WO2017121258A1 - 信息投放方法、装置、服务器和存储介质 - Google Patents

信息投放方法、装置、服务器和存储介质 Download PDF

Info

Publication number
WO2017121258A1
WO2017121258A1 PCT/CN2016/113842 CN2016113842W WO2017121258A1 WO 2017121258 A1 WO2017121258 A1 WO 2017121258A1 CN 2016113842 W CN2016113842 W CN 2016113842W WO 2017121258 A1 WO2017121258 A1 WO 2017121258A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
ranking
support
candidate
type
Prior art date
Application number
PCT/CN2016/113842
Other languages
English (en)
French (fr)
Inventor
刘大鹏
肖磊
Original Assignee
腾讯科技（深圳）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 腾讯科技（深圳）有限公司 filed Critical 腾讯科技（深圳）有限公司
Priority to JP2017559381A priority Critical patent/JP6522160B2/ja
Priority to EP16884809.1A priority patent/EP3404607A4/en
Priority to KR1020177032826A priority patent/KR101999469B1/ko
Publication of WO2017121258A1 publication Critical patent/WO2017121258A1/zh
Priority to US15/804,559 priority patent/US11144950B2/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0242—Determining effectiveness of advertisements
    • G06Q30/0243—Comparative campaigns
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0242—Determining effectiveness of advertisements
    • G06Q30/0244—Optimization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00—Computing arrangements based on specific mathematical models
    • G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0242—Determining effectiveness of advertisements
    • G06Q30/0246—Traffic
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0251—Targeted advertisements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0277—Online advertisement
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0282—Rating or review of business operators or products
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning

Definitions

  • the present invention relates to the field of information delivery technologies, and in particular, to an information delivery method, apparatus, system, and storage medium.
  • the information delivery system cooperates with the e-commerce website, and the information system must ensure that 20% of the traffic is used to place the seller's advertisement on the e-commerce website, so that the products sold on the e-commerce website can be used as a kind of stable traffic, which is beneficial to the product transportation. dimension.
  • e-commerce advertising is gradually developed, explore the application of the development platform (English full name: application, English abbreviation: app), but there are fewer advertisers in the initial app category, no advertising experience, if not Protective measures, which may fluctuate greatly in terms of traffic and effects, hinder the expansion of advertisers.
  • it is necessary to support the app-type advertising introduce sufficient app-like ads, and wait until the ecology is gradually formed, and then completely freely compete with other types of advertising.
  • the existing information delivery solution lacks flexible control over the support requirements, so a new delivery solution is needed.
  • the embodiment of the invention provides an information delivery method and device, which can support the information belonging to the support type on the basis of ensuring the effect of the information delivery.
  • an embodiment of the present invention provides a method for information delivery, including:
  • the candidate information list including a plurality of information
  • an embodiment of the present invention further provides an information delivery apparatus, including:
  • a primary selection module configured to select a candidate information list according to a page request sent by the user, where the candidate information list includes multiple pieces of information
  • a ranking prediction module configured to generate a ranking of each information in the candidate information list
  • a current request calculation module configured to acquire, according to a ranking of each information in the candidate information list, a current average ranking of information of a specified type in the candidate information list;
  • the information delivery selection module is configured to determine whether the information of the specified type is used as the designated delivery information based on a current average ranking of the specified type of information and a pre-stored historical average ranking corresponding to the information of the specified type.
  • a candidate information list is first selected according to a page request sent by a user, where the candidate information list includes: a plurality of information, and then the ranking order of each information in the candidate information list is predicted according to the page request and the configured prediction model. , thereby generating a ranking of each information in the candidate information list, and then obtaining a current average ranking of information of a specified type in the candidate information list according to the ranking of each information in the candidate information list, and finally according to the current average ranking of the information of the specified type and Specify the historical support average position corresponding to the type, and select the specified delivery information from the specified type of information.
  • the sort order prediction can obtain the current average rank of the information belonging to the support type in the candidate information list, and finally select from the information of the support type according to the current average rank of the information belonging to the support type and the historical support average rank corresponding to the support type.
  • the selected supported information is determined according to the current average ranking and the historical support average ranking corresponding to the support type, the selected supported information is higher in all the information of the candidate information list than the user who sends the page request. Matching information so that it can match the user's interest in love Well, to ensure the effectiveness of information delivery, while the information to be supported is selected from the information of the type of support, so as to meet the needs of specific types of information support.
  • FIG. 1 is a schematic block diagram of a method for information delivery according to an embodiment of the present invention
  • FIG. 2 is a schematic diagram of an implementation scenario of an information delivery apparatus according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of an implementation scenario of an information delivery method according to an embodiment of the present invention.
  • FIG. 4 is a schematic structural diagram of an information delivery apparatus according to an embodiment of the present disclosure.
  • FIG. 4 is a schematic structural diagram of another information delivery apparatus according to an embodiment of the present disclosure.
  • FIG. 4 is a schematic structural diagram of a structure of an information delivery selection module according to an embodiment of the present invention.
  • FIG. 4 is a schematic structural diagram of another information delivery device according to an embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of a composition of an information delivery method applied to a server according to an embodiment of the present invention.
  • the existing technical solution is to configure a certain proportion of exclusive traffic, the exclusive traffic is only allowed to deliver certain types of supported information, and the control method is usually randomly controlled according to the user's page request.
  • this technical solution is simple and easy to implement and the traffic is stable, there are the following problems: some users in the randomly generated traffic are not interested in such supported information, but the random control may just put the information and some users. Interested in such information that is supported, but may not be placed in this type of information when randomly controlled. Therefore, although the information support strategy adopted by such a scheme can control the proportion of traffic, the information delivery effect is not good.
  • the embodiment of the invention provides an information delivery method and device, which can support the information belonging to the support type on the basis of ensuring the effect of the information delivery.
  • an embodiment of the information delivery method of the present invention may be specifically applied to a specific type of information support scenario.
  • the information may be an advertisement, and the information described in the embodiment of the present invention may also be a resource that needs to be supported.
  • the content, data and other forms are not limited.
  • an information delivery method provided by an embodiment of the present invention may include steps 101 to 104.
  • a candidate information list is selected according to a page request sent by the user, and the candidate information list includes a plurality of information.
  • the user when the user accesses the page, the user may send a page request, and the information delivery device provided by the embodiment of the present invention may receive the page request sent by the user, and the page request sent by the user may obtain the relevant information of the user.
  • the information about the user can be used to determine which types of information the user is interested in or to determine which information the user wants to view, such as determining which aspect of the advertisement the user is willing to watch.
  • the page request sent by the user may carry the device information, the access context information, and the user attribute information of the user operation terminal, and the information delivery device may obtain the information content from the page request, and the information delivery device may All the information to be broadcasted is pre-selected, and a candidate information list matching the user is obtained.
  • All the information in the candidate information list is information obtained after the information delivery device has performed the primary selection, and all the information in the candidate information list is The type of information that may meet the needs of the user's hobbies, and the primary selection of information according to the page request may improve the targeting of the user group for information delivery, which is completely different from the prior art in that random play is not guaranteed to ensure the effect of the information.
  • the primary selection of the advertisement according to the page request may improve the targeting of the user group for the advertisement, which is completely different from the manner in which the random play is not guaranteed in the prior art.
  • the object-oriented retrieval of the information may be performed according to the page request of the user and the delivery condition of the information main delivery information, The problem of poor information delivery caused by the broadcast of information that does not match the user is reduced.
  • the information is specifically for the advertisement.
  • the advertisement produced by the advertiser is a maternal and child-like product, and the advertiser is determined to target the female person aged 25 to 40, and the object of the advertisement can be targeted according to the demand of the advertiser. If the user who sends the page request is a male, the targeted search can be used to determine that the advertisement of the maternal and child products is not delivered to the male user, and the advertisement of the maternal and child supplies does not need to be included in the candidate advertisement list.
  • the candidate information list is set for the user who sends the page request, and the candidate information list includes multiple information, and the information may belong to the same information type or may be different.
  • Type of information and some of the information included in the candidate information list may belong to a specified type of information, such as information requiring a support type.
  • an app type advertisement may be an advertisement requiring a support type, and other specific types of advertisements may also be advertisements requiring a support type.
  • the information belonging to the support type in the embodiment of the present invention is included in the candidate information list, but whether the information belonging to the support type is determined to be supported may also be determined in combination with the information delivery method described in the subsequent embodiments.
  • step 102 a plurality of pieces of information in the candidate information list are sorted according to a predetermined rule, and a ranking of each piece of information in the candidate information list is generated.
  • the ranking order of each information in the candidate information list may be predicted according to the page request and the configured prediction model, thereby generating a ranking of each information in the candidate information list.
  • the candidate information list is determined by the primary selection manner, and then the ranking order of all the information in the candidate information list is predicted according to the page request sent by the user and the configured prediction model, and the candidate information list includes M (M is If there is a non-zero natural number) information, then the ranking order of the first information, the second information, ..., the Mth information in the candidate information list needs to be separately predicted, thereby generating the ranking of each information in the candidate information list. .
  • the ranking of each information determines the ranking of each information in the candidate information list. By ranking the predictions of all the information in the candidate information list, it is possible to predict between each information in the candidate information list and the user who sends the page request. The degree of matching, the higher the ranking of the information in the candidate information list, the more the user who sends the page request matches.
  • the ranking prediction of each information in the candidate information list may be determined according to the user's page request and the configured prediction model.
  • the information placing apparatus may predict the ranking order of each information according to three factors: the user, the context scenario, and the information.
  • the specific prediction model may be a logistic regression algorithm or a specific prediction algorithm implemented by a decision tree algorithm or the like.
  • step 102 predicts the ranking order of each information in the candidate information list according to the page request and the configured prediction model, so as to generate a ranking of each information in the candidate information list, which may specifically include the following steps. :
  • A1 Acquire a click rate prediction value of each information in the candidate information list according to the page request and the preset prediction model, and obtain a ranking of each information in the candidate information list according to the click rate prediction value of each information.
  • the information delivery device may predict the click rate (English full name: Click Through Rate, English abbreviation: CTR) of all the information in the candidate information list according to the page request and the preset prediction model, and the CTR may reflect the user's information (such as an advertisement).
  • the CTR of each information in the candidate information list can be predicted, thereby generating a CTR prediction value of each information, and the size of the CTR prediction value can be used as each information in the candidate information list.
  • the ranking is based on which the ranking of each information in the candidate information list is obtained.
  • the calculation of the click rate prediction value of each information may be used as a basis for the information ranking, but the embodiment of the present invention is not limited to the method of using the click rate prediction value, and may be adopted in practical applications. Other ways are used as the basis for ranking information.
  • the ranking of each information in the candidate information list according to the click rate prediction value of each information in step A1 may include steps A11 to A12.
  • step A11 the revenue prediction value of each information is obtained according to the click rate prediction value of each information
  • step A12 the revenue prediction values of all the information in the candidate information list are sequentially arranged in the order of the magnitude of the values, and the ranking of each information in the candidate information list is obtained.
  • the candidate information list can be The revenue prediction values of all the information in the order are ranked in order of magnitude, and the ranking of each information in the candidate information list is obtained.
  • the specified type of information may be information belonging to a type of support, that is, information that is predetermined to be supported.
  • step 102 predicts the ranking order of each information in the candidate information list according to the page request and the configured prediction model, thereby generating the ranking of each information in the candidate information list, and the embodiment of the present invention
  • the information delivery method provided may further include:
  • step 103 Determining whether the information belonging to the support type in the candidate information list satisfies the preset condition, and if the information belonging to the support type satisfies the condition, triggering the following step 103: obtaining the specified type in the candidate information list according to the ranking of each information in the candidate information list The current average ranking of the information.
  • the information belonging to the support type in the candidate information list may be analyzed, and the preset support conditions are used to determine whether information belonging to the support type is needed, if necessary If the information belonging to the support type is supported, the subsequent step 103 is triggered, and after the step 103 is performed, the step 104 can be performed. If it is not necessary to support the information belonging to the support type, the subsequent information support process does not need to be performed, the process can be ended, the fair competition is performed according to all the information in the candidate information list, and finally the information broadcasted to the user is determined.
  • the support conditions may be set as needed, and embodiments of the present invention are not limited in this regard.
  • the support conditions can be determined according to the business rules. For example, if the information belonging to the support type is ranked in the last 20% of the entire list in the candidate information list, the information belonging to the support type matches the user. The degree is low, and the information belonging to the support type can be not supported to avoid the user's satisfaction with the information.
  • the support condition can be set to the user sending the request in the wireless local area network (English name: Wireless Fidelity, English abbreviation: WiFi) environment, according to the page request sent by the user, it is determined that the user does not use the information support when the user is not using the WiFi, so as to avoid bringing the user The resulting flow loss.
  • the wireless local area network English name: Wireless Fidelity, English abbreviation: WiFi
  • step 103 the current average ranking of the information belonging to the specified type in the candidate information list is obtained according to the ranking of each information in the candidate information list.
  • the ranking of each information in the candidate information list may be generated, and then the ranking of each information in the candidate information list may be calculated according to the ranking of each information in the candidate information list.
  • the current average ranking of the information of the type specified in the candidate information list For example, taking the specified type as the required support type as an example, taking the support type as the app type as an example, and taking the information as the advertisement as an example, the support type advertisement may be an app type advertisement, and the calculation app type advertisement is in the candidate advertisement list.
  • the current average position if the app type of ad 1 ranked candidate ad list as t 1, if the app type ad 2 ranking in the candidate list of advertisements to T 2, then the app types of ads current average position (t 1 + t 2 )/2.
  • the ranking of the advertisement in the candidate advertisement list is the current average ranking of the advertisements belonging to the support type.
  • the specified delivery information is selected from the information of the specified type according to the current average ranking of the information of the specified type and the historical average ranking corresponding to the information of the specified type.
  • the specified delivery information is information that is delivered in a supported form
  • the historical average ranking is a historical support average ranking
  • the matching degree of the information of the specified type and the user may be determined by the current average ranking, and the specified type may be determined according to the historical average ranking corresponding to the supporting type.
  • the current average ranking of the information and the historical average ranking corresponding to the specified type, thereby obtaining the relationship between the current average ranking and the historical average ranking, and according to the relationship between the two, the specified delivery information can be selected from the specified types of information.
  • the historical average ranking is a ranking average obtained by historical data for the type of information that needs to be supported, and the historical support average ranking can be used to measure the current specified type of information.
  • the reference amount should be supported.
  • the historical support average ranking corresponding to the current average ranking and the support type of the information belonging to the support type may not be limited to the above manner when selecting the supported information, and may be combined with other parameter factors for determining the supported information. .
  • step 104 selects the specified delivery information from the specified type of information according to the current average ranking of the information of the specified type and the historical average ranking corresponding to the support type. Specifically, it may include steps C1 to C4.
  • step C1 a preferred probability of information belonging to the support type is calculated according to the current average ranking of the information belonging to the support type and the historical support average ranking corresponding to the support type;
  • step C2 the current support satisfaction degree is calculated according to the selected support number of the information belonging to the support type, the total number of requests sent by the user, and the preset support ratio;
  • step C3 the support probability of the information belonging to the support type is calculated according to the preferred probability of the information belonging to the support type and the current support satisfaction degree;
  • step C4 it is determined whether the information belonging to the support type is selected and output according to the support probability of the information belonging to the support type, and the information selected for output is used as the supported information.
  • Steps C1 to C4 provide a specific implementation method for selecting the supported information.
  • the preferred probability of the information belonging to the support type needs to be calculated first.
  • the preferred probability refers to the information selected to be supported.
  • the probability is selected, and the preferred probability can be calculated by the current average ranking of the information belonging to the support type and the historical support average ranking corresponding to the support type.
  • the normal distribution probability calculation method can calculate the preferred probability of the information belonging to the support type, that is, the standard normal distribution between the current average ranking of the information belonging to the support type and the historical support average ranking corresponding to the support type can be calculated,
  • the value of the preferred probability is determined by a standard normal distribution. It is to be understood that the above-described examples are merely illustrative of the embodiments of the present invention and are not intended to limit the invention.
  • the selected support number of the information belonging to the support type refers to the number of information that has been supported in the current request situation, and the total number of requests sent by the user refers to the page request that the information delivery device has received.
  • the number of support is a product rule, which can be pre-configured, and is an external parameter of the information delivery device.
  • the information delivery device has the support function.
  • the calculated current support satisfaction degree refers to the degree of satisfaction of the information support in this request. If the current support satisfaction degree is too large, it indicates that it is necessary to reduce the support for the information belonging to the support type, if the current support satisfaction value is If it is too small, it means that it is necessary to increase the support for the information belonging to the type of support.
  • the current support satisfaction degree can be implemented in multiple ways according to the selected support number of the information belonging to the support type, the total number of requests sent by the user, and the preset support ratio.
  • An example is as follows: firstly, according to the selected support number of the information belonging to the support type, the total supported number sent by the user, the current supported percentage is calculated, and then the current supported percentage and the preset support ratio are compared numerically, and the current support is satisfied. degree.
  • step C3 the preferred probability of the information belonging to the support type and the current support satisfaction degree are obtained.
  • the support probability of the information belonging to the support type can be determined, and the support probability refers to the probability parameter that the information belonging to the support type can be supported.
  • the preferred probability of the information belonging to the support type may be divided by the current support satisfaction degree to obtain the support probability of the information belonging to the support type.
  • the above implementation is only an example. In practical applications, other methods may be used to calculate the support probability of the information belonging to the support type, for example, the preferred probability of the information belonging to the support type is divided by the current support satisfaction degree.
  • the support factor can be a proportional correction parameter, and the specific implementation manner is not limited.
  • After obtaining the support probability of the information belonging to the support type it may be determined according to the support probability of the information belonging to the support type whether the information belonging to the support type is selected for output, and the information selected for output is used as the supported information, that is, the support can be calculated. Probability to determine the supported information to ensure the stability of the support flow.
  • the information provided by the embodiment of the present invention is performed after the specified delivery information is selected from the information of the specified type according to the current average ranking of the information of the specified type and the historical average ranking corresponding to the type of the support.
  • the method can also include the following steps:
  • the historical average ranking corresponding to the support type is updated according to the current average ranking of the information belonging to the support type, and the historical average ranking corresponding to the updated support type is saved.
  • the historical average ranking may be a historically supported average ranking.
  • the historical support average ranking corresponding to the support type may be a dynamically updated historical ranking average. After the current request has determined the supported information, the historical support average ranking may be updated in real time to provide updated support for the next request.
  • the historical support average ranking corresponding to the type ensures that the historical support average ranking is a real-time latest average to ensure the balance of supporting flows.
  • the candidate information list is first selected according to the page request sent by the user.
  • the candidate information list includes: a plurality of information, and then each information in the candidate information list is determined according to the page request and the configured prediction model.
  • the ranking order is predicted to generate a ranking of each information in the candidate information list, and then the current average ranking of the information belonging to the support type in the candidate information list is obtained according to the ranking of each information in the candidate information list, and finally according to the type of support
  • the current average ranking of the information and the historical support average ranking corresponding to the support type, and the supported information is selected from the information of the support type.
  • the surface request is to perform primary selection on all the information, and obtain a candidate information list that matches the user, thereby improving the targeting of the user group for information delivery, and predicting the ranking order of all the information in the candidate information list can be supported in the candidate information list.
  • the current average ranking of the type information is finally selected according to the current average ranking of the information belonging to the support type and the historical support average ranking corresponding to the support type, and the supported information is selected from the information belonging to the support type. Since the selected supported information is determined according to the current average ranking and the historical support average ranking corresponding to the support type, the selected supported information is higher in all the information of the candidate information list than the user who sends the page request.
  • the information of the matching degree can match the interests of the users and ensure the effect of the information delivery.
  • the supported information is selected from the information of the supporting type, so as to satisfy the demand for the specific type of information.
  • FIG. 2 is a schematic diagram of an implementation scenario of an information delivery device according to an embodiment of the present invention
  • FIG. 3 is a schematic diagram of the present invention.
  • the embodiment of the invention introduces a preferential support scheme for real-time calculation, and determines in real time that the ranking of the pending advertisements in each request is compared with the average ranking of the selected support advertisements, and the candidate advertisement ranks higher, the higher the support probability is selected, otherwise the election is selected.
  • the probability of support is low.
  • the program not only ensures that the number of support meets the demand, but also guarantees the quality of support and improves the effectiveness of advertising.
  • FIG. 2 an implementation scenario of the advertisement support device provided by the embodiment of the present invention is described.
  • the process of the advertisement support generally includes: a targeted search, an advertisement primary selection (a selection of hundreds of tens of thousands of advertisements) Advertising), accurate prediction of advertisements and rearrangement of several links, followed by detailed explanation.
  • the user can send a page request through the display page, the advertisement supporting device receives the page request of the user, and the advertisement in the order advertisement table records all the advertisements that can be broadcasted, taking tens of thousands of advertisements as an example, after the directional retrieval, the remaining number can be left Thousands of ads, and then advertised, the remaining hundreds of ads can form a list of candidate ads, and then accurately predict the click-through rate of all ads in the list of candidates, and then rearrange, for example, according to ECPM rearrangement, the most Excellent advertising, support strategies are usually implemented in the primary election, because if it is implemented in the following links, the ads that may need to be supported have been filtered by the primary election.
  • the largest (English name: ToP) N advertisements for broadcasting to the user are selected from the candidate advertisement list, and the supported advertisements and the normally competitive advertisements may be included in the ToP N advertisements.
  • ToP N advertisements for broadcasting to the user are selected from the candidate advertisement list, and the supported advertisements and the normally competitive advertisements may be included in the ToP N advertisements.
  • a detailed process of the advertisement support provided by the embodiment of the present invention is given as an example. For instructions, see Figure 3. Next, we will introduce the ad selection process for the primary selection.
  • CTR calculation is the basis of the process operation.
  • the click rate prediction value of each advertisement in the candidate advertisement list of this request is calculated in real time by the model.
  • the specific prediction model can use algorithms such as logistic regression and decision tree.
  • Judgment of support conditions According to the business rules, it is determined whether the request is to enter the support process. There are a variety of support conditions. For example, there must be at least N advertisements of this type in this request before entering the support process. Otherwise, the normal free competition process, the number of N is determined according to the business scenario, if the type of advertisement is too small, follow-up It may be filtered out, resulting in no requests for this request. Another example is when supporting the app, it is judged that the request must be a WiFi environment, and the support is not supported. The conversion rate in the non-WiFi environment is too low to support.
  • This request ranking calculation (the result obtained is represented by rank_this): Calculate the current average ranking of the advertisements of the type of support in this request.
  • the ranking may not be the average of all ads of that type, but the average of top M ads, because only the top ad will be exposed in one request, such as a carousel, determining top N based on the carousel rules, non-rotating class The ad is only top1.
  • Calculated rankings are also ranked by CTR ranking or ECPM based on business needs.
  • the historical support average ranking (the results obtained with rank_avg):
  • rank_avg use the historical data statistics to support the average ranking of the type of advertising.
  • rank_this uses rank_this to update the ranking summary value and the number of support in the support list, and recalculate the historical support average ranking.
  • the embodiment of the invention can calculate the historical support average ranking for each request in real time, thereby ensuring the real-time update of the support type ranking mean.
  • the input conditions can be rank_this, rank_avg, total number of requests, number of supported supports, configured support ratio, and whether the request needs to be selected for support. It has the following three parts:
  • rank_this is less than rank_avg, the current average ranking of this request is before the historical support average ranking, indicating that the matching of the ad type of the request and the user is higher, then the probability of support is large.
  • the method for calculating the probability is not limited.
  • rank_avg as the mean value
  • Cover area a Select the data within 2 times variance to be credible.
  • the preferred probability p_match is calculated as follows:
  • P_match reflects the user's preference for the support type advertisement.
  • the traffic that enters support is the top user of this type of advertisement, that is to say, the user who has a high desire to click on this type of advertisement guarantees the advertisement with higher matching quality in the quality. Therefore, the strategy of preferential support is achieved in the case of guaranteeing the proportion of support.
  • the user ranking is adjusted in real time, and the flow is stabilized.
  • dynamically selecting a traffic with a high matching degree to support a certain type of advertisement, and supporting the amount of the guarantee can significantly improve the click rate effect.
  • an information delivery device 400 may be specifically applied to a specific type of information delivery scenario.
  • the information delivery device 400 may be an advertisement support device.
  • the information described in the embodiment of the present invention may also be a form of resources, content, data, and the like that need to be supported, and is not limited in specific terms.
  • the information delivery device 400 may include: a primary selection module 401, a ranking prediction module 402, a current request calculation module 403, and an information delivery selection module 404, where
  • the primary selection module 401 is configured to select a candidate information list according to a page request sent by the user, where the candidate information list includes: multiple pieces of information;
  • the ranking prediction module 402 is configured to predict a ranking order of each information in the candidate information list according to the page request and the configured prediction model, thereby generating a ranking of each information in the candidate information list;
  • the current request calculation module 403 is configured to obtain, according to the ranking of each information in the candidate information list, a current average ranking of information of a specified type in the candidate information list;
  • the information delivery selection module 404 is configured to determine, according to the current average ranking of the information of the specified type and the historical average ranking corresponding to the support type, whether the information of the specified type is used as the designated delivery information.
  • the specified type of information is information belonging to a support type
  • the designated delivery information is information that is delivered in a supported form
  • the historical average ranking is a historical support average ranking
  • the ranking prediction module 402 is specifically configured to acquire, according to the page request and the preset prediction model, a click rate prediction value of each information in the candidate information list, and according to the The click rate prediction value of each information acquires the ranking of each piece of information in the candidate information list.
  • the ranking prediction module 402 is specifically configured to acquire, according to the click rate prediction value of each information, a revenue prediction value of each information; and all information in the candidate information list.
  • the revenue prediction values are sequentially arranged in order of magnitude, and the ranking of each information in the candidate information list is obtained.
  • the information placing apparatus 400 further includes: a support condition determining module 405, where
  • the support condition determination module 405 is configured to: the ranking prediction module 402 predicts a ranking order of each information in the candidate information list according to the page request and the configured prediction model, thereby generating the candidate information list. After the ranking of each information, it is judged whether the information of the specified type in the candidate information list satisfies the preset condition, and if the information of the specified type satisfies the condition, the current request calculation module 403 is triggered to execute.
  • the specified type of information may be information that requires a type of support.
  • the information delivery selection module 404 includes:
  • the first calculating unit 4041 is configured to calculate, according to the current average ranking of the information belonging to the support type and the historical support average ranking corresponding to the support type, a preferred probability of the information belonging to the support type;
  • the second calculating unit 4042 is configured to calculate a current support satisfaction degree according to the selected support number of the information belonging to the support type, the total number of requests sent by the user, and the preset support ratio;
  • the third calculating unit 4043 is configured to calculate, according to the preferred probability of the information belonging to the support type and the current support satisfaction degree, the support probability of the information belonging to the support type;
  • the information support determining unit 4044 is configured to determine, according to the support probability of the information belonging to the support type, whether the information belonging to the support type is selected for output, and use the information selected for output as the supported information.
  • the information delivery device 400 further includes: a history update module 406, where
  • the history update module 406 is configured to: after determining the specified type of information as the specified delivery information, update the pre-stored historical average ranking corresponding to the specified type of information according to the current average ranking of the specified type of information. And saving the historical average ranking corresponding to the updated specified type of information.
  • the candidate information list is first selected according to the page request sent by the user.
  • the candidate information list includes: a plurality of information, and then each information in the candidate information list is determined according to the page request and the configured prediction model.
  • the ranking order is predicted to generate a ranking of each information in the candidate information list, and then the current average ranking of the information belonging to the support type in the candidate information list is obtained according to the ranking of each information in the candidate information list, and finally according to the type of support
  • the current average ranking of the information and the historical support average ranking corresponding to the support type, and the supported information is selected from the information of the support type.
  • the sort order prediction can obtain the current average rank of the information belonging to the support type in the candidate information list, and finally select from the information of the support type according to the current average rank of the information belonging to the support type and the historical support average rank corresponding to the support type.
  • the selected supported information is determined according to the current average ranking and the historical support average ranking corresponding to the support type, the selected supported information is higher in all the information of the candidate information list than the user who sends the page request.
  • the information of the matching degree can match the interests of the users and ensure the effect of the information delivery.
  • the supported information is selected from the information of the supporting type, so as to satisfy the demand for the specific type of information.
  • FIG. 5 is a schematic structural diagram of a server according to an embodiment of the present invention.
  • the server 500 may have a large difference due to different configurations or performances, and may include one or more central processing units (CPUs) 522 (for example, One or more processors and memory 532, one or more storage media 530 storing application 542 or data 544 (eg, one or one storage device in Shanghai).
  • the memory 532 and the storage medium 530 may be short-term storage or persistent storage.
  • the program stored on storage medium 530 may include one or more modules (not shown), each of which may include a series of instruction operations in the server.
  • central processor 522 can be configured to communicate with storage medium 530, executing a series of instruction operations in storage medium 530 on server 500.
  • Server 500 may also include one or more power supplies 526, one or more wired or wireless network interfaces 550, one or more input and output interfaces 558, and/or one or more operating systems 541, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM and more.
  • the steps performed by the server in the above embodiment may be based on the server structure shown in FIG.
  • the server may be the information delivery device described in the foregoing embodiment, and the server may perform the information delivery method described in the foregoing embodiment. For details, refer to the description of the foregoing embodiment.
  • the device embodiments described above are merely illustrative, wherein the The units described for the separate components may or may not be physically separate, and the components displayed as the units may or may not be physical units, ie may be located in one place, or may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
  • the connection relationship between the modules indicates that there is a communication connection between them, and specifically, one or more communication buses or signal lines can be realized. Those of ordinary skill in the art can understand and implement without any creative effort.
  • the present invention can be implemented by means of software plus necessary general hardware, and of course, dedicated hardware, dedicated CPU, dedicated memory, dedicated memory, Special components and so on.
  • functions performed by computer programs can be easily implemented with the corresponding hardware, and the specific hardware structure used to implement the same function can be various, such as analog circuits, digital circuits, or dedicated circuits. Circuits, etc.
  • software program implementation is a better implementation in more cases.
  • the technical solution of the present invention which is essential or contributes to the prior art, can be embodied in the form of a software product stored in a readable storage medium, such as a floppy disk of a computer.
  • U disk mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., including a number of instructions to make a computer device (may be A personal computer, server, or network device, etc.) performs the methods described in various embodiments of the present invention.
  • a computer device may be A personal computer, server, or network device, etc.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Marketing (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Algebra (AREA)
  • Probability & Statistics with Applications (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Transfer Between Computers (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

一种信息扶持方法和装置。所述方法包括：根据用户发送的页面请求选择候选信息列表，所述候选信息列表包括多个信息（101）；按照预定规则对所述候选信息列表中所述多个信息进行排序，以生成所述候选信息列表中每个信息的排名（102）；根据所述候选信息列表中每个信息的排名获取所述候选信息列表中指定类型的信息的当前平均排名（103）；以及根据所述指定类型的信息的当前平均排名和所述指定类型的信息所对应的历史平均排名，从所述指定类型的信息中选择指定投放信息（104）。

Description

信息投放方法、装置、服务器和存储介质
本申请要求于2016年1月12日提交中国专利局、申请号为201610017753.2、发明名称为“一种信息扶持方法和装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及信息投放技术领域，尤其涉及一种信息投放方法、装置、***和存储介质。
背景技术
在信息投放的某些时期，常有对某类信息进行扶持的需求。比如信息投放***和电商网站合作，约定信息***需保证20％流量用于投放该电商网站的卖家广告，这样对电商网站上挂售的产品可以作为一种稳定流量，有利于产品运维。又比如，在电商广告逐渐发展较为完善时，探索引入开发平台的应用程序(英文全称:application，英文简称:app)类广告，但初期app类广告主较少，没有广告投放经验，如果没有保护性措施，可能流量和效果波动很大，妨碍广告主的拓展。为创建稳定的app类广告生态，需要对app类广告扶持，在引入充足的app类广告，等到逐渐形成生态时，再完全放开与其他类型广告的自由竞争。
现有的信息投放方案缺乏对扶持需求的灵活控制，因此亟需一种新的投放方案。
发明内容
本发明实施例提供了一种信息投放方法和装置，能够在保证信息投放效果的基础上实现对属于扶持类型的信息的扶持。
第一方面，本发明实施例提供一种信息投放方法，包括：
根据用户发送的页面请求选择候选信息列表，所述候选信息列表包括多个信息；
生成所述候选信息列表中每个信息的排名；
根据所述候选信息列表中每个信息的排名获取所述候选信息列表中指定类型的信息的当前平均排名；
基于所述指定类型的信息的当前平均排名和所述指定类型的信息所对应的预存历史平均排名，确定所述指定类型的信息是否作为指定投放信息。
第二方面，本发明实施例还提供一种信息投放装置，包括：
初选模块，用于根据用户发送的页面请求选择候选信息列表，所述候选信息列表包括多个信息；
排名预测模块，用于生成所述候选信息列表中每个信息的排名；
当前请求计算模块，用于根据所述候选信息列表中每个信息的排名获取所述候选信息列表中指定类型的信息的当前平均排名；
信息投放选择模块，用于基于所述指定类型的信息的当前平均排名和所述指定类型的信息所对应的预存历史平均排名，确定所述指定类型的信息是否作为指定投放信息。
从以上技术方案可以看出，本发明实施例具有以下优点：
在本发明实施例中，首先根据用户发送的页面请求选择候选信息列表，候选信息列表包括：多个信息，然后根据页面请求和配置的预测模型对候选信息列表中每个信息的排名顺序进行预测，从而生成候选信息列表中每个信息的排名，接下来根据候选信息列表中每个信息的排名获取候选信息列表中指定类型的信息的当前平均排名，最后根据指定类型的信息的当前平均排名和指定类型对应的历史扶持平均排名，从指定类型的信息中选择出指定投放信息。本发明实施例中用户发送页面请求之后根据该页面请求对所有信息进行初选，得到和该用户匹配的候选信息列表，从而提高信息投放的用户群体针对性，通过对候选信息列表中所有信息的排序顺序预测可以得到在候选信息列表中属于扶持类型的信息的当前平均排名，最后根据属于扶持类型的信息的当前平均排名和该扶持类型对应的历史扶持平均排名，从属于扶持类型的信息中选择出被扶持信息。由于选择出的被扶持信息是根据当前平均排名和该扶持类型对应的历史扶持平均排名确定的，因此选择出的被扶持信息在候选信息列表的所有信息中是与发送页面请求的用户具有较高匹配度的信息，从而可以匹配用户的兴趣爱 好，保证信息投放效果，同时被扶持信息是从属于扶持类型的信息中选择出的，从而可满足对特定类型的信息扶持需求。
附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域的技术人员来讲，还可以根据这些附图获得其他的附图。
图1为本发明实施例提供的一种信息投放方法的流程方框示意图；
图2为本发明实施例中信息投放装置的一种实现场景示意图；
图3为本发明实施例中信息投放方法的一种实现场景示意图；
图4-a为本发明实施例提供的一种信息投放装置的组成结构示意图；
图4-b为本发明实施例提供的另一种信息投放装置的组成结构示意图；
图4-c为本发明实施例提供的一种信息投放选择模块的组成结构示意图；
图4-d为本发明实施例提供的另一种信息投放装置的组成结构示意图；以及
图5为本发明实施例提供的信息投放方法应用于服务器的组成结构示意图。
具体实施方式
为使得本发明的发明目的、特征、优点能够更加的明显和易懂，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，下面所描述的实施例仅仅是本发明一部分实施例，而非全部实施例。基于本发明中的实施例，本领域的技术人员所获得的所有其他实施例，都属于本发明保护的范围。
本发明的说明书和权利要求书及上述附图中的术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排他的包含，以便包含一系列单元的过程、方法、***、产品或设备不必限于那些单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它单元。
针对某类信息的扶持需求，现有的技术方案是配置一定比例的专属流量，该专属流量只允许投放某类受扶持的信息，控制方法通常是根据用户的页面请求按比例随机控制。这种技术方案虽然简单容易实现并且流量稳定，但是存在如下的问题:通过随机产生的流量中有些用户对受扶持的这类信息不感兴趣，但是随机控制可能刚好投放了该类信息，也有部分用户对受扶持的这类信息感兴趣，但在随机控制时可能恰好没有投放到该类信息。因此，这样的方案采用的信息扶持策略虽然能控制流量占比，但信息投放效果不佳。
本发明实施例提供了信息投放方法和装置，能够在保证信息投放效果的基础上实现对属于扶持类型的信息的扶持。
本发明信息投放方法的一个实施例，具体可以应用于对特定类型的信息扶持场景中，具体的，该信息可以为广告，另外本发明实施例中所述的信息也可以是需要扶持的资源、内容、数据等形式，具体不做限定。请参阅图1所示，本发明一个实施例提供的信息投放方法，可以包括步骤101至步骤104。
在步骤101中，根据用户发送的页面请求选择候选信息列表，候选信息列表包括多个信息。
在本发明实施例中，当用户访问页面时，用户可以发送页面请求，本发明实施例提供的信息投放装置可以接收到用户发送的页面请求，从用户发送的页面请求可以获取到该用户的相关信息，该用户的相关信息可以用于确定出该用户对哪些类型的信息感兴趣或者确定出该用户对哪方面的信息有观看意愿，例如确定用户对哪方面的广告有观看意愿。具体的，用户发送的页面请求可以携带用户操作终端的设备信息、访问上下文信息和用户属性信息，则信息投放装置可以从该页面请求中获取到上述信息内容，信息投放装置根据该页面请求可以对所有待定播出的信息进行初选，得到和该用户匹配的候选信息列表，该候选信息列表中的所有信息是信息投放装置进行过初选之后得到的信息，候选信息列表中的所有信息都是可能满足用户兴趣爱好需要的信息类型，根据页面请求进行信息初选可以提高信息投放的用户群体针对性，这与现有技术中采用随机播放而无法保证信息效果的方式是完全不同的实现方式，例如本发明实施例可以根据页面请求进行广告初选可以提高广告投放的用户群体针对性，这与现有技术中采用随机播放而无法保证广告效果的方式是完全不同的实现方式。
在本发明的一些实施例中，接收到用户发送的页面请求之后，在生成候选信息列表时还可以根据该用户的页面请求以及信息主投放信息的投放条件对信息的面向对象进行定向检索，以减少和用户不匹配的信息被播出导致的信息投放效果差的问题。举例说明如下，以信息具体为广告为例，广告主制作的广告为母婴类用品，广告主确定的投放对象为25至40岁的女性，则可以按照广告主的需求来对广告的面向对象进行定向检索，若发送页面请求的用户为男性，则可以通过该定向检索，确定母婴类用品的广告不向男性用户投放，该母婴类用品的广告不需要包括在候选广告列表中。
需要说明的是，在本发明实施例中，候选信息列表是针对发送页面请求的用户来制定的，该候选信息列表中包括有多个信息，这些信息可以属于同一个信息类型，也可以属于不同的信息类型。并且在候选信息列表中包括的所有信息中有的信息可以属于指定类型的信息,例如需要扶持类型的信息。由背景技术可知，例如，app类型广告可以是需要扶持类型的广告，其它特定类型的广告也可以是需要扶持类型的广告。本发明实施例中属于扶持类型的信息是包括在候选信息列表中，但是属于扶持类型的信息是否确定被扶持还可以结合后续实施例中描述的信息投放方法来确定。
在步骤102中，按照预定规则对候选信息列表中的多个信息进行排序，生成候选信息列表中每个信息的排名。
在本发明的一些实施例中，可以根据页面请求和配置的预测模型对候选信息列表中每个信息的排名顺序进行预测，从而生成候选信息列表中每个信息的排名。
在本发明实施例中，通过初选方式确定了候选信息列表，然后根据用户发送的页面请求和配置的预测模型预测候选信息列表中所有信息的排名顺序，以候选信息列表中包括M(M为非零自然数)个信息，那么就需要对第1个信息、第2个信息、…、第M个信息在候选信息列表中的排名顺序分别进行预测，从而生成候选信息列表中每个信息的排名。其中每个信息的排名确定了在候选信息列表中每个信息的排名先后，通过对候选信息列表中所有信息的排名预测，可以预测中候选信息列表中每个信息与发送页面请求的用户之间的匹配程度，在候选信息列表中排名越靠前的信息与发送页面请求的用户越匹配。
需要说明的是，在实际应用中，候选信息列表中每个信息的排名预测可以根据用户的页面请求和配置的预测模型来确定。例如，信息投放装置可以根据用户、上下文场景、信息三方面因素预测各个信息的排名顺序，具体使用到的预测模型可以是逻辑回归算法，也可以是决策树算法等具体的预测算法实现的模型。
在本发明的一些实施例中，步骤102根据页面请求和配置的预测模型对候选信息列表中每个信息的排名顺序进行预测，从而生成候选信息列表中每个信息的排名，具体可以包括如下步骤：
A1、根据页面请求和预置的预测模型获取候选信息列表中每个信息的点击率预测值，并根据每个信息的点击率预测值获取候选信息列表中每个信息的排名。
其中，信息投放装置可以根据页面请求和预置的预测模型对候选信息列表中所有信息的点击率(英文全称：Click Through Rate，英文简称：CTR)进行预测，CTR可以反映用户对信息(例如广告)的喜好，根据用户的页面请求和预测模型可以预测候选信息列表中每个信息的CTR，从而生成每个信息的CTR预测值，根据CTR预测值的大小可以作为候选信息列表中每个信息的排名依据，从而得到候选信息列表中每个信息的排名。需要说明的是，计算每个信息的点击率预测值可以作为信息排名的一种依据，但是本发明实施例中并不局限于使用点击率预测值这种方式，在实际应用中，还可以采用其它方式作为信息排名的依据。例如，步骤A1中的根据每个信息的点击率预测值获取候选信息列表中每个信息的排名，可以包括步骤A11至步骤A12。
在步骤A11中，根据每个信息的点击率预测值获取每个信息的收益预测值；
在步骤A12中，对候选信息列表中所有信息的收益预测值按照取值大小顺序依次排列，得到候选信息列表中每个信息的排名。
其中，步骤A11中，得到信息的点击率预测值之后，可以根据信息的每次点击收益(英文全称：Cost Per Click，英文简称：CPC)计算出期望的千次曝光收益(英文全称：Expect Cost Per Mile，英文简称：ECPM)，即ECPM＝CTR*CPC。计算出每个信息的收益预测值之后，可以对候选信息列表 中所有信息的收益预测值按照取值大小顺序依次排列，得到候选信息列表中每个信息的排名。
在本发明的一些实施例中，所述指定类型的信息可以是属于扶持类型的信息，也即预先确定需要受扶持类型的信息。在本发明的这些实施例中，步骤102根据页面请求和配置的预测模型对候选信息列表中每个信息的排名顺序进行预测，从而生成候选信息列表中每个信息的排名之后，本发明实施例提供的信息投放方法还可以包括：
判断候选信息列表中属于扶持类型的信息是否满足预置条件，若属于扶持类型的信息满足所述条件，触发如下步骤103执行：根据候选信息列表中每个信息的排名获取候选信息列表中指定类型的信息的当前平均排名。
也就是说，在生成候选信息列表中每个信息的排名之后，可以对候选信息列表中属于扶持类型的信息进行分析，结合预置的扶持条件确定是否需要对属于扶持类型的信息，若需要对属于扶持类型的信息进行扶持，则再触发执行后续步骤103，并在步骤103执行之后可以执行步骤104。若不需要对属于扶持类型的信息进行扶持，则无需执行后续的信息扶持流程，可以结束流程，按照候选信息列表中所有信息进行公平竞争，最后确定出向用户播出的信息。
应该理解，可以根据需要设定不同的扶持条件，并且本发明实施例在此方面不受限制。在实际应用中，扶持条件可以根据业务规则来确定，举例说明如下，若属于扶持类型的信息在候选信息列表中的排名位于整个列表的最后20％，则说明属于扶持类型的信息与用户的匹配程度较低，可以不对属于扶持类型的信息进行扶持，以避免用户对信息满意度的下降。另外扶持条件可以设置为用户发送请求处于无线局域网(英文全称：Wireless Fidelity，英文简称：WiFi)环境，则根据用户发送的页面请求确定用户不是使用WiFi时可以不进行信息扶持，以避免给用户带来的流量损耗。
在步骤103中，根据候选信息列表中每个信息的排名获取候选信息列表中属于指定类型的信息的当前平均排名。
在本发明实施例中，通过前述步骤102中对候选信息列表中所有信息的排名预测，可以生成候选信息列表中每个信息的排名，接下来根据候选信息列表中每个信息的排名可以计算出在候选信息列表中指定类型的信息的当前平均 排名。例如，以指定类型为需要扶持类型为例,以需要扶持类型为app类型为例，以信息具体为广告为例，属于扶持类型的广告可以是app类型广告，计算app类型广告在候选广告列表中的当前平均排名，若app类型广告1在候选广告列表中排名为t1，若app类型广告2在候选广告列表中排名为t2，则app类型广告的当前平均排名为(t1+t2)/2。可以理解的是，若属于扶持类型的广告为1个，则该广告在候选广告列表中的排名就是属于扶持类型的广告的当前平均排名。可以理解，虽然上面示例性地描述了计算排名的方式，但是本发明在此方面不受限制，可以根据需要选择现有或者后续开发的算法。
在步骤104中，根据指定类型的信息的当前平均排名和指定类型的信息所对应的历史平均排名，从指定类型的信息中选择出指定投放信息。
根据本发明的一些实施例，所述指定投放信息是以受扶持形式投放的信息，所述历史平均排名为历史扶持平均排名。
在本发明实施例中，获取到指定类型的信息的当前平均排名之后，通过该当前平均排名可以确定指定类型的信息与用户的匹配程度，结合该扶持类型对应的历史平均排名，可以判断指定类型的信息的当前平均排名和指定类型对应的历史平均排名，从而得到当前平均排名与历史平均排名的关系，依据这两者的关系可以从指定类型的信息中选择出指定投放信息。例如，对于指定类型的信息是确定需要扶持的信息，历史平均排名是针对需要扶持的信息类型通过历史数据累计得出的一个排名均值，该历史扶持平均排名可以用于衡量当前的指定类型的信息是否应该被扶持的参照量。在实际应用中，对于当前平均排名和历史扶持平均排名的分析可以有多种方式，从而依据两者的关系确定属于扶持类型的某个信息或者某些信息是否作为被扶持信息。举例说明如下，若当前平均排名小于历史扶持平均排名，则说明属于扶持类型的信息与用户的匹配程度高，可以将属于扶持类型的信息作为被扶持信息。不限定的是，根据属于扶持类型的信息的当前平均排名和扶持类型对应的历史扶持平均排名在选择被扶持信息时可以不局限于上述方式，还可以结合其他的参量因子用于确定被扶持信息。
在本发明的一些实施例中，步骤104根据指定类型的信息的当前平均排名和扶持类型对应的历史平均排名，从指定类型的信息中选择出指定投放信息， 具体可以包括步骤C1至步骤C4。
在步骤C1中，根据属于扶持类型的信息的当前平均排名和扶持类型对应的历史扶持平均排名计算出属于扶持类型的信息的优选概率；
在步骤C2中，根据属于扶持类型的信息的已选择扶持数、用户发送的总请求数、预置的扶持比例计算出当前扶持满足度；
在步骤C3中，根据属于扶持类型的信息的优选概率和当前扶持满足度计算属于扶持类型的信息的扶持概率；
在步骤C4中，根据属于扶持类型的信息的扶持概率确定属于扶持类型的信息是否被选中输出，将被选中输出的信息作为被扶持信息。
其中，步骤C1至步骤C4给出一种具体的选择被扶持信息的实现方式，这种实现方式中需要先计算出属于扶持类型的信息的优选概率，优选概率指的是信息被选中进行扶持的选中概率，通过属于扶持类型的信息的当前平均排名和扶持类型对应的历史扶持平均排名可以计算出该优选概率。例如，通过正态分布计算概率法可以计算出属于扶持类型的信息的优选概率，即可以计算属于扶持类型的信息的当前平均排名和扶持类型对应的历史扶持平均排名之间的标准正态分布，通过标准正态分布确定优选概率的取值。可以理解的是，上述举例只是本发明实施例中的一种可实现方式，不能作为对本发明的限定。
在步骤C2中，属于扶持类型的信息的已选择扶持数指的是在当前的请求情况下已经被扶持的信息个数，用户发送的总请求数是指信息投放装置已经接收到的页面请求的个数，扶持比例属于产品规则，是可以预先配置理决定，是信息投放装置的外部参数，信息投放装置具备该扶持功能。计算出的当前扶持满足度是指在本次请求中信息扶持的满足程度，若当前扶持满足度取值过大，则说明需要减少对属于扶持类型的信息的扶持，若当前扶持满足度取值过小，则说明需要加大对属于扶持类型的信息的扶持。在实际应用中，根据属于扶持类型的信息的已选择扶持数、用户发送的总请求数、预置的扶持比例计算出当前扶持满足度可以有多种实现。举例说明如下：先根据属于扶持类型的信息的已选择扶持数、用户发送的总请求数计算出当前已扶持百分比，再根据当前已扶持百分比和预置的扶持比例进行数值比较，得到当前扶持满足度。
在步骤C3中，得到属于扶持类型的信息的优选概率和当前扶持满足度之 后，根据属于扶持类型的信息的优选概率和当前扶持满足度可以确定属于扶持类型的信息的扶持概率，扶持概率指的是属于扶持类型的信息可以进行扶持的概率参量。举例说明，可以将属于扶持类型的信息的优选概率除以当前扶持满足度得到属于扶持类型的信息的扶持概率。需要说明的是，上述实现情况只是举例，在实际应用中，还可以采用其它的方式来计算属于扶持类型的信息的扶持概率，例如，将属于扶持类型的信息的优选概率除以当前扶持满足度，再对相除之后得到的结果按照扶持因子进行修正从而得到属于扶持类型的信息的扶持概率。扶持因子可以是一个比例校正参量，具体实现方式不做限定。在得到属于扶持类型的信息的扶持概率之后，可以根据属于扶持类型的信息的扶持概率确定属于扶持类型的信息是否被选中输出，将被选中输出的信息作为被扶持信息，即可以计算出的扶持概率来确定被扶持信息，从而保证扶持流量的稳定性。
在本发明的一些实施例中，步骤104根据指定类型的信息的当前平均排名和扶持类型对应的历史平均排名，从指定类型的信息中选择出指定投放信息之后，本发明实施例提供的信息投放方法还可以包括如下步骤：
根据属于扶持类型的信息的当前平均排名对该扶持类型对应的历史平均排名进行更新，并保存更新后的扶持类型对应的历史平均排名。
特别地，所述历史平均排名可以是历史扶持平均排名。扶持类型对应的历史扶持平均排名可以是一个动态更新的历史排名均值，在当前的请求已经确定被扶持信息之后，可以对该历史扶持平均排名进行实时更新，从而为下一个请求提供更新后的扶持类型对应的历史扶持平均排名，保证历史扶持平均排名是一个实时的最新均值，以保证扶持流量的均衡。
通过以上实施例对本发明实施例的描述可知，首先根据用户发送的页面请求选择候选信息列表，候选信息列表包括：多个信息，然后根据页面请求和配置的预测模型对候选信息列表中每个信息的排名顺序进行预测，从而生成候选信息列表中每个信息的排名，接下来根据候选信息列表中每个信息的排名获取候选信息列表中属于扶持类型的信息的当前平均排名，最后根据属于扶持类型的信息的当前平均排名和扶持类型对应的历史扶持平均排名，从属于扶持类型的信息中选择出被扶持信息。本发明实施例中用户发送页面请求之后根据该页 面请求对所有信息进行初选，得到和该用户匹配的候选信息列表，从而提高信息投放的用户群体针对性，通过对候选信息列表中所有信息的排序顺序预测可以得到在候选信息列表中属于扶持类型的信息的当前平均排名，最后根据属于扶持类型的信息的当前平均排名和该扶持类型对应的历史扶持平均排名，从属于扶持类型的信息中选择出被扶持信息。由于选择出的被扶持信息是根据当前平均排名和该扶持类型对应的历史扶持平均排名确定的，因此选择出的被扶持信息在候选信息列表的所有信息中是与发送页面请求的用户具有较高匹配度的信息，从而可以匹配用户的兴趣爱好，保证信息投放效果，同时被扶持信息是从属于扶持类型的信息中选择出的，从而可满足对特定类型的信息扶持需求。
为便于更好的理解和实施本发明实施例的上述方案，下面举例相应的应用场景来进行具体说明。接下来以指定的需要扶持的信息具体为广告为例进行说明，请参阅图2和图3所示，图2为本发明实施例中信息投放装置的一种实现场景示意图，图3为本发明实施例中信息投放方法的一种实现场景示意图。本发明实施例介绍一种实时计算的择优扶持方案，实时判定每次请求中待定广告的排名与已选扶持广告平均排名比较，候选广告排名靠前，则选入扶持概率越高，否则选入扶持概率低。该方案不仅保证扶持数量符合需求，而且保证扶持质量优质，提升广告投放效果。如图2所示，对本发明实施例提供的广告扶持装置的一种实现场景进行说明，广告扶持的流程中通常包含：定向检索、广告初选(上万个广告中选择较优的几百个广告)、广告精准预测和重排几个环节，接下来分别进行详细说明。用户可以通过展示页面下发页面请求，广告扶持装置接收到该用户的页面请求，订单广告表中记录了所有可以播出的广告，以几万条广告为例，经过定向检索之后，可以剩余几千条广告，然后进行广告初选，可以剩余几百条广告组成候选广告列表，然后对候选广告列表中所有广告的点击率进行精准预测，最后再经过重排，例如根据ECPM重排，投放最优广告，扶持策略通常会在初选就实施，因为如果在后面的环节实施的话，可能需要扶持的广告已经被初选过滤。从候选广告列表中选择出用于向用户播出的最大(英文名称：ToP)N条广告，在ToP N条广告中可以包括被扶持的广告以及正常竞争的广告。接下来对本发明实施例提供的广告扶持的详细流程进行举例 说明，请参阅如图3所示。接下来介绍初选中的广告择优流程。
1、CTR计算：CTR计算是流程运行的基础，通过模型实时计算本次请求的候选广告列表中每个广告的点击率预测值，具体预测模型可使用逻辑回归、决策树等算法。
2、扶持条件判断：根据业务规则判定该次请求是否要进入扶持流程。扶持条件可以有多种，比如，本次请求中至少要有N个该类型广告，才进入扶持流程，否则正常自由竞争流程，N的个数根据业务场景决定，如果该类型广告太少，后续可能会被过滤完，导致本次请求无广告可展示。又如对app广告扶持时，判断本次请求必须是WiFi环境，才进行扶持，非WiFi环境下的转化率太低，不能扶持。
3、本次请求排名计算(得到的结果用rank_this表示)：计算本次请求中需扶持类型的广告的当前平均排名。该排名可以不是所有该类型所有广告的均值，而是top M个广告的均值，因为一次请求只有top广告会被曝光，例如轮播类广告，根据轮播页面规则确定top N，非轮播类广告只有top1。计算排名也根据业务需求按CTR排名或ECPM排名。
4、历史扶持平均排名(得到的结果用rank_avg表示)：在初始化运行时，使用历史数据统计需扶持类型广告的平均排名。当新来一条符合扶持的流量时，使用rank_this更新扶持列表的排名汇总值和扶持个数，重新计算出历史扶持平均排名。本发明实施例可以对每条请求实时计算历史扶持平均排名，从而保证了扶持类型排名均值的实时更新。
5、扶持计算：其中输入条件可以是rank_this、rank_avg、总请求数、已选择扶持数、配置的扶持比例，输出本次请求是否需要被选中扶持。有如下3部分组成：
1)通过rank_this和rank_avg，计算优选概率。
如果rank_this小于rank_avg，本次请求的当前平均排名在历史扶持平均排名之前，说明本次请求的属于扶持类型的广告与用户的匹配度更高，则选择扶持概率大。具体计算概率的方法不限，接下来举例说明，一种正态分布计算概率法。以rank_avg为均值，通过扶持列表实时更新方差σ，转化成标准正态分布u＝(rank_avg–rank_this)/σ，通过正态分布概率映射表，由u查询获取 覆盖面积a。选择2倍方差内数据可信，优选概率p_match计算方法如下：
p_match＝a     -2<z<2时；
p_match＝1     z>＝2；
p_match＝0     z<＝-2。
p_match反映了本次请求用户对扶持类型广告的喜好度。
2)通过当前扶持数、总请求数、配置的扶持比例，计算当前扶持满足度。
p_now＝当前扶持数/总请求数；
p_need＝扶持比例；
p_satisfy＝p_now/p_need。
如果p_satisfy>1，说明扶持过多，需要减少扶持，反之说明扶持力度太小，需要加大扶持。
3)计算扶持概率p_select＝p_match/p_satisfy。当前请求匹配度越高，当前满足度越低，则被选中的概率越大。可以设置p_select大于1，则被选中，小于1，则不选。也可以再设计个p_select的概率函数，通过概率计算是否被选中。满足上述扶持概率的流量，最后流入专属扶持流量。整个流程是个实时更新的闭环，每条数据都会修正当前系数，影响下一条请求被选中的概率，扶持占比实时调整，围绕着配置扶持比例波动，在数量上保证了扶持流量的稳定性。进入扶持的流量是该类广告排名靠前的用户，也就说是对该类广告点击意愿高的用户，在质量上保证了投放匹配度更高的广告。从而达到了在保证扶持比例的情况下，择优扶持的策略。
本发明实施例中，实时计算调整用户排名，扶持流量稳定。基于用户对广告的点击率预测基础上，动态选择匹配度高的流量扶持某类广告，扶持保量的基础上，可以明显提升点击率效果。
需要说明的是，对于前述的各方法实施例，为了简单描述，故将其都表述为一系列的动作组合，但是本领域技术人员应该知悉，本发明并不受所描述的动作顺序的限制，因为依据本发明，某些步骤可以采用其他顺序或者同时进行。其次，本领域技术人员也应该知悉，说明书中所描述的实施例均属于优选实施例，所涉及的动作和模块并不一定是本发明所必须的。
为便于更好的实施本发明实施例的上述方案，下面还提供用于实施上 述方案的相关装置。
请参阅图4-a所示，本发明实施例提供的一种信息投放装置400，具体可以应用于对特定类型的信息投放场景中，具体的，信息投放装置400可以为广告扶持装置，另外本发明实施例中所述的信息也可以是需要扶持的资源、内容、数据等形式，具体不做限定。信息投放装置400，可以包括：初选模块401、排名预测模块402、当前请求计算模块403、信息投放选择模块404，其中，
初选模块401，用于根据用户发送的页面请求选择候选信息列表，所述候选信息列表包括：多个信息；
排名预测模块402，用于根据所述页面请求和配置的预测模型对所述候选信息列表中每个信息的排名顺序进行预测，从而生成所述候选信息列表中每个信息的排名；
当前请求计算模块403，用于根据所述候选信息列表中每个信息的排名获取所述候选信息列表中指定类型的信息的当前平均排名；
信息投放选择模块404，用于根据所述指定类型的信息的当前平均排名和所述扶持类型对应的历史平均排名，确定所述指定类型的信息是否作为指定投放信息。
根据本发明的一些实施例，所述指定类型的信息是属于扶持类型的信息，所述指定投放信息是以受扶持形式投放的信息，所述历史平均排名为历史扶持平均排名。
在本发明的一些实施例中，所述排名预测模块402，具体用于根据所述页面请求和预置的预测模型获取所述候选信息列表中每个信息的点击率预测值，并根据所述每个信息的点击率预测值获取所述候选信息列表中每个信息的排名。
在本发明的一些实施例中，所述排名预测模块402，具体用于根据所述每个信息的点击率预测值获取所述每个信息的收益预测值；对所述候选信息列表中所有信息的收益预测值按照取值大小顺序依次排列，得到所述候选信息列表中每个信息的排名。
在本发明的一些实施例中，请参阅如图4-b所示，所述信息投放装置400还包括：扶持条件判断模块405，其中，
所述扶持条件判断模块405，用于所述排名预测模块402根据所述页面请求和配置的预测模型对所述候选信息列表中每个信息的排名顺序进行预测，从而生成所述候选信息列表中每个信息的排名之后，判断所述候选信息列表中指定类型的信息是否满足预置的条件，若所述指定类型的信息满足所述条件，触发当前请求计算模块403执行。
在本发明的一些实施例中，所述指定类型的信息可以是需要扶持类型的信息。在本发明的这些实施例中，请参阅如图4-c所示，所述信息投放选择模块404，包括：
第一计算单元4041，用于根据所述属于扶持类型的信息的当前平均排名和所述扶持类型对应的历史扶持平均排名计算出所述属于扶持类型的信息的优选概率；
第二计算单元4042，用于根据所述属于扶持类型的信息的已选择扶持数、用户发送的总请求数、预置的扶持比例计算出当前扶持满足度；
第三计算单元4043，用于根据所述属于扶持类型的信息的优选概率和所述当前扶持满足度计算所述属于扶持类型的信息的扶持概率；
信息扶持确定单元4044，用于根据所述属于扶持类型的信息的扶持概率确定所述属于扶持类型的信息是否被选中输出，将被选中输出的信息作为所述被扶持信息。
在本发明的一些实施例中，请参阅如图4-d所示，所述信息投放装置400还包括：历史更新模块406，其中，
所述历史更新模块406，用于在确定所述指定类型的信息作为指定投放信息之后，根据所述指定类型的信息的当前平均排名对所述指定类型的信息所对应的预存历史平均排名进行更新，并保存更新后的所述指定类型的信息所对应的历史平均排名。
通过以上实施例对本发明实施例的描述可知，首先根据用户发送的页面请求选择候选信息列表，候选信息列表包括：多个信息，然后根据页面请求和配置的预测模型对候选信息列表中每个信息的排名顺序进行预测，从而生成候选信息列表中每个信息的排名，接下来根据候选信息列表中每个信息的排名获取候选信息列表中属于扶持类型的信息的当前平均排名，最后根据属于扶持类型 的信息的当前平均排名和扶持类型对应的历史扶持平均排名，从属于扶持类型的信息中选择出被扶持信息。本发明实施例中用户发送页面请求之后根据该页面请求对所有信息进行初选，得到和该用户匹配的候选信息列表，从而提高信息投放的用户群体针对性，通过对候选信息列表中所有信息的排序顺序预测可以得到在候选信息列表中属于扶持类型的信息的当前平均排名，最后根据属于扶持类型的信息的当前平均排名和该扶持类型对应的历史扶持平均排名，从属于扶持类型的信息中选择出被扶持信息。由于选择出的被扶持信息是根据当前平均排名和该扶持类型对应的历史扶持平均排名确定的，因此选择出的被扶持信息在候选信息列表的所有信息中是与发送页面请求的用户具有较高匹配度的信息，从而可以匹配用户的兴趣爱好，保证信息投放效果，同时被扶持信息是从属于扶持类型的信息中选择出的，从而可满足对特定类型的信息扶持需求。
图5是本发明实施例提供的一种服务器结构示意图，该服务器500可因配置或性能不同而产生比较大的差异，可以包括一个或一个以***处理器(central processing units，CPU)522(例如，一个或一个以上处理器)和存储器532，一个或一个以上存储应用程序542或数据544的存储介质530(例如一个或一个以上海量存储设备)。其中，存储器532和存储介质530可以是短暂存储或持久存储。存储在存储介质530的程序可以包括一个或一个以上模块(图示没标出)，每个模块可以包括对服务器中的一系列指令操作。更进一步地，中央处理器522可以设置为与存储介质530通信，在服务器500上执行存储介质530中的一系列指令操作。
服务器500还可以包括一个或一个以上电源526，一个或一个以上有线或无线网络接口550，一个或一个以上输入输出接口558，和/或，一个或一个以上操作***541，例如Windows ServerTM，Mac OS XTM，UnixTM,LinuxTM，FreeBSDTM等等。
上述实施例中由服务器所执行的步骤可以基于该图5所示的服务器结构。该服务器可以是前述实施例中所述的信息投放装置，该服务器可以执行前述实施例中描述的信息投放方法，具体详见前述实施例的描述说明。
另外需说明的是，以上所描述的装置实施例仅仅是示意性的，其中所述作 为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。另外，本发明提供的装置实施例附图中，模块之间的连接关系表示它们之间具有通信连接，具体可以实现为一条或多条通信总线或信号线。本领域普通技术人员在不付出创造性劳动的情况下，即可以理解并实施。
通过以上的实施方式的描述，所属领域的技术人员可以清楚地了解到本发明可借助软件加必需的通用硬件的方式来实现，当然也可以通过专用硬件包括专用集成电路、专用CPU、专用存储器、专用元器件等来实现。一般情况下，凡由计算机程序完成的功能都可以很容易地用相应的硬件来实现，而且，用来实现同一功能的具体硬件结构也可以是多种多样的，例如模拟电路、数字电路或专用电路等。但是，对本发明而言更多情况下软件程序实现是更佳的实施方式。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在可读取的存储介质中，如计算机的软盘，U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本发明各个实施例所述的方法。
综上所述，以上实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照上述实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对上述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。

Claims (20)

  1. 一种信息投放方法，包括:
    根据用户发送的页面请求选择候选信息列表，所述候选信息列表包括多个信息；
    按照预定规则对所述候选信息列表中所述多个信息进行排序，以生成所述候选信息列表中每个信息的排名；
    根据所述候选信息列表中每个信息的排名获取所述候选信息列表中指定类型的信息的当前平均排名；以及
    根据所述指定类型的信息的当前平均排名和所述指定类型的信息所对应的历史平均排名，从所述指定类型的信息中选择指定投放信息。
  2. 根据权利要求1所述的方法，其中，所述指定类型的信息是属于扶持类型的信息，所述指定投放信息是以受扶持形式投放的信息，所述历史平均排名为历史扶持平均排名。
  3. 根据权利要求1所述的方法，其中，按照预定规则对所述候选信息列表中所述多个信息进行排序包括根据所述页面请求和配置的预测模型对所述候选信息列表中每个信息的排名顺序进行预测，以生成所述候选信息列表中每个信息的排名。
  4. 根据权利要求3所述的方法，其中，所述根据所述页面请求和配置的预测模型对所述候选信息列表中每个信息的排名顺序进行预测，生成所述候选信息列表中每个信息的排名，包括：
    根据所述页面请求和预置的预测模型获取所述候选信息列表中每个信息的点击率预测值，并根据所述每个信息的点击率预测值获取所述候选信息列表中每个信息的排名。
  5. 根据权利要求4所述的方法，其中，所述根据所述每个信息的点击率预测值获取所述候选信息列表中每个信息的排名，包括：
    根据所述每个信息的点击率预测值获取所述每个信息的收益预测值；以及
    对所述候选信息列表中所有信息的收益预测值按照取值大小顺序依次排列，得到所述候选信息列表中每个信息的排名。
  6. 根据权利要求1所述的方法，其中，生成所述候选信息列表中每个信息的排名之后，所述方法还包括：
    判断所述候选信息列表中指定类型的信息是否满足预置条件，若所述指定类型的信息满足所述预置条件，触发如下步骤执行：根据所述候选信息列表中每个信息的排名获取所述候选信息列表中指定类型的信息的当前平均排名。
  7. 根据权利要求2所述的方法，其中，所述根据所述指定类型的信息的当前平均排名和所述指定类型的信息所对应的历史平均排名，从所述指定类型的信息中选择指定投放信息，包括：
    根据所述属于扶持类型的信息的当前平均排名和所述扶持类型对应的历史扶持平均排名计算出所述属于扶持类型的信息的优选概率；
    根据所述属于扶持类型的信息的已选择扶持数、用户发送的总请求数、预置的扶持比例计算出当前扶持满足度；
    根据所述属于扶持类型的信息的优选概率和所述当前扶持满足度计算所述属于扶持类型的信息的扶持概率；以及
    根据所述属于扶持类型的信息的扶持概率确定所述属于扶持类型的信息是否被选中为所述指定投放信息。
  8. 根据权利要求2所述的方法，其中，在从所述指定类型的信息中选择指定投放信息之后，所述方法还包括：
    根据所述指定类型的信息的当前平均排名对所述指定类型的信息所对应的历史平均排名进行更新，并保存更新后的所述指定类型的信息所对应的历史平均排名。
  9. 根据权利要求1至8中任一项所述的方法，其中，所述信息为广告。
  10. 一种信息投放装置，包括:
    初选模块，用于根据用户发送的页面请求选择候选信息列表，所述候选信息列表包括多个信息；
    排名预测模块，用于按照预定规则对所述候选信息列表中所述多个信息进行排序，以生成所述候选信息列表中每个信息的排名；
    当前请求计算模块，用于根据所述候选信息列表中每个信息的排名获取所述候选信息列表中指定类型的信息的当前平均排名；以及
    信息投放选择模块，用于根据所述指定类型的信息的当前平均排名和所述指定类型的信息所对应的历史平均排名，从所述指定类型的信息中选择指定投放信息。
  11. 根据权利要求10所述的装置，其中，所述指定类型的信息是属于扶持类型的信息，所述指定投放信息是以受扶持形式投放的信息，所述历史平均排名为历史扶持平均排名。
  12. 根据权利要求10所述的装置，其中，所述排名预测模块根据所述页面请求和配置的预测模型对所述候选信息列表中每个信息的排名顺序进行预测以生成所述候选信息列表中每个信息的排名。
  13. 根据权利要求12所述的装置，其中，所述排名预测模块根据所述页面请求和预置的预测模型获取所述候选信息列表中每个信息的点击率预测值，并根据所述每个信息的点击率预测值获取所述候选信息列表中每个信息的排名。
  14. 根据权利要求13所述的装置，其中，所述排名预测模块根据所述每个信息的点击率预测值获取所述每个信息的收益预测值；对所述候选信息列表中所有信息的收益预测值按照取值大小顺序依次排列，得到所述候选信息列表中每个信息的排名。
  15. 根据权利要求10所述的装置，其中，所述信息投放装置还包括：条件判断模块，其中，
    所述条件判断模块，用于在生成所述候选信息列表中每个信息的排名之后判断所述候选信息列表中指定类型的信息是否满足预置的条件，若所述指定类型的信息满足所述条件，触发当前请求计算模块执行。
  16. 根据权利要求11所述的装置，其中，所述信息投放选择模块，包括：
    第一计算单元，用于根据所述属于扶持类型的信息的当前平均排名和所述扶持类型对应的历史扶持平均排名计算出所述属于扶持类型的信息的优选概率；
    第二计算单元，用于根据所述属于扶持类型的信息的已选择扶持数、用户发送的总请求数、预置的扶持比例计算出当前扶持满足度；
    第三计算单元，用于根据所述属于扶持类型的信息的优选概率和所述当前 扶持满足度计算所述属于扶持类型的信息的扶持概率；以及
    信息扶持确定单元，用于根据所述属于扶持类型的信息的扶持概率确定所述属于扶持类型的信息是否被选中为所述被扶持信息。
  17. 根据权利要求11所述的装置，其中，所述信息投放装置还包括：历史更新模块，用于在从所述指定类型的信息中选择指定投放信息之后，根据所述指定类型的信息的当前平均排名对所述指定类型的信息所对应的历史平均排名进行更新，并保存更新后的所述指定类型的信息所对应的历史平均排名。
  18. 根据权利要求10至17中任一项所述的装置，其中，所述信息为广告。
  19. 一种服务器，包括处理器和存储有指令的存储器，所述处理器执行所述指令时，被配置为：
    根据用户发送的页面请求选择候选信息列表，所述候选信息列表包括多个信息；
    按照预定规则对所述候选信息列表中所述多个信息进行排序，以生成所述候选信息列表中每个信息的排名；
    根据所述候选信息列表中每个信息的排名获取所述候选信息列表中指定类型的信息的当前平均排名；
    根据所述指定类型的信息的当前平均排名和所述指定类型的信息所对应的历史平均排名，从所述指定类型的信息中选择指定投放信息。
  20. 一种计算机可读存储介质，存储有机器可执行指令，所述指令被配置为使能机器执行以下操作：
    根据用户发送的页面请求选择候选信息列表，所述候选信息列表包括多个信息；
    按照预定规则对所述候选信息列表中所述多个信息进行排序，以生成所述候选信息列表中每个信息的排名；
    根据所述候选信息列表中每个信息的排名获取所述候选信息列表中指定类型的信息的当前平均排名；
    根据所述指定类型的信息的当前平均排名和所述指定类型的信息所对应的历史平均排名，从所述指定类型的信息中选择指定投放信息。
PCT/CN2016/113842 2016-01-12 2016-12-30 信息投放方法、装置、服务器和存储介质 WO2017121258A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
JP2017559381A JP6522160B2 (ja) 2016-01-12 2016-12-30 情報配信方法、ならびに装置、サーバ、および記憶媒体
EP16884809.1A EP3404607A4 (en) 2016-01-12 2016-12-30 METHOD AND APPARATUS FOR DISTRIBUTING INFORMATION, SERVER AND INFORMATION CARRIER
KR1020177032826A KR101999469B1 (ko) 2016-01-12 2016-12-30 정보 전달 방법 및 장치, 서버 및 저장 매체
US15/804,559 US11144950B2 (en) 2016-01-12 2017-11-06 Information delivery method and apparatus, server, and storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610017753.2 2016-01-12
CN201610017753.2A CN105678586B (zh) 2016-01-12 2016-01-12 一种信息扶持方法和装置

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US15/804,559 Continuation US11144950B2 (en) 2016-01-12 2017-11-06 Information delivery method and apparatus, server, and storage medium

Publications (1)

Publication Number Publication Date
WO2017121258A1 true WO2017121258A1 (zh) 2017-07-20

Family

ID=56300070

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/113842 WO2017121258A1 (zh) 2016-01-12 2016-12-30 信息投放方法、装置、服务器和存储介质

Country Status (6)

Country Link
US (1) US11144950B2 (zh)
EP (1) EP3404607A4 (zh)
JP (1) JP6522160B2 (zh)
KR (1) KR101999469B1 (zh)
CN (1) CN105678586B (zh)
WO (1) WO2017121258A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111666488A (zh) * 2020-04-21 2020-09-15 上海晶赞融宣科技有限公司 信息发布方法及装置、存储介质、终端
JP2021502651A (ja) * 2017-12-29 2021-01-28 オッポ広東移動通信有限公司Ｇｕａｎｇｄｏｎｇ Ｏｐｐｏ Ｍｏｂｉｌｅ Ｔｅｌｅｃｏｍｍｕｎｉｃａｔｉｏｎｓ Ｃｏｒｐ．， Ｌｔｄ． 融合データ処理方法及び情報推薦システム
TWI756967B (zh) * 2020-12-04 2022-03-01 萬里雲互聯網路有限公司 內容點擊數預測裝置及其方法

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105678586B (zh) 2016-01-12 2020-09-29 腾讯科技（深圳）有限公司 一种信息扶持方法和装置
CN109325825B (zh) * 2018-09-04 2020-06-30 口口相传(北京)网络技术有限公司 流量数据的调控方法及装置
JP2020087270A (ja) * 2018-11-30 2020-06-04 大日本印刷株式会社 広告配信装置、制御方法及びプログラム
CN111831838B (zh) * 2019-04-17 2024-03-29 腾讯科技（深圳）有限公司 一种多媒体信息推荐方法及装置、服务器存储介质
CN111242689B (zh) * 2020-01-14 2023-11-10 喜大（上海）网络科技有限公司 广告投放方法、装置、服务器及存储介质
CN112199523A (zh) * 2020-09-29 2021-01-08 北京字节跳动网络技术有限公司 一种作品分发方法、装置、计算机设备及存储介质
CN112488747A (zh) * 2020-11-03 2021-03-12 广州易起行信息技术有限公司 基于大数据技术的广告投放方法、电子设备、存储介质
CN114710686A (zh) * 2022-03-11 2022-07-05 武汉斗鱼鱼乐网络科技有限公司 一种直播间的曝光方法、装置、介质及设备

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080033810A1 (en) * 2006-08-02 2008-02-07 Yahoo! Inc. System and method for forecasting the performance of advertisements using fuzzy systems
CN101673385A (zh) * 2009-09-28 2010-03-17 百度在线网络技术（北京）有限公司 消费预估方法和装置
CN102346899A (zh) * 2011-10-08 2012-02-08 亿赞普(北京)科技有限公司 一种基于用户行为的广告点击率预测方法和装置
CN102929939A (zh) * 2012-09-28 2013-02-13 北京奇虎科技有限公司 个性化信息的提供方法及装置
CN104933591A (zh) * 2014-03-17 2015-09-23 北京学之途网络科技有限公司 一种基于实时数据实现广告投放的方法及装置
CN104965890A (zh) * 2015-06-17 2015-10-07 深圳市腾讯计算机***有限公司 广告推荐的方法和装置
CN105678586A (zh) * 2016-01-12 2016-06-15 腾讯科技（深圳）有限公司 一种信息扶持方法和装置

Family Cites Families (53)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1998058334A1 (en) 1997-06-16 1998-12-23 Doubleclick Inc. Method and apparatus for automatic placement of advertising
WO2000008802A2 (en) 1998-08-03 2000-02-17 Doubleclick Inc. Network for distribution of re-targeted advertising
US7778872B2 (en) 2001-09-06 2010-08-17 Google, Inc. Methods and apparatus for ordering advertisements based on performance information and price information
US20160162931A1 (en) 2001-09-06 2016-06-09 Georges Harik Methods and apparatus for ordering advertisements based on performance information
US8560493B2 (en) 2003-10-01 2013-10-15 Google Inc. Determining and/or using end user local time information in an ad system
US7743045B2 (en) 2005-08-10 2010-06-22 Google Inc. Detecting spam related and biased contexts for programmable search engines
US20110264508A1 (en) 2002-03-29 2011-10-27 Harik George R Scoring, modifying scores of, and/or filtering advertisements using advertiser information
US8229957B2 (en) 2005-04-22 2012-07-24 Google, Inc. Categorizing objects, such as documents and/or clusters, with respect to a taxonomy and data structures derived from such categorization
US20060149624A1 (en) 2004-12-30 2006-07-06 Shumeet Baluja Generating and/or serving local area advertisements, such as advertisements for devices with call functionality
US7693830B2 (en) 2005-08-10 2010-04-06 Google Inc. Programmable search engine
US20050114198A1 (en) 2003-11-24 2005-05-26 Ross Koningstein Using concepts for ad targeting
US8352499B2 (en) 2003-06-02 2013-01-08 Google Inc. Serving advertisements using user request information and user information
US7136875B2 (en) 2002-09-24 2006-11-14 Google, Inc. Serving advertisements based on content
US7346606B2 (en) 2003-06-30 2008-03-18 Google, Inc. Rendering advertisements with documents having one or more topics using user topic interest
US20060069616A1 (en) 2004-09-30 2006-03-30 David Bau Determining advertisements using user behavior information such as past navigation information
US7680796B2 (en) 2003-09-03 2010-03-16 Google, Inc. Determining and/or using location information in an ad system
US7716199B2 (en) 2005-08-10 2010-05-11 Google Inc. Aggregating context data for programmable search engines
US7716161B2 (en) 2002-09-24 2010-05-11 Google, Inc, Methods and apparatus for serving relevant advertisements
US9235849B2 (en) * 2003-12-31 2016-01-12 Google Inc. Generating user information for use in targeted advertising
US20050222901A1 (en) 2004-03-31 2005-10-06 Sumit Agarwal Determining ad targeting information and/or ad creative information using past search queries
US20070038603A1 (en) 2005-08-10 2007-02-15 Guha Ramanathan V Sharing context data across programmable search engines
US20070038614A1 (en) 2005-08-10 2007-02-15 Guha Ramanathan V Generating and presenting advertisements based on context data for programmable search engines
US7844493B1 (en) 2002-11-08 2010-11-30 Google, Inc. Automated price maintenance for use with a system in which advertisements are rendered with relative preference based on performance information and price information
US20050021397A1 (en) 2003-07-22 2005-01-27 Cui Yingwei Claire Content-targeted advertising using collected user behavior data
US8086559B2 (en) 2002-09-24 2011-12-27 Google, Inc. Serving content-relevant advertisements with client-side device support
US20100100437A1 (en) 2002-09-24 2010-04-22 Google, Inc. Suggesting and/or providing ad serving constraint information
US20040059712A1 (en) 2002-09-24 2004-03-25 Dean Jeffrey A. Serving advertisements using information associated with e-mail
US20050144069A1 (en) * 2003-12-23 2005-06-30 Wiseman Leora R. Method and system for providing targeted graphical advertisements
US20170372348A1 (en) 2016-06-22 2017-12-28 Google Inc. Device configuration-based function delivery
US20170237801A1 (en) 2004-06-30 2017-08-17 Google Inc. Device configuration-based function delivery
US10032452B1 (en) 2016-12-30 2018-07-24 Google Llc Multimodal transmission of packetized data
US20060004627A1 (en) 2004-06-30 2006-01-05 Shumeet Baluja Advertisements for devices with call functionality, such as mobile phones
US10510043B2 (en) * 2005-06-13 2019-12-17 Skyword Inc. Computer method and apparatus for targeting advertising
US20080256056A1 (en) * 2007-04-10 2008-10-16 Yahoo! Inc. System for building a data structure representing a network of users and advertisers
KR100892846B1 (ko) * 2007-04-17 2009-04-10 엔에이치엔(주) 순위별 광고 노출 확률 계산 방법 및 시스템
US20110231241A1 (en) * 2010-03-18 2011-09-22 Yahoo! Inc. Real-time personalization of sponsored search based on predicted click propensity
US20130179271A1 (en) * 2012-01-11 2013-07-11 Paul Adams Grouping and Ordering Advertising Units Based on User Activity
JP5491556B2 (ja) * 2012-03-05 2014-05-14 ヤフー株式会社 広告配信装置、広告配信方法及びプログラム
JP5507607B2 (ja) * 2012-04-06 2014-05-28 ヤフー株式会社 コンテンツ提供装置、低ランク近似行列生成装置、コンテンツ提供方法、低ランク近似行列生成方法およびプログラム
US10152723B2 (en) 2012-05-23 2018-12-11 Google Llc Methods and systems for identifying new computers and providing matching services
US10776830B2 (en) 2012-05-23 2020-09-15 Google Llc Methods and systems for identifying new computers and providing matching services
CN103514178A (zh) * 2012-06-18 2014-01-15 阿里巴巴集团控股有限公司 一种基于点击率的搜索排序方法及装置
CN103002051B (zh) * 2012-12-24 2016-03-16 百度在线网络技术（北京）有限公司 移动终端的信息推送方法、***及云端服务器
US10650066B2 (en) 2013-01-31 2020-05-12 Google Llc Enhancing sitelinks with creative content
US10735552B2 (en) 2013-01-31 2020-08-04 Google Llc Secondary transmissions of packetized data
WO2014155682A1 (ja) * 2013-03-29 2014-10-02 楽天株式会社 情報処理装置、情報処理方法、及び情報処理プログラム
US8788338B1 (en) * 2013-07-01 2014-07-22 Yahoo! Inc. Unified marketplace for advertisements and content in an online system
CN104636940A (zh) * 2013-11-07 2015-05-20 深圳市腾讯计算机***有限公司 广告投放方法及***、广告筛选方法及装置、服务器
CN104866515A (zh) * 2014-02-23 2015-08-26 祁勇 一种分类发布和接收社交网络内容的方法
CN104883585B (zh) * 2014-02-28 2019-06-14 腾讯科技（北京）有限公司 显示媒体数据的方法、设备及***
WO2015141931A1 (ko) * 2014-03-19 2015-09-24 에스케이플래닛 주식회사 광고 제공 장치 및 방법
CN106664463A (zh) * 2014-10-20 2017-05-10 谷歌公司 任意大小内容条目生成
US10708313B2 (en) 2016-12-30 2020-07-07 Google Llc Multimodal transmission of packetized data

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080033810A1 (en) * 2006-08-02 2008-02-07 Yahoo! Inc. System and method for forecasting the performance of advertisements using fuzzy systems
CN101673385A (zh) * 2009-09-28 2010-03-17 百度在线网络技术（北京）有限公司 消费预估方法和装置
CN102346899A (zh) * 2011-10-08 2012-02-08 亿赞普(北京)科技有限公司 一种基于用户行为的广告点击率预测方法和装置
CN102929939A (zh) * 2012-09-28 2013-02-13 北京奇虎科技有限公司 个性化信息的提供方法及装置
CN104933591A (zh) * 2014-03-17 2015-09-23 北京学之途网络科技有限公司 一种基于实时数据实现广告投放的方法及装置
CN104965890A (zh) * 2015-06-17 2015-10-07 深圳市腾讯计算机***有限公司 广告推荐的方法和装置
CN105678586A (zh) * 2016-01-12 2016-06-15 腾讯科技（深圳）有限公司 一种信息扶持方法和装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3404607A4 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2021502651A (ja) * 2017-12-29 2021-01-28 オッポ広東移動通信有限公司Ｇｕａｎｇｄｏｎｇ Ｏｐｐｏ Ｍｏｂｉｌｅ Ｔｅｌｅｃｏｍｍｕｎｉｃａｔｉｏｎｓ Ｃｏｒｐ．， Ｌｔｄ． 融合データ処理方法及び情報推薦システム
US11061966B2 (en) 2017-12-29 2021-07-13 Guangdong Oppo Mobile Telecommunications Corp., Ltd. Method for processing fusion data and information recommendation system
JP7052037B2 (ja) 2017-12-29 2022-04-11 オッポ広東移動通信有限公司 融合データ処理方法及び情報推薦システム
CN111666488A (zh) * 2020-04-21 2020-09-15 上海晶赞融宣科技有限公司 信息发布方法及装置、存储介质、终端
TWI756967B (zh) * 2020-12-04 2022-03-01 萬里雲互聯網路有限公司 內容點擊數預測裝置及其方法

Also Published As

Publication number Publication date
US11144950B2 (en) 2021-10-12
KR20170137845A (ko) 2017-12-13
EP3404607A1 (en) 2018-11-21
KR101999469B1 (ko) 2019-07-11
CN105678586A (zh) 2016-06-15
EP3404607A4 (en) 2019-07-17
JP6522160B2 (ja) 2019-05-29
JP2018519572A (ja) 2018-07-19
CN105678586B (zh) 2020-09-29
US20180060903A1 (en) 2018-03-01

Similar Documents

Publication Publication Date Title
WO2017121258A1 (zh) 信息投放方法、装置、服务器和存储介质
JP6600703B2 (ja) 動的コンテンツアイテム作成
US10776829B2 (en) Systems and methods for dynamic ad selection of multiple ads or ad campaigns on devices
JP6009568B2 (ja) 広告キャッシュ保守の方法および装置
US11210701B2 (en) Dynamic content item creation
WO2015158149A1 (zh) 一种用于对特征信息的变化进行预测的方法和装置
JP4621793B2 (ja) 広告配信サーバ
TW201426592A (zh) 投放資訊搜索的方法和裝置
US20080228571A1 (en) Automated recommendation of targeting criteria
CN106296247B (zh) 网络信息资源在线排序方法和装置
WO2018121209A1 (zh) 资源竞争参数阈值的动态调整方法及装置和服务器
CN111566684A (zh) 用于优化广告的用户分组的方法和***
JP6946082B2 (ja) 広告配信支援装置、広告配信支援方法、およびプログラム
JP6181817B1 (ja) 決定装置、決定方法、及び決定プログラム
JP5699233B1 (ja) 生成装置、生成方法および生成プログラム
US10262333B2 (en) Linear programming approach for querying a trie data structure
US20180268443A1 (en) Determination method, determination apparatus, and non-transitory computer-readable storage medium
KR102564364B1 (ko) 고객 단말의 정보를 이용하여 온라인 마켓 및 판매 상품을 추천하고 판매 상품의 가격을 결정하는 방법
CN112308579A (zh) 一种信息投放方法、装置及存储介质
KR102573336B1 (ko) 고객 단말의 구독 정보 및 고객 정보에 기초하여 온라인 마켓을 추천하는 방법
US20120158495A1 (en) Inventory allocation in advertising systems
KR102564355B1 (ko) 고객 정보를 이용하여 온라인 마켓의 판매 상품을 추천하는 방법
CN113129070B (zh) 消息投放方法、装置、服务器及存储介质
CN113392340A (zh) 生成展示内容的方法、装置、设备和计算机可读介质
JP2021120855A (ja) 予測及び観測された位置イベントに基づいて情報キャンペーンをペーシングするシステム及び方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16884809

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 20177032826

Country of ref document: KR

Kind code of ref document: A

Ref document number: 2017559381

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE