WO2018133566A1 - 推广信息新鲜度确定方法及装置、电子设备及存储介质 - Google Patents

推广信息新鲜度确定方法及装置、电子设备及存储介质 Download PDF

Info

Publication number
WO2018133566A1
WO2018133566A1 PCT/CN2017/114633 CN2017114633W WO2018133566A1 WO 2018133566 A1 WO2018133566 A1 WO 2018133566A1 CN 2017114633 W CN2017114633 W CN 2017114633W WO 2018133566 A1 WO2018133566 A1 WO 2018133566A1
Authority
WO
WIPO (PCT)
Prior art keywords
memory
promotion information
freshness
user
determining
Prior art date
Application number
PCT/CN2017/114633
Other languages
English (en)
French (fr)
Inventor
周蓝珺
陈功
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 WO2018133566A1 publication Critical patent/WO2018133566A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/35—Clustering; Classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation

Definitions

  • the present invention relates to the field of information technology, and in particular, to a method and device for determining the freshness of information, an electronic device, and a computer storage medium.
  • Promotion information is information content that is actively pushed to the user for viewing or viewing, rather than content that the user actively requests to view.
  • Common promotion information may include advertisements, government announcements, and event introduction information.
  • the embodiments of the present invention are expected to provide a method and device for determining the freshness of the promotion information, an electronic device, and a computer storage medium, so as to solve the problem that the promotion effect of the promotion information is poor and/or the user's experience of viewing the promotion information is poor.
  • the first aspect of the embodiments of the present invention provides a method for determining the freshness of the promotion information, which can be applied to an information processing platform, including:
  • the second promotion information is information currently to be promoted
  • the first promotion information is the promotion information that the user has viewed
  • the memory strength is used to indicate that the user memorizes the strength of the first promotion information
  • a second aspect of the embodiments of the present invention provides a freshness determining apparatus for promoting information, including:
  • the obtaining unit is configured to obtain historical data of the user viewing the promotion information
  • a first determining unit configured to determine, according to the historical data, a memory strength of the first promotion information, where the first promotion information and the second promotion information are associated in a predetermined dimension; the second promotion information is current The information to be promoted; the first promotion information is promotion information that has been viewed by the user; and the memory strength is used to indicate that the user memorizes the strength of the first promotion information;
  • the second determining unit is configured to determine freshness of the second promotion information in the predetermined dimension according to the memory strength, wherein the freshness is used to indicate a degree of strangeness of the second promotion information.
  • a third aspect of the embodiments of the present invention provides an electronic device, including: a memory and a processor;
  • the memory is configured to store computer executable instructions
  • the processor is coupled to the memory and configured to read and execute the computer executable instructions to implement a freshness determination method of the promotion information provided by the foregoing technical solution.
  • a fourth aspect of the embodiments of the present invention provides a computer storage medium, where the computer storage medium
  • the computer executable instructions are stored in the quality, and the computer executable instructions are used to execute the freshness determination method of the promotion information provided by the above technical solution.
  • the method and device for determining the freshness of the promotion information, the electronic device and the storage medium provided by the embodiment of the present invention obtain the historical data of the user viewing the promotion information, and determine the relevance of the second promotion information to be promoted according to the historical data.
  • the memory intensity of the first promotion information, and according to the memory intensity and the attenuation factor, the freshness of the second promotion information in a predetermined dimension is obtained, and the freshness can be used as a parameter for filtering the second promotion information to the user, and the freshness is adopted.
  • the calculation can reduce the same or the same type of information to the same user, and the promotion effect is poor and the user satisfaction is low, so that the promotion information is sent to the user or is not filtered. Promotion effect and user satisfaction.
  • FIG. 1 is a schematic flowchart diagram of a method for determining freshness of a first promotion information according to an embodiment of the present invention
  • FIG. 2 is a schematic structural diagram of a freshness determining apparatus for providing first promotion information according to an embodiment of the present invention
  • FIG. 3 is a schematic flowchart diagram of a method for determining freshness of a second promotion information according to an embodiment of the present invention
  • FIG. 4 is a schematic diagram of a correspondence relationship between a predetermined dimension and a memory content according to an embodiment of the present invention
  • FIG. 5 is a schematic structural diagram of a freshness determining apparatus for providing second promotion information according to an embodiment of the present invention
  • FIG. 6 is a schematic structural diagram of a promotion information system according to an embodiment of the present invention.
  • this embodiment provides a method for determining freshness of promotion information, including:
  • Step S110 Acquire historical data of the user viewing the promotion information
  • Step S120 Determine, according to the historical data, a memory strength of the first promotion information, where the first promotion information and the second promotion information are associated in a predetermined dimension; and the second promotion information is information currently to be promoted.
  • the first promotion information is promotion information that the user has viewed; the memory strength is used to indicate that the user memorizes the strength of the first promotion information;
  • Step S130 determining the freshness of the second promotion information in the predetermined dimension according to the memory strength, wherein the freshness is used to indicate the degree of strangeness of the second promotion information, which is current with the user.
  • the depth of memory of a certain information is negatively correlated.
  • the freshness determination method described in this embodiment may be an information processing method applied to various electronic devices having information processing functions, for example, may be applied to an advertisement server, an information search server, etc., including one or more servers.
  • the information processing method in the information processing platform that is, the steps S110 to S130 can be performed by an electronic device, for example, an advertisement server or a push server in the promotion platform.
  • the promotional information may include information such as advertisements, various types of announcements or events, and the advertisements herein may include commercial advertisements and non-commercial advertisements.
  • the non-commercial advertisements may generally include charity advertisements and the like.
  • the freshness may be a basis for determining whether to push the second promotion information to the user, for example, when the value of the freshness is within a specified range, sending the second promotion information to the user. Otherwise, the second promotion information is not sent.
  • the freshness may be a parameter related to the receiving status of the first promotion information received by the client where the user is located.
  • the receiving status information may include parameters such as the number of times of receiving the first promotion information, the frequency of reception, and/or the time of receiving.
  • the freshness may be display status information of the first promotion information displayed by the client held by the user, where the display status information may include: the number of times of display, Display information such as time and/or display frequency.
  • the historical data is historical data for the user to view the promotion information
  • the historical data may include: the information name of the promotion information viewed by the user, the promotion object of the promotion information, the source, the material involved in the promotion information, the viewing time, and the viewing time.
  • Various parameters such as the number of views.
  • the material may include: various data constituting the promotion information, such as a picture, text, and/or video used in the promotion information.
  • the historical data may further include: a duration of display of the first promotion information on the user terminal, a number of displays, a duration of the user's operation of viewing the user information in the first promotion information, and/or a display time distance from the current time.
  • the user operation may include: a forwarding operation, a click operation, a favorite operation of collecting to a favorite directory, a recommended operation recommended to a friend, or a copy operation, and other user operations other than simple viewing, such operations may be referred to as viewing operations. Scheduled operation.
  • the memory strength when determining the memory strength, is determined not only according to whether the user has the parameters such as viewing, viewing times, viewing time, etc., but also according to a predetermined operation other than the above-mentioned viewing operation accompanying the viewing process. If a predetermined operation other than the above viewing operation occurs, based on the memory strength obtained by the viewing operation, the memory strength may be recalculated by using the enhancement coefficient according to the enhancement coefficient corresponding to the user operation other than the viewing operation, so that the memory strength is recalculated. Will increase.
  • the enhancement factor can be a positive number greater than one.
  • the longer the viewing time point is from the current time point the lower the memory intensity of the user's promotion information, the less the number of viewings, and the lower the memory intensity.
  • the memory intensity may represent the degree of memory of the user for a certain promotion information, for example, may include the memory percentage or memory definition of a certain promotion information remembered by the user.
  • a slogan includes M words, where the memory strength can include the total number of words that the user remembers, and/or the total number of words that are remembered correctly.
  • the memory ratio may be N1/M; if the user remembers the correct total number of words N2, the memory resolution may be N2/M.
  • the memory intensity may be represented by the memory ratio, or may be represented by the memory definition, or the memory duration may be represented, or combined with the memory ratio and/or memory clear, A comprehensive memory intensity.
  • the integrated memory intensity may be the product of the memory ratio and the proportional weight plus the product of the memory definition and the clear scale weight.
  • the value of the memory intensity may be between 0 and 1.
  • the freshness is a parameter that exhibits a negative correlation with the memory intensity, for example, the freshness is equal to the reciprocal of the memory intensity, and the like.
  • the freshness calculated in this embodiment may be used to determine whether to send the second promotion information to a client held by the user, or to determine when to send the second promotion information to the first user. For example, the freshness calculated by the current time is very low, indicating that the user may have just read the promotion information, and may determine that the current promotion time does not send the second promotion information to the user according to the freshness, and may also be used to calculate the current time.
  • the second promotion information is pushed to the user at a later time.
  • the second promotion information and the first promotion information are two promotion information having relevance, and the first promotion information is that the promotion information that the user has viewed may be determined according to the historical data of the user.
  • the second promotion information and the first promotion information are related, and the related features may be reflected in these aspects:
  • the first aspect may be reflected in the fact that the second promotion information and the first promotion information have the same attribute in a predetermined dimension.
  • the second promotion information is the same as the industry involved in the first promotion information, for example, a car advertisement or a mobile phone advertisement, or the same brand or advertiser.
  • the second promotion information is Apple's mobile phone advertisement
  • the first promotion information is Apple's tablet advertisement.
  • the first promotion information may include the second promotion information itself.
  • the user may send the second promotion information to the user to determine the freshness of the second promotion information, and then according to the calculated freshness, whether to continue to The user sends the second promotion information and/or when the second promotion information is sent to the user.
  • the second aspect is a first aspect:
  • the second promotion information and the first promotion information are related, and the second promotion information and the first promotion information have similarities in a predetermined dimension.
  • the materials used in advertisement A and advertisement B have similarities.
  • advertisement A and advertisement B use similar pictures and/or background music, and the like.
  • the user may be represented by various user identifiers.
  • the user identifier may be a social account of a social application, and, for example, in a news application, the user identifier may be a login. The application account for this news app.
  • the sending the second promotion information to the user, where the second promotion information is sent to the client running the user identifier such as the social account or the application account corresponding to the user, and the social account of the user is The promotion information can be seen on the application interface of the user identification.
  • the embodiment by using the calculation of the freshness, it is possible to determine which promotion information is to be sent or not to be sent to the user, and to avoid repeatedly sending the same or the same type of promotion information to the user, and the promotion effect of the promotion information is poor and effective.
  • the problem of poor transmission rate and low user satisfaction has improved the promotion effect, effective transmission rate and user satisfaction of promotion information.
  • an attenuation factor is also introduced, where the attenuation factor is used to indicate the degree of memory attenuation of the user moving over time.
  • the attenuation factor and the viewing time to the current time.
  • the duration is negatively correlated. That is, the larger the duration, the smaller the attenuation factor, indicating that the greater the memory intensity decay, the higher the freshness. Therefore, the step S130 may include determining, according to the memory strength and the attenuation factor, the freshness of the second promotion information in the predetermined dimension.
  • the freshness can be calculated more accurately by the introduction of the attenuation factor.
  • step S130 There are various implementations of the step S130.
  • An optional method is provided below:
  • Determining, according to the memory strength and the attenuation factor, the freshness of the second promotion information in the predetermined dimension including:
  • the ⁇ is a fitting coefficient
  • the CalMemory is the memory intensity
  • the ⁇ is an attenuation factor
  • the present embodiment provides a freshness calculation method.
  • the freshness is quantified, and various electronic devices are conveniently determined to perform operations such as pushing the second promotion information to the user according to the quantized parameters.
  • the step S120 may include:
  • the memory intensity is a specified value when the first promotion information has not been viewed within the predetermined time range.
  • the predetermined time range may correspond to a time window with an end time of the current time. For example, if the time window has a duration of one month, the predetermined time window may be one month before the current time. Of course, if periodic statistics are performed in this embodiment, the predetermined time range may be the previous period.
  • the memory strength in the embodiment is a specified value.
  • the specified value may be 0 or a positive real number within a predetermined value range, indicating that the user has not seen it. Obviously, even if there is no memory, the memory strength is not involved or the user can view the first promotion information in other scenes, so the specified value can be given in advance, or a value can be randomly selected from the predetermined range of values. As the specified value. If the second information is viewed, the memory strength may be calculated according to a predetermined functional relationship.
  • the memory intensity is positively correlated with the number of times the first promotion information is viewed, and is negatively correlated with the duration of the viewing time point from the current time point; functions satisfying this relationship can be used to calculate the memory. strength.
  • the step S120 may include:
  • the memory strength is determined according to the number of views and the viewing time.
  • the client displays one of the first promotion information at time A in the predetermined time range, and the time A is the viewing time.
  • the method further includes:
  • the step S130 can include:
  • the freshness is determined based on the memory intensity and the number of attenuation factors.
  • the aforementioned attenuation factor may be an integrated memory attenuation factor
  • the integrated memory attenuation factor may be any parameter that is negatively correlated with the duration of the current time of viewing time, and may be a statically set value, specifically When used, the duration is used as the query index, and the table is determined by looking up the table.
  • the attenuation factor is a number attenuation factor. It can be calculated using a specific functional relationship. The following provides a calculation function relation of the number of times of attenuation factor, which is not limited to the following functional relationships in specific use:
  • the CalDesc(x) is a calculated function value, which is the memory attenuation factor; and the x is the number of times of the first promotion information viewed in the predetermined time range.
  • the determining the memory strength according to the number of views and the viewing time comprises:
  • the memory intensity is determined according to the number of views, the viewing time, and the memory quality parameter.
  • the memory quality parameter here is to reflect the user's memory strength, for example, it can represent the degree parameter of the duration of the information after watching a message. Generally, if a person's memory is better, the memory quality parameter is larger.
  • the big data analysis can be used, and all users adopt a unified default value.
  • the default value can be used to represent the value 60 of the ordinary person's memory quality parameter.
  • the memory strength parameter in order to accurately calculate the freshness, will be determined in a targeted manner according to user attributes.
  • the memory quality parameter is dynamically determined based on a reference factor such as the user's gender and/or age and/or education level.
  • the age can be divided into different age groups, and each age group is given a memory parameter.
  • the memory quality parameters of users aged 20 to 30 may be larger.
  • gender it is possible to set the memory quality parameter of the female user to be greater than the memory quality parameter of the male user.
  • the degree of education the memory quality parameter of the user with high education level can be made larger than the memory quality parameter of the user with low education level.
  • the above is only a rough classification based on user attributes to determine the memory quality parameters.
  • the memory pros and cons parameters may also be dynamically determined based on the personal characteristics of each user. For example, some users can determine that the memory of the user is stronger than the memory of the ordinary user through the social activities or participating games and some intelligent tests, and the specific memory quality parameters can be given according to the memory strength of the user. Instead of passing The user's other attribute analysis yields one of the memory strength parameters.
  • the unified memory value of the above-mentioned memory power quality parameter can be uniformly distributed, or it can be a different value for different users, and can be a static value or a dynamic value.
  • the determining the memory quality parameter of the user includes: acquiring a user attribute of the user; and determining the memory quality parameter according to the user attribute.
  • the user attributes herein may include the age, gender, occupation, education level, preferences, or personal behavior habits.
  • the user's preferences can also determine how much the user remembers a certain piece of information. For example, if the user is interested in the information A, then the memory of the information A may continue to be much higher than the information B that is not of interest.
  • the personal behavior habits although some users have strong memory, but in some cases, they will automatically block the memory, and in some cases, they will actively remember and produce an inherent effect, so in this embodiment, The determination of the parameters of memory quality can be made according to personal behavior habits.
  • the memory quality parameter is also related to the time point of memory generation. For example, some users have better memory in the morning than the memory in the afternoon, so the value of the memory parameter can not only be Depending on the user attribute, it may also depend on when the memory occurs. Therefore, the memory quality parameter may also be determined according to the time of occurrence of the memory, and/or the user attribute.
  • the method further includes:
  • the comprehensive freshness is determined according to the freshness of the plurality of predetermined dimensions, wherein, optionally, the comprehensive freshness is used to determine whether to send the second promotion information to the user.
  • the calculation of the freshness is performed from multiple dimensions, and the comprehensive freshness is determined, thereby achieving more accurate freshness calculation and accurate promotion of the promotion information.
  • the predetermined dimension may include various information such as the promotion information itself, the source of the promotion information, the promotion object related to the promotion information, the industry or brand involved in the promotion information, and the material included in the promotion information.
  • the determining the comprehensive freshness according to the freshness of the plurality of the predetermined dimensions including:
  • FreshScore ⁇ w i ⁇ FreshScore i
  • the FreshScore i is the freshness of the i-th predetermined dimension; the w i is the weight of the i-th predetermined dimension; wherein the value of the i is an integer not less than 1.
  • weights of different predetermined dimensions indicate the degree of influence of the freshness of the predetermined dimension on the comprehensive freshness.
  • the weight is a preset static value.
  • the static value may be a value that is determined according to big data statistics or simulation or experience, and can reflect the influence of the dimension on the degree of strangeness of the user to a certain promotion information.
  • a predetermined dimension corresponding to an advertisement may include two dimensions of an advertiser and a promotion.
  • the user's attention to the advertiser may not be highly concerned with the promotion itself, and when calculating the comprehensive freshness, the weight corresponding to the advertiser may be set to be lower than the promotion. value.
  • the value of the weight value in the embodiment may be positively correlated with the degree of attention of the user to the dimension.
  • the value of i starts from 1, and the maximum value is the total predetermined number of dimensions M.
  • determining the memory strength according to the number of views and the viewing time including:
  • the memory intensity is determined in conjunction with the overall memory intensity and the attenuation factor.
  • a method of specifically calculating the memory intensity is provided below, but it is worth noting that the specific implementation is not limited to the algorithm.
  • the t is the viewing time; the impressCount is the number of times the user views the first promotion information, and the value is 0 or 1 or an integer greater than 1.
  • the s is the memory quality parameter.
  • the CalMemory(t, impressCount) is the memory intensity of the viewing time t, that is, the memory intensity of the user at the current time T after viewing the first promotion information at time t. In the specific calculation, it is necessary to combine the memory intensity at each time point to obtain the overall memory intensity in the predetermined time range. For example, the overall memory strength can be calculated using the following functional relationship:
  • the T is the current time. If the T1 is the earliest time point within the predetermined time range, the T2 is the latest time point within the predetermined time range, and if the T2 is the earliest time point within the predetermined time range, then The T1 is the latest time point within the predetermined time range.
  • the step S130 may include:
  • the ⁇ is a fitting coefficient
  • the CalMemory is the overall memory intensity
  • the ⁇ is a memory attenuation factor
  • ⁇ can be calculated by the following functional relationship:
  • the embodiment provides a freshness determining apparatus for promoting information, including:
  • the obtaining unit 110 is configured to acquire historical data of the user viewing the promotion information
  • the first determining unit 120 is configured to determine, according to the historical data, a memory strength of the first promotion information, where the first promotion information and the second promotion information are associated in a predetermined dimension; the second promotion information is The information to be promoted; the first promotion information is the promotion information that the user has viewed; the memory strength is used to indicate that the user memorizes the strength of the first promotion information;
  • the second determining unit 130 is configured to determine freshness of the second promotion information in the predetermined dimension according to the memory strength, wherein the freshness is used to indicate a degree of strangeness of the second promotion information.
  • the freshness determining apparatus provided in this embodiment can be applied to a providing server or a search server or a push server or a processing platform for promoting information of various promotion information.
  • the obtaining unit 110 may correspond to a processor or a processing circuit, and may read the historical data from a local storage medium of the device, and may also receive or query the historical data from a peripheral device corresponding to the communication interface, for example, from A network platform such as a cloud platform queries or receives the historical data.
  • a network platform such as a cloud platform queries or receives the historical data.
  • the first determining unit 120 and the second determining unit 130 correspond to a processor or a processing circuit.
  • the processor can include a central processing unit, a digital signal processor, an application processor, a microprocessor, an application processor or a programmable array, and the like.
  • the processing circuit can include an application specific integrated circuit.
  • the processor or processing circuit can implement the operations of the first determining unit 120 and the second determining unit 130 by execution of a predetermined code.
  • the first determining unit 120 and the second determining unit 130 may jointly correspond to the same processor or processing circuit, or may correspond to different processors or processing circuits.
  • the second determining unit 130 is configured to determine the freshness of the second promotion information in the predetermined dimension according to the memory strength and the attenuation factor.
  • the freshness determining apparatus may be used to calculate the freshness of the second promotion information to be promoted for one or more users, and facilitate the subsequent promotion server or the promotion information providing server according to the freshness selectivity.
  • the promotion information is sent to the user, and the promotion effect of the promotion information and the satisfaction of the user's information are improved.
  • the apparatus further includes:
  • a third determining unit configured to determine the comprehensive freshness according to the freshness of the plurality of the predetermined dimensions, wherein the comprehensive freshness is used at least to determine whether to send the second promotion information to the user.
  • the third determining unit is also introduced in the embodiment.
  • the specific structure of the third determining unit in the embodiment is the same as or similar to the structure of the first determining unit 120 and the second determining unit 130, but
  • the third determining unit is a comprehensive calculating unit of a plurality of predetermined dimensions of freshness, and can combine the freshness of the plurality of predetermined dimensions to obtain a comprehensive freshness for transmitting the selection basis of the promotion information to the user.
  • the third determining unit is configured to calculate the comprehensive freshness FreshScore according to the following formula:
  • FreshScore ⁇ w i ⁇ FreshScore i
  • the FreshScore i is the freshness of the i-th predetermined dimension; the w i is the weight of the i-th predetermined dimension; wherein the value of the i is an integer not less than 1.
  • the third determining unit in this embodiment may correspond to a calculator or a processor having a computing function.
  • the third determining unit calculates the comprehensive freshness by using the above function relationship, and conveniently selects promotion information suitable for sending to the current user from multiple dimensions, and ensures accurate determination of the user receiving the promotion information to further enhance promotion.
  • the effect ensures that the user receives and/or views the satisfaction of the promotional information.
  • the first determining unit 120 is configured to determine, according to the historical data, whether the user has viewed the first promotion information within a predetermined time range; If the first promotion information has not been viewed within a certain time range, the memory intensity is a specified value.
  • the first determining unit 120 first determines the memory according to the viewing record of the first promotion information related to the second promotion information in a predetermined dimension within the predetermined time range. strength.
  • the correlation between the second promotion information and the first promotion information in the predetermined dimension in this embodiment can be referred to the foregoing embodiment, and will not be repeated here.
  • the first determining unit 120 is further configured to: when the first promotion information is viewed within the predetermined time range, determine a number of views; and determine, according to the number of views and the viewing time, The memory intensity.
  • the number of times of viewing the first promotion information within a predetermined time range is determined, and the memory intensity of the first promotion information is determined according to the number of views and the viewing time.
  • the memory intensity here can be expressed in parameters such as memory ratio and/or memory definition.
  • the apparatus further includes:
  • a fourth determining unit configured to determine a frequency attenuation factor according to the number of times of viewing
  • the second determining unit 130 is configured to determine the freshness according to the memory strength and the number of times attenuation factor.
  • the fourth determining unit is further included in the embodiment, where the specific hardware structure of the fourth determining unit is similar to that of the foregoing first determining unit 120, the second determining unit 130, and/or the third determining unit. Corresponds to the processor or processing circuitry.
  • the attenuation factor is a number attenuation factor, and in some embodiments, the attenuation factor may also be an attenuation factor of a single memory intensity that is uncorrelated with the number of times.
  • the fourth determining unit may calculate the number of times of attenuation by the function in the foregoing embodiment, calculate the freshness of each predetermined dimension, and further obtain the comprehensive freshness based on the freshness of each predetermined dimension.
  • the second determining unit 130 is configured to determine the memory of the user. a parameter of the strength and the inferiority; determining the memory intensity according to the number of views, the viewing time, and the memory quality parameter.
  • the second determining unit 130 uses the view count, the viewing time, and the memory quality parameter as input parameters, and can determine the memory strength by looking up the table or the foregoing function relationship.
  • the second determining unit 130 is configured to acquire a user attribute of the user, and determine the memory quality parameter according to the user attribute.
  • the memory quality parameter is determined according to the user memory strength parameter, and the memory strength of different users is determined, so that the promotion information can be filtered according to the user's personal characteristics, so as to achieve accurate push and promote promotion. At the same time, it can also be used to improve user satisfaction.
  • the second determining unit 130 is configured to determine a single memory intensity of a single view according to the time difference between the viewing time and the current time; and determine an overall memory intensity according to the single memory intensity; Determining an attenuation factor within the predetermined time range according to the number of views; determining the memory intensity in combination with the overall memory intensity and the attenuation factor.
  • the second determining unit 130 is configured to calculate the freshness FreshScore i of the i th the predetermined dimension according to the following formula
  • the ⁇ is a fitting coefficient
  • the CalMemory is the memory intensity
  • the ⁇ is a memory attenuation factor
  • the freshness can be accurately and simply calculated.
  • Freshness final score FreshScore is a fusion of freshness with multiple predetermined dimensions Cheng, we used a linear weighting method to fuse the scores of each freshness dimension:
  • FreshScore ⁇ w i ⁇ FreshScore i
  • FreshScore i Freshness values for each dimension, satisfying FreshScore i ⁇ [0,1]
  • ⁇ fitting coefficient, the default value is 0.1, which can be adjusted according to actual needs
  • ⁇ The value range [0,1], the user repeatedly receives the same dimension advertisement, and imposes additional punishment on the freshness score.
  • CalMemory The range of values [0, 1] indicates the user's memory of the memory content of the dimension.
  • lastImpressTime How much time has elapsed since the last time I saw the industry ad
  • N The number of days between the users of the rule setting to pull the same industry advertisement
  • the time difference between the T1 and T2 may be a predetermined prescribed value, for example, 30 days or 15 days or 10 days, and the like.
  • the x user receives the number of views of the same industry advertisement.
  • CalMemory(t, impressCount) is a calculation function of memory intensity, and its calculation method is as follows:
  • the Ebbinghaus Forgetting Curve is used, and the default value of the memory parameter is set to 60.
  • the impressCount in CalCountFactor(impressCount) is the time t user receives several advertisements for the industry, and the impressCount takes an integer of 0 or 1 or greater than 1.
  • the Ebbinghaus Forgetting Curve is a curve used to express the forgetting rate of medium- and long-term memory in memory. This curve was first proposed by psychologist Hermann Ebbinghaus through his own experiments. In this experiment, Ebbinghaus used some meaningless combinations of letters. By remembering these words The mother combination and checking the forgetting rate after a series of time intervals yields this curve. Therefore, this curve is also called Ebbinghaus Forgetting Curve.
  • the standard form of the forgetting curve is usually:
  • R is the memory intensity
  • S is the memory quality parameter
  • t is the viewing time
  • T is the current time
  • Ebbinghaus forgetting curve can be replaced by other functional forms, not necessarily Such a form can also be approximated using polynomials, power functions, and other functions.
  • the attenuation factor can be selected according to the actual situation or not, depending on the degree of protection to the user experience.
  • the calculation of the memory intensity may also use only the time when the content was last seen to simplify the calculation. Compared to using "memory content" that sees the content for a while in the past.
  • the global freshness score can be obtained according to the weight of each dimension.
  • the predetermined dimension is many, and is not limited to the above-mentioned industries.
  • Figure 4 shows a list of predetermined dimensions and memory contents.
  • the predetermined dimensions include advertisements, advertisers, advertisements, industries, materials, and the like.
  • an advertisement of the mobile phone A is taken as an example for description.
  • the subject matter of the advertisement may include an advertising service in addition to the advertisement.
  • the advertisements generally correspond to physical items, which may be various services that do not involve physical items, such as cleaning services.
  • the advertisement is used as the promotion information and the industry dimension is taken as an example for description.
  • this example provides a method for sending promotion information, including:
  • Step S1 a request to pull an advertisement
  • Step S2 obtaining a candidate advertisement list
  • Step S3 determining whether there is still an advertisement that does not calculate freshness in the candidate advertisement list, if not, the process ends, and if so, proceeds to step S4;
  • Step S4 selecting a candidate advertisement that does not calculate freshness
  • Step S5 Extracting historical data of the user viewing the advertisement, where the historical data may include the industry involved in the advertisement viewed by the user.
  • Step S6.1 calculating the number of times of attenuation
  • Step S6.2 calculating user memory strength
  • Step S7 Calculate the freshness of the industry.
  • the present example provides an optional hardware structure diagram of a freshness determining apparatus for promoting information, including a processor 11, an input/output interface 13 (eg, a display screen, a touch screen, a speaker), and a storage medium 14 And a network interface 12, the components can be connected to communicate via the system bus 15.
  • the freshness determining method executable instructions for executing the promotion information described in the embodiment of the present invention are stored in the storage medium 14.
  • Each hardware module shown in FIG. 5 may be partially implemented, implemented, or implemented in other hardware modules as needed.
  • the number of hardware modules of each type may be one or more, and each hardware module may be implemented in the same geographical location or distributed in Different geographical locations may be used to perform at least one of the freshness determination methods of the promotion information shown in FIG. 1 or FIG. 3 above.
  • this example illustrates an advertisement as a promotion information, and provides an advertisement promotion system.
  • the advertising system includes:
  • the advertisement requesting end may correspond to the foregoing user or user terminal, and may be used to send an advertisement request;
  • Access server cluster which can be used for access to the ad requesting end
  • Retrieving the server cluster obtaining historical data from the storage server cluster, performing freshness calculation based on historical data, and searching for advertisements according to freshness and retrieval parameters, and returning the advertisements to the advertisement server data through the access server cluster Ad request side.
  • the cluster is a fleet of at least multiple servers.
  • the embodiment of the present invention further provides an electronic device, where the electronic device can be a server, and the server can be located on the network side, and can be used to send the promotion information to the terminal.
  • the electronic device can be a server
  • the server can be located on the network side, and can be used to send the promotion information to the terminal.
  • the server can include: a memory and a processor.
  • the memory may be various types of memory, such as a random access memory, a read only memory, a flash memory, etc., and may be configured to store various information, such as computer executable instructions such as a computer program;
  • the processor may be connected to the memory through an integrated circuit bus or the like, and may read and store the executable instructions stored in the computer, and may implement the one or more technical solutions to provide the promotion information by executing the computer executable instructions. Degree determination method.
  • the processor may be an application processor (AP), a central processing unit (CPU), a digital signal processor (DSP), a programmable gate array (FPGA, Field Programmable Gate). Array) or an information processing module or component with information processing such as an application specific integrated circuit.
  • AP application processor
  • CPU central processing unit
  • DSP digital signal processor
  • FPGA Field Programmable Gate
  • Array an information processing module or component with information processing such as an application specific integrated circuit.
  • the processor can implement the steps by executing computer-executable instructions, which can include the following steps:
  • the second promotion information is information currently to be promoted
  • the first promotion information is promotion information that the user has viewed
  • the memory is strong Degree is used to instruct the user to memorize the strength of the first promotion information
  • the processor may further implement the following steps by executing computer executable instructions:
  • the processor may further implement the following steps by executing computer executable instructions:
  • Determining, according to the memory strength and the attenuation factor, the freshness of the second promotion information in the predetermined dimension including:
  • the ⁇ is a fitting coefficient
  • the CalMemory is the memory intensity
  • the ⁇ is an attenuation factor
  • the processor may further implement the following steps by executing computer executable instructions:
  • the memory strength is a specified value
  • the memory strength is determined according to the number of views and the viewing time.
  • the embodiment of the present invention further provides a computer storage medium, where the computer storage medium stores computer executable instructions, and after the computer executable instructions are executed, the freshness of the promotion information provided by the one or more technical solutions may be implemented.
  • Determine the method for example, at least the following steps:
  • the second promotion information is information currently to be promoted
  • the first promotion information is the promotion information that the user has viewed
  • the memory strength is used to indicate that the user memorizes the strength of the first promotion information
  • the determining, according to the memory strength, the freshness of the second promotion information in the predetermined dimension including:
  • the determining, according to the memory strength and the attenuation factor, the freshness of the second promotion information in the predetermined dimension including:
  • the ⁇ is a fitting coefficient
  • the CalMemory is the memory intensity
  • the ⁇ is an attenuation factor
  • the computer storage medium may include: a removable storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like.
  • the computer storage medium can be a non-transitory storage medium or a non-volatile storage medium.
  • the disclosed apparatus and method may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of the unit is only a logical function division.
  • there may be another division manner such as: multiple units or components may be combined, or Can be integrated into another system, or some features can be ignored or not executed.
  • the coupling, or direct coupling, or communication connection of the components shown or discussed may be indirect coupling or communication connection through some interfaces, devices or units, and may be electrical, mechanical or other forms. of.
  • the units described above as separate components may or may not be physically separated, and the components displayed as the unit may or may not be physical units, that is, may be located in one place or distributed to multiple network units; Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
  • each functional unit in each embodiment of the present invention may be integrated into one processing module, or each unit may be separately used as one unit, or two or more units may be integrated into one unit; the above integration
  • the unit can be implemented in the form of hardware or in the form of hardware plus software functional units.
  • the historical data of the first promotion information is obtained by the user, and the memory intensity of the first promotion information is determined according to the historical data of the first promotion information, and is based on The memory strength further calculates the freshness of the second promotion information that the user is to promote, and the freshness can represent the user's strangeness to the second promotion information.

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种推广信息的新鲜度确定方法及装置，涉及信息技术领域，方法包括：获取用户查看推广信息的历史数据；根据所述历史数据，确定第一推广信息的记忆强度，其中，所述第一推广信息与第二推广信息在预定维度内相关联；所述第二推广信息为当前待推广的信息；所述第一推广信息为用户已查看的推广信息；所述记忆强度用于指示所述用户记忆所述第一推广信息的强度；根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，其中，所述新鲜度用于指示所述第二推广信息的陌生程度。

Description

推广信息新鲜度确定方法及装置、电子设备及存储介质
本申请基于申请号为201710036301.3、申请日为2017年01月17日的中国专利申请提出，并要求该中国专利申请的优先权，该中国专利申请的全部内容在此引入本申请作为参考。
技术领域
本发明涉及信息技术领域，尤其涉及一种推广信息新鲜度确定方法及装置、电子设备及计算机存储介质。
背景技术
推广信息为主动推送给用户观看或查阅的信息内容，而非是用户主动要求查看的内容，常见的推广信息可包括广告、政府公告以及活动介绍信息等。
在现有技术中，通常一个推广信息会重复推送给同一个用户观看，或相类似的推广信息可能集中的推送给某一个或多个用户观看，但是这种推广信息的推送方法的推广效果并不好，显然这种信息的有效传播率就低了。此外，反复的向同一个或同一类用户推送同一个推广信息或类似的推广信息，可能会招致用户的不满，甚至引起投诉，或关闭删除某些应用，从而导致用户流失的问题。
发明内容
有鉴于此，本发明实施例期望提供一种推广信息新鲜度确定方法及装置、电子设备及计算机存储介质，以解决推广信息的推广效果差和/或用户接收查看推广信息的体验差的问题。
本发明实施例的技术方案是这样实现的：
本发明实施例第一方面提供一种推广信息的新鲜度确定方法，可应用于信息处理平台，包括：
获取用户查看推广信息的历史数据；
根据所述历史数据，确定第一推广信息的记忆强度，其中，所述第一推广信息与第二推广信息在预定维度内相关联；所述第二推广信息为当前待推广的信息；所述第一推广信息为用户已查看的推广信息；所述记忆强度用于指示所述用户记忆所述第一推广信息的强度；
根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，其中，所述新鲜度用于指示所述第二推广信息的陌生程度。
本发明实施例第二方面提供一种推广信息的新鲜度确定装置，包括：
获取单元，配置为获取用户查看推广信息的历史数据；
第一确定单元，配置为根据所述历史数据，确定第一推广信息的记忆强度，其中，所述第一推广信息与第二推广信息在预定维度内相关联；所述第二推广信息为当前待推广的信息；所述第一推广信息为用户已查看的推广信息；所述记忆强度用于指示所述用户记忆所述第一推广信息的强度；
第二确定单元，配置为根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，其中，所述新鲜度用于指示所述第二推广信息的陌生程度。
本发明实施例第三方面提供一种电子设备，其中，包括：存储器及处理器；
所述存储器，配置为存储有计算机可执行指令；
所述处理器，与所述存储器连接，配置为读取并执行所述计算机可执行指令，以实现上述技术方案提供的推广信息的新鲜度确定方法。
本发明实施例第四方面提供一种计算机存储介质，所述计算机存储介 质中存储有计算机可执行指令，所述计算机可执行指令用于执行上述技术方案提供的推广信息的新鲜度确定方法。
本发明实施例提供的推广信息新鲜度确定方法及装置、电子设备及存储介质，会获取用户查看推广信息的历史数据，将根据历史数据确定出与当前待推广的第二推广信息具有关联性的第一推广信息的记忆强度，并根据记忆强度及衰减因子，会得到第二推广信息在预定维度的新鲜度，而新鲜度可以作为筛选向用户发送第二推广信息的依据参数，通过新鲜度的计算，相对于随机向用户推广信息或不加筛选的向用户发送推广信息，可以减少同一条或同一类的信息反复发送给同一个用户，导致的推广效果差及用户满意度低的问题，提升了推广效果及用户满意度。
附图说明
图1为本发明实施例提供第一种推广信息的新鲜度确定方法的流程示意图；
图2为本发明实施例提供第一种推广信息的新鲜度确定装置的结构示意图；
图3为本发明实施例提供第二种推广信息的新鲜度确定方法的流程示意图；
图4为本发明实施例提供的一种预定维度和记忆内容的对应关系示意图标；
图5为本发明实施例提供第二种推广信息的新鲜度确定装置的结构示意图；
图6为本发明实施例提供的一种推广信息系统的结构示意图。
具体实施方式
以下结合说明书附图及具体实施例对本发明的技术方案做进一步的详 细阐述，应当理解，以下所说明的优选实施例仅用于说明和解释本发明，并不用于限定本发明。
如图1所示，本实施例提供一种推广信息的新鲜度确定方法，包括：
步骤S110：获取用户查看推广信息的历史数据；
步骤S120：根据所述历史数据，确定第一推广信息的记忆强度，其中，所述第一推广信息与第二推广信息在预定维度内相关联；所述第二推广信息为当前待推广的信息；所述第一推广信息为用户已查看的推广信息；所述记忆强度用于指示所述用户记忆所述第一推广信息的强度；
步骤S130：根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，其中，所述新鲜度用于指示所述第二推广信息的陌生程度，是与用户当前对某一个信息的记忆深刻程度呈现负相关的。
本实施例所述的新鲜度确定方法，可为应用于各种具有信息处理功能的电子设备中的信息处理方法，例如，可为应用于广告服务器、信息搜索服务器等包括一个或多个服务器组成的信息处理平台中的信息处理方法，即所述步骤S110至步骤S130均可以由电子设备，例如，推广平台中的广告服务器或推送服务器等服务器执行。所述推广信息可包括广告、各种类型的公告或活动等信息，这里的广告可包括商业广告和非商业的广告。所述非商业的广告通常可包括慈善广告等。
所述新鲜度可为确定是否向所述用户推送所述第二推广信息的依据，例如，当所述新鲜度的取值位于指定范围内时，则向所述用户发送所述第二推广信息，否则不发送所述第二推广信息。在本实施例中所述新鲜度可为与所述用户所在客户端接收所述第一推广信息的接收状况相关的参数。所述接收状况信息可包括第一推广信息的接收次数、接收频次和/或接收时间等参数。当然，所述新鲜度可为所述用户所持有客户端是否显示过所述第一推广信息的显示状况信息，这里的显示状况信息可包括：显示次数、 显示时间和/或显示频次等信息。
所述历史数据为用户查看推广信息的历史数据，所述历史数据可包括：用户查看过的推广信息的信息名称、推广信息的推广物、来源、推广信息中涉及的素材、查看时间、查看时长、查看的次数等各种参数。所述素材可包括：推广信息中使用过的图片、文本和/或视频等各种构成所述推广信息的数据。
可选地，所述历史数据还可包括：第一推广信息在用户终端上的显示时长、显示次数、用户查看该第一推广信息中伴随的用户操作和/或显示时间距离当前时间的时长。所述用户操作可包括：转发操作、点击操作、收藏到收藏目录的收藏操作、推荐给朋友的推荐操作或复制操作等除单纯查看以外的其他用户操作，这种操作可以称为查看操作以外的预定操作。
这些用户操作都能够加强用户的记忆，能够提升用户的记忆强度。故在本实施例中在确定记忆强度时，不仅会根据用户是否有查看、查看次数、查看时长等参数确定记忆强度，还会根据查看过程中伴随的上述查看操作以外的预定操作来确定。若发生上述查看操作以外的预定操作，则可以在基于查看操作得到的记忆强度的基础上，根据查看操作以外的用户操作对应的增强系数，利用增强系数重新计算记忆强度，这样重新计算得到记忆强度将增大。所述增强系数可为大于1的正数。
通过用户观看过某一个推广信息，观看时间点距离当前时间点越久，用户对该推广信息的记忆强度就越低，观看的次数越少，记忆强度也越低。
在本实施例中所述记忆强度可表征用户对某一个推广信息的记忆程度，例如，可包括用户记住的某一个推广信息的记忆百分比或记忆清晰度。例如，一条广告语包括M个字，这里的记忆强度可包括用户记住的总字数，和/或，记忆正确的总字数。例如，若用户记住了N1个字，所述记忆比例可为N1/M；若用户记忆正确的总字数N2，则所述记忆清晰度可为N2/M。 在本实施例中所述记忆强度可以用所述记忆比例来代表，也可以用所述记忆清晰度来代表，或所述记忆持续时长代表，或结合所述记忆比例和/或记忆清晰，得到一个综合记忆强度。所述综合记忆强度可为记忆比例与比例权值的乘积加上所述记忆清晰度与清晰比例权值的乘积。总之，表征所述记忆强度的方式有很多种，在本实施例中就不再一一举例了。
在本实施例中所述记忆强度的取值可为0到1之间。在本实施例中所述新鲜度为与所述记忆强度呈现负相关的参数，例如，所述新鲜度等于所述记忆强度的倒数等。
在本实施例中计算出的所述新鲜度，可以用于确定是否向用户所持的客户端发送所述第二推广信息，或确定何时向所述第一用户发送所述第二推广信息。例如，当前时间计算出的新鲜度很低，表示用户可能刚刚看过该推广信息，可以根据该新鲜度确定当前时刻不向用户发送所述第二推广信息，还可以用于计算出在当前时刻之后的某一个时刻向用户推送所述第二推广信息。
在本实施例中所述第二推广信息和所述第一推广信息是具有关联性的两个推广信息，第一推广信息为根据用户的历史数据可确定知道用户查看过的推广信息。在本实施例中所述第二推广信息和第一推广信息具有关联性，这些关联性可以体现在这些方面：
第一方面：这种关联性可体现在第二推广信息和第一推广信息在预定维度的属性相同。例如，第二推广信息和第一推广信息涉及的行业相同，例如，都是汽车广告或手机广告，或品牌或广告主相同。又例如，第二推广信息是苹果的手机广告，第一推广信息为苹果的平板电脑广告等。一个特殊示例为，所述第一推广信息可以包括所述第二推广信息自身。例如，当前有一条广告A，前两天刚给用户推广该广告A，用户也查看了广告A，则广告A的查看记录会呈现在所述历史数据中。故在本实施例中，想再次 向用户发送所述第二推广信息时，可以根据之前向用户发送第二推广信息的历史数据，确定出用户对第二推广信息的新鲜度，然后根据计算出的新鲜度，可确定是否继续向用户发送所述第二推广信息和/或何时向所述用户发送所述第二推广信息。
第二方面：
所述第二推广信息和所述第一推广信息具有关联性，还体现在所述第二推广信息与第一推广信息在预定维度具有相似性。例如，在广告A和广告B使用的素材具有相似性。例如，广告A和广告B使用了类似的图片和/或背景音乐等。
在本实施例中所述用户可以用各种用户标识表示，例如，在社交应用中，所述用户标识可以为社交应用的社交账号，又例如，在新闻应用中，所述用户标识可为登录该新闻应用的应用账号。这里的向所述用户发送所述第二推广信息，可为向所述用户对应的社交账号或应用账号等用户标识所运行的客户端发送所述第二推广信息，则在该用户的社交账号等用户标识的应用界面上就可以看到该条推广信息。
在本实施例中通过所述新鲜度的计算，可以确定向用户发送哪些推广或不发送哪些推广信息，避免反复向用户发送同一个或同一类推广信息，导致的推广信息的推广效果差、有效传播率差及用户满意度低的问题，从而提升了推广信息的推广效果、有效传播率及用户满意度。
在本实施例中在计算所述新鲜度时，还引入了衰减因子，这里的衰减因子用于指示用户随时间迁移的记忆衰减程度，一般情况下，所述衰减因子与观看时间到当前时间的时长呈负相关。即，时长越大，则所述衰减因子越小，表示记忆强度衰减越大，则所述新鲜度越高。故所述步骤S130可包括：根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度。
通过衰减因子的引入，可以更加精确计算出所述新鲜度。
所述步骤S130的实现方式有多种，以下提供一种可选方式：
所述根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度，包括：
依据如下公式计算所述预定维度上的新鲜度FreshScore，
Figure PCTCN2017114633-appb-000001
其中，所述α为拟合系数，所述CalMemory为所述记忆强度；所述β为衰减因子。
这样的话，首先本实施例提供了一种新鲜度计算方法，其次，实现了新鲜度的量化，方便各种电子设备根据量化的参数，确定是否向用户推送所述第二推广信息等操作。
在一些实施例中，所述步骤S120可，包括：
根据所述历史数据，确定在预定时间范围内用户是否有查看过所述第一推广信息；
当在所述预定时间范围内未有查看过所述第一推广信息，则所述记忆强度为指定值。
所述预定时间范围可对应于一个以当前时间为结束时间的时间窗。例如，所述时间窗的时长为1个月，则所述预定时间窗可为当前时间之前的一个月。当然，在本实施例中若进行周期性统计，则所述预定时间范围则可为上一个周期。
在本实施例中，首先，先确定预定时间范围内是否有查看过与第二推广信息在预定维度有关联的第一推广信息。若通过查询历史数据发现，在预定时间范围内，用户没有查看过在该预定维度上与第二推广信息相关联的第一推广信息，故在本实施例中所述记忆强度就为指定值，在本实施例中所述指定值可为0或不为预定取值范围内的正实数，表示用户没有看过， 显然连记忆都没有，就不涉及记忆强度了或者考虑到用户可以在其他场景中查看到所述第一推广信息，故可以预先赋予指定值，或随机从所述预定取值范围内选择一个值作为所述指定值。若有查看过所述第二信息，则可以依据预定函数关系，计算所述记忆强度。
在实施例中所述记忆强度，将与观看第一推广信息的次数呈正相关，且与观看时间点距离当前时间点的时长呈负相关；满足这种关系的函数都可以用来计算所述记忆强度。
在一些实施例中，所述步骤S120可包括：
当在所述预定时间范围内有查看过所述第一推广信息，则确定查看次数及每一次的查看时间；
根据所述查看次数及所述查看时间，确定所述记忆强度。
例如，客户端在预定时间范围内的A时刻显示过一条所述第一推广信息，则所述时刻A为所述查看时间。
进一步地，所述方法还包括：
根据所述查看次数，确定次数衰减因子；
所述步骤S130即可包括：
根据所述记忆强度及所述次数衰减因子，确定所述新鲜度。
在一些实施例中前述提到的衰减因子可为综合记忆衰减因子，该综合记忆衰减因子可以是任意一个与查看时间距离当前时间的时长呈负相关的参数，可以为静态设置的值，在具体使用时，以所述时长为查询索引，通过查表确定。在本实施例中所述衰减因子为次数衰减因子。可以采用特定的函数关系来计算。以下提供一个所述次数衰减因子的计算函数关系式，在具体使用时不限于下列函数关系式：
Figure PCTCN2017114633-appb-000002
其中，所述CalDesc(x)为计算得到的函数值，为所述记忆衰减因子；所述x为在所述预定时间范围内观看的第一推广信息的次数。
在一些实施例中，所述根据所述查看次数及查看时间，确定所述记忆强度，包括：
确定所述用户的记忆力优劣参数；
根据所述查看次数、查看时间及所述记忆力优劣参数，确定所述记忆强度。
这里的记忆力优劣参数为，反映用户的记忆强弱，例如，可以表征观看过一个信息之后记住该信息的持续时长的程度参数。通常，若一个人的记忆力越好，则所述记忆力优劣参数越大。
在进行计算时，可以通过大数据分析，所有的用户采用统一默认值，例如，在本实施例中可以默认取值为表征普通人记忆力优劣参数的取值60。
但是在一些实施例中为了精确计算所述新鲜度，将根据用户属性，针对性确定所述记忆力优劣参数。例如，根据用户的性别和/或年纪和/或受教育程度等参考因素，动态确定所述记忆力优劣参数。例如，可以将年纪分为不同年龄段，每一个年龄段赋予一个记忆力优劣参数，例如，相对于50到60岁的用户，可能20到30岁用户的记忆力优劣参数的取值更大。针对于性别，可以设置使得女性用户的记忆力优劣参数大于男性用户的记忆力优劣参数。针对于受教育程度，可以使得受教育程度高的用户的记忆力优劣参数大于受教育程度低的用户的记忆力优劣参数。当然，以上仅是根据用户属性进行大致分类，来确定记忆力优劣参数。
在一些实施例中，还可以根据每一个用户的个人特点来动态确定所述记忆力优劣参数。例如，有些用户通过其参与的社交活动或参与的游戏以及一些智能测试等，可以确定出该用户的记忆力强于普通用户的记忆力，可以根据该用户的记忆力强度，给出具体的记忆力优劣参数，而不是通过 该用户的其他属性分析得到一个所述记忆力优劣参数。
故综合上述所述记忆力优劣参数可以统一分配的统一值，也可以是针对不同用户的不同值，可以是静态值，也可以是动态确定值。
例如，所述确定所述用户的记忆力优劣参数，包括：获取所述用户的用户属性；根据所述用户属性，确定所述记忆力优劣参数。这里的用户属性就可包括所述年龄、性别、职业、受教育程度、喜好或个人行为习惯。用户的喜好也可以决定用户对某一信息的记忆程度。例如，用户对信息A感兴趣，则可能对信息A的记忆持续大高于其不感兴趣的信息B。所述个人行为习惯，有的用户虽然记忆力很强，但是在某些情况下是会自动屏蔽记忆的，而有些情况下就会主动记忆，产生过目不忘的效果，故在本实施例中还可以根据个人行为习惯来进行记忆力优劣参数的确定。
当然，在一些实施例中，所述记忆力优劣参数，还与记忆产生的时间点相关，例如，有些用户早上的记忆力明显好于下午的记忆力，故所述记忆力优劣参数的取值不仅可以取决于所述用户属性，还可取决于记忆的发生时间。故所述记忆力优劣参数也可以根据所述记忆的发生时间，和/或，所述用户属性进行确定。
在一些实施例中，所述方法还包括：
根据多个所述预定维度的新鲜度，确定综合新鲜度，其中，可选的，所述综合新鲜度，至少用于确定是否向所述用户发送所述第二推广信息。
在本实施例中将从多维度进行所述新鲜度的计算，并确定出综合新鲜度，从而实现更精确的新鲜度计算，和推广信息的精准推广。
所述预定维度可包括推广信息自身、推广信息的来源、推广信息涉及的推广物、推广信息涉及的行业或品牌、推广信息中包含的素材等各种信息。
分别计算至少两个预定维度的新鲜度，再结合各个预定维度的新鲜度， 得到一个综合新鲜度，最终根据这个综合新鲜度确定是否向用户发送所述第二推广信息和/或何时向所述用户发送所述第二推广信息。
例如，所述根据多个所述预定维度的新鲜度，确定综合新鲜度，包括：
根据如下公式计算所述综合新鲜度FreshScore
FreshScore＝∑wi×FreshScorei
其中，所述FreshScorei为第i个所述预定维度的新鲜度；所述wi为第i个所述预定维度的权值；其中，所述i的取值为不小于1的整数。
不同预定维度的权值不同，则表示该预定维度的新鲜度对综合新鲜度的影响度。通常情况下，所述权值为预先设定的静态值。所述静态值可为根据大数据统计或仿真或经验总结，确定的能够反映该维度对用户对某一个推广信息的陌生程度的影响值。
例如，针对于广告对应的预定维度可包括广告主和推广物这两个维度。通常情况下，用户对广告主的关注度可能没有对推广物本身的关注度高，则在计算综合新鲜度时，可以将所述广告主对应的权值设置为低于所述推广物的权值。总之，本实施例中所述权值的取值大小，可与用户对该维度的关注度呈正相关。所述i的取值从1开始，最大值为总预定维度数M。
在一些实施例中，根据所述查看次数及查看时间，确定所述记忆强度，包括：
根据所述查看时间与当前时间的时间差，确定单次查看的单次记忆强度；
根据所述单次记忆强度，确定整体记忆强度；
根据所述查看次数，确定在所述预定时间范围内的衰减因子；
结合所述整体记忆强度及所述衰减因子，确定所述记忆强度。以下提供一种具体计算所述记忆强度的方法，但是值得注意的是，具体实现时不局限于该算法。
CalMemory(t,impressCount)＝CalTimeFactor(t)×CalCountFactor(impressCount)
Figure PCTCN2017114633-appb-000003
如impressCount≠0，则CalCountFactor(impressCount)＝1，否则CalCountFactor(impressCount)＝0；
所述t为查看时间；所述impressCount为时间t用户查看第一推广信息的次数，取值为0或1或大于1的整数；所述s为所述记忆力优劣参数。所述CalMemory(t,impressCount)为查看时间t的记忆强度，也即是用户在时间t查看所述第一推广信息后在当前时间T的记忆强度。在具体计算时，需要结合各个时间点的记忆强度，得到所述预定时间范围内整体记忆强度。例如，可以采用如下函数关系式计算所述整体记忆强度：
Figure PCTCN2017114633-appb-000004
所述T为当前时刻。若所述T1为所述预定时间范围内的最早时刻点，则所述T2为所述预定时间范围内的最晚时刻点，若所述T2为所述预定时间范围内的最早时刻点，则所述T1为所述预定时间范围内的最晚时刻点。
在一些实施例中，所述步骤S130可包括：
依据如下公式计算第i个的新鲜度FreshScorei
Figure PCTCN2017114633-appb-000005
其中，所述α为拟合系数，所述CalMemory为所述整体记忆强度；所述β为记忆衰减因子。
在本实施例中所述β可以通过如下函数关系式进行计算：
Figure PCTCN2017114633-appb-000006
这里的各个参数的定义可以参见前述实施例对应符号表示的参数。
如图2所示，本实施例提供一种推广信息的新鲜度确定装置，包括：
获取单元110，配置为获取用户查看推广信息的历史数据；
第一确定单元120，配置为根据所述历史数据，确定第一推广信息的记忆强度，其中，所述第一推广信息与第二推广信息在预定维度内相关联；所述第二推广信息为当前待推广的信息；所述第一推广信息为用户已查看的推广信息；所述记忆强度用于指示所述用户记忆所述第一推广信息的强度；
第二确定单元130，配置为根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，其中，所述新鲜度用于指示所述第二推广信息的陌生程度。
本实施例提供的新鲜度确定装置可以应用于各种推广信息的提供服务器或搜索服务器或推送服务器或推广信息的处理平台中。
所述获取单元110可对应于处理器或处理电路，可以从设备的本地存储介质中读取所述历史数据，还可对应于通信接口，从外设接收或查询所述历史数据，例如，从云平台等网络平台查询或接收所述历史数据。
所述第一确定单元120及所述第二确定单元130对应于的处理器或处理电路。所述处理器可包括中央处理器、数字信号处理器、应用处理器、微处理器、应用处理器或可编程阵列等。所述处理电路可包括专用集成电路。
所述处理器或处理电路可通过预定代码的执行，实现所述第一确定单元120及第二确定单元130的操作。
所述第一确定单元120及第二确定单元130可以共同对应于同一个处理器或处理电路，也可以对应于不同的处理器或处理电路。
所述第二确定单元130，配置为根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度。
总之，本实施例提供的新鲜度确定装置，可用于计算待推广的第二推广信息对于某一个或多个用户的新鲜度，方便后续推广服务器或推广信息的提供服务器根据所述新鲜度选择性的向用户发送推广信息，提升推广信息的推广效果及用户的信息查阅满意度。
在一些实施例中，所述装置还包括：
第三确定单元，配置为根据多个所述预定维度的新鲜度，确定综合新鲜度，其中，所述综合新鲜度至少用于确定是否向所述用户发送所述第二推广信息。
在本实施例中还引入了第三确定单元，在本实施例中所述第三确定单元的具体的结构，与所述第一确定单元120及第二确定单元130的结构相同或类似，但是第三确定单元为多个预定维度的新鲜度的综合计算单元，能够结合多个预定维度的新鲜度，得到一个综合新鲜度，用于向用户发送推广信息的选择依据。
在一些实施例中，所述第三确定单元，配置为根据如下公式计算所述综合新鲜度FreshScore：
FreshScore＝∑wi×FreshScorei
其中，所述FreshScorei为第i个所述预定维度的新鲜度；所述wi为第i个所述预定维度的权值；其中，所述i的取值为不小于1的整数。
在本实施例中所述第三确定单元可对应于计算器或具有计算功能的处理器。所述第三确定单元，利用上述函数关系计算所述综合新鲜度，方便从多个维度选择出适合向当前用户发送的推广信息，一方面确保接收推广信息的用户的精确确定，以进一步提升推广效果，另一方面能够确保用户接收和/或查看推广信息的满意度。
在一些实施例中，所述第一确定单元120，配置为根据所述历史数据，确定在预定时间范围内用户是否有查看过所述第一推广信息；当在所述预 定时间范围内未有查看过所述第一推广信息，则所述记忆强度为指定值。
在本实施例中所述第一确定单元120，首先将依据划定的预定时间范围内的与第二推广信息在某一个预定维度内相关的第一推广信息的查看记录，确定出所述记忆强度。
在本实施例中所述第二推广信息和第一推广信息在预定维度内的相关性，可参见前述实施例，在此就不再重复了。
在一些实施例中，所述第一确定单元120，还配置为当在所述预定时间范围内有查看过所述第一推广信息，则确定查看次数；根据所述查看次数及查看时间，确定所述记忆强度。
在本实施例中将确定在预定时间范围内的查看第一推广信息的查看次数，并将根据查看次数和查看时间，确定对该第一推广信息的记忆强度。这里的记忆强度可体现在记忆比例和/或记忆清晰度等参量。
在一些实施例中，所述装置还包括：
第四确定单元，配置为根据所述查看次数，确定次数衰减因子；
所述第二确定单元130，配置为根据所述记忆强度及所述次数衰减因子，确定所述新鲜度。
在本实施例中还包括第四确定单元，这里的第四确定单元的具体的硬件结构，与前述第一确定单元120、第二确定单元130和/或第三确定单元的结构类似，都可对应于处理器或处理电路。
在本实施例中所述衰减因子为次数衰减因子，在一些实施例中，所述衰减因子还可为与次数不相关的单次记忆强度的衰减因子。
所述第四确定单元，可通过前述实施例中的函数来计算次数衰减因子，可计算出每一个预定维度的新鲜度，并可进一步基于每一个预定维度的新鲜度得到所述综合新鲜度。
在一些实施例中，所述第二确定单元130，配置为确定所述用户的记忆 力优劣参数；根据所述查看次数、查看时间及所述记忆力优劣参数，确定所述记忆强度。
在本实施例中所述第二确定单元130，将以所述查看次数、查看时间及记忆力优劣参数为输入参数，通过查表或前述函数关系，可以确定记忆强度。
在一些实施例中，所述第二确定单元130，配置为获取所述用户的用户属性；根据所述用户属性，确定所述记忆力优劣参数。
在本实施例中所述记忆力优劣参数，是根据所述用户记忆力优劣参数来确定不同用户的记忆强度，从而可以根据用户的个人特点，筛选发送的推广信息，以实现精准推送，提升推广效果的同时，还可以用于提升用户满意度。
在一些实施例中，所述第二确定单元130，配置为根据所述查看时间与当前时间的时间差，确定单次查看的单次记忆强度；根据所述单次记忆强度，确定整体记忆强度；根据所述查看次数，确定在所述预定时间范围内的衰减因子；结合所述整体记忆强度及所述衰减因子，确定所述记忆强度。
在一些实施例中，所述第二确定单元130，配置为依据如下公式计算第i个所述预定维度的新鲜度FreshScorei
Figure PCTCN2017114633-appb-000007
其中，所述α为拟合系数，所述CalMemory为所述记忆强度；所述β为记忆衰减因子。
在本实施例中利用上述函数关系，可以精确并简便的计算出所述新鲜度。
以下结合上述实施例提供几个具体示例：
示例一：
新鲜度最终的分值FreshScore是由多个预定维度的新鲜度度融合而 成，我们采用了线性加权的方式来融合各个新鲜度维度的分值：
FreshScore＝∑wi×FreshScorei
其中：
wi：各维度权重，满足∑wi＝1，wi≥0，该权重可以通过人工或者机器学习来指定。
FreshScorei：各维度新鲜度值，满足FreshScorei∈[0，1]
对于每个维度的新鲜度值，其计算如以下公式所示：
Figure PCTCN2017114633-appb-000008
α：拟合系数，默认值为0.1，可以根据实际需要调整
β：取值范围[0,1]，用户反复收到相同维度广告，对新鲜度分值进行额外惩罚。
CalMemory：取值范围[0,1]，表示用户对该维度的记忆内容的记忆度。
各维度新鲜度计算
以行业维度为例，我们使用如下的流程计算用户的候选广告列表中，每条广告在行业上的新鲜度：
针对某用户(user)的一条候选广告，假设其行业为industry，行业新鲜度的计算公式如下：
如lastImpressTime≤60*24*N，则FreshScoreindustry(user,industry)＝0，否则，
Figure PCTCN2017114633-appb-000009
其中：
lastImpressTime：上一次看到该行业广告之后过去了多少时间；
N：规则设定的用户拉取相同行业广告的间隔天数；
t：查看时间；
所述T1和T2之间的时间差可为预定的规定值，例如，30天或15天或10天等。
α：拟合系数，默认值为0.1。
利用如下公式计算次数衰减因子：
Figure PCTCN2017114633-appb-000010
所述x用户收到相同行业广告的查看次数。
再利用如下公式进行计算：
CalMemory(t,impressCount)＝CalTimeFactor(t)×CalCountFactor(impressCount)
Figure PCTCN2017114633-appb-000011
若impressCount≠0，则CalCountFactor(impressCount)＝1，否则CalCountFactor(impressCount)＝0；
CalMemory(t,impressCount)：为记忆强度的计算函数，其计算方法如下：
其中：
Figure PCTCN2017114633-appb-000012
使用了艾宾浩斯遗忘曲线，记忆优劣参数的默认值设置为60。
CalCountFactor(impressCount)中impressCount是时间t用户收到了几次该行业的广告，所述impressCount取值为0或1或大于1的整数。
艾宾浩斯遗忘曲线是用于表述记忆中的中长期记忆的遗忘率的一种曲线。这一曲线最早由心理学家赫尔曼·艾宾浩斯通过自己的实验提出。在这一实验中，艾宾浩斯使用了一些毫无意义的字母组合。通过记忆这些字 母组合，并在一系列时间间隔后检查遗忘率，得到了这一曲线。因此，这一曲线又叫艾宾浩斯遗忘曲线。
遗忘曲线的标准形式通常为：
Figure PCTCN2017114633-appb-000013
其中：R表示被记忆强度；S为记忆优劣参数；t是查看时间，T为当前时间。从上述公式不难看出，记忆强度越高，被记忆的内容随时间的衰减就越慢。记忆时间越长，被记忆内容就越低。
值得注意的是：艾宾浩斯遗忘曲线可以用其他的函数形式来替代，不一定是
Figure PCTCN2017114633-appb-000014
这样的形式，也可以使用多项式，幂函数等其他函数来逼近。所述衰减因子可以根据实际情况选择是否使用，取决于对用户体验的保护程度。
所述记忆强度的计算，也可以仅使用最近一次看到该内容的时间以简化计算。相比使用过去一段时间所有看到该内容的“记忆内容”加和。
各个维度的新鲜度分值计算出来之后，就可以根据各维度的权重，得到全局的新鲜度分值。
以上是以行业为预定维度，在一些示例中，所述预定维度很多，不局限于上述行业。
图4所示的为预定维度和记忆内容的列表。预定维度包括广告、广告主、广告物、行业、素材等。在图4中是以手机A的广告为例进行说明的。
在一些示例中，广告的标的物除了广告物还可包括广告服务。广告物通常对应于实体物品，所述广告服务可为各种不涉及实体物品的服务，例如，保洁服务等。
示例二：
在本实施例中以广告作为推广信息及以行业维度为例进行说明。如图3所示，本示例提供一种推广信息的发送方法，包括：
步骤S1：拉取广告的请求；
步骤S2：获取候选广告列表；
步骤S3：判断候选广告列表中是否还有未计算新鲜度的广告，若无，则流程结束，若有，则进入步骤S4；
步骤S4：选取一条未计算新鲜度的候选广告；
步骤S5：提取用户查看广告的历史数据，这里的历史数据可包括用户查看过的广告所涉及的行业。
步骤S6.1：计算次数衰减；
步骤S6.2：计算用户记忆强度；
步骤S7：计算行业新鲜度。
示例三
如图5所示，本示例提供一种推广信息的新鲜度确定装置的一个可选的硬件结构示意图，包括处理器11、输入/输出接口13(例如显示屏、触摸屏、扬声器)，存储介质14以及网络接口12，组件可以经系统总线15连接通信。相应地，存储介质14中均存储有用于执行本发明实施例记载的推广信息的新鲜度确定方法可执行指令。图5中示出的各硬件模块可根据需要部分实施、全部实施或实施其他的硬件模块，各类型硬件模块的数量可以为一个或多个，各硬件模块可以在同一地理位置实施，或者分布在不同的地理位置实施，可以用于执行上述图1或图3所示的推广信息的新鲜度确定方法的至少之一。
示例四：
如图6所示，本示例以广告作为推广信息进行举例说明，提供一种广告推广系统。
所述广告系统包括：
广告请求端，可对应于前述用户或用户终端，可用于发送广告请求；
接入服务器集群，可用于供广告请求端的接入；
检索服务器集群，与所述接入服务器集群连接，可接收接入服务器集群根据广告请求提供的检索参数；
检索服务器集群，将从存储服务器集群获取历史数据，将根据历史数据进行新鲜度计算，并根据新鲜度及检索参数，检索广告，将检索到广告以广告数据的方式，通过接入服务器集群返回给广告请求端。
在本示例中所述集群为包括至少多台服务器的机群。
本发明实施例还提供一种电子设备，该电子设备可为服务器，该服务器可位于网络侧，可以用于向终端发送推广信息。
所述服务器可包括：存储器及处理器。
所述存储器可为各种类型的存储器，例如，随机存储器、只读存储器、闪存等存储器件，可配置为存储各种信息，例如，存储计算机程序等计算机可执行指令；
所述处理器，可通过集成电路总线等与所述存储器连接，可以读取存储在计算机可执行指令，可通过执行所述计算机可执行指令，实现前述一个或多个技术方案提供推广信息的新鲜度确定方法。
所述处理器可为应用处理器AP(AP，Application Processor)、中央处理器(CPU，Central Processing Unit)、数字信号处理器(DSP，Digital Signal Processor)、可编程门阵列(FPGA，Field Programmable Gate Array)或专用集成电路等具有信息处理的信息处理模组或组件。
所述处理器通过执行可计算机可执行指令，可实现步骤可包括以下步骤：
获取用户查看推广信息的历史数据；
根据所述历史数据，确定第一推广信息的记忆强度，其中，所述第一推广信息与第二推广信息在预定维度内相关联；所述第二推广信息为当前待推广的信息；所述第一推广信息为用户已查看的推广信息；所述记忆强 度用于指示所述用户记忆所述第一推广信息的强度；
根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，其中，所述新鲜度用于指示所述第二推广信息的陌生程度。
可选地，所述处理器还可以通过执行计算机可执行指令，实现以下步骤：
根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度。
可选地，所述处理器还可以通过执行计算机可执行指令，实现以下步骤：
所述根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度，包括：
依据如下公式计算所述预定维度上的新鲜度FreshScore，
Figure PCTCN2017114633-appb-000015
其中，所述α为拟合系数，所述CalMemory为所述记忆强度；所述β为衰减因子。
可选地，所述处理器还可以通过执行计算机可执行指令，实现以下步骤：
根据所述历史数据，确定在预定时间范围是否有查看过所述第一推广信息；
当在所述预定时间范围内未查看过所述第一推广信息，则所述记忆强度为指定值；
或者，
当在所述预定时间范围内有查看过所述第一推广信息时，确定所述预定时间范围内查看所述第一推广信息的查看时间及查看次数；
根据所述查看次数及所述查看时间，确定所述记忆强度。
本发明实施例还提供一种计算机存储介质，所述计算机存储介质存储有计算机可执行指令，所述计算机可执行指令被执行后，可以实现前述一个或多个技术方案提供的推广信息的新鲜度确定方法，例如，至少可执行以下步骤：
获取用户查看推广信息的历史数据；
根据所述历史数据，确定第一推广信息的记忆强度，其中，所述第一推广信息与第二推广信息在预定维度内相关联；所述第二推广信息为当前待推广的信息；所述第一推广信息为用户已查看的推广信息；所述记忆强度用于指示所述用户记忆所述第一推广信息的强度；
根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，其中，所述新鲜度用于指示所述第二推广信息的陌生程度。
可选地，所述根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，包括：
根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度。
可选地，所述根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度，包括：
依据如下公式计算所述预定维度上的新鲜度FreshScore，
Figure PCTCN2017114633-appb-000016
其中，所述α为拟合系数，所述CalMemory为所述记忆强度；所述β为衰减因子。
所述计算机存储介质可包括：移动存储设备、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。所述计算机存储介质可为非瞬间存储介质或非易失性存储介质。
在本申请所提供的几个实施例中，应该理解到，所揭露的设备和方法，可以通过其它的方式实现。以上所描述的设备实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，如：多个单元或组件可以结合，或可以集成到另一个系统，或一些特征可以忽略，或不执行。另外，所显示或讨论的各组成部分相互之间的耦合、或直接耦合、或通信连接可以是通过一些接口，设备或单元的间接耦合或通信连接，可以是电性的、机械的或其它形式的。
上述作为分离部件说明的单元可以是、或也可以不是物理上分开的，作为单元显示的部件可以是、或也可以不是物理单元，即可以位于一个地方，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全部单元来实现本实施例方案的目的。
另外，在本发明各实施例中的各功能单元可以全部集成在一个处理模块中，也可以是各单元分别单独作为一个单元，也可以两个或两个以上单元集成在一个单元中；上述集成的单元既可以采用硬件的形式实现，也可以采用硬件加软件功能单元的形式实现。
本领域普通技术人员可以理解：实现上述方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成，前述的程序可以存储于一计算机可读取存储介质中，该程序在执行时，执行包括上述方法实施例的步骤。
以上所述，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应以所述权利要求的保护范围为准。
工业实用性
本发明实施例中在向用户所持有的终端之前，获取用户之前查看第一推广信息的历史数据，根据第一推广信息的历史数据，确定出用户对第一推广信息的记忆强度，并基于该记忆强度进一步计算出用户对待推广的第二推广信息的新鲜度，这种新鲜度可表征用户对第二推广信息的陌生程度。这样的话，就可以根据该新鲜度确定是否需要向对应的终端推送第二推广信息，和/或推送第二推广信息的时间等推广参数。这样的话，显然可以减少同一个或同一类推广信息的反复推送，并减少用户对推广信息的反感，从而提升用户体验及用户使用满意度；故具有积极的工业效果。与此同时，具有实现简便的特点，可在工业上大规模推广应用。

Claims (17)

  1. 一种推广信息的新鲜度确定方法，应用于信息处理平台，包括：
    获取用户查看推广信息的历史数据；
    根据所述历史数据，确定第一推广信息的记忆强度，其中，所述第一推广信息与第二推广信息在预定维度内相关联；所述第二推广信息为当前待推广的信息；所述第一推广信息为用户已查看的推广信息；所述记忆强度用于指示所述用户记忆所述第一推广信息的强度；
    根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，其中，所述新鲜度用于指示所述第二推广信息的陌生程度。
  2. 根据权利要求1所述的方法，其中，
    所述根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，包括：
    根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度。
  3. 根据权利要求2所述的方法，其中，
    所述根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度，包括：
    依据如下公式计算所述预定维度上的新鲜度FreshScore，
    Figure PCTCN2017114633-appb-100001
    其中，所述α为拟合系数，所述CalMemory为所述记忆强度；所述β为衰减因子。
  4. 根据权利1至3任一项所述的方法，其中，
    所述根据所述历史数据，确定所述第一推广信息的记忆强度，包括：
    根据所述历史数据，确定在预定时间范围是否有查看过所述第一推广 信息；
    当在所述预定时间范围内未查看过所述第一推广信息，则所述记忆强度为指定值；
    或者，
    当在所述预定时间范围内有查看过所述第一推广信息时，确定所述预定时间范围内查看所述第一推广信息的查看时间及查看次数；
    根据所述查看次数及所述查看时间，确定所述记忆强度。
  5. 根据权利要求4所述的方法，其中，
    所述根据所述查看次数及所述查看时间，确定所述记忆强度，包括：
    确定所述用户的记忆力优劣参数；
    根据所述查看次数、所述查看时间及所述记忆力优劣参数，确定所述记忆强度；
    或者，
    根据所述查看时间与当前时间的时间差，确定单次查看的单次记忆强度；
    根据所述单次记忆强度，确定整体记忆强度；
    根据所述查看次数，确定在所述预定时间范围内的衰减因子；
    结合所述整体记忆强度及所述衰减因子，确定所述记忆强度。
  6. 根据权利要求5所述的方法，其中，
    所述确定所述用户的记忆力优劣参数，包括：
    获取所述用户的用户属性；
    根据所述用户属性，确定所述记忆力优劣参数。
  7. 根据权利要求1至3任一项所述的方法，其中，
    所述方法还包括：
    根据多个所述预定维度的新鲜度，确定综合新鲜度。
  8. 根据权利要求7所述的方法，其中，
    所述根据多个所述预定维度的新鲜度，确定综合新鲜度，包括：
    根据如下公式计算所述综合新鲜度FreshScore：
    FreshScore＝∑wi×FreshScorei
    其中，所述FreshScorei为第i个所述预定维度的新鲜度；所述wi为第i个所述预定维度的权值；其中，所述i的取值为不小于1的整数。
  9. 一种推广信息的新鲜度确定装置，包括：
    获取单元，配置为获取用户查看推广信息的历史数据；
    第一确定单元，配置为根据所述历史数据，确定第一推广信息的记忆强度，其中，所述第一推广信息与第二推广信息在预定维度内相关联；所述第二推广信息为当前待推广的信息；所述第一推广信息为用户已查看的推广信息；所述记忆强度用于指示所述用户记忆所述第一推广信息的强度；
    第二确定单元，配置为根据所述记忆强度，确定所述第二推广信息在所述预定维度上的新鲜度，其中，所述新鲜度用于指示所述第二推广信息的陌生程度。
  10. 根据权利要求9所述的装置，其中，
    所述第二确定单元，配置为根据所述记忆强度及衰减因子，确定所述第二推广信息在所述预定维度上的新鲜度。
  11. 根据权利要求10所述的装置，其中，
    所述第二确定单元，配置为依据如下公式计算所述预定维度上的新鲜度FreshScore，
    Figure PCTCN2017114633-appb-100002
    其中，所述α为拟合系数，所述CalMemory为所述记忆强度；所述β为衰减因子。
  12. 根据权利要求9至11任一项所述的装置，其中，
    所述第一确定单元，配置为根据所述历史数据，确定在预定时间范围是否有查看过所述第一推广信息；当在所述预定时间范围内未查看过所述第一推广信息，则所述记忆强度为指定值；或者，
    所述第一确定单元，配置为当在所述预定时间范围内有查看过所述第一推广信息时，确定所述预定时间范围内查看所述第一推广信息的查看时间及查看次数；根据所述查看次数及所述查看时间，确定所述记忆强度。
  13. 根据权利要求12所述的装置，其中，
    所述第二确定单元，配置为确定所述用户的记忆力优劣参数；根据所述查看次数、所述查看时间及所述记忆力优劣参数，确定所述记忆强度；
    或者，
    所述第二确定单元，配置为根据所述查看时间与当前时间的时间差，确定单次查看的单次记忆强度；根据所述单次记忆强度，确定整体记忆强度；根据所述查看次数，确定在所述预定时间范围内的衰减因子；结合所述整体记忆强度及所述衰减因子，确定所述记忆强度。
  14. 根据权利要求13所述的装置，其中，
    所述第二确定单元，配置为获取所述用户的用户属性；根据所述用户属性，确定所述记忆力优劣参数。
  15. 根据权利要求9至11任一项所述的装置，其中，
    所述装置还包括：
    第三确定单元，配置为根据多个所述预定维度的新鲜度，确定综合新鲜度。
  16. 一种电子设备，其中，包括：存储器及处理器；
    所述存储器，配置为存储有计算机可执行指令；
    所述处理器，与所述存储器连接，配置为读取并执行所述计算机可执行指令，以实现权利要求1至8任一项提供的推广信息的新鲜度确定方法。
  17. 一种计算机存储介质，所述计算机存储介质中存储有计算机可执行指令，所述计算机可执行指令用于执行权利要求1至8任一项提供的推广信息的新鲜度确定方法。
PCT/CN2017/114633 2017-01-17 2017-12-05 推广信息新鲜度确定方法及装置、电子设备及存储介质 WO2018133566A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710036301.3 2017-01-17
CN201710036301.3A CN106777354A (zh) 2017-01-17 2017-01-17 推广信息新鲜度确定方法及装置

Publications (1)

Publication Number Publication Date
WO2018133566A1 true WO2018133566A1 (zh) 2018-07-26

Family

ID=58944486

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/114633 WO2018133566A1 (zh) 2017-01-17 2017-12-05 推广信息新鲜度确定方法及装置、电子设备及存储介质

Country Status (2)

Country Link
CN (1) CN106777354A (zh)
WO (1) WO2018133566A1 (zh)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106777354A (zh) * 2017-01-17 2017-05-31 腾讯科技（深圳）有限公司 推广信息新鲜度确定方法及装置
CN110147803B (zh) * 2018-02-08 2022-02-18 北大方正集团有限公司 用户流失预警处理方法与装置
CN108596711B (zh) * 2018-03-28 2021-04-23 阿里巴巴（中国）有限公司 应用推荐方法、装置及电子设备
CN111062734B (zh) * 2018-10-16 2023-08-25 北京字节跳动网络技术有限公司 一种投放信息重排序的方法、装置、电子设备及可读介质
CN110389801B (zh) * 2019-06-05 2022-08-30 北京梧桐车联科技有限责任公司 一种信息处理方法、装置、车载设备及存储介质
CN112291297B (zh) * 2020-09-04 2022-04-26 腾讯科技（深圳）有限公司 资讯数据的处理方法、装置、存储介质以及电子设备
CN117235380B (zh) * 2023-10-07 2024-05-14 广州禧闻信息技术有限公司 一种文物数字云展全网普及度分析系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105005582A (zh) * 2015-06-17 2015-10-28 深圳市腾讯计算机系统有限公司 多媒体信息的推荐方法及装置
CN105787061A (zh) * 2016-02-29 2016-07-20 广东顺德中山大学卡内基梅隆大学国际联合研究院 信息推送方法
CN105843902A (zh) * 2016-03-23 2016-08-10 乐视网信息技术（北京）股份有限公司 交互信息排序方法及装置
US20160315904A1 (en) * 2014-05-23 2016-10-27 Tencent Technology (Shenzhen) Company Limited Control method and apparatus for data display
CN106777354A (zh) * 2017-01-17 2017-05-31 腾讯科技（深圳）有限公司 推广信息新鲜度确定方法及装置

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
AU2003210929A1 (en) * 2002-02-06 2003-09-02 Mintel Learning Technologies, Inc. A system and method to optimize human associative memory and to enhance human memory
WO2013118198A1 (ja) * 2012-02-09 2013-08-15 パナソニック株式会社 お薦めコンテンツ提供装置、お薦めコンテンツ提供プログラムおよびお薦めコンテンツ提供方法
CN103368992B (zh) * 2012-03-28 2017-05-03 阿里巴巴集团控股有限公司 一种信息推送方法及装置
CN102779310A (zh) * 2012-06-25 2012-11-14 亿赞普(北京)科技有限公司 一种广告播放方法及系统
CN103853965A (zh) * 2014-03-11 2014-06-11 广州杰赛科技股份有限公司 基于指纹验证的娱乐系统
CN105827510B (zh) * 2016-03-31 2019-03-01 维沃移动通信有限公司 用于移动终端的推送消息处理方法以及移动终端

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160315904A1 (en) * 2014-05-23 2016-10-27 Tencent Technology (Shenzhen) Company Limited Control method and apparatus for data display
CN105005582A (zh) * 2015-06-17 2015-10-28 深圳市腾讯计算机系统有限公司 多媒体信息的推荐方法及装置
CN105787061A (zh) * 2016-02-29 2016-07-20 广东顺德中山大学卡内基梅隆大学国际联合研究院 信息推送方法
CN105843902A (zh) * 2016-03-23 2016-08-10 乐视网信息技术（北京）股份有限公司 交互信息排序方法及装置
CN106777354A (zh) * 2017-01-17 2017-05-31 腾讯科技（深圳）有限公司 推广信息新鲜度确定方法及装置

Also Published As

Publication number Publication date
CN106777354A (zh) 2017-05-31

Similar Documents

Publication Publication Date Title
WO2018133566A1 (zh) 推广信息新鲜度确定方法及装置、电子设备及存储介质
US11315142B2 (en) Method and system for correlating social media conversions
US20210110428A1 (en) Click-Through Prediction for Targeted Content
CN108763502B (zh) 信息推荐方法和系统
US10635710B2 (en) Methods, systems, and media for presenting recommended content based on social cues
US11620678B2 (en) Advertising method, device and system, and computer-readable storage medium
CN107222566B (zh) 信息推送方法、装置及服务器
US11533356B2 (en) Establishing a trust association
KR102347083B1 (ko) 소셜 미디어를 이용하는 사용자의 통계를 추정하는 방법 및 장치
US9356806B2 (en) Prioritizing messages within a message network
US20150289021A1 (en) System and method for collecting viewer opinion information
CN106341312B (zh) 社交信息显示方法、系统和服务器
US9571597B2 (en) Multimedia file push method and apparatus
US20230089961A1 (en) Optimizing content distribution using a model
CN105095465B (zh) 一种信息推荐方法、系统及装置
CN114747227A (zh) 跨人口统计群体估计普查级受众规模和总印象持续时间的方法、系统和装置
CN103997662A (zh) 节目推送的方法及系统
CN110782286A (zh) 广告推送方法、装置、服务器及计算机可读存储介质
CN110866040A (zh) 用户画像生成方法、装置和系统
JP5813052B2 (ja) 情報処理装置、方法及びプログラム
CN108710656B (zh) 一种内容推送方法和装置
CN113051481B (zh) 内容推荐方法、装置、电子设备和介质
CN107341172B (zh) 视频收益计算建模装置与方法及视频推荐装置与方法
CN111125501B (zh) 用于处理信息的方法和装置
US20140337118A1 (en) Method and system for generating social signal vocabularies

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: 17892482

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: 17892482

Country of ref document: EP

Kind code of ref document: A1