining that initial model training is complete according to the value of loss function At, can by training complete initial model be determined as information prediction model.Wherein it is possible to be set in advance in the value of loss function When meeting preset condition, determine that initial model training is completed.

It, can be based on the value of loss function, using anti-if determining that initial model complete by training according to the value of loss function To the parameter propagated with gradient descent algorithm adjustment initial model.Meanwhile it can be concentrated from training sample and choose at least one again A training sample, and initial model adjusted is determined as initial model, continue to train.

Generally, completion could be trained by requiring repeatedly successive ignition to the training of initial model.Wherein it is possible to be based on The average value of the value for the corresponding loss function of each training sample chosen every time, adjusts the parameter of initial model.

Optionally, information prediction model can be trained as follows and be obtained: obtain training sample set.Wherein, training Training sample in sample set may include sample matches information and sample operations information.Wherein, sample matches information can wrap Include the match information of the corresponding history pushed information of target user.Sample operations information can serve to indicate that target user to correspondence History pushed information performance objective operation probability.Sample operations information can be labeled in advance by technical staff.Later, The method that curve matching can be used, is carried out curve fitting based on training sample set, and using obtained matched curve as letter Cease prediction model.

Wherein it is possible to using the existing various methods that can be used in curve matching (such as based on the fitting of least square method Algorithm etc.) or application, it is carried out curve fitting based on training sample set.

Step 203, it according to corresponding first operation information, is concentrated from candidate pushed information and chooses candidate pushed information.

In the present embodiment, since the first operation information can be used for characterizing based on the first match information, target user couple The probability of candidate pushed information performance objective operation.Therefore, can according to the sequence of corresponding first operation information from big to small, It is concentrated from candidate pushed information and chooses the candidate candidate pushed information of pushed information target numbers.

Wherein, target numbers can be preset by technical staff, (such as can also will according to fixed condition determination really is arranged The 10 of the total number for the candidate pushed information that candidate pushed information collection includes are determined as target numbers etc.).

Optionally, it can also be concentrated from candidate pushed information first according to preset probability threshold value and choose corresponding first behaviour The probability for making information representation is greater than the candidate pushed information of probability threshold value, obtains candidate push subset.Later, it can be pushed away from candidate It send and randomly selects the candidate pushed information of preset number in subset.

Step 204, the candidate pushed information selected to the corresponding terminal device push of target user.

In some optional implementations of the present embodiment, in response to detecting the predetermined registration operation of target user, to end The selected candidate pushed information of end equipment push, and control show selected candidate pushed information on the terminal device. Wherein, predetermined registration operation can be used for requesting push information relevant to the position that first location information indicates.

Wherein, predetermined registration operation may include by the pre-set any user's operation of technical staff.For example, predetermined registration operation can To include refresh operation, click the operation of preset control on the page that shows of terminal device etc..

Such as Such analysis, due to the number to the interested user of information relevant to the position that first location information indicates It is relatively fewer, therefore can be relevant in the push request for receiving user, then to user's push position corresponding to the user Information.It is possible thereby to the resource consumption for avoiding unnecessary information from pushing, reducing in process of information push as far as possible.

Optionally, terminal can also uniformly be concentrated in the push request for receiving user in this way Candidate pushed information is shown in equipment, in order to which user quickly browses to relevant information.For example, it can be browsed in user Two buttons are set on the page of information, for switching the different display pages.One button can control for showing and using The relevant information in the corresponding position in family.Another button can control for showing other pushed informations.

Based on this so that when user it is expected to browse the relevant information in position corresponding to the user, can quickly with It is shown on the corresponding terminal device in family, in order to which user checks.

The method provided by the above embodiment of the disclosure based on user at the position where it, to candidate pushed information Preference predicts probability that user operates candidate pushed information performance objective.In turn, it can be selected according to the probability of prediction Take pushed information to be pushed to the user, thus facilitate promoted pushed information accuracy, avoid to corresponding position The relevant uninterested user of information pushes relevant information, reduces the resource consumption in process of information push.On this basis, It can also be requested, push the relevant information in position corresponding to the user to user and controlled corresponding in user according to the push of user Terminal device on shown, can be with to realize when user wants the relevant information in browsing position corresponding to the user Timely view relevant information.

With further reference to Fig. 3, it illustrates the processes 300 of another embodiment of the method for pushed information.The use In the process 300 of the method for pushed information, comprising the following steps:

Step 301, the first location information for indicating the position of target user is obtained, and obtains candidate pushed information Collection.

The specific implementation procedure of this step can refer to the related description of the step 201 in Fig. 2 corresponding embodiment, herein not It repeats again.

Step 302, according to first location information, at least one second location information is determined.

In the present embodiment, it can preset and be stored between first location information and at least one second location information Corresponding relationship mapping table.At this point, can be searched and first in mapping table after determining first location information Confidence ceases at least one corresponding second location information.

In some optional implementations of the present embodiment, can according to first location information and it is preset at least one Distance threshold determines at least one second location information.Wherein, distance threshold can be used for characterizing first location information instruction The distance between the position of position and second location information instruction.

Wherein, distance threshold can be preset by technical staff.For example, distance threshold can be ten kilometers, that It can will be within ten kilometers of the position that first location information indicates and the position of the position comprising first location information instruction It sets and is determined as a second position, so as to obtain second location information.

Generally, due to which location information is that have certain hierarchical relationship.For example, the city XX, the XX province area the XX street XX.Its In, XX province can be a location information, and the city XX is also possible to a location information, and similarly, the area XX and the street XX are equally all It is location information.I.e. location information can have varigrained expression.Therefore, in practical applications, how position to be selected to believe The expression granularity of breath is also the one aspect for needing to consider.

For example, the position of user is obtained in the area XX, but user to the relevant information in the area XX and may lose interest in, The relevant information in the city Er Dui XX is interested, or may also be interested in the relevant information in the street XX.Similarly, one A pushed information can be information relevant to the area XX, also can be the relevant information in the street XX simultaneously.

Therefore, after determining the position of target user, it may be considered that the position of determining user is extensive to different grains The multiple positions indicated are spent, and then target user can be measured to candidate's push letter from multiple position Synthesis that granularity indicates are not had to The operation information of breath further promotes the accuracy of the operation information of prediction.

It should be noted that in the disclosure, different positions, decibel are named as first position, for ease of description Two positions etc., it should be appreciated by those skilled in the art that the first or second therein does not constitute the particular determination to position.

Step 303, the candidate pushed information concentrated for candidate pushed information executes following determining step:

Step 3031, the first match information of candidate's pushed information is determined.

Step 3032, according to the first match information, corresponding first operation information of candidate's pushed information is determined.

The specific implementation procedure of above-mentioned steps 3031 and 3032 can refer to step 2021 in Fig. 2 corresponding embodiment and 2022 related description, details are not described herein.

Step 3033, for the second location information at least one second location information, following steps are executed:

Step 30331, the second match information of candidate's pushed information is determined.

In this step, the second match information can be used for characterizing based on the second location information, and target user is to the time Select the preference of pushed information.Preference can weigh according to the various relevant informations of user and candidate's pushed information Amount.

Similarly with the first match information, can it is different with application demand flexible choice according to actual application scenarios, One or more can be characterized based on second location information, and target user carrys out the information of the preference of candidate's pushed information As the second match information.

Optionally, the second match information may include that there are user some history of higher preference to push away within a preset period of time It delivers letters the similarity of breath and candidate's pushed information.Wherein, user has some history of higher preference within a preset period of time Pushed information can be according within a preset period of time, and historical behavior data of the user at the position that second location information indicates are come It determines.

For example, can be pushed away within a preset period of time, and when user is at the position that second location information indicates to user The history pushed information that user clicks or commented on is chosen in each history pushed information sent has higher preference as user History pushed information.

Generally, there are some history pushed informations of higher preference and the candidate to push within a preset period of time and believe by user The similarity of breath is higher, user can be indicated at the position that second location information indicates, to the preference of candidate's pushed information Degree is higher.

Optionally, the second match information is included at least one of the following: for characterizing target user couple and the second confidence Cease the information of the preference of the relevant information in position of instruction, for characterizing candidate's pushed information and the second location information The information of the degree of correlation of the position of instruction, for characterize the second location information instruction position attention rate information.

Wherein, the determination method of the second match information can with reference in the step 2021 in Fig. 2 corresponding embodiment about Determine the related description of the first match information, details are not described herein.

Step 30332, according to the second match information, corresponding second operation information of candidate's pushed information is determined.

In this step, the second operation information can be used for characterizing based on the second match information, and target user is to the candidate The probability of pushed information performance objective operation.

In some optional implementations of the present embodiment, the second match information can be input to information prediction mould Type obtains corresponding second operation information of candidate's pushed information.

Wherein it is determined that the detailed process and information prediction model of the second operation information can refer to Fig. 2 corresponding embodiment In step 2022 in related description, details are not described herein.

Step 304, according to corresponding first operation information and at least one second operation information, from candidate pushed information collection The candidate pushed information of middle selection.

In some optional implementations of the present embodiment, for candidate pushed information concentrate candidate pushed information, In response to determining that the corresponding destination probability of candidate's pushed information is greater than preset probability threshold value, the candidate can be chosen and push letter Breath.Wherein, destination probability can be used to indicate that in maximum value or probability set in the corresponding probability set of candidate's pushed information The average value of each probability.Wherein, probability set can be by the probability of the corresponding first operation information characterization of candidate's pushed information The probability composition characterized respectively at least one corresponding second operation information.

It in the present embodiment, can be according to first when obtaining the first operation information and at least one second operation information The probability that operation information and at least one second operation information characterize respectively is as considering whether choose the one of candidate pushed information A aspect.In application scenes, needs whether to choose a candidate pushed information and need the factor considered can be by very much Aspect.At this point it is possible to which the probability that the first operation information and at least one second operation information are characterized respectively is as one of them Aspect, in conjunction with other aspects, whether comprehensive consideration chooses each candidate pushed information.

For example, in some cases, it is desirable to consider the format of candidate pushed information, for example, candidate pushed information be video also It is text, or when candidate pushed information is text, whether the image for including in candidate pushed information more etc..So as to tie The network state of target user is closed, suitable candidate pushed information is chosen, when using data network, is received to avoid in user Video or the text information etc. more comprising picture and text.

Step 305, the candidate pushed information selected to the corresponding terminal device push of target user.

The specific implementation procedure of this step can refer to the related description of the step 204 in Fig. 2 corresponding embodiment, herein not It repeats again.

With continued reference to the signal that Fig. 4, Fig. 4 are according to the application scenarios of the method for pushed information of the present embodiment Figure 40 0.In the application scenarios of Fig. 4, above-mentioned executing subject can first obtain the location information of the position of user based on GPS technology 402 be " five road junction of Haidian District, Beijing City ".It then can be extensive to varigrained two location informations by location information 402 401 and 403.As shown in the figure, location information 401 is " Haidian District, Beijing City ", and location information 403 is " Beijing ".

Above-mentioned executing subject can also from the database of connection 404 obtain information collection 405.Then for information collection 405 In an information, can determine respectively based on three varigrained location informations, clicking rate of the user to the information.Below with An information 4051 in information collection 405 is illustrated as example.

For location information 401, i.e., " Haidian District, Beijing City ", can determine that user pushes history with " Beijing sea The clicking rate of the relevant information in shallow lake area " characterizes the preference of user couple information relevant to " Haidian District, Beijing City ".Later, The content of information 4051 be can analyze to determine the degree of correlation of information 4051 Yu " Haidian District, Beijing City ".Later, available pre- If " Haidian District, Beijing City " corresponding attention rate.It is then possible to by determining " Haidian District, Beijing City " corresponding clicking rate, phase Guan Du, attention rate form match information 406.Then, match information 406 can be input to clicking rate prediction mould trained in advance Type 407 obtains under " Haidian District, Beijing City " this location information, and the probability 408 of user's click information 4051 of prediction is 0.9。

For location information 402, i.e., " five road junction of Haidian District, Beijing City ", can determine that user pushes history with " north The clicking rate of the relevant information in five road junction of the Haidian District Jing Shi " characterizes user couple letter relevant to " five road junction of Haidian District, Beijing City " The preference of breath.Later, it can analyze the content of information 4051 to determine information 4051 and " five road junction of Haidian District, Beijing City " The degree of correlation.Later, available preset " five road junction of Haidian District, Beijing City " corresponding attention rate.It is then possible to by determining " five road junction of Haidian District, Beijing City " corresponding clicking rate, the degree of correlation, attention rate form match information 409.It then, can general It is input to clicking rate prediction model 407 trained in advance with information 409, obtains in " five road junction of Haidian District, Beijing City " this position Under confidence breath, the probability 410 of user's click information 4051 of prediction is 0.8.

For location information 403, i.e., " Beijing ", it can determine the letter relevant to " Beijing " that user pushes history The clicking rate of breath characterizes the preference of user couple information relevant to " Beijing ".Later, it can analyze information 4051 Content is to determine the degree of correlation of information 4051 Yu " Beijing ".Later, available preset " Beijing " corresponding attention rate. It is then possible to form match information 411 by determining " Beijing " corresponding clicking rate, the degree of correlation, attention rate.It then, can be with Match information 411 is input to clicking rate prediction model 407 trained in advance, is obtained in " Beijing " this location information Under, the probability 412 of user's click information 4051 of prediction is 0.2.

Later, it is 0.9 that maximum value can be chosen from the corresponding upper prediction probability of information 4051.It is greater than due to 0.9 Preset threshold value 0.6, it is possible to be shown to corresponding 413 pushed information 4051 of terminal device of user, and on terminal device 413 Show information 4051.

The method provided by the above embodiment of the disclosure is extensive by the position of user by after obtaining the position of user To multiple positions of different granularities, hence for the position of each granularity, predict user to the behaviour of candidate pushed information respectively Make information, whether push candidate pushed information so as to comprehensively consider, facilitates the accuracy for further promoting pushed information.

With further reference to Fig. 5, as the realization to method shown in above-mentioned each figure, present disclose provides for pushed information One embodiment of device, the Installation practice is corresponding with embodiment of the method shown in Fig. 2, which specifically can be applied to In various electronic equipments.

As shown in figure 5, the device 500 provided in this embodiment for pushed information includes acquiring unit 501, operation information Determination unit 502, selection unit 503 and push unit 504.Wherein, acquiring unit 501 is configured to obtain for indicating target The first location information of the position of user, and obtain candidate pushed information collection；Operation information determination unit 502 is configured to pair In the candidate pushed information that candidate pushed information is concentrated, executes following determining step: determining first of candidate's pushed information With information, wherein the first match information is based on first location information for characterizing, and target user is inclined to candidate's pushed information Good degree；According to the first match information, corresponding first operation information of candidate's pushed information is determined, wherein the first operation letter Breath is based on the first match information, the probability that target user operates candidate's pushed information performance objective for characterizing；It chooses single Member 503 is configured to concentrate from candidate pushed information according to corresponding first operation information and choose candidate pushed information；Push is single Member 504 is configured to the candidate pushed information selected to the corresponding terminal device push of target user.

In the present embodiment, in the device of pushed information 500: acquiring unit 501, operation information determination unit 502, The specific processing of selection unit 503 and push unit 504 and its brought technical effect can refer to Fig. 2 corresponding embodiment respectively In step 201, step 202, the related description of step 203 and step 204, details are not described herein.

In some optional implementations of the present embodiment, the above-mentioned device 500 for pushed information further include: position Information determination unit (not shown) is configured to determine at least one second location information according to first location information；With And aforesaid operations information determination unit 502, it is further configured to for the second position at least one second location information Information determines the second match information of candidate's pushed information, wherein the second match information is based on the second position for characterizing Information, preference of the target user to candidate's pushed information；According to the second match information, candidate's pushed information pair is determined The second operation information answered, wherein the second operation information is based on the second match information for characterizing, and target user pushes away the candidate The probability of breath performance objective of delivering letters operation；And above-mentioned selection unit 503 is further configured to according to corresponding first operation Information and at least one second operation information are concentrated from candidate pushed information and choose candidate pushed information.

In some optional implementations of the present embodiment, above-mentioned location information determination unit is further configured to root According to first location information, determine at least one second location information, comprising: according to first location information and it is preset at least one Distance threshold determines at least one second location information, wherein distance threshold is used to characterize the position of first location information instruction With the distance between the position of second location information instruction.

In some optional implementations of the present embodiment, aforesaid operations information determination unit 502 is further configured At the first match information to be input to information prediction model trained in advance, corresponding first operation of candidate's pushed information is obtained Information, wherein information prediction model is used to characterize the corresponding relationship between match information and operation information；And aforesaid operations letter Breath determines that single 502 are further configured to: the second match information being input to information prediction model, obtains candidate's pushed information Corresponding second operation information.

In some optional implementations of the present embodiment, the first match information is included at least one of the following: for table Levy the information of the preference of target user couple information relevant to the position that first location information indicates, for characterizing the candidate The information of the degree of correlation of the position of pushed information and first location information instruction, the position for characterizing first location information instruction Attention rate information.

In some optional implementations of the present embodiment, training obtains information prediction model as follows: obtaining Take training sample set, wherein the training sample that training sample is concentrated includes sample matches information and sample instruction information, wherein Sample matches information includes the match information of the corresponding history pushed information of target user, and sample instruction information is used to indicate target Whether user executed object run to corresponding history pushed information；Obtain the initial model pre-established；Utilize engineering The method of habit is based on training sample set and preset loss function, is trained to initial model, and will be initial after training Model is determined as information prediction model.

In some optional implementations of the present embodiment, above-mentioned selection unit 503 is further configured to for waiting The candidate pushed information for selecting pushed information to concentrate, in response to determining that it is preset that the corresponding destination probability of candidate's pushed information is greater than Probability threshold value chooses candidate's pushed information, wherein destination probability is for indicating in the corresponding probability set of candidate's pushed information Maximum value or each probability in probability set average value, wherein probability set by candidate's pushed information it is corresponding first behaviour The probability composition that the probability and at least one corresponding second operation information for making information representation characterize respectively.

In some optional implementations of the present embodiment, above-mentioned acquiring unit 501 is further configured to according to One location information obtains candidate pushed information collection, wherein the candidate pushed information and first position that candidate pushed information is concentrated are believed The position for ceasing instruction is related.

In some optional implementations of the present embodiment, above-mentioned push unit 503 be further configured in response to The predetermined registration operation for detecting target user, the candidate pushed information selected to terminal device push, and control are set in terminal Candidate pushed information selected by standby upper display, wherein predetermined registration operation is used to request the position of push with first location information instruction Set relevant information.

The device provided by the above embodiment of the disclosure obtains the position for indicating target user by acquiring unit First location information, and obtain candidate pushed information collection；The time that operation information determination unit concentrates candidate pushed information Pushed information is selected, following determining step is executed: determining the first match information of candidate's pushed information, wherein the first matching letter Breath is based on first location information, preference of the target user to candidate's pushed information for characterizing；According to the first matching letter Breath, determines corresponding first operation information of candidate's pushed information, wherein the first operation information is for characterizing based on the first matching Information, the probability that target user operates candidate's pushed information performance objective；Selection unit is believed according to corresponding first operation Breath is concentrated from candidate pushed information and chooses candidate pushed information；Push unit pushes institute to the corresponding terminal device of target user The candidate pushed information of selection avoids to facilitate the accuracy of promotion pushed information to relevant to corresponding position The uninterested user of information pushes relevant information, reduces the resource consumption in process of information push.

Below with reference to Fig. 6, it illustrates the electronic equipment that is suitable for being used to realize embodiment of the disclosure, (example is as shown in figure 1 Server) 600 structural schematic diagram.Server shown in Fig. 6 is only an example, should not be to the function of embodiment of the disclosure Any restrictions can be brought with use scope.

As shown in fig. 6, electronic equipment 600 may include processing unit (such as central processing unit, graphics processor etc.) 601, random access can be loaded into according to the program being stored in read-only memory (ROM) 602 or from storage device 608 Program in memory (RAM) 603 and execute various movements appropriate and processing.In RAM 603, it is also stored with electronic equipment Various programs and data needed for 600 operations.Processing unit 601, ROM 602 and RAM603 are connected with each other by bus 604. Input/output (I/O) interface 605 is also connected to bus 604.

In general, following device can connect to I/O interface 605: including such as touch screen, touch tablet, keyboard, mouse, taking the photograph As the input unit 606 of head, microphone, accelerometer, gyroscope etc.；Including such as liquid crystal display (LCD), loudspeaker, vibration The output device 607 of dynamic device etc.；And communication device 609.Communication device 609 can permit electronic equipment 600 and other equipment It is wirelessly or non-wirelessly communicated to exchange data.Although Fig. 6 shows the electronic equipment 600 with various devices, should manage Solution be, it is not required that implement or have all devices shown.It can alternatively implement or have more or fewer devices. Each box shown in Fig. 6 can represent a device, also can according to need and represent multiple devices.

Particularly, in accordance with an embodiment of the present disclosure, it may be implemented as computer above with reference to the process of flow chart description Software program.For example, embodiment of the disclosure includes a kind of computer program product comprising be carried on computer-readable medium On computer program, which includes the program code for method shown in execution flow chart.In such reality It applies in example, which can be downloaded and installed from network by communication device 609, or from storage device 608 It is mounted, or is mounted from ROM 602.When the computer program is executed by processing unit 601, the implementation of the disclosure is executed The above-mentioned function of being limited in the method for example.

It should be noted that computer-readable medium described in embodiment of the disclosure can be computer-readable signal Medium or computer readable storage medium either the two any combination.Computer readable storage medium for example can be with System, device or the device of --- but being not limited to --- electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor, or it is any more than Combination.The more specific example of computer readable storage medium can include but is not limited to: have one or more conducting wires Electrical connection, portable computer diskette, hard disk, random access storage device (RAM), read-only memory (ROM), erasable type are programmable Read-only memory (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic are deposited Memory device or above-mentioned any appropriate combination.In embodiment of the disclosure, computer readable storage medium, which can be, appoints What include or the tangible medium of storage program that the program can be commanded execution system, device or device use or and its It is used in combination.And in embodiment of the disclosure, computer-readable signal media may include in a base band or as carrier wave The data-signal that a part is propagated, wherein carrying computer-readable program code.The data-signal of this propagation can be adopted With diversified forms, including but not limited to electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal is situated between Matter can also be any computer-readable medium other than computer readable storage medium, which can be with It sends, propagate or transmits for by the use of instruction execution system, device or device or program in connection.Meter The program code for including on calculation machine readable medium can transmit with any suitable medium, including but not limited to: electric wire, optical cable, RF (radio frequency) etc. or above-mentioned any appropriate combination.

Above-mentioned computer-readable medium can be included in above-mentioned server；It is also possible to individualism, and without It is incorporated in the server.Above-mentioned computer-readable medium carries one or more program, when said one or multiple journeys When sequence is executed by the server, so that the server: the first location information for indicating the position of target user is obtained, and Obtain candidate pushed information collection；For the candidate pushed information that candidate pushed information is concentrated, execute following determining step: determining should First match information of candidate pushed information, wherein the first match information is based on first location information, target user for characterizing To the preference of candidate's pushed information；According to the first match information, corresponding first operation of candidate's pushed information is determined Information, wherein the first operation information is based on the first match information for characterizing, and target user executes mesh to candidate's pushed information Mark the probability of operation；According to corresponding first operation information, is concentrated from candidate pushed information and choose candidate pushed information；To target The selected candidate pushed information of the corresponding terminal device push of user.

The behaviour for executing embodiment of the disclosure can be write with one or more programming languages or combinations thereof The computer program code of work, programming language include object oriented program language-such as Java, Smalltalk, C++ further include conventional procedural programming language-such as " C " language or similar program design language Speech.Program code can be executed fully on the user computer, partly be executed on the user computer, as an independence Software package execute, part on the user computer part execute on the remote computer or completely in remote computer or It is executed on server.In situations involving remote computers, remote computer can pass through the network of any kind --- packet Include local area network (LAN) or wide area network (WAN) --- it is connected to subscriber computer, or, it may be connected to outer computer (such as It is connected using ISP by internet).

Flow chart and block diagram in attached drawing are illustrated according to the system of the various embodiments of the disclosure, method and computer journey The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation A part of one module, program segment or code of table, a part of the module, program segment or code include one or more use The executable instruction of the logic function as defined in realizing.It should also be noted that in some implementations as replacements, being marked in box The function of note can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are actually It can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it to infuse Meaning, the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart can be with holding The dedicated hardware based system of functions or operations as defined in row is realized, or can use specialized hardware and computer instruction Combination realize.

Being described in unit involved in embodiment of the disclosure can be realized by way of software, can also be passed through The mode of hardware is realized.Described unit also can be set in the processor, for example, can be described as: a kind of processor Including acquiring unit, operation information determination unit, selection unit and push unit.Wherein, the title of these units is in certain feelings The restriction to the unit itself is not constituted under condition, for example, acquiring unit is also described as " obtaining for indicating that target is used The first location information of the position at family, and obtain the unit of candidate pushed information collection ".

Above description is only the preferred embodiment of the disclosure and the explanation to institute's application technology principle.Those skilled in the art Member it should be appreciated that embodiment of the disclosure involved in invention scope, however it is not limited to the specific combination of above-mentioned technical characteristic and At technical solution, while should also cover do not depart from foregoing invention design in the case where, by above-mentioned technical characteristic or its be equal Feature carries out any combination and other technical solutions for being formed.Such as disclosed in features described above and embodiment of the disclosure (but It is not limited to) technical characteristic with similar functions is replaced mutually and the technical solution that is formed.

Claims

1. a kind of method for pushed information, comprising:

The first location information for indicating the position of target user is obtained, and obtains candidate pushed information collection；

For the candidate pushed information that the candidate pushed information is concentrated, executes following determining step: determining that the candidate pushes letter First match information of breath, wherein first match information is based on the first location information for characterizing, and the target is used Preference of the family to candidate's pushed information；According to first match information, candidate's pushed information corresponding is determined One operation information, wherein first operation information is based on first match information for characterizing, and the target user is to this The probability of candidate pushed information performance objective operation；

According to corresponding first operation information, is concentrated from the candidate pushed information and choose candidate pushed information；

The candidate pushed information selected to the corresponding terminal device push of the target user.

2. according to the method described in claim 1, wherein, the method also includes:

According to the first location information, at least one second location information is determined；And

The determining step further include:

For the second location information at least one described second location information, the second matching of candidate's pushed information is determined Information, wherein second match information is based on the second location information for characterizing, and the target user pushes the candidate The preference of information；According to second match information, corresponding second operation information of candidate's pushed information is determined, In, second operation information is based on second match information for characterizing, and the target user is to candidate's pushed information The probability of performance objective operation；And

It is described according to corresponding first operation information, concentrated from the candidate pushed information and choose candidate pushed information, comprising:

According to corresponding first operation information and at least one second operation information, concentrates to choose from the candidate pushed information and wait Select pushed information.

3. described according to the first location information according to the method described in claim 2, wherein, determine at least one second Location information, comprising:

According to the first location information and at least one preset distance threshold, at least one second location information is determined, In, distance threshold be used to characterize between the position of first location information instruction and the position of second location information instruction away from From.

4. it is described according to first match information according to the method described in claim 2, wherein, determine that the candidate pushes letter Cease corresponding first operation information, comprising:

First match information is input to in advance trained information prediction model, obtains candidate's pushed information corresponding the One operation information, wherein the information prediction model is used to characterize the corresponding relationship between match information and operation information；And

It is described according to second match information, determine corresponding second operation information of candidate's pushed information, comprising:

Second match information is input to the information prediction model, obtains corresponding second operation of candidate's pushed information Information.

5. according to the method described in claim 1, wherein, first match information is included at least one of the following: for characterizing The information of the preference of the target user couple information relevant to the position that the first location information indicates, for characterizing Candidate's pushed information and the first location information instruction position the degree of correlation information, for characterizing the first position The information of the attention rate of the position of information instruction.

6. according to the method described in claim 2, wherein, second match information is included at least one of the following: for characterizing The information of the preference of the target user couple information relevant to the position that the second location information indicates, for characterizing this The information of the degree of correlation for the position that candidate pushed information is indicated with the second location information refers to for characterizing the second location information The information of the attention rate of the position shown.

7. according to the method described in claim 4, wherein, training obtains the information prediction model as follows:

Obtain training sample set, wherein the training sample that the training sample is concentrated includes sample matches information and sample instruction Information, wherein sample matches information includes the match information of the corresponding history pushed information of the target user, sample instruction letter Breath is used to indicate whether the target user executed the object run to corresponding history pushed information；

Obtain the initial model pre-established；

Using the method for machine learning, it is based on the training sample set and preset loss function, the initial model is carried out Training, and the initial model after training is determined as information prediction model.

8. described according to corresponding first operation information and at least one second behaviour according to the method described in claim 2, wherein Make information, concentrated from the candidate pushed information and choose candidate pushed information, comprising:

For the candidate pushed information that the candidate pushed information is concentrated, in response to determining the corresponding target of candidate's pushed information Probability is greater than preset probability threshold value, chooses candidate's pushed information, wherein the destination probability is for indicating that the candidate pushes The average value of each probability in maximum value or the probability set in the corresponding probability set of information, wherein the probability set by The probability of the corresponding first operation information characterization of candidate's pushed information and at least one corresponding second operation information distinguish table The probability of sign forms.

9. according to the method described in claim 1, wherein, the candidate pushed information collection of the acquisition, comprising:

According to the first location information, the candidate pushed information collection is obtained, wherein the time that candidate's pushed information is concentrated The position for selecting pushed information to indicate to the first location information is related.

10. method described in one of -9 according to claim 1, wherein described to be pushed away to the corresponding terminal device of the target user Send selected candidate pushed information, comprising:

In response to detecting the predetermined registration operation of the target user, the selected candidate push letter of Xiang Suoshu terminal device push Breath, and control show selected candidate pushed information on the terminal device, wherein the predetermined registration operation is for requesting Push information relevant to the position that the first location information indicates.

11. a kind of device for pushed information, comprising:

Acquiring unit is configured to obtain the first location information for indicating the position of target user, and obtains candidate and push away It delivers letters breath collection；

Operation information determination unit is configured to the candidate pushed information concentrated for the candidate pushed information, executes as follows It determines step: determining the first match information of candidate's pushed information, wherein first match information is based on institute for characterizing State first location information, preference of the target user to candidate's pushed information；According to first match information, really Fixed corresponding first operation information of candidate's pushed information, wherein first operation information is based on described first for characterizing Match information, the probability that the target user operates candidate's pushed information performance objective；

Selection unit, is configured to according to corresponding first operation information, concentrates selection candidate to push away from the candidate pushed information It delivers letters breath；

Push unit is configured to the candidate pushed information selected to the corresponding terminal device push of the target user.

12. device according to claim 11, wherein described device further include:

Location information determination unit is configured to determine at least one second location information according to the first location information；With And

The operation information determination unit is further configured to:

The selection unit is further configured to:

13. device according to claim 12, wherein the location information determination unit is further configured to:

14. device according to claim 12, wherein the operation information determination unit is further configured to:

The operation information determination unit is further configured to:

15. device according to claim 11, wherein first match information is included at least one of the following: for table The information of the preference for information relevant to the position that the first location information indicates of levying the target user couple, for table Levy the information of the degree of correlation of the position of candidate's pushed information and first location information instruction, for characterizing described first The information of the attention rate of the position of confidence breath instruction.

16. device according to claim 12, wherein second match information is included at least one of the following: for table The information of the preference for information relevant to the position that the second location information indicates of levying the target user couple, for characterizing Candidate's pushed information and the second location information instruction position the degree of correlation information, for characterizing the second location information The information of the attention rate of the position of instruction.

17. device according to claim 14, wherein training obtains the information prediction model as follows:

Obtain the initial model pre-established；

18. device according to claim 12, wherein the selection unit is further configured to:

19. device according to claim 11, wherein the acquiring unit is further configured to:

20. device described in one of 1-19 according to claim 1, wherein the push unit is further configured to:

21. a kind of server, comprising:

One or more processors；

Storage device is stored thereon with one or more programs；

When one or more of programs are executed by one or more of processors, so that one or more of processors are real The now method as described in any in claim 1-10.

22. a kind of computer-readable medium, is stored thereon with computer program, wherein the realization when program is executed by processor Method as described in any in claim 1-10.