WO2018121209A1 - 资源竞争参数阈值的动态调整方法及装置和服务器 - Google Patents

资源竞争参数阈值的动态调整方法及装置和服务器 Download PDF

Info

Publication number
WO2018121209A1
WO2018121209A1 PCT/CN2017/115019 CN2017115019W WO2018121209A1 WO 2018121209 A1 WO2018121209 A1 WO 2018121209A1 CN 2017115019 W CN2017115019 W CN 2017115019W WO 2018121209 A1 WO2018121209 A1 WO 2018121209A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
competition
contention
resource
parameter threshold
Prior art date
Application number
PCT/CN2017/115019
Other languages
English (en)
French (fr)
Inventor
向卓林
邹正勇
石瑞超
李学凯
苏麒匀
林世飞
赵艳
黄磊
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
Publication of WO2018121209A1 publication Critical patent/WO2018121209A1/zh
Priority to US16/384,203 priority Critical patent/US11277352B2/en

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/70Admission control; Resource allocation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/06Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063Operations research, analysis or management
    • G06Q10/0631Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06312Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce
    • G06Q30/02Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241Advertisements
    • G06Q30/0273Determination of fees for advertising
    • G06Q30/0275Auctions
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/01Social networking

Definitions

  • the present invention relates to the field of computer technologies, and in particular, to a dynamic adjustment method, apparatus, and server for resource contention parameter thresholds.
  • each terminal contends for a channel.
  • each advertiser competes for display resources for advertising.
  • the resource competition participants can send different resource competition parameters to the system according to their own conditions.
  • the system can filter out some competitors based on a certain fixed threshold, and the remaining competitors (which can be called admissionrs) are eligible to participate in the resources. competition.
  • the value of the threshold is too high, there may be cases where no access or access is less than the number of resources, which is not conducive to the rational use of resources. If the value of the threshold is too low, filtering may not be achieved.
  • An object of the embodiments of the present invention is to provide a method, a device, and a server for dynamically adjusting a resource contention parameter threshold to dynamically adjust a resource contention parameter threshold.
  • the embodiment of the present invention provides the following solutions:
  • an embodiment of the present invention provides a dynamic adjustment method for a resource contention parameter threshold in resource competition, including:
  • the first contention data includes a resource competition participant participating in the competition for the placement position in the first specified time period Competitive behavior data;
  • the second competition data includes participating in the competition for the second designated time period
  • the resource of the location competes with the participant's competitive behavior data
  • the end time of the second specified time period is the current time
  • the end time of the first specified time period is not the current time
  • the third specified time period includes at least the second specified time period. portion.
  • an embodiment of the present application provides a method for dynamically adjusting a resource contention parameter threshold, including:
  • first contention data of the display location and obtaining a first optimal resource contention parameter threshold according to the first contention data a value, the first contention data comprising competition behavior data of a resource competition participant participating in the competition for the first specified time period;
  • the end time of the second specified time period is the current time
  • the end time of the first specified time period is earlier than the current time
  • the third specified time period includes at least the second specified time period. Part time.
  • an embodiment of the present invention provides a dynamic adjustment device for a resource contention parameter threshold in resource competition, including:
  • a first data analysis module configured to obtain first contention data of the display location, and obtain a first optimal resource contention parameter threshold according to the acquired first contention data;
  • the first contention data includes: Competitive behavior data of competitors competing for resources competing for the placement;
  • a second data analysis module for:
  • the end time of the second specified time period is the current time
  • the end time of the first specified time period is not the current time
  • the third specified time period includes at least the second specified time period. portion.
  • an embodiment of the present invention provides a dynamic adjustment device for a resource contention parameter threshold, including:
  • a first data analysis module configured to acquire first contention data of the display location, and obtain a first optimal resource contention parameter threshold according to the first contention data, where the first contention data includes: participating in the first specified time period Competitive behavior data of competitors competing for resources of the placement;
  • a second data analysis module for:
  • the end time of the second specified time period is the current time
  • the end time of the first specified time period is earlier than the current time
  • the third specified time period includes at least the second specified time period. Part time.
  • an embodiment of the present invention provides a computer readable storage medium having instructions stored therein that, when run on a computer, cause the computer to perform the methods described in the above aspects.
  • an embodiment of the present invention provides a computer program product comprising instructions that, when run on a computer, cause the computer to perform the methods described in the above aspects.
  • the embodiment of the present invention further provides a server, where the server includes: a processor, a memory; the memory is configured to store an instruction; the processor is configured to execute the instruction in the memory, so that the The server performs the method of any of the preceding aspects.
  • the first optimal resource contention parameter threshold (referred to as a first optimal threshold) is obtained by using the first contention data of the display resource. And acquiring second competition data of the display resource multiple times, and obtaining a second optimal resource competition parameter threshold (referred to as a second optimal threshold) according to the second competition data. And in a case that the second optimal threshold and the first optimal threshold are excessively different, the first optimal resource contention parameter threshold is obtained by using the third contention data of the display resource. In this way, the threshold is always adjusted with the real-time conditions, and the threshold is optimized.
  • FIG. 1 is a schematic diagram of an application scenario according to an embodiment of the present disclosure
  • FIG. 2 is an exemplary structural diagram of an adjustment apparatus or a server according to an embodiment of the present invention
  • FIG. 3 is an exemplary flowchart of an adjustment method according to an embodiment of the present invention.
  • FIG. 4a is another exemplary flowchart of an adjustment method according to an embodiment of the present invention.
  • FIG. 4b is another exemplary flowchart of an adjustment method according to an embodiment of the present invention.
  • FIG. 5 is another exemplary structural diagram of an adjustment apparatus or a server according to an embodiment of the present disclosure.
  • FIG. 6 is a schematic diagram of a composition structure of a server for dynamically adjusting a resource contention parameter threshold according to an embodiment of the present invention.
  • the embodiment of the invention provides a method, a device and a server for dynamically adjusting a resource contention parameter threshold, and adjusts a resource competition parameter threshold by dynamic adaptiveness.
  • FIG. 1 shows an exemplary application scenario of the above-described adjustment device, in which the server 101 (including the adjustment device) and the terminal devices C1, C2, C3 are included.
  • the above adjustment means can be applied to the server 101 in a software or hardware manner.
  • the terminal devices C1, C2, C3, etc. may be various handheld devices with communication functions, in-vehicle devices, wearable devices, computing devices, positioning devices or other processing devices connected to the wireless modem, and various forms of user devices.
  • User Equipment UE for short
  • MS Mobile Station
  • mobile phone tablet computer
  • desktop computer PDA (Personal Digital Assistant, personal digital assistant) and so on.
  • FIG. 1 exemplarily shows three terminal devices. In an application scenario, the number of terminal devices is not limited to three, and may be fewer or more.
  • Clients can be deployed on each of the above terminal devices, such as a social application client, a news client, and the like.
  • an ordinary mass user or an advertisement-pushing audience sends an information acquisition request (such as an advertisement pull request) to the server 101 when logging in to the client, or when browsing the client webpage or even clicking a link.
  • the request server 101 pushes a media file (e.g., an advertisement) thereto.
  • Media files include, but are not limited to, video files, audio files, picture files, text files, etc., or any combination of several files.
  • each open client's advertising space has been auctioned by multiple advertisers (also known as resource competitors), and the winning advertiser's advertisement can be displayed.
  • the reserve price is similar to the threshold, and the principle of bidding can be explained by the following simple example. Assuming the reserve price is 3 yuan, the bids (bid) of advertisers 1, advertisers 2, advertisers 3, and advertisers 4 participating in a certain display resource (for example, the advertisement space 1 of the first user A at the time T0) are respectively 1 yuan. , 2 yuan, 3 yuan, 4 yuan, the advertisers 1, 2 will be eliminated, and advertisers 3 and 4 will be allowed to participate in the competition, assuming the final advertiser 4 wins, the advertiser 4 ads will be displayed .
  • server 101 will serve an ad that wins the advertiser to an ad slot of User A's WeChat client.
  • server 101 will serve an ad that wins the advertiser to an ad slot of User A's WeChat client.
  • news client as user B browses the number of news, there will be multiple ad slots available for ad exposure, and each ad slot will be auctioned by multiple advertisers, each ad.
  • the ads served on the site are the ones that win the advertiser.
  • the server 101 or the adjustment device may be an advertisement server or a server cluster/cloud platform composed of multiple advertisement servers.
  • FIG. 2 is a diagram showing an example of the configuration of the above-described adjusting apparatus or server 101. As shown in FIG. 2, it may include a bus, a processor 1, a memory 2, a communication interface 3, an input device 4, and an output device 5.
  • the processor 1, the memory 2, the communication interface 3, the input device 4, and the output device 5 are connected to each other through a bus. among them:
  • the bus can include a path for communicating information between various components of the computer system.
  • the processor 1 may be a general-purpose processor, such as a general-purpose central processing unit (CPU), a network processor (NP Processor, NP for short, a microprocessor, etc., or an application-specific integrated circuit (ASIC). Or one or more integrated circuits for controlling the execution of the program of the embodiments of the present invention. It can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components.
  • DSP digital signal processor
  • ASIC application specific integrated circuit
  • FPGA off-the-shelf programmable gate array
  • the memory 2 stores a program for executing the technical solution of the embodiment of the present invention, and can also store an operating system and other key services.
  • the program can include program code, the program code including computer operating instructions.
  • the memory 2 may include a read-only memory (ROM), other types of static storage devices that can store static information and instructions, random access memory (RAM), storable information, and Other types of dynamic storage devices, disk storage, flash, and so on.
  • Communication interface 3 may include devices that use any type of transceiver to communicate with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), and the like.
  • RAN Radio Access Network
  • WLAN Wireless Local Area Network
  • Input device 4 may include means for receiving data and information input by a user, such as a keyboard, mouse, camera, scanner, light pen, voice input device, touch screen, pedometer or gravity sensor, and the like.
  • Output device 5 may include devices that allow output of information to the user, such as a display screen, speakers, and the like.
  • the processor 1 of the adjustment device or server 101 executes the program stored in the memory 2, and calls other devices, A dynamic adjustment method for implementing a resource contention parameter threshold provided by an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of an exemplary interaction of a dynamic adjustment method of a resource contention parameter threshold according to an embodiment of the present invention.
  • the processor 1 of the foregoing server 101 or the adjustment apparatus interacts with other internal or external devices.
  • Processor 1 obtains first contention data for the placement.
  • a placement can refer to an ad slot.
  • the first competition data may include competition behavior data of the resource competition participants participating in the competition for the first specified time period. It can also be said that the first competition data characterizes the competition behavior of the resource competition participants for the placement at the first specified time period.
  • the length and the start time of the first specified time period described above can be flexibly designed according to the situation, for example, the time period is 1 hour, 1 day, one week, two weeks, and the like. For example, taking the length of the time period as 1 hour as an example, the competition data between 12:00 and 13:00 is obtained at 13:00.
  • the placement will be associated with multiple impression assets over a period of time.
  • the above-mentioned "competitive behavior data of the resource competition participants participating in the competition for the placement” may further include competition behavior data of the resource competition participants participating in the competition for each display resource.
  • the competition behavior data of each resource competition participant may include a user identifier of the resource competition participant, a media file identifier associated with the user identifier, and a resource competition parameter. It should be noted that the user identity of a resource competition participant may be associated with one or more media file identifiers.
  • the presentation resource may include: an identification identifier of the audience (eg, a user identification identifier of the news client), a placement identifier (eg, an advertisement slot identifier of the news client), and a presentation moment, for example, The time at which the advertisement client identifies the advertisement to the audience corresponding to the identification identifier.
  • an identification identifier of the audience eg, a user identification identifier of the news client
  • a placement identifier eg, an advertisement slot identifier of the news client
  • a presentation moment for example, The time at which the advertisement client identifies the advertisement to the audience corresponding to the identification identifier.
  • each open client's advertising space has been auctioned by multiple advertisers, and the winning advertiser's advertisement can be displayed.
  • the first contention data may illustratively include the following contents:
  • resource 1 (advertising space 1 displayed to audience A at 12:00), multiple participating resource competition participant identification (advertiser) corresponding to resource 1, media file identification (advertising), advertising bid (resource Competitive parameter);
  • the resource competition participant is the advertiser
  • the media file associated with it is the advertisement file that the advertiser wants to serve
  • the resource competition parameter is the bid of the advertiser.
  • the advertiser M1 wants to place a car advertisement (identified as P1) with a bid of 1 yuan; the advertiser M2 wants to put cosmetics on it.
  • the advertisement (identified as P2) has a bid of 0.5 yuan; the advertiser M3 wants to place a leather shoe advertisement (identified as P3) with a bid of 0.1 yuan;
  • the advertiser M4 wants to put a health product advertisement (identified as P4) with a bid of 0.5 yuan; the advertiser M5 wants to advertise the electronic product to it ( Mark as P5), bid It is 0.5 yuan.
  • the display resources and their corresponding competitive behavior data may be acquired by means of extraction.
  • the sampling frequency can be extracted at a certain time. For example, if the first specified time period is one day, then every other hour is taken during this day.
  • the resource contention parameter threshold may include a system floor price, which is not disclosed to the advertiser.
  • the advertiser can automatically send a bid to the server 101 through the automatic bidding system to bid.
  • advertisers need to set a bid for each ad exposure in advance.
  • the above bids and the like are stored in a log manner, and therefore, in an exemplary illustration, the processor 1 obtains the first contention data by extracting the logs.
  • the processor 1 obtains a first optimal resource contention parameter threshold according to the first contention data.
  • the first optimal resource competition parameter threshold may include a first optimal reserve price, that is, an initial optimal reserve price. Different ad slots may correspond to different first optimal reserve prices.
  • the resource competition participants participating in the real-time competition of the display resources will be filtered using the first optimal reserve price obtained this time.
  • the first optimal reserve price obtained at 13:00 is 0.2 yuan
  • the first optimal reserve price is 0.1 yuan at 15:00
  • the first optimal reserve price is 0.3 yuan at 21:00.
  • obtaining the foregoing first optimal resource contention parameter threshold may be obtained as follows:
  • the processor 1 acquires the second contention data of the above placement location multiple times, and obtains the second optimal resource contention parameter threshold based on the second contention data.
  • the second competition data may include competition behavior data of the resource competition participants participating in the competition for the second specified time period. Or it can be said that the second competition data represents the competition behavior of the resource competition participants for the placement position in the second specified time period.
  • the second competition data represents the competition behavior of the resource competition participants for the placement position in the second specified time period.
  • the end time of the second specified time period is the current time.
  • the current time is the acquisition time of the second competition data described above.
  • the end time of the first specified time period is not the current time.
  • the end time of the first specified time period is earlier than the end time of the second specified time period. Assuming that the current time is 13:00, the end time of the first specified time period is earlier than 13:00.
  • the second contention data may be acquired at fixed time intervals during the second specified time period.
  • the time interval can be designed to be one second or several seconds, and the second contention data is acquired once every second or several seconds.
  • the time interval can be designed to be one minute, one hour, and the like.
  • the second contention data may also be randomly acquired multiple times in the second specified time period, and the second optimal resource contention parameter threshold may be calculated.
  • the duration of the second specified time period can be designed to be one second, several seconds, one minute, one hour, and the like.
  • the duration of the second specified time period may be less than the duration of the first specified time period.
  • the second optimal resource contention parameter threshold may be calculated differently from the first optimal resource contention parameter threshold.
  • the probability distribution and the probability density of the quality factor corresponding to the resource competition participant may be calculated according to the second contention data, and the second optimal resource contention parameter threshold is calculated according to the quality factor probability distribution and the probability density.
  • This paper will further detail how to calculate the second optimal resource competition parameter threshold based on the quality factor probability distribution and probability density.
  • the quality factor is a comprehensive value. According to the actual situation, the advertisers with high bids generally have higher quality of advertisements, and the probability of attracting users is higher. It should be noted that the relationship between bid height and quality factor is non-linear.
  • the quality factor can be a factor obtained by fitting the prices of different advertisers. More specifically, in the advertisement bidding application scenario, the second optimal resource contention parameter threshold may include a second optimal reserve price, and the probability distribution and probability density of the quality factor may be obtained by an advertiser's bid.
  • the number of advertisers bidding 4 yuan in the second competition data is A
  • the number of advertisers biding 3 yuan is B
  • the number of advertisers biding 5 yuan is C
  • the processor 1 determines whether the difference between the second optimal resource contention parameter threshold obtained by the current calculation and the first optimal resource contention parameter threshold is greater than a difference threshold, and if yes, enters part 304, and if not, returns to part 302. .
  • the difference value can be calculated in various ways: for example, it is assumed that the second optimal resource contention parameter threshold is represented by Y1, and the first optimal resource contention parameter threshold is represented by Y2, and
  • the difference threshold can be flexibly designed according to different actual scenarios, and will not be described here.
  • section 304 Obtain third competition data for the above placement.
  • the third competition data may include competition behavior data of the resource competition participants participating in the competition for the third specified time period. Or it can be said that the third competition data represents the resource competition participants at the third designation The competitive behavior of the placement of the segments.
  • the end time of the third specified time period may be the acquisition time of the third competition data.
  • the third specified time period includes at least a part of the second specified time period. Assuming that the time for acquiring the second contention data is 13:00 and the duration of the third specified time period is 30 minutes, the third specified time period may be 12:30-13:00, or 13:00-13:30, or 12:50-13:20 and so on.
  • the duration of the third specified time period can be designed to be one second, several seconds, one minute, one hour, one day, and the like.
  • the first optimal resource contention parameter threshold is updated according to the third contention data.
  • a portion 302 may also be returned.
  • the first optimal resource contention parameter threshold may be recalculated based on the third contention data. That is, the 305 portion is similar to the 301 portion.
  • the first specified time period is 1 hour and the third specified time period is 2 hours.
  • the competition data first competition data
  • the first optimal reserve price was obtained.
  • the first optimal reserve price is used in subsequent advertising auctions.
  • obtaining the second competitive base data multiple times to calculate the second optimal reserve price.
  • the difference between the second optimal reserve price and the first reserve price is greater than the difference threshold, then the current time (17:00) is the end time, and the competition data within two hours before 17:00 is obtained. , that is, the competition data between 15:00-17:00 (ie the third competition data). Based on the third competitive data, the first optimal reserve price is recalculated and used in subsequent advertising auctions. This cycle is executed.
  • the reason for this loop execution is that the behavior of the advertiser is dynamically changing.
  • the threshold of the first optimal resource competition parameter needs to be correlated. Corrected.
  • the dynamic adjustment method of the resource contention parameter threshold provided by the embodiment of the present invention is different for each advertisement position, and the advertisement space is different from the display resource, and the display resource also covers the audience ID and the display time. Therefore, for each advertisement bit, the first contention data in the first specified time period is obtained, and the first optimal resource contention parameter threshold corresponding to the advertisement bit is obtained according to the first contention data. Then, the second contention data of the advertisement space is obtained multiple times, the second optimal resource contention parameter threshold is calculated according to the second contention data, and the second optimal resource contention parameter threshold and the first optimal corresponding to the advertisement bit are obtained. When the difference of the resource competition parameter threshold is too large, the third contention data is obtained, and the first optimal resource contention parameter threshold corresponding to the advertisement bit is updated.
  • the first optimal resource contention parameter threshold (referred to as the first optimal threshold) is obtained by using the first contention data of the display resource.
  • the updated first optimal resource contention parameter threshold is obtained by using the third contention data of the display resource. In this way, the threshold is always adjusted with the real-time conditions, and the threshold is optimized.
  • FIG. 4a and FIG. 4b are another exemplary schematic diagram of an adjustment method according to an embodiment of the present invention.
  • the processor 1 of the foregoing server 101 or the adjustment apparatus is implemented by interacting with other devices.
  • the advertisement space A is taken as an example for description.
  • the above interaction process includes:
  • Processor 1 obtains first contention data for ad slot A.
  • the first contention data is divided into N data groups, and N is a positive integer greater than one.
  • Each data group may include multiple display resources, and the competition behavior data of the competition participants participating in the competition for each display resource.
  • the competition behavior data of the competition participants participating in the competition for each display resource For details, refer to section 300 in the foregoing embodiment, and details are not described herein.
  • Section 401 Processor 1 assigns different resource contention parameter thresholds (base prices) to N data sets.
  • processor 1 allocates five reserve prices for five data sets, which are 0.1, 0.2, 0.3, 0.4, and 0.5, respectively.
  • Processor 1 determines a candidate data set from the N data sets described above.
  • the resource contention parameter threshold corresponding to the candidate data group includes: a target resource contention parameter threshold.
  • it may be determined whether the ith data set is a candidate array, and i is a positive integer less than N.
  • Step A Calculate the revenue impact related data of the i-th data group based on the i-th resource competition parameter threshold.
  • the above-mentioned revenue impact related data may include the predicted CTR (click rate), the number of affected advertisers (filtered advertisers), the number of advertisements affected, and the like.
  • Step B Determine whether the above-mentioned income affects whether the relevant data meets the constraint condition.
  • the minimum threshold of the CTR, the maximum number of affected advertisers, and the maximum number of affected advertisements may be set, if the predicted CTR is greater than the minimum threshold of the CTR, and the predicted number of affected advertisers is less than the number of affected advertisers Threshold, and the predicted number of affected ads is less than the maximum number of affected ads, then the constraint is considered to be met.
  • Step C If it is determined that the income impact related data meets the above constraint condition, the ith data group is used as a candidate data group.
  • Section 403 and subsequent operations will not be performed.
  • the processor 1 calculates an overall competitive return value of the candidate data set based on a resource contention parameter threshold corresponding to each candidate data set.
  • the competitive revenue value corresponding to each display resource may be calculated based on the target resource contention parameter threshold corresponding to the candidate data group, and then the competitive income value corresponding to all the display resources is summed to obtain the candidate data group. Overall competitive return value.
  • a candidate data set has 100 display resources, and the expected return value of the 100 display resources can be separately calculated, and then the sum operation is performed to obtain the overall competitive return value of the entire set.
  • calculating the competitive revenue value corresponding to any display resource may include the following steps:
  • Step A1 Filtering the competition behavior data of the resource competition participants competing for any of the above display resources to obtain a subset of the competition behavior data.
  • the resource contention parameter threshold corresponding to the candidate data group includes: a target resource contention parameter threshold; and the contention data subset includes the contention competition data of the resource competition participant whose resource competition parameter is greater than or equal to the target resource contention parameter threshold.
  • Step B1 Calculate the thousand display advertising revenue (ecpm) of each media file (advertisement) in the competitive behavior data subset.
  • Step C1 Calculate an expected revenue value (or a revenue value) for the largest media file of ecpm, and use the calculated expected revenue value as the competitive revenue value corresponding to any of the display resources.
  • the ads in the subset of competing behavioral data can be sorted by ecpm from large to small.
  • the subset of competitive behavior data includes competitive behavior data for 40 advertisers. A total of 50 ads associated with these 40 advertisers are calculated for the ecpm of the 50 ads, and the ads are sorted by ecpm from large to small.
  • the expected revenue value of the fourth advertisement is calculated as the competitive revenue value of the display resource 1.
  • GSP Generalized Second Price
  • the deduction is calculated by the bid for the second ad + fixed value (for example, 0.1);
  • the first-ranked advertisement has a bid of 1 yuan
  • the deduction fee is 0.8 yuan
  • the allocated reserve price is 0.2
  • the expected return value is calculated according to 0.8.
  • the top bid is 1 yuan
  • the deduction is 0.18 yuan
  • the allocated reserve price is 0.2
  • the expected return value is calculated according to 0.18.
  • the processor 1 uses the resource contention parameter threshold corresponding to the maximum overall competitive revenue value as the first optimal resource contention parameter threshold.
  • candidate data sets AD are determined among the five data sets, wherein the candidate data set D has the largest overall competitive return value, and the candidate data set D corresponds to the resource contention parameter threshold of 0.4, and 0.4 is the first most Excellent resource competition parameter threshold.
  • the overall competitive return value of each data group may be directly calculated, and the resource competition parameter threshold corresponding to the largest overall competitive return value may be selected as the first optimal resource contention parameter threshold.
  • the processor 1 acquires the second contention data multiple times, and obtains the second optimal resource contention parameter threshold based on the second contention data.
  • the second optimal resource contention parameter threshold can be calculated based on the quality factor probability distribution and the probability density.
  • the second optimal resource contention parameter threshold can be calculated according to the following formula:
  • p * represents the second optimal resource contention parameter threshold
  • S * represents the value of the quality factor S (which maximizes the benefit)
  • e represents the click rate
  • e is the average click rate of the ad slot (eg, ad slot A) for the second specified time period or the current click rate of the ad slot.
  • S * is the order A value equal to zero.
  • S represents the quality factor
  • F(S) and f(S) are the probability distribution function and probability density function of the aforementioned quality factor.
  • the second optimal resource contention parameter threshold may also be calculated in other manners, and details are not described herein.
  • the processor 1 determines whether the difference between the second optimal resource contention parameter threshold and the first optimal resource contention parameter threshold obtained by the current calculation is greater than a difference threshold. If yes, the process proceeds to section 407, and if not, returns to section 405.
  • Section 406 is similar to Section 303 and will not be described here.
  • FIG. 5 is a schematic diagram showing another possible structure of the dynamic adjustment apparatus involved in the above embodiment, including:
  • the first data analysis module 501 is configured to obtain first contention data of the display location, and obtain a first optimal resource contention parameter threshold according to the first contention data, where the first contention data includes participating in the first specified time period Competition data competing for participants in the competition for the placement;
  • the second data analysis module 502 is configured to:
  • Second contention data of the display location and calculating, by the second contention data, a second optimal resource contention parameter threshold, where the second contention data includes resources participating in the competition for the placement location in a second specified time period Competitive behavior data of competing participants;
  • the end time of the second specified time period is the current time
  • the end time of the first specified time period is earlier than the current time
  • the third specified time period includes at least the second specified time period. Part time.
  • the second data analysis module 502 may be further configured to: return the second contention data of the display location multiple times, and An operation of calculating a second optimal resource contention parameter threshold based on the second contention data.
  • the first data analysis module 501 is specifically configured to: obtain the first optimal resource contention parameter threshold according to the first contention data:
  • the first data analysis module 501 is further configured to: before the competitive value of the competition corresponding to the display, the first data analysis module 501 is further configured to:
  • the first contention data is divided into N data groups, and the N is a positive integer greater than 1.
  • the first data analysis module 501 is specifically configured to use, in the calculating, based on the first contention data, a competitive revenue value corresponding to the display location under a different resource contention parameter threshold.
  • the first data analysis module is specifically configured to: in the determining, by using the resource competition parameter threshold that maximizes the competitive benefit value, the first data analysis module threshold is used to:
  • the resource competition parameter threshold corresponding to the largest overall competitive benefit value is used as the first optimal resource competition parameter threshold.
  • the first data analysis module 501 is configured to: in the determining the candidate data group from the N data groups, the:
  • the i-th data group is used as a candidate data group.
  • the second data analysis module 502 in the calculating the second optimal resource contention parameter threshold based on the acquired second contention data, is specifically configured to:
  • the first data analysis module 501 can be used to perform the 300-301 portion of the embodiment shown in FIG. 3; in addition, portions 400-404 of the embodiment shown in Figures 4a and 4b can also be performed.
  • the second data analysis module 502 can be used to perform portions 302-305 of the embodiment illustrated in Figure 3; in addition, portions 405-407 of the embodiment illustrated in Figures 4a and 4b can also be performed.
  • FIG. 6 is a schematic structural diagram of a server according to an embodiment of the present invention.
  • the server 1100 may have a large difference due to different configurations or performances, and may include one or more central processing units (CPUs) 1122 (for example, One or more processors and memory 1132, one or more storage media 1130 that store application 1142 or data 1144 (eg, one or one storage device in Shanghai).
  • the memory 1132 and the storage medium 1130 may be short-term storage or persistent storage.
  • the program stored on storage medium 1130 may include one or more modules (not shown), each of which may include a series of instruction operations in the server.
  • central processor 1122 can be configured to communicate with storage medium 1130, executing a series of instruction operations in storage medium 1130 on server 1100.
  • Server 1100 may also include one or more power sources 1126, one or more wired or wireless network interfaces 1150, one or more input and output interfaces 1158, and/or one or more operating systems 1141, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM and more.
  • operating systems 1141 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. 6.
  • the steps of a method or algorithm described in connection with the present disclosure may be implemented in a hardware, or may be implemented by a processor executing software instructions.
  • the software instructions may be comprised of corresponding software modules that may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable hard disk, CD-ROM, or any other form of storage well known in the art.
  • An exemplary storage medium is coupled to the processor to enable the processor to read information from, and write information to, the storage medium.
  • the storage medium can also be an integral part of the processor.
  • the processor and the storage medium can be located in an ASIC. Additionally, the ASIC can be located in the user equipment.
  • the processor and the storage medium may also reside as discrete components in the user equipment.
  • the functions described herein can be implemented in hardware, software, firmware, or any combination thereof.
  • the functions may be stored in a computer readable medium or transmitted as one or more instructions or code on a computer readable medium.
  • Computer readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one location to another.
  • a storage medium may be any available media that can be accessed by a general purpose or special purpose computer.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Marketing (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Educational Administration (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Transfer Between Computers (AREA)

Abstract

一种资源竞争参数阈值的动态调整方法及装置和服务器。方法包括如下步骤:使用展示资源的第一竞争数据得到第一最优资源竞争参数阈值。多次获取展示资源的第二竞争数据,根据第二竞争数据得到第二最优资源竞争参数阈值。在第二最优资源竞争参数阈值和第一最优资源竞争参数阈值差异过大的情况下,使用展示资源的第三竞争数据得到第一最优资源竞争参数阈值。

Description

资源竞争参数阈值的动态调整方法及装置和服务器
本申请要求于2016年12月29日提交中国专利局、申请号为2016112456461、发明名称为“资源竞争中资源竞争参数阈值的动态调整方法及装置”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
本发明涉及计算机技术领域,特别是涉及资源竞争参数阈值的动态调整方法及装置和服务器。
背景技术
很多场景下均会有竞争资源的情况。例如,通信场景下,各终端竞争信道。再例如,广告竞价场景下,各广告主竞争展示资源进行广告投放。
具体的,资源竞争参与者可根据自己的情况向系统发送不同的资源竞争参数,系统可基于某一固定阈值过滤掉一部分竞争者,剩余竞争者(可称为准入者)才有资格参与资源竞争。
现有技术中,上述阈值的取值若过高,可能会出现无准入者或者准入者少于资源数的情况,从而不利于资源的合理利用。而若该阈值的取值过低,可能起不到过滤的作用。
因此,如何优化阈值的取值,是目前需要解决的技术问题。
发明内容
本发明实施例的目的在于提供资源竞争参数阈值的动态调整方法及装置和服务器,以动态自适应调整资源竞争参数阈值。
为实现上述目的,本发明实施例提供了如下方案:
第一方面,本发明实施例提供一种资源竞争中资源竞争参数阈值的动态调整方法,包括:
获取展示位置的第一竞争数据,根据第一竞争数据得到第一最优资源竞争参数阈值;所述第一竞争数据包括在第一指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
多次获取所述展示位置的第二竞争数据,并基于所述第二竞争数据计算得到第二最优资源竞争参数阈值;所述第二竞争数据包括在第二指定时间段参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
在第二最优资源竞争参数阈值与第一最优资源竞争参数阈值间的差异值大于差异阈值时,获取所述展示位置的第三竞争数据;所述第三竞争数据包括在第三指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
根据所述第三竞争数据更新所述第一最优资源竞争参数阈值,返回所述多次获取所述展示位置的第二竞争数据,并基于所述第二竞争数据计算得到第二最优资源竞争参数阈值的操作;
其中,所述第二指定时间段的结束时刻为当前时刻,所述第一指定时间段的结束时刻不为所述当前时刻,所述第三指定时间段包括所述第二指定时间段的至少一部分。
第二方面,本申请的实施例提供一种资源竞争参数阈值的动态调整方法,包括:
获取展示位置的第一竞争数据,根据所述第一竞争数据得到第一最优资源竞争参数阈 值,所述第一竞争数据包括在第一指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
获取所述展示位置的第二竞争数据,以及基于所述第二竞争数据计算得到第二最优资源竞争参数阈值,所述第二竞争数据包括在第二指定时间段参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
在所述第二最优资源竞争参数阈值与所述第一最优资源竞争参数阈值间的差异值大于差异阈值时,获取所述展示位置的第三竞争数据,所述第三竞争数据包括在第三指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
根据所述第三竞争数据更新第一最优资源竞争参数阈值;
其中,所述第二指定时间段的结束时刻为当前时刻,所述第一指定时间段的结束时刻早于所述当前时刻,所述第三指定时间段包括所述第二指定时间段的至少一部分时间。
第三方面,本发明实施例提供了一种资源竞争中资源竞争参数阈值的动态调整装置,包括:
第一数据分析模块,用于获取展示位置的第一竞争数据,根据获取到的第一竞争数据得到第一最优资源竞争参数阈值;所述第一竞争数据包括:在第一指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
第二数据分析模块,用于:
多次获取所述展示位置的第二竞争数据,基于所述第二竞争数据计算得到第二最优资源竞争参数阈值;所述第二竞争数据包括在第二指定时间段参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
在第二最优资源竞争参数阈值与第一最优资源竞争参数阈值间的差异值大于差异阈值时,获取所述展示位置的第三竞争数据;所述第三竞争数据包括在第三指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
根据所述第三竞争数据更新第一最优资源竞争参数阈值,返回所述多次获取所述展示位置的第二竞争数据,并基于所述第二竞争数据计算得到第二最优资源竞争参数阈值的操作;
其中,所述第二指定时间段的结束时刻为当前时刻,所述第一指定时间段的结束时刻不为所述当前时刻,所述第三指定时间段包括所述第二指定时间段的至少一部分。
第四方面,本发明实施例提供了一种资源竞争参数阈值的动态调整装置,包括:
第一数据分析模块,用于获取展示位置的第一竞争数据,根据所述第一竞争数据得到第一最优资源竞争参数阈值,所述第一竞争数据包括:在第一指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
第二数据分析模块,用于:
获取所述展示位置的第二竞争数据,以及基于所述第二竞争数据计算得到第二最优资源竞争参数阈值,所述第二竞争数据包括在第二指定时间段参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
在所述第二最优资源竞争参数阈值与所述第一最优资源竞争参数阈值间的差异值大于差异阈值时,获取所述展示位置的第三竞争数据,所述第三竞争数据包括在第三指定时 间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
根据所述第三竞争数据更新第一最优资源竞争参数阈值;
其中,所述第二指定时间段的结束时刻为当前时刻,所述第一指定时间段的结束时刻早于所述当前时刻,所述第三指定时间段包括所述第二指定时间段的至少一部分时间。
第五方面,本发明实施例提供了一种计算机可读存储介质,所述计算机可读存储介质中存储有指令,当其在计算机上运行时,使得计算机执行上述各方面所述的方法。
第六方面,本发明实施例提供了一种包含指令的计算机程序产品,当其在计算机上运行时,使得计算机执行上述各方面所述的方法。
第七方面,本发明实施例还提供一种服务器,所述服务器包括:处理器、存储器;所述存储器用于存储指令;所述处理器用于执行所述存储器中的所述指令,使得所述服务器执行如前述各方面中任一项所述的方法。
在本发明实施例提供的调整方案中,使用展示资源的第一竞争数据得到第一最优资源竞争参数阈值(简称第一最优阈值)。并多次获取展示资源的第二竞争数据,根据第二竞争数据得到第二最优资源竞争参数阈值(简称第二最优阈值)。并在第二最优阈值和第一最优阈值差异过大的情况下,使用展示资源的第三竞争数据得到第一最优资源竞争参数阈值。这样,可保证阈值始终随着实时情况进行自适性调整,实现了对阈值的优化。
附图说明
图1为本发明实施例提供的应用场景示意图;
图2为本发明实施例提供的调整装置或服务器的一种示例性结构图;
图3为本发明实施例提供的调整方法的一种示例性流程图;
图4a为本发明实施例提供的调整方法的另一种示例性流程图;
图4b为本发明实施例提供的调整方法的另一种示例性流程图;
图5为本发明实施例提供的调整装置或服务器的另一种示例性结构图;
图6为本发明实施例提供的资源竞争参数阈值的动态调整方法应用于服务器的组成结构示意图。
具体实施方式
本发明实施例提供了资源竞争参数阈值的动态调整方法及装置和服务器,以动态自适性调整资源竞争参数阈值。
图1示出了上述调整装置的一种示例性应用场景,在该场景下包括服务器101(包含调整装置)和终端设备C1、C2、C3。
上述调整装置可以软件或硬件的方式应用于服务器101中。
其中,终端设备C1、C2、C3等可以是各种具有通信功能的手持设备、车载设备、可穿戴设备、计算设备、定位设备或连接到无线调制解调器的其它处理设备,以及各种形式的用户设备(User Equipment,简称UE)、移动台(Mobile station,简称MS)、手机、平板电脑、台式电脑、PDA(Personal Digital Assistant,个人数字助理)等等。需要说明的是,图1示例性的显示了3个终端设备,在应用场景中,终端设备数目并不仅局限于3个,其可以更少或更多。
上述各终端设备上可部署客户端,例如社交应用客户端、新闻客户端等等。
在广告投放场景下,普通大众用户或广告推送的受众在登陆客户端时,或者在浏览客户端网页乃至点击某一链接时,会向服务器101发送信息获取请求(例如广告拉取请求),以请求服务器101向其推送媒体文件(例如广告)。媒体文件包括但不限于视频文件、音频文件、图片文件、文本文件等,或几种文件的任意组合。
以某新闻客户端的应用场景为例,可能同一时刻有上万普通大众用户打开或浏览新闻客户端,因兴趣爱好等不同,不同客户端看到的广告内容不同。该时刻每一打开的客户端的广告位都经过了多位广告主(也可称为资源竞争参与者)参与竞拍,胜出的广告主的广告才可以展示。
底价类似于门槛,竞拍的原理可通过如下的简单举例进行解释。假定底价为3元,参与竞争某一展示资源(例如T0时刻第一用户A的广告位1)的广告主1、广告主2、广告主3、广告主4的出价(bid)分别为1元、2元、3元、4元,则广告主1、2会被淘汰掉,而广告主3和4会被准入参与竞争,假定最终广告主4胜出,该广告主4的广告将予以展示。
一个示例中,对于用户A的微信朋友圈而言,一般具有一个广告位,则服务器101会向用户A的微信客户端的一个广告位投放胜出广告主的广告。对于其他应用场景,例如,新闻客户端,随着用户B浏览新闻条数的增多,会有多个广告位可用于广告曝光,则每一广告位都经过多位广告主参与竞拍,每一广告位上投放的广告均是胜出广告主的广告。
服务器101或调整装置可以是一台广告服务器,或多台广告服务器组成的服务器集群/云平台。
图2是上述调整装置或服务器101的一种结构示例图,如图2所示,可包括总线、处理器1、存储器2、通信接口3、输入设备4和输出设备5。处理器1、存储器2、通信接口3、输入设备4和输出设备5通过总线相互连接。其中:
总线可包括一通路,在计算机系统各个部件之间传送信息。
处理器1可以是通用处理器,例如通用中央处理器(CPU)、网络处理器(Network Processor,简称NP)、微处理器等,也可以是特定应用集成电路(application-specific integrated circuit,ASIC),或一个或多个用于控制本发明实施例方案程序执行的集成电路。还可以是数字信号处理器(DSP)、专用集成电路(ASIC)、现成可编程门阵列(FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。
存储器2中保存有执行本发明实施例技术方案的程序,还可以保存有操作系统和其他关键业务。具体地,程序可以包括程序代码,程序代码包括计算机操作指令。更具体的,存储器2可以包括只读存储器(read-only memory,ROM)、可存储静态信息和指令的其他类型的静态存储设备、随机存取存储器(random access memory,RAM)、可存储信息和指令的其他类型的动态存储设备、磁盘存储器、flash等等。
通信接口3可包括使用任何收发器一类的装置,以便与其他设备或通信网络通信,如以太网,无线接入网(RAN),无线局域网(WLAN)等。
输入设备4可包括接收用户输入的数据和信息的装置,例如键盘、鼠标、摄像头、扫描仪、光笔、语音输入装置、触摸屏、计步器或重力感应器等。
输出设备5可包括允许输出信息给用户的装置,例如显示屏、扬声器等。
调整装置或服务器101的处理器1执行存储器2中所存放的程序,以及调用其他设备, 可用于实现本发明实施例所提供的资源竞争参数阈值的动态调整方法。
下面将基于上面所述的本发明实施例涉及的共性方面,进行进一步详细说明。
图3为本发明实施例提供的资源竞争参数阈值的动态调整方法的一种示例性交互示意图,由前述的服务器101或调整装置的处理器1与其他内部或外部设备交互实现。
上述示例性流程包括:
在300部分:处理器1获取展示位置的第一竞争数据。
在广告竞价应用场景下,展示位置可以指的是广告位。
上述第一竞争数据可包括在第一指定时间段内参与竞争上述展示位置的资源竞争参与者的竞争行为数据。也可以说,第一竞争数据表征了资源竞争参与者对在第一指定时间段的展示位置的竞争行为。
上述第一指定时间段的长度和始末时刻可根据情况进行灵活设计,例如,时间段长度为1小时、1天、一周、两周等等。例如,以时间段长度为1小时为例,在13:00获取的是12:00-13:00之间的竞争数据。
可选的,展示位置在一个时间段内会与多个展示资源相关联。则上述“参与竞争所述展示位置的资源竞争参与者的竞争行为数据”可进一步包括参与竞争各展示资源的资源竞争参与者的竞争行为数据。
其中,每一资源竞争参与者的竞争行为数据可包括该资源竞争参与者的用户标识、与用户标识相关联的媒体文件标识及资源竞争参数。需要说明的是,一个资源竞争参与者的用户标识可能与一个或多个媒体文件标识相关联。
在本发明的一些实施例中,展示资源则可包括:受众的身份识别标识(例如新闻客户端的用户身份识别标识)、展示位置标识(例如新闻客户端的广告位标识)和展示时刻,例如,在新闻客户端上述广告位标识处向上述身份识别标识所对应的受众展示广告的时刻。
在本发明的一些实施例中,可能同一时刻有上万受众打开或浏览新闻客户端,因兴趣爱好等不同,不同受众看到的广告内容不同。每一时刻每一打开的客户端的广告位都经过了多位广告主参与竞拍,胜出的广告主的广告才可以展示。
以新闻客户端为例,假定,获取了12:00-13:00的第一竞争数据,则第一竞争数据可示例性的包括如下内容:
展示资源1(12:00时刻展示给受众A的广告位1),展示资源1对应的多个参与竞争的资源竞争参与者标识(广告主),媒体文件标识(广告),广告的出价(资源竞争参数);
展示资源2(12:00时刻展示给受众B的广告位1),展示资源2对应的多个参与竞争的资源竞争参与者,媒体文件(广告),广告的出价(bid)。
举例说明,在广告竞价应用场景下,资源竞争参与者为广告主,与之相关联的媒体文件为广告主欲投放的广告文件,而资源竞争参数则为广告主的出价。
举例来讲,假定针对新闻客户端的受众Lily的第一个广告位,在T0时刻,广告主M1欲向其投放汽车广告(标识为P1),出价为1元;广告主M2欲向其投放化妆品广告(标识为P2),出价为0.5元;广告主M3欲向其投放皮鞋广告(标识为P3),出价为0.1元;
而针对新闻客户端的受众Lucy的第一个广告位,在T0时刻,广告主M4欲向其投放保健品广告(标识为P4),出价为0.5元;广告主M5欲向其投放电子产品广告(标识为P5),出价 为0.5元。
则竞争行为数据与展示资源间的对应关系可示例性得包括下表所示内容:
Figure PCTCN2017115019-appb-000001
由于每一时刻可能均有海量的展示资源,为了减少处理负担,可采用抽取的方式获取展示资源及其对应的竞争行为数据。
具体的,可以一定的时间抽样频率抽取。例如,若第一指定时间段时长为一天,则在这一天中,每隔一小时抽取一次。
在广告竞价应用场景中,资源竞争参数阈值可包括系统底价,该底价是不向广告主公示的。广告主可通过自动竞拍系统向服务器101自动发送出价进行竞拍。当然,广告主需要事先设定每次广告曝光的出价。
上述出价等信息会以日志的方式存储,因此,在一个示例说明中,处理器1通过提取日志可获得第一竞争数据。
在301部分:处理器1根据第一竞争数据,得到第一最优资源竞争参数阈值。
在广告竞价应用场景下,上述第一最优资源竞争参数阈值可包括第一最优底价,即初定最优底价。不同的广告位可能对应不同的第一最优底价。
在获得下一个第一最优底价(或称为更新第一最优底价)之前,将一直使用本次获得的第一最优底价对参与展示资源的实时竞争的资源竞争参与者进行过滤。
举例来讲,假定在13:00时刻获得的第一最优底价为0.2元,在15:00更新第一最优底价为0.1元,在21:00又更新第一最优底价为0.3元。
则在13:00-15:00之间,采用0.2元过滤掉出价小于0.2元的广告主;在15:00-21:00之间,采用0.1元过滤掉出价小于0.1元的广告主,在21:00之后,采用0.3元过滤掉出价小于0.3元的广告主。
在一个示例中,可通过如下方式得到获取上述第一最优资源竞争参数阈值:
基于第一竞争数据进行展示资源竞争模拟,计算不同的资源竞争参数阈值下该展示位置对应的竞争收益值,然后,从上述不同的资源竞争参数阈值中确定出令竞争收益值最大化的第一最优资源竞争参数阈值。本发明的后续实施例中还将进行更为详细地介绍。
在302部分:处理器1多次获取上述展示位置的第二竞争数据,并基于第二竞争数据得到第二最优资源竞争参数阈值。
上述第二竞争数据可包括在第二指定时间段内参与竞争上述展示位置的资源竞争参与者的竞争行为数据。或者也可说,第二竞争数据表征了资源竞争参与者对在第二指定时间段的展示位置的竞争行为。其他相关内容请参见前述300部分的介绍,在此不作赘述。
需要说明的是,第二指定时间段的结束时刻为当前时刻。当前时刻为上述第二竞争数据的获取时刻。
而第一指定时间段的结束时刻则不为当前时刻,例如该第一指定时间段的结束时刻早于该第二指定时间段的结束时刻。假定当前时刻为13:00,则第一指定时间段的结束时刻要早于13:00。
在一示例中,可以在第二指定时间段内以固定时间间隔获取第二竞争数据。本领域技术人员可灵活设计时间间隔,例如,可以设计时间间隔为一秒或几秒,则每隔一秒或几秒获取一次第二竞争数据。再例如,可设计时间间隔为一分钟、一小时等等。
在另一个示例中,也可在第二指定时间段内随机多次获取第二竞争数据,并计算得到第二最优资源竞争参数阈值。
此外,本领域技术人员亦可灵活设计第二指定时间段的时长,例如,可以设计第二指定时间段的时长为一秒、几秒、一分钟、一小时等等。
第二指定时间段的时长可以小于第一指定时间段的时长。
第二最优资源竞争参数阈值的计算方式可以不同于第一最优资源竞争参数阈值的确定方式。
在一个示例中,可根据上述第二竞争数据,计算资源竞争参与者对应的质量因子的概率分布和概率密度,并根据质量因子概率分布和概率密度,计算出第二最优资源竞争参数阈值。本文后续还将详细介绍如何根据质量因子概率分布和概率密度计算第二最优资源竞争参数阈值。
其中,质量因子是一个综合数值,根据实际情况,出价高的广告主一般广告质量比较优质,吸引用户点击的概率较高,需要说明的是,出价高低与质量因子之间的关系并非线性,该质量因子可以是通过不同广告主所出价格进行拟合后得到的因子。更具体的,在广告竞价应用场景下,上述第二最优资源竞争参数阈值可包括第二最优底价,而质量因子的概率分布和概率密度可通过广告主的出价来得到。
例如,通过统计得到,在第二竞争数据中出价4元的广告主的数量为A,出价3元的广告主的数量为B,出价5元的广告主的数量为C,则可根据上述统计结果,得到出价不同的广告主的概率分布函数和概率密度函数,作为质量因子的概率分布和概率密度。
在303部分:处理器1判断本次计算得到的第二最优资源竞争参数阈值与上述第一最优资源竞争参数阈值的差异值是否大于差异阈值,若是,进入304部分,若否返回302部分。
差异值可有多种计算方式:例如,假定第二最优资源竞争参数阈值用Y1表示,第一最优资源竞争参数阈值用Y2表示,可取|Y2-Y1|作为差异值。
或者,将采用公式
Figure PCTCN2017115019-appb-000002
来计算差异值。
差异阈值可根据不同的实际场景进行灵活设计,在此不作赘述。
在304部分:获取上述展示位置的第三竞争数据。
上述第三竞争数据可包括在第三指定时间段内参与竞争上述展示位置的资源竞争参与者的竞争行为数据。或者也可说,第三竞争数据表征了资源竞争参与者对在第三指定时 间段的展示位置的竞争行为。
第三指定时间段的结束时刻可为上述第三竞争数据的获取时刻。
需要说明的是,第三指定时间段包括第二指定时间段的至少一部分。假定获取第二竞争数据的时刻为13:00,第三指定时间段的时长为30分钟,则第三指定时间段可为12:30-13:00,或者13:00-13:30,或者12:50-13:20等等。
本领域技术人员亦可灵活设计第三指定时间段的时长,例如,可以设计第三指定时间段的时长为一秒、几秒、一分钟、一小时、一天等等。
在305部分:根据上述第三竞争数据更新第一最优资源竞争参数阈值。
在本发明的另一些实施例中,在305部分根据上述第三竞争数据更新第一最优资源竞争参数阈值之后,还可以返回302部分。
在一个示例中,可根据第三竞争数据重新计算得到第一最优资源竞争参数阈值。也即,305部分与301部分相类似。
为便于理解本方案,可举个简单的例子:假定第一指定时间段时长为1小时,第三指定时间段时长为2小时。在13:00时刻获取了12:00-13:00间的竞争数据(第一竞争数据),得到了第一最优底价。将第一最优底价用至后续的广告竞拍中。并多次获取第二竞争数据计算第二最优底价。
假定在17:00时刻,发现第二最优底价与第一底价的差异值大于差异阈值了,则以当前时刻(17:00)为结束时刻,获取17:00之前两个小时内的竞争数据,也即15:00-17:00间的竞争数据(即第三竞争数据)。根据的第三竞争数据,重新计算第一最优底价,再将其用至后续的广告竞拍中。如此循环执行。
这样循环执行的原因是,广告主的行为是在动态发生变化的,当参与竞争的广告主有增减,或者广告主的出价有变化的时候,需要对第一最优资源竞争参数阈值进行相关修正。
需要说明的是,本发明实施例提供的资源竞争参数阈值的动态调整方法是针对每一广告位的,广告位与展示资源不同,展示资源还涵盖了受众ID和展示时刻。因此,对于每一广告位,均获取第一指定时间段内的第一竞争数据,根据第一竞争数据得到该广告位对应的第一最优资源竞争参数阈值。然后,多次获取该广告位的第二竞争数据,根据第二竞争数据计算出第二最优资源竞争参数阈值,并在该广告位对应的第二最优资源竞争参数阈值和第一最优资源竞争参数阈值的差异过大时,获取第三竞争数据,并更新该广告位对应的第一最优资源竞争参数阈值。
可见,在本发明实施例提供的调整方案中,使用展示资源的第一竞争数据得到第一最优资源竞争参数阈值(简称第一最优阈值)。多次获取展示资源的第二竞争数据,根据第二竞争数据得到第二最优资源竞争参数阈值(简称第二最优阈值)。在第二最优阈值和第一最优阈值差异过大的情况下,再使用展示资源的第三竞争数据得到更新后的第一最优资源竞争参数阈值。这样,可保证阈值始终随着实时情况进行自适性调整,实现了对阈值的优化。
下面,将以广告竞价为应用场景,对本发明的技术方案进行更详细的介绍。
请参见图4a和图4b,4a和图4b为本发明实施例提供的调整方法的另一种示例性示意图,由前述服务器101或调整装置的处理器1与其他设备交互实现。
在本实施例中,以广告位A为例进行说明。
上述交互流程包括:
在400部分:处理器1获取广告位A的第一竞争数据。
第一竞争数据的相关介绍请参见前述300部分,在此不作赘述。
上述第一竞争数据被划分为N个数据组,N为大于1的正整数。
例如,请参见图4b,分成了5个数据组。
其中,每一数据组可包括多个展示资源,参与竞争每一展示资源的资源竞争参与者的竞争行为数据。具体细节请参见前述实施例中的300部分,在此不作赘述。
在401部分:处理器1为N个数据组分配不同的资源竞争参数阈值(底价)。
例如,请参见图4b,处理器1为5个数据组分配了5个底价,分别为0.1、0.2、0.3、0.4、0.5。
在402部分:处理器1从上述N个数据组中确定出候选数据组。
所述候选数据组对应的资源竞争参数阈值包括:目标资源竞争参数阈值。
在一个示例中,可采用如下方式确定出第i个数据组是否为候选数组,i为小于N的正整数。
步骤A:基于第i个资源竞争参数阈值,计算第i个数据组的收益影响相关数据。
上述收益影响相关数据可包括预测的CTR(点击率)、受影响的广告主数(被过滤掉的广告主)、受影响的广告数等。
需要说明的是,不同的底价是会影响到CTR、受影响的广告主数和广告数等的。
例如,假定50个广告主中,有3个出价高于0.5,5个出价高于0.4小于0.5,6个出价高于0.3但小于0.4,17个出价高于0.2但小于0.3,10个出价高于0.1但小于0.2,9个出价低于0.1。如果底价为0.1,则受影响的广告主数为9,但如果底价为0.2,则受影响的广告主数为19。
同时底价越高,被过滤掉的广告主和广告数越多,则曝光的广告的种类则相对越单一,这样,可能导致受众的点击率不高。
步骤B:判断上述收益影响相关数据是否符合约束条件。
可设置CTR的最小阈值、受影响的广告主数最大阈值以及受影响的广告数最大阈值,如果预测的CTR大于CTR的最小阈值、预测的受影响的广告主数小于受影响的广告主数最大阈值,并且预测的受影响的广告数小于受影响的广告数最大阈值,则认为符合约束条件。
步骤C:若判定上述收益影响相关数据符合上述约束条件,将上述第i个数据组作为候选数据组。
对于不符合上述约束条件的数据组,将不会进行403部分的操作以及后续操作。
在403部分:处理器1基于每一候选数据组对应的资源竞争参数阈值,计算该候选数据组的整体竞争收益值。
在一个示例中,可基于候选数据组对应的目标资源竞争参数阈值,计算各展示资源对应的竞争收益值,然后,对所有展示资源对应的竞争收益值进行求和运算,得到该候选数据组的整体竞争收益值。
举例来讲,假定某候选数据组有100个展示资源,可分别计算这100个展示资源的预期收益值,再进行求和运算,即可得到整组的整体竞争收益值。
更具体的,计算任一展示资源对应的竞争收益值可包括如下步骤:
步骤A1:对竞争上述任一展示资源的资源竞争参与者的竞争行为数据进行过滤,得到竞争行为数据子集。
所述候选数据组对应的资源竞争参数阈值包括:目标资源竞争参数阈值;上述竞争行为数据子集包括资源竞争参数大于等于目标资源竞争参数阈值的资源竞争参与者的竞争行为数据。
举例来讲,假定数据组2为候选数据组,为其分配的底价为0.2,在该组中,有50个广告主竞争展示资源1,其中10个广告主的出价低于0.2,则过滤掉这10个出低于0.2的广告主。剩下40个广告主的竞争行为数据构成一个竞争行为数据子集。
步骤B1:计算竞争行为数据子集中每一媒体文件(广告)的千次展示广告收入(ecpm)。
步骤C1:对ecpm最大的媒体文件计算预期收益值(或称为收益值),将计算出的预期收益值作为上述任一展示资源对应的竞争收益值。
在一个示例中,可按照ecpm由大至小对竞争行为数据子集中的广告进行排序。沿用前例,假定竞争行为数据子集中包括40个广告主的竞争行为数据。与这40个广告主相关联的广告共50个,则计算这50个广告的ecpm,并按ecpm由大到小对广告进行排序。
假定第4个广告排在首位,则计算第4个广告的预期收益值作为展示资源1的竞争收益值。
需要说明的是,对于不同的计费方式可有不同的预期收益计算逻辑,从而得到不同的预期收益值。
例如对于广义二阶价格(Generalized Second Price,GSP)模式,可采用如下计算逻辑来计算其预期收益值:
(1),在GSP模式中,若排在首位的广告的出价(bid)大于底价,并且排在首位的广告的扣费也大于底价,则根据扣费来计算预期收益值。
扣费的计算方式是:排第二位的广告的出价+固定值(例如0.1);
举例来讲,对于展示资源1,排首位的广告的出价为1元,扣费为0.8元,分配的底价为0.2,则根据0.8来计算预期收益值。
(2),若排在首位的广告的出价大于>底价,但是,排在首位的广告的扣费小于底价,则按照底价进行扣费,计算预期收益值。
举例来讲,对于展示资源1,排首位的广告的出价为1元,扣费为0.18元,分配的底价为0.2,则根据0.18来计算预期收益值。
在404部分:处理器1将最大的整体竞争收益值所对应的资源竞争参数阈值,作为第一最优资源竞争参数阈值。
假定5个数据组中确定出了4个候选数据组A-D,其中,候选数据组D的整体竞争收益值最大,而候选数据组D对应的资源竞争参数阈值为0.4,则将0.4作为第一最优资源竞争参数阈值。
此外,在本发明其他实施例中,也可直接计算各数据组的整体竞争收益值,并从中选取最大的整体竞争收益值所对应的资源竞争参数阈值,作为上述第一最优资源竞争参数阈值。
在405部分:处理器1多次获取第二竞争数据,并基于第二竞争数据得到第二最优资源竞争参数阈值。
前述提及了,可根据质量因子概率分布和概率密度,计算出第二最优资源竞争参数阈值。
更具体的,可根据下述公式来计算第二最优资源竞争参数阈值:
Figure PCTCN2017115019-appb-000003
其中,p*表示第二最优资源竞争参数阈值,S*表示(令收益最大化的)质量因子S的取值,e表示点击率。
在一个示例中,在广告竞价应用场景下,e为广告位(例如广告位A)在第二指定时间段的平均点击率或广告位的当前点击率。
而S*是令
Figure PCTCN2017115019-appb-000004
等于0的数值。其中S表示质量因子,F(S)和f(S)即为前述的质量因子的概率分布函数和概率密度函数。
当然,在其他实施例中,也可采用其他的方式计算第二最优资源竞争参数阈值,在此不作赘述。
在406部分:处理器1判断本次计算得到的第二最优资源竞争参数阈值与第一最优资源竞争参数阈值的差异值是否大于差异阈值,若是,进入407部分,若否返回405部分。
406部分与303部分相类似,在此不作赘述。
在407部分:获取上述展示位置的第三竞争数据,根据上述第三竞争数据更新第一最优资源竞争参数阈值,返回405部分。
第三竞争数据相关介绍请参见前述304部分的介绍,在此不作赘述。
在更新第一最优资源竞争参数阈值时,可采用类似于401-404部分的方式。
图5示出了上述实施例中所涉及的动态调整装置的另一种可能的结构示意图,包括:
第一数据分析模块501,用于获取展示位置的第一竞争数据,根据所述第一竞争数据得到第一最优资源竞争参数阈值,所述第一竞争数据包括在第一指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争数据;
第二数据分析模块502,用于:
获取所述展示位置的第二竞争数据,以及所述第二竞争数据计算得到第二最优资源竞争参数阈值,所述第二竞争数据包括在第二指定时间段参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
在第二最优资源竞争参数阈值与第一最优资源竞争参数阈值间的差异值大于差异阈值时,获取所述展示位置的第三竞争数据,所述第三竞争数据包括在第三指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
根据所述第三竞争数据更新第一最优资源竞争参数阈值。
其中,所述第二指定时间段的结束时刻为当前时刻,所述第一指定时间段的结束时刻早于所述当前时刻,所述第三指定时间段包括所述第二指定时间段的至少一部分时间。
进一步的,第二数据分析模块502,根据所述第三竞争数据更新第一最优资源竞争参数阈值之后,还可以用于:返回所述多次获取所述展示位置的第二竞争数据,并基于所述第二竞争数据计算得到第二最优资源竞争参数阈值的操作。
在本发明的一些实施例中,在根据第一竞争数据得到第一最优资源竞争参数阈值方面,所述第一数据分析模块501具体用于:
基于所述第一竞争数据,计算在不同的资源竞争参数阈值的情况下所述展示位置对应的竞争收益值;
从所述不同的资源竞争参数阈值中确定出令竞争收益值最大化的资源竞争参数阈值作为所述第一最优资源竞争参数阈值。
在本发明的一些实施例中,在所述计算不同的资源竞争参数阈值下,所述展示位置对应的竞争收益值之前,所述第一数据分析模块501还用于:
将所述第一竞争数据划分为N个数据组,所述N为大于1的正整数。
在本发明的一些实施例中,在所述基于所述第一竞争数据,计算不同的资源竞争参数阈值下所述展示位置对应的竞争收益值方面,所述第一数据分析模块501具体用于:
为所述N个数据组分配不同的资源竞争参数阈值;
从所述N个数据组中确定出候选数据组;
基于每一候选数据组分别对应的资源竞争参数阈值,计算该候选数据组的整体竞争收益值;
在所述从不同的资源竞争参数阈值中确定出令竞争收益值最大化的资源竞争参数阈值作为所述第一最优资源竞争参数阈值方面,所述第一数据分析模块具体用于:
将最大的整体竞争收益值所对应的资源竞争参数阈值,作为所述第一最优资源竞争参数阈值。
在本发明的一些实施例中,在所述从所述N个数据组中确定出候选数据组方面,所述第一数据分析模块501用于:
基于第i个资源竞争参数阈值,计算第i个数据组的收益影响相关数据,所述i为小于N的正整数;
判断所述第i个数据组的收益影响相关数据是否符合约束条件;
若判定所述第i个数据组的收益影响相关数据符合所述约束条件,将所述第i个数据组作为候选数据组。
在本发明的一些实施例中,在所述基于获取到的所述第二竞争数据计算得到第二最优资源竞争参数阈值方面,所述第二数据分析模块502具体用于:
根据所述第二竞争数据计算资源竞争参与者对应的质量因子的概率分布和概率密度;
根据所述质量因子概率分布和概率密度,计算出当前第二最优资源竞争参数阈值。
具体细节请参见本文前述记载,在此不作赘述。
其中,第一数据分析模块501可用于执行图3所示实施例的300-301部分;此外,还可执行图4a和图4b所示实施例的400-404部分。
第二数据分析模块502可用于执行图3所示实施例的302-305部分;此外,还可执行图4a和图4b所示实施例的405-407部分。
图6是本发明实施例提供的一种服务器结构示意图,该服务器1100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上中央处理器(central processing units,CPU)1122(例如,一个或一个以上处理器)和存储器1132,一个或一个以上存储应用程序1142或数据1144的存储介质1130(例如一个或一个以上海量存储设备)。其中,存储器1132和存储介质1130可以是短暂存储或持久存储。存储在存储介质1130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1122可以设置为与存储介质1130通信,在服务器1100上执行存储介质1130中的一系列指令操作。
服务器1100还可以包括一个或一个以上电源1126,一个或一个以上有线或无线网络接口1150,一个或一个以上输入输出接口1158,和/或,一个或一个以上操作系统1141,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6所示的服务器结构。
结合本发明公开内容所描述的方法或者算法的步骤可以硬件的方式来实现,也可以是由处理器执行软件指令的方式来实现。软件指令可以由相应的软件模块组成,软件模块可以被存放于RAM存储器、闪存、ROM存储器、EPROM存储器、EEPROM存储器、寄存器、硬盘、移动硬盘、CD-ROM或者本领域熟知的任何其它形式的存储介质中。一种示例性的存储介质耦合至处理器,从而使处理器能够从该存储介质读取信息,且可向该存储介质写入信息。当然,存储介质也可以是处理器的组成部分。处理器和存储介质可以位于ASIC中。另外,该ASIC可以位于用户设备中。当然,处理器和存储介质也可以作为分立组件存在于用户设备中。
本领域技术人员应该可以意识到,在上述一个或多个示例中,本发明所描述的功能可以用硬件、软件、固件或它们的任意组合来实现。当使用软件实现时,可以将这些功能存储在计算机可读介质中或者作为计算机可读介质上的一个或多个指令或代码进行传输。计算机可读介质包括计算机存储介质和通信介质,其中通信介质包括便于从一个地方向另一个地方传送计算机程序的任何介质。存储介质可以是通用或专用计算机能够存取的任何可用介质。
以上所述的具体实施方式,对本发明的目的、技术方案和有益效果进行了进一步详细说明,所应理解的是,以上所述仅为本发明的具体实施方式而已,并不用于限定本发明的保护范围,凡在本发明的技术方案的基础之上,所做的任何修改、等同替换、改进等,均应包括在本发明的保护范围之内。

Claims (20)

  1. 一种资源竞争中资源竞争参数阈值的动态调整方法,其特征在于,包括:
    获取展示位置的第一竞争数据,根据第一竞争数据得到第一最优资源竞争参数阈值;所述第一竞争数据包括在第一指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    多次获取所述展示位置的第二竞争数据,并基于所述第二竞争数据计算得到第二最优资源竞争参数阈值;所述第二竞争数据包括在第二指定时间段参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    在第二最优资源竞争参数阈值与第一最优资源竞争参数阈值间的差异值大于差异阈值时,获取所述展示位置的第三竞争数据;所述第三竞争数据包括在第三指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    根据所述第三竞争数据更新所述第一最优资源竞争参数阈值,返回所述多次获取所述展示位置的第二竞争数据,并基于所述第二竞争数据计算得到第二最优资源竞争参数阈值的操作;
    其中,所述第二指定时间段的结束时刻为当前时刻,所述第一指定时间段的结束时刻不为所述当前时刻,所述第三指定时间段包括所述第二指定时间段的至少一部分。
  2. 一种资源竞争参数阈值的动态调整方法,其特征在于,所述方法应用于服务器,所述方法包括:
    获取展示位置的第一竞争数据,根据所述第一竞争数据得到第一最优资源竞争参数阈值,所述第一竞争数据包括在第一指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    获取所述展示位置的第二竞争数据,以及基于所述第二竞争数据计算得到第二最优资源竞争参数阈值,所述第二竞争数据包括在第二指定时间段参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    在所述第二最优资源竞争参数阈值与所述第一最优资源竞争参数阈值间的差异值大于差异阈值时,获取所述展示位置的第三竞争数据,所述第三竞争数据包括在第三指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    根据所述第三竞争数据更新所述第一最优资源竞争参数阈值;
    其中,所述第二指定时间段的结束时刻为当前时刻,所述第一指定时间段的结束时刻早于所述当前时刻,所述第三指定时间段包括所述第二指定时间段的至少一部分时间。
  3. 如权利要求2所述的方法,其特征在于,所述根据所述第一竞争数据得到第一最优资源竞争参数阈值,包括:
    基于所述第一竞争数据,计算在不同的资源竞争参数阈值的情况下所述展示位置对应的竞争收益值;
    从所述不同的资源竞争参数阈值中确定出令所述竞争收益值最大化的资源竞争参数阈值作为所述第一最优资源竞争参数阈值。
  4. 如权利要求3所述的方法,其特征在于,所述计算在不同的资源竞争参数阈值的情况下所述展示位置对应的竞争收益值之前,所述方法还包括:
    将所述第一竞争数据划分为N个数据组,所述N为大于1的正整数。
  5. 如权利要求4所述的方法,其特征在于,
    所述基于所述第一竞争数据,计算在不同的资源竞争参数阈值的情况下所述展示位置对应的竞争收益值,包括:
    为所述N个数据组分配不同的资源竞争参数阈值;
    从所述N个数据组中确定出候选数据组;
    基于每一候选数据组分别对应的资源竞争参数阈值,计算该候选数据组的整体竞争收益值;
    所述从所述不同的资源竞争参数阈值中确定出令所述竞争收益值最大化的资源竞争参数阈值作为所述第一最优资源竞争参数阈值,包括:
    将最大的整体竞争收益值所对应的资源竞争参数阈值,作为所述第一最优资源竞争参数阈值。
  6. 如权利要求5所述的方法,其特征在于,所述从所述N个数据组中确定出候选数据组包括:
    基于第i个资源竞争参数阈值,计算第i个数据组的收益影响相关数据,所述i为小于N的正整数;
    若所述第i个数据组的收益影响相关数据符合所述约束条件,将所述第i个数据组作为候选数据组。
  7. 如权利要求6所述的方法,其特征在于,
    所述展示位置在所述第一指定时间段内关联有多个展示资源;
    所述参与竞争所述展示位置的资源竞争参与者的竞争行为数据包括:参与竞争所述多个展示资源的资源竞争参与者的竞争行为数据;
    每一所述资源竞争参与者的竞争行为数据包括:所述资源竞争参与者的用户标识、与所述用户标识相关联的媒体文件标识、资源竞争参数。
  8. 如权利要求7所述的方法,其特征在于,
    所述基于每一候选数据组分别对应的资源竞争参数阈值,计算该候选数据组的整体竞争收益值,包括:
    基于所述候选数据组对应的资源竞争参数阈值,计算所述多个展示资源对应的竞争收益值;
    对所有展示资源对应的竞争收益值进行求和运算,得到所述候选数据组的整体竞争收益值。
  9. 如权利要求8所述的方法,其特征在,
    所述候选数据组对应的资源竞争参数阈值包括:目标资源竞争参数阈值;
    所述计算所述多个展示资源对应的竞争收益值包括:
    对竞争所述任一展示资源的资源竞争参与者的竞争行为数据进行过滤,得到竞争行为数据子集,所述竞争行为数据子集包括:资源竞争参数大于或等于所述目标资源竞争参数阈值的资源竞争参与者的竞争行为数据;
    计算所述竞争行为数据子集中每一媒体文件的千次展示广告收入ecpm;
    对ecpm最大的媒体文件计算预期收益值,将计算出的所述预期收益值作为所述任一展示资源对应的竞争收益值。
  10. 如权利要求4所述的方法,其特征在于,
    所述基于所述第一竞争数据,计算在不同的资源竞争参数阈值的情况下所述展示位置对应的竞争收益值,包括:
    为所述N个数据组分配不同的资源竞争参数阈值;
    基于第i个资源竞争参数阈值,计算第i个数据组的整体竞争收益值,所述i为小于N的正整数;
    所述从所述不同的资源竞争参数阈值中确定出令所述竞争收益值最大化的资源竞争参数阈值作为所述第一最优资源竞争参数阈值,包括:
    将最大的整体竞争收益值所对应的资源竞争参数阈值,作为所述第一最优资源竞争参数阈值。
  11. 如权利要求2所述的方法,其特征在于,所述基于所述第二竞争数据计算得到第二最优资源竞争参数阈值,包括:
    根据所述第二竞争数据计算资源竞争参与者对应的质量因子的概率分布和概率密度;
    根据所述概率分布和概率密度,计算出所述第二最优资源竞争参数阈值。
  12. 一种资源竞争中资源竞争参数阈值的动态调整装置,其特征在于,包括:
    第一数据分析模块,用于获取展示位置的第一竞争数据,根据获取到的第一竞争数据得到第一最优资源竞争参数阈值;所述第一竞争数据包括:在第一指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    第二数据分析模块,用于:
    多次获取所述展示位置的第二竞争数据,基于所述第二竞争数据计算得到第二最优资源竞争参数阈值;所述第二竞争数据包括在第二指定时间段参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    在第二最优资源竞争参数阈值与第一最优资源竞争参数阈值间的差异值大于差异阈值时,获取所述展示位置的第三竞争数据;所述第三竞争数据包括在第三指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    根据所述第三竞争数据更新第一最优资源竞争参数阈值,返回所述多次获取所述展示位置的第二竞争数据,并基于所述第二竞争数据计算得到第二最优资源竞争参数阈值的操作;
    其中,所述第二指定时间段的结束时刻为当前时刻,所述第一指定时间段的结束时刻不为所述当前时刻,所述第三指定时间段包括所述第二指定时间段的至少一部分。
  13. 一种资源竞争参数阈值的动态调整装置,其特征在于,包括:
    第一数据分析模块,用于获取展示位置的第一竞争数据,根据所述第一竞争数据得到第一最优资源竞争参数阈值,所述第一竞争数据包括:在第一指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    第二数据分析模块,用于:
    获取所述展示位置的第二竞争数据,以及基于所述第二竞争数据计算得到第二最优资 源竞争参数阈值,所述第二竞争数据包括在第二指定时间段参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    在所述第二最优资源竞争参数阈值与所述第一最优资源竞争参数阈值间的差异值大于差异阈值时,获取所述展示位置的第三竞争数据,所述第三竞争数据包括在第三指定时间段内参与竞争所述展示位置的资源竞争参与者的竞争行为数据;
    根据所述第三竞争数据更新第一最优资源竞争参数阈值;
    其中,所述第二指定时间段的结束时刻为当前时刻,所述第一指定时间段的结束时刻早于所述当前时刻,所述第三指定时间段包括所述第二指定时间段的至少一部分时间。
  14. 如权利要求13所述的装置,其特征在于,所述第一数据分析模块用于:
    基于所述第一竞争数据,计算在不同的资源竞争参数阈值的情况下所述展示位置对应的竞争收益值;
    从所述不同的资源竞争参数阈值中确定出令竞争收益值最大化的资源竞争参数阈值作为所述第一最优资源竞争参数阈值。
  15. 如权利要求14所述的装置,其特征在于,所述第一数据分析模块还用于:
    所述计算在不同的资源竞争参数阈值的情况下所述展示位置对应的竞争收益值之前,将所述第一竞争数据划分为N个数据组,所述N为大于1的正整数。
  16. 如权利要求15所述的装置,其特征在于,所述第一数据分析模块用于:
    为所述N个数据组分配不同的资源竞争参数阈值;
    从所述N个数据组中确定出候选数据组;
    基于每一候选数据组分别对应的资源竞争参数阈值,计算该候选数据组的整体竞争收益值;
    将最大的整体竞争收益值所对应的资源竞争参数阈值,作为所述第一最优资源竞争参数阈值。
  17. 如权利要求16所述的装置,其特征在于,所述第一数据分析模块用于:
    基于第i个资源竞争参数阈值,计算第i个数据组的收益影响相关数据,所述i为小于N的正整数;
    若所述第i个数据组的收益影响相关数据符合所述约束条件,将所述第i个数据组作为候选数据组。
  18. 如权利要求13所述的装置,其特征在于,所述第二数据分析模块用于:
    根据所述第二竞争数据计算资源竞争参与者对应的质量因子的概率分布和概率密度;
    根据所述质量因子概率分布和概率密度,计算出当前第二最优资源竞争参数阈值。
  19. 一种计算机可读存储介质,包括指令,当其在计算机上运行时,使得计算机执行如权利要求1至11任意一项所述的方法。
  20. 一种服务器,其特征在于,所述服务器包括:处理器和存储器;
    所述存储器,用于存储指令;
    所述处理器,用于执行所述存储器中的所述指令,执行如权利要求1至11中任一项所述的方法。
PCT/CN2017/115019 2016-12-29 2017-12-07 资源竞争参数阈值的动态调整方法及装置和服务器 WO2018121209A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/384,203 US11277352B2 (en) 2016-12-29 2019-04-15 Method, apparatus, and server for dynamically adjusting resource competition thresholds

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201611245646.1A CN107067142A (zh) 2016-12-29 2016-12-29 资源竞争中资源竞争参数阈值的动态调整方法及装置
CN201611245646.1 2016-12-29

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/384,203 Continuation US11277352B2 (en) 2016-12-29 2019-04-15 Method, apparatus, and server for dynamically adjusting resource competition thresholds

Publications (1)

Publication Number Publication Date
WO2018121209A1 true WO2018121209A1 (zh) 2018-07-05

Family

ID=59623286

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/115019 WO2018121209A1 (zh) 2016-12-29 2017-12-07 资源竞争参数阈值的动态调整方法及装置和服务器

Country Status (3)

Country Link
US (1) US11277352B2 (zh)
CN (1) CN107067142A (zh)
WO (1) WO2018121209A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107067142A (zh) * 2016-12-29 2017-08-18 腾讯科技(深圳)有限公司 资源竞争中资源竞争参数阈值的动态调整方法及装置
CN109033343B (zh) * 2018-07-24 2023-01-03 百度在线网络技术(北京)有限公司 用于生成信息的方法和装置
CN111489182B (zh) * 2019-01-28 2023-06-23 腾讯科技(深圳)有限公司 信息展示的竞争资源控制方法、装置、计算机设备及存储介质
CN113553485B (zh) * 2021-07-29 2024-02-06 北京达佳互联信息技术有限公司 多媒体资源的展示方法、装置、设备及存储介质
CN113949678B (zh) * 2021-09-15 2023-09-01 北京三快在线科技有限公司 流量控制方法、装置、电子设备及计算机可读存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150088665A1 (en) * 2013-09-26 2015-03-26 Aol Advertising Inc. Computerized systems and methods related to controlled content optimization
CN105338649A (zh) * 2014-08-01 2016-02-17 展讯通信(上海)有限公司 资源竞争的控制方法及装置
CN106096999A (zh) * 2016-06-02 2016-11-09 腾讯科技(深圳)有限公司 媒体文件的投放方法和装置
CN107067142A (zh) * 2016-12-29 2017-08-18 腾讯科技(深圳)有限公司 资源竞争中资源竞争参数阈值的动态调整方法及装置

Family Cites Families (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060122879A1 (en) * 2004-12-07 2006-06-08 O'kelley Brian Method and system for pricing electronic advertisements
US7895116B2 (en) * 2007-07-25 2011-02-22 Mukesh Chatter Seller automated engine architecture and methodology for optimized pricing strategies in automated real-time iterative reverse auctions over the internet and the like for the purchase and sale of goods and services
US8522270B2 (en) * 2007-10-16 2013-08-27 Mukesh Chatter System for and method of automatic optimizing quantitative business objectives of sellers (advertisers) with synergistic pricing, promotions and advertisements, while simultaneously minimizing expenditure discovery and optimizing allocation of advertising channels that optimize such objectives
US10157387B2 (en) * 2008-12-17 2018-12-18 OpenX Technologies, Inc. Method and system for establishing a reserve price for a publisher's ad space inventory offered via a real-time bidding market
US8478643B1 (en) * 2009-10-12 2013-07-02 Google Inc. Optimal pricing and advertisement slot allocation
US9289063B2 (en) * 2010-02-18 2016-03-22 Lamplight Development, Llc Convertible headboard table apparatus and method of use
US20130097028A1 (en) * 2011-10-14 2013-04-18 Microsoft Corporation Dynamic Floor Pricing for Managing Exchange Monetization
US9991998B2 (en) * 2011-11-25 2018-06-05 Avago Technologies General Ip (Singapore) Pte. Ltd. Ratio resource sharing and contention scheme for device-to-device communication in white space spectrum bands
EP2813072A4 (en) * 2012-02-07 2015-09-16 Visible World Inc DYNAMIC CONTENT AND OPTIMIZATION
US10853848B2 (en) * 2012-08-20 2020-12-01 OpenX Technologies, Inc. System and methods for generating dynamic market pricing for use in real-time auctions
US20140256326A1 (en) * 2013-03-06 2014-09-11 Motorola Mobility Llc Handoff prioritization in multi-rab calls
WO2014160854A1 (en) * 2013-03-27 2014-10-02 Sandholm Tuomas W Method and system for media advertising campaign sales and allocation
US20140337120A1 (en) * 2013-05-08 2014-11-13 Radiate Media Holding Company Integrating media analytics to configure an advertising engine
US10366413B2 (en) * 2013-06-25 2019-07-30 Microsoft Technology Licensing, Llc Sponsored online content management using query clusters
US10134053B2 (en) * 2013-11-19 2018-11-20 Excalibur Ip, Llc User engagement-based contextually-dependent automated pricing for non-guaranteed delivery
US20150332349A1 (en) * 2013-12-18 2015-11-19 MaxPoint Interactive, Inc. System and Method for Controlling Purchasing Pace in a Real-Time Bidding Environment Using Proportional-Integral-Derivative (PID) Control
US20150178790A1 (en) * 2013-12-20 2015-06-25 Yahoo! Inc. User Engagement-Based Dynamic Reserve Price for Non-Guaranteed Delivery Advertising Auction
US10565626B2 (en) * 2015-03-18 2020-02-18 Xandr Inc. Methods and systems for dynamic auction floors
US10565613B2 (en) * 2015-03-27 2020-02-18 Facebook, Inc. Simulating advertising campaigns
EP3125564A1 (en) * 2015-07-27 2017-02-01 Palantir Technologies, Inc. Computer-based optimized insertion of non-program media items in media programs

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150088665A1 (en) * 2013-09-26 2015-03-26 Aol Advertising Inc. Computerized systems and methods related to controlled content optimization
CN105338649A (zh) * 2014-08-01 2016-02-17 展讯通信(上海)有限公司 资源竞争的控制方法及装置
CN106096999A (zh) * 2016-06-02 2016-11-09 腾讯科技(深圳)有限公司 媒体文件的投放方法和装置
CN107067142A (zh) * 2016-12-29 2017-08-18 腾讯科技(深圳)有限公司 资源竞争中资源竞争参数阈值的动态调整方法及装置

Also Published As

Publication number Publication date
CN107067142A (zh) 2017-08-18
US20190280987A1 (en) 2019-09-12
US11277352B2 (en) 2022-03-15

Similar Documents

Publication Publication Date Title
WO2018121209A1 (zh) 资源竞争参数阈值的动态调整方法及装置和服务器
US9147200B2 (en) Frequency capping of content across multiple devices
JP6707136B2 (ja) プロモーション情報のための配信方法、装置及びシステム
US20160275545A1 (en) Methods and systems for using device connectivity information in on-line advertising
WO2018103622A1 (zh) 信息投放控制方法、装置及存储介质
WO2017121258A1 (zh) 信息投放方法、装置、服务器和存储介质
AU2017261494A1 (en) Exchange server method and system
US8554621B2 (en) Impression-trend techniques for providing a display advertising supply forecast
CN105608125B (zh) 一种信息处理方法及服务器
CN107111654A (zh) 基于事件的内容分发
US11107130B1 (en) Content offers based on social influences
US10694225B2 (en) Customizing supplemental content delivery
CN111026969B (zh) 一种内容推荐方法、装置及存储介质和服务器
US20180165717A1 (en) Keyword Bidding based on Search Traffic on Online Social Networks
US9858589B2 (en) Measuring search lift resulted by online advertisement
US10672040B1 (en) Ordering of fill requests to send to online ad networks to fill an ad unit in an electronic publisher application with an online advertising impression
CN115375339A (zh) 多媒体信息推荐方法、装置和设备及计算机存储介质
US10607254B1 (en) Attribution modeling using withheld or near impressions
US11080755B1 (en) Native advertisements
CN111815368A (zh) 广告推送方法、装置、终端及存储介质
US10497031B1 (en) Conditional bids in an auction
JP5699233B1 (ja) 生成装置、生成方法および生成プログラム
WO2015096791A1 (en) Method and system for releasing media content
JP6956564B2 (ja) 情報処理装置、情報処理方法、およびプログラム
US20130268354A1 (en) Selecting Content Items for Display in a Content Stream

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: 17889330

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17889330

Country of ref document: EP

Kind code of ref document: A1