CN112650921A - Object recommendation method, device, equipment and storage medium - Google Patents

Object recommendation method, device, equipment and storage medium Download PDF

Info

Publication number
CN112650921A
CN112650921A CN202011523881.7A CN202011523881A CN112650921A CN 112650921 A CN112650921 A CN 112650921A CN 202011523881 A CN202011523881 A CN 202011523881A CN 112650921 A CN112650921 A CN 112650921A
Authority
CN
China
Prior art keywords
target
user behavior
target object
recommendation
browsing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202011523881.7A
Other languages
Chinese (zh)
Other versions
CN112650921B (en
Inventor
王帅帅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ping An Life Insurance Company of China Ltd
Original Assignee
Ping An Life Insurance Company of China Ltd
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 Ping An Life Insurance Company of China Ltd filed Critical Ping An Life Insurance Company of China Ltd
Priority to CN202011523881.7A priority Critical patent/CN112650921B/en
Publication of CN112650921A publication Critical patent/CN112650921A/en
Application granted granted Critical
Publication of CN112650921B publication Critical patent/CN112650921B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G06F16/90—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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08—Insurance

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Technology Law (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The embodiment of the invention discloses an object recommendation method, an object recommendation device, object recommendation equipment and a storage medium, wherein the object recommendation method comprises the steps of collecting user behavior data from a client; performing statistical processing on the user behavior data, and storing at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, wherein each user behavior record comprises a browsing statistical value aiming at one object in an object set; acquiring structural characteristics of a target behavior list, and determining a target analysis strategy aiming at least one user behavior record based on the structural characteristics; calculating at least one user behavior record based on a target analysis strategy to obtain an interest score for a target object; and determining a recommendation strategy aiming at the target object according to the interest score, and recommending the target object to the client based on the recommendation strategy. By implementing the method, the object recommendation process has pertinence, and the object recommendation effect is improved.

Description

Object recommendation method, device, equipment and storage medium
Technical Field
The invention relates to the technical field of data analysis, in particular to an object recommendation method, device, equipment and storage medium.
Background
Object recommendation is generally used for recommending different objects (such as insurance products, fund products and the like) to various users, so that the users can acquire detailed information of the objects, and an object recommendation mode related in an object recommendation system (such as a product sale system) is generally full recommendation, namely, the objects are recommended to all the users; for example, in the insurance product selling process, all users are recommended insurance products that need to be sold.
However, the above-mentioned total recommendation process needs to consume a large amount of recommendation resources (such as human resources, network resources, etc.), and since the recommended object has no pertinence, the degree of matching between the recommended object and the object expected by the user is often low, resulting in poor recommendation efficiency.
Disclosure of Invention
The embodiment of the invention provides an object recommendation method, device, equipment and storage medium, which can make the object recommendation process have pertinence and improve the object recommendation effect.
In one aspect, an embodiment of the present invention provides an object recommendation method, where the object processing method includes:
collecting user behavior data from a client, wherein the user behavior data is generated by browsing behaviors of a user operating the client on each object in an object set;
performing statistical processing on the user behavior data, and storing at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, wherein each user behavior record comprises a browsing statistical value aiming at one object in an object set;
acquiring structural characteristics of a target behavior list, and determining a target analysis strategy aiming at least one user behavior record based on the structural characteristics;
calculating at least one user behavior record based on a target analysis strategy to obtain an interest score aiming at a target object, wherein the target object is any object in an object set;
and determining a recommendation strategy aiming at the target object according to the interest score, and recommending the target object to the client based on the recommendation strategy.
In one aspect, an embodiment of the present invention provides an object recommendation apparatus, where the object processing apparatus includes:
the acquisition module is used for acquiring user behavior data from the client, wherein the user behavior data is generated by browsing behaviors of a user operating the client on each object in the object set;
the processing module is used for carrying out statistical processing on the user behavior data;
the storage module is used for storing at least one user behavior record obtained based on user behavior data statistics in a preset behavior list to obtain a target behavior list, wherein each user behavior record comprises a browsing statistic value aiming at one object in an object set;
the acquisition module is also used for acquiring the structural characteristics of the target behavior list;
a determination module for determining a target analysis strategy for at least one user behavior record based on the structural features;
the calculation module is used for calculating at least one user behavior record based on a target analysis strategy to obtain an interest score aiming at a target object, wherein the target object is any one object in an object set;
a determination module for determining a recommendation strategy for the target object according to the interest score;
and the recommending module is used for recommending the target object to the client based on the recommending strategy.
In one aspect, an embodiment of the present invention provides an object processing apparatus, where the object processing apparatus includes a processor, an input interface, an output interface, and a memory, where the processor, the input interface, the output interface, and the memory are connected to each other, where the memory is used to store a computer program, and the computer program includes program instructions, and the processor is configured to call the program instructions to execute the object recommendation method.
In one aspect, an embodiment of the present invention provides a computer-readable storage medium, in which a computer program is stored, the computer program including program instructions, which, when executed by a processor, cause the processor to execute the object recommendation method.
In the embodiment of the invention, user behavior data are collected from a client, and the user behavior data are generated by browsing behaviors of a user operating the client on each object in an object set; performing statistical processing on the user behavior data, and storing at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, wherein each user behavior record comprises a browsing statistical value aiming at one object in an object set; acquiring structural characteristics of a target behavior list, and determining a target analysis strategy aiming at least one user behavior record based on the structural characteristics; calculating at least one user behavior record based on a target analysis strategy to obtain an interest score aiming at a target object, wherein the target object is any object in an object set; and determining a recommendation strategy aiming at the target object according to the interest score, and recommending the target object to the client based on the recommendation strategy. By implementing the method, the interest score aiming at the target object can be determined from the user behavior data, the recommendation strategy aiming at the target object is determined based on the interest score, and the target object is recommended based on the recommendation strategy, so that the object recommendation process has pertinence, and the object recommendation effect is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a schematic diagram of an architecture of an object recommendation system according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating an object recommendation method according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating another object recommendation method according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an object recommendation apparatus according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an object recommendation device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention relates to user behavior data, wherein the user behavior data is generated according to browsing behaviors of a user operating a client to each object in an object set; wherein, the client may include any one of the following: application independent (APP), API (Application Programming Interface), SDK (Software Development Kit); for example, the user behavior data of the user a is generated according to the browsing behavior of the user a operating the insurance product APP on each insurance product in the set of insurance products.
In the embodiment of the invention, the user behavior data comprises at least one operation behavior record, and each operation behavior record comprises a starting time point and an ending time point of a browsing behavior of a user operating a client to an object in an object set; as shown in table 1, the user behavior data includes 5 operation behavior records, each operation behavior record includes a start time point and an end time point of a browsing behavior of the user a operating the client with respect to one object in the object set; for example, a first operational behavior record includes a start time point and an end time point of a browsing behavior of the user a for insurance product a, a second operational behavior record includes a start time point and an end time point of a browsing behavior of the user a for insurance product B, and a fourth operational behavior record includes a start time point and an end time point of a browsing behavior of the user a for insurance product C. The user behavior data may be used to indicate the degree of interest of the user in each object, indicate that the user has a higher interest in the target object if the user browses the target object (any object in the object set) more times or for a longer browsing duration within a target time range (e.g., a day, a week, a month, etc.), and indicate that the user has a lower interest in the target object if the user browses the target object within the target time range less times or for a shorter browsing duration; for example, if the browsing times of the user a browsing the insurance product a in one day is 10 times, the user a is indicated to be very interested in the insurance product a, if the browsing times of the user a browsing the insurance product a in one day is 5 times, the user a is indicated to have a higher interest in the insurance product a, and if the browsing times of the user a browsing the insurance product a in one day is 1 time, the user a is indicated to have a lower interest in the insurance product a; for another example, if the browsing time of the user a browsing the insurance product a in one day is 10 minutes, the user a is indicated to be very interested in the insurance product a, if the browsing time of the user a browsing the insurance product a in one day is 5 minutes, the user a is indicated to have a higher interest in the insurance product a, and if the browsing time of the user a browsing the insurance product a in one day is 1 minute, the user a is indicated to have a lower interest in the insurance product a.
TABLE 1 user behavior data
Object Starting time point of browsing behavior End time point of browsing behavior
Insurance products A 2020-09-01 13:45 2020-09-01 13:48
Insurance product B 2020-09-01 13:49 2020-09-01 13:52
Insurance products A 2020-09-01 14:21 2020-09-01 14:23
Insurance product C 2020-09-01 14:25 2020-09-01 14:26
Insurance products A 2020-09-01 14:27 2020-09-01 14:28
The embodiment of the invention provides an object recommendation scheme, wherein the object recommendation scheme is used for analyzing and processing user behavior data to obtain interest scores of all objects in an object set, formulating a recommendation strategy for all objects based on the interest scores of all objects in the object set, and recommending the objects based on the recommendation strategies of all objects, so that the object recommendation process has pertinence, and the object recommendation effect is improved.
Fig. 1 is a schematic architecture diagram of an object recommendation system in an embodiment of the present invention, and as shown in fig. 1, an object recommendation system 10 may include a client 101 and a server 102. The client is installed and operated in a terminal, and the terminal can comprise electronic equipment such as a smart phone, a tablet computer, a digital audio and video player, an electronic reader, a handheld game console or vehicle-mounted electronic equipment; the server may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing cloud computing services, and the embodiment of the present invention is not limited herein.
In an object recommendation system composed of a client and a server, the object recommendation scheme may specifically be: the server collects user behavior data from the client, performs statistical processing on the user behavior data, and stores at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, wherein each user behavior record comprises a browsing statistic value aiming at one object in an object set; the server acquires the structural characteristics of the target behavior list, and determines a target analysis strategy aiming at least one user behavior record based on the structural characteristics, wherein the structural characteristics of the target behavior list comprise the number of lines or columns of the target behavior list; the server calculates at least one user behavior record based on a target analysis strategy to obtain an interest score aiming at a target object, wherein the target object is any object in an object set; and the terminal determines a recommendation strategy for the target object according to the interest score and recommends the target object to the client based on the recommendation strategy. In the implementation mode, the server determines the interest score aiming at the target object from the user behavior data, determines the recommendation strategy aiming at the target object based on the interest score, and recommends the target object based on the recommendation strategy, so that the object recommendation process has pertinence, and the object recommendation effect is improved.
Fig. 2 is a schematic flowchart of an object recommendation method according to an embodiment of the present invention, and as shown in fig. 2, the flow of the object recommendation method according to the embodiment may include the following steps S201 to S205:
s201, the server collects user behavior data from the client.
In the embodiment of the invention, the client can generate user behavior data according to the browsing behavior of a user operating the client on each object in the object set, and record the generated user behavior data in a log file of the client; and the server acquires the log file of the client and further acquires the user behavior data from the client.
S202, the server carries out statistical processing on the user behavior data, and at least one user behavior record obtained based on the user behavior data statistics is stored in a preset behavior list to obtain a target behavior list.
In the embodiment of the invention, after the server collects the user behavior data from the client, the server carries out statistical processing on the user behavior data, and at least one user behavior record obtained based on the user behavior data statistics is stored in a preset behavior list to obtain a target behavior list. Each user behavior record in the target behavior list comprises a browsing statistic value for one object in the object set, the target object is any one object in the object set, the target user behavior record is any one user behavior record in the target behavior list, and the target user behavior record comprises the browsing statistic value for the target object.
In an implementation manner, the specific implementation manner of the server performing statistical processing on the user behavior data to obtain the target user behavior record may be: the server determines a statistical interval, wherein the statistical interval is a target time range for counting the user behavior data; the server determines the number of operation behavior records aiming at the target object in the statistical interval from the user behavior data, and determines the number of the operation behavior records aiming at the target object as the browsing times of the target object; the server determines the browsing time of the target object according to the starting time point and the ending time point of the browsing behavior in the operation behavior record aiming at the target object; the server acquires a first weight corresponding to the browsing times and a second weight corresponding to the browsing duration, wherein the first weight and the second weight can be set according to an empirical value; the server carries out weighting processing on the browsing times of the target object based on the first weight to obtain the weighted browsing times of the target object; the server carries out weighting processing on the browsing duration of the target object based on the second weight to obtain the weighted browsing duration of the target object; and the server sums the weighted browsing duration and the weighted browsing times to obtain a browsing statistic value of the target object.
For example, the statistical interval is one day, and the server performs statistical processing on the user behavior data shown in table 1 as follows: the server determines that the number of the operation behavior records aiming at the insurance product A in one day is 3, and the browsing times of the insurance product A are 3 times; the server determines that the browsing time of the insurance product a is 3 minutes according to the start time point and the end time point of the browsing behavior of the insurance product a included in the first operation behavior record in table 1, determines that the browsing time of the insurance product a is 2 minutes according to the start time point and the end time point of the browsing behavior of the insurance product a included in the third operation behavior record in table 1, determines that the browsing time of the insurance product a is 1 minute according to the start time point and the end time point of the browsing behavior of the insurance product a included in the fifth operation behavior record in table 1, and determines that the browsing duration of the insurance product a is 6 minutes according to the 3 operation behavior records of the insurance product a in table 1 (3 minutes +2 minutes +1 minutes is 6 minutes); the server acquires that a first weight corresponding to the browsing times is 0.4 and a second weight corresponding to the browsing duration is 0.6; the server performs weighting processing on the browsing times of the insurance product A based on the first weight to obtain the weighted browsing times of the insurance product A, wherein the weighted browsing times of the insurance product A are 1.2 times (3 times multiplied by 0.4 times which are 1.2 times); the server performs weighting processing on the browsing duration of the insurance product A based on the second weight to obtain the weighted browsing duration of the insurance product A, which is 3.6 minutes (6 minutes × 0.6-3.6 minutes); and the server sums the weighted browsing duration and the weighted browsing times to obtain a browsing statistic value 4.8(1.2+3.6 is 4.8) of the insurance product A. According to the method, the server counts the user behavior data shown in the table 1 to obtain the browsing times and browsing duration of the insurance product B and the browsing times and browsing duration of the insurance product C, the table 2 shows a browsing statistical table obtained according to the user behavior data shown in the table 1, the browsing statistical table comprises the browsing times and browsing duration of 3 insurance products, the table 3 shows a target behavior list obtained by further counting the browsing statistical table shown in the table 2, and the target behavior list comprises the browsing statistical values of 3 insurance products.
TABLE 2 browse statistics
Object Number of browsing (times) Duration of browsing (minutes)
Insurance products A 3 6
Insurance product B 1 3
Insurance product C 1 1
TABLE 3 target behavior List
Object Browsing statistics
Insurance products A 4.8
Insurance product B 2.2
Insurance product C 0.6
S203, the server obtains the structural characteristics of the target behavior list and determines a target analysis strategy aiming at least one user behavior record based on the structural characteristics.
In the embodiment of the invention, after the server performs statistical processing on the user behavior data to obtain the target behavior list, the structural characteristics of the target behavior list are obtained, and the target analysis strategy for at least one user behavior record is determined based on the structural characteristics. The structural characteristics of the target behavior list include row and column information of the target behavior list, and the row and column information may include the number of rows or columns. The rank information of the target behavior list can be used for indicating the wide interest degree of the user; if the row and column information comprises more rows or columns, indicating that the interest of the user is wider; if the row and column information comprises fewer rows or columns, the interest of the user is indicated to be concentrated. For example, the number of rows in the target behavior list is 1-5 rows, indicating that the interest of the user is relatively concentrated; the line number of the target behavior list is 6-10 lines, and the interest of the user is indicated to be wide; the number of rows of the target behavior list is 10 or more, indicating that the interest of the user is very wide. The scoring model mentioned in the embodiment of the present invention may include at least one of: the system comprises an interest wide scoring model, an interest wide scoring model and an interest concentrated scoring model, wherein the 3 scoring models are all used for calculating at least one user behavior record included in a target behavior list to obtain an interest score; due to the difference of network structures, the scoring effects of the 3 scoring models are different, the scoring accuracy of the interest concentrated scoring model when scoring the target behavior list with concentrated interests and hobbies of the indicated user is high, the scoring accuracy of the wide interest scoring model when scoring the target behavior list with wide interests and hobbies of the indicated user is high, and the scoring accuracy of the wide interest scoring model when scoring the target behavior list with wide interests of the indicated user is high.
In one implementation, the objective analysis policy indicates a first scoring model and a second scoring model for computing the at least one user behavior record, and a first model weight corresponding to the first scoring model and a second model weight corresponding to the second scoring model. The specific implementation manner of the server determining the target analysis policy for the at least one user behavior record based on the structural features may be: and the server acquires a first scoring model and a second scoring model corresponding to the row and column information of the target behavior list from the 3 scores, and acquires a first model weight corresponding to the first scoring model and a second model weight corresponding to the second scoring model. Specifically, there is a correspondence between the rank information of the target behavior list and the scoring model, the model weight, taking table 4 as an example, table 4 shows a correspondence table between the rank information and the scoring model; if the row and column information of the target behavior list indicates that the row number (column number) of the target behavior list is 1-5 rows (columns), the first scoring model is an interest centralized scoring model, the second scoring model is a wider interest scoring model, and the first model weight corresponding to the first scoring model is greater than the second model weight corresponding to the second scoring model; if the row and column information of the target behavior list indicates that the row number (column number) of the target behavior list is 6-10 rows (columns), the first scoring model is an interest centralized scoring model, the second scoring model is a wider interest scoring model, and the first model weight corresponding to the first scoring model is smaller than the second model weight corresponding to the second scoring model; if the row and column information of the target behavior list indicates that the row number (column number) of the target behavior list is more than 10 rows (columns), the first scoring model is a wide interest scoring model, the second scoring model is a wide interest scoring model, and the weight of the first model corresponding to the first scoring model is smaller than the weight of the second model corresponding to the second scoring model. The server determines a model fusion mode of the first scoring model and the second scoring model based on the first model weight and the second model weight; and the server makes a target analysis strategy aiming at least one user behavior record based on a model fusion mode.
TABLE 4-corresponding relationship table between rank information and rating model
Figure BDA0002846581770000081
S204, the server calculates at least one user behavior record based on the target analysis strategy to obtain an interest score aiming at the target object.
In the embodiment of the invention, after the server determines the target analysis strategy aiming at the at least one user behavior record, the server calculates the at least one user behavior record based on the target analysis strategy to obtain the interest score aiming at the target object.
In an implementation manner, the specific implementation manner of calculating, by the server, at least one user behavior record based on the target analysis policy to obtain the interest score for the target object may be: the server calls a first scoring model to calculate at least one user behavior record to obtain a first interest score aiming at the target object; the server calls a second scoring model to calculate at least one user behavior record to obtain a second interest score aiming at the target object; if the first interest score and the second interest score both meet preset screening conditions, the first interest score is weighted by adopting the first model weight to obtain a first weighted interest score, the second interest score is weighted by adopting the second model weight to obtain a second weighted interest score, and the server sums the first weighted interest score and the second interest score to obtain an interest score for the target object. The preset screening condition means that the absolute value of the difference value between the first interest score and the second interest score is smaller than a difference threshold value; the absolute value of the difference between the first interest score and the second interest score is smaller than a difference threshold value, which indicates that the difference between the first interest score and the second interest score is smaller, and no calculation error exists when the first scoring model or the second scoring model calculates at least one user behavior record; the absolute value of the difference between the first interest score and the second interest score is greater than or equal to the difference threshold, indicating that the difference between the first interest score and the second interest score is large, and a calculation error may exist when at least one user behavior record is calculated by either the first scoring model or the second scoring model.
For example, the target behavior list is shown in table 3, the row and column information of the target behavior list is 3 rows, the server determines that the first scoring model is an interest centralized scoring model, the second scoring model is a broader scoring model, the first model weight of the first scoring model is 0.7, and the model weight of the second model is 0.3 from the 3 scoring models; the server calls a first scoring model to calculate that a first interest score for the target object is 0.63, and calls a second scoring model to calculate that a second interest score for the target object is 0.6; the absolute value 0.03 of the difference between the first interest score and the second interest score is smaller than a difference threshold value 0.05, the first interest score and the second interest score both meet preset screening conditions, the server performs weighting processing on the first interest score by adopting a first model weight 0.7 to obtain a first weighted interest score of 0.44, performs weighting processing on the second interest score by adopting a second model weight 0.3 to obtain a second weighted interest score of 0.18, and the server performs summation processing on the first weighted interest score and the second interest score to obtain an interest score of 0.62 for the target object.
S205, the server determines a recommendation strategy for the target object according to the interest score, and recommends the target object to the client based on the recommendation strategy.
In the embodiment of the invention, after the server obtains the interest score aiming at the target object, the recommendation strategy aiming at the target object is determined according to the interest score, and the target object is recommended to the client side based on the recommendation strategy. Wherein the recommendation policy of the target object may include at least one of: the recommendation frequency of the target object and the recommendation path of the target object; the recommendation frequency of the target object refers to the number of times of recommending the target object to the client within the target time range, and the recommendation path of the target object is a transmission path when the target object is recommended to the client.
In one implementation, the recommendation policy for the target object includes a recommendation frequency for the target object. When the interest score is larger than a score threshold value, the server determines that the recommendation frequency for the target object is a first recommendation frequency; when the interest score is smaller than or equal to the score threshold value, the server determines that the recommendation frequency aiming at the target object is a second recommendation frequency, and the first recommendation frequency is higher than the second recommendation frequency. For example, the interest score of 0.62 for the target object is greater than the score threshold of 0.5, and the server determines that the recommended frequency for the target object is 3 times per day; the interest score of the target object of 0.34 is less than the score threshold of 0.5, and the server determines that the recommendation frequency for the target object is 1 time per day.
In another implementation, the recommendation policy for the target object includes a recommendation path for the target object. When the interest score is larger than a score threshold value, determining that a recommended path for the target object is a first transmission path; and when the interest score is smaller than or equal to the score threshold value, determining that the recommended path for the target object is a second transmission path, wherein the transmission efficiency of the first transmission path is higher than that of the second transmission path. For example, the interest score of 0.62 of the target object is greater than the score threshold value of 0.5, and the server determines to recommend the target object by means of instant messaging; the interest score of the target object is 0.34 smaller than the score threshold value of 0.5, the server determines that the recommendation is carried out in a short message mode, and the transmission efficiency of the instant messaging message is higher than that of the short message.
In another implementation, the recommendation policy for the target object includes a recommendation frequency of the target object and a transmission path of the target object. When the interest score is larger than a score threshold value, the server determines that the recommendation frequency for the target object is a first recommendation frequency and the recommendation path for the target object is a first transmission path; when the interest score is smaller than or equal to the score threshold, the server determines that the recommendation frequency for the target object is a second recommendation frequency and the recommendation path for the target object is a second transmission path, the first recommendation frequency is higher than the second recommendation frequency, and the transmission efficiency of the first transmission path is higher than that of the second transmission path.
In the embodiment of the invention, a server collects user behavior data from a client, wherein the user behavior data is generated by browsing behaviors of a user operating the client on each object in an object set; the server carries out statistical processing on the user behavior data, and stores at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, wherein each user behavior record comprises a browsing statistic value aiming at one object in an object set; the server acquires the structural characteristics of the target behavior list and determines a target analysis strategy aiming at least one user behavior record based on the structural characteristics; the server calculates at least one user behavior record based on a target analysis strategy to obtain an interest score aiming at a target object, wherein the target object is any object in an object set; and the server determines a recommendation strategy aiming at the target object according to the interest score and recommends the target object to the client based on the recommendation strategy. By implementing the method, the interest score aiming at the target object can be determined from the user behavior data, the recommendation strategy aiming at the target object is determined based on the interest score, and the target object is recommended based on the recommendation strategy, so that the object recommendation process has pertinence, and the object recommendation effect is improved.
Fig. 3 is a schematic flowchart of another object recommendation method provided in an embodiment of the present invention, and as shown in fig. 3, the flowchart of the object recommendation method in the embodiment may include the following steps S301 to S308:
s301, the server collects user behavior data from the client.
S302, the server carries out statistical processing on the user behavior data, and at least one user behavior record obtained based on the user behavior data statistics is stored in a preset behavior list to obtain a target behavior list.
In the embodiment of the present invention, the execution process of step S301 is the same as the execution process of step S201 in the embodiment shown in fig. 2, the execution process of step S302 is the same as the execution process of step S202 in the embodiment shown in fig. 2, and the specific execution process may refer to the embodiment shown in fig. 2, which is not described herein again.
S303, the server acquires the structural characteristics of the target behavior list, wherein the structural characteristics of the target behavior list comprise row and column information of the target behavior list.
In the embodiment of the invention, after the server carries out statistical processing on the user behavior data to obtain the target behavior list, the structural characteristics of the target behavior list are obtained, the structural characteristics of the target behavior list comprise row and column information of the target behavior list, and the row and column information of the target behavior list comprises the row number or column number of the target behavior list. The rank information of the target behavior list can be used for indicating the wide interest degree of the user; if the row and column information comprises more rows or columns, indicating that the interest of the user is wider; if the row and column information comprises fewer rows or columns, the interest of the user is indicated to be concentrated. For example, the number of rows in the target behavior list is 1-5 rows, indicating that the interest of the user is relatively concentrated; the line number of the target behavior list is 6-10 lines, and the interest of the user is indicated to be wide; the number of rows of the target behavior list is 10 or more, indicating that the interest of the user is very wide.
S304, the server obtains the initial weight corresponding to the rank information.
In the embodiment of the invention, after the server acquires the structural characteristics (namely row and column information) of the target behavior list, the server acquires the initial weight corresponding to the row and column information; the row and column information of the target behavior list is different, the initial weight corresponding to the target behavior list is also different, table 5 shows a corresponding relationship table between the row and column information and the initial weight, as shown in table 5, if the row and column information of the target behavior list indicates that the row number (column number) of the target behavior list is 1-5 rows (columns), the initial weight corresponding to the row and column information is 0.9; if the row and column information of the target behavior list indicates that the row number (column number) of the target behavior list is 6-10 rows (columns), the initial weight corresponding to the row and column information is 0.8; if the row and column information of the target behavior list indicates that the row number (column number) of the target behavior list is more than 10 rows (columns), the initial weight corresponding to the row and column information is 0.7.
TABLE 5-correspondence between rank information and initial weight Table
Line and column information (number of lines/column) Initial weight
1-5 0.9
6-10 0.8
≥10 0.7
S305, the server determines a weighting mode for each browsing statistic value in at least one user behavior record based on the initial weight.
In the embodiment of the invention, after the server acquires the initial weight corresponding to the rank information, the server determines the weighting mode aiming at each browsing statistic value in at least one user behavior record based on the initial weight, and the weighting mode indicates the initial weight for calculating each browsing statistic value in at least one user behavior record.
S306, the server makes a target analysis strategy aiming at least one user behavior record based on a weighting mode.
In the embodiment of the invention, after determining the weighting mode for each browsing statistic value in at least one user behavior record based on the initial weight, the server makes a target analysis strategy for at least one user behavior record based on the weighting mode, wherein the target analysis strategy indicates that each browsing statistic value in at least one user behavior record is weighted by adopting the initial weight indicated by the weighting mode.
S307, the server calculates at least one user behavior record based on the target analysis strategy to obtain an interest score aiming at the target object.
In the embodiment of the invention, after the server makes a target analysis strategy for at least one user behavior record based on a weighting mode, the server calculates the at least one user behavior record based on the target analysis strategy to obtain the interest score for the target object.
In an implementation manner, the specific implementation manner of calculating, by the server, at least one user behavior record based on the target analysis policy to obtain the interest score for the target object may be: the server acquires a target user behavior record aiming at the target object from the target behavior list and determines a browsing statistic value aiming at the target object in the target user behavior record; the server acquires a target weighting mode of the browsing statistic value for the target object from the target analysis strategy, and performs weighting processing on the browsing statistic value of the target object based on the target weighting mode to obtain a weighted browsing statistic value for the target object; the server determines the total browsing statistics value of each object in the object set from the target behavior list; the server determines an interest score for the target object based on a ratio between the weighted browsing statistics and the sum of the browsing statistics.
For example, the target behavior list is shown in table 3, the server obtains a target user behavior record for the target object (insurance product a) from the target behavior list, and determines the browsing statistics value 4.8 for the target object in the target user behavior record; the row and column information of the target behavior list is 3 rows, and the target weighting mode aiming at the browsing statistic value of the target object indicates that the initial weight for calculating the browsing statistic value of the target object is 0.9; the server carries out weighting processing on the browsing statistic value of the target object based on a target weighting mode to obtain a weighted browsing statistic value 4.32 aiming at the target object; the server determines from the target behavior list a total of browsing statistics 7.6(4.8+2.2+0.6 ═ 7.6) for each object (3 insurance products) in the object set; the server determines to get an interest score of 0.59 for the target object based on the ratio between the weighted view statistics and the sum of the view statistics.
S308, the server determines a recommendation strategy for the target object according to the interest score, and recommends the target object to the client based on the recommendation strategy.
In the embodiment of the present invention, the execution manner of step S308 is the same as the execution process of step S205 in the embodiment shown in fig. 2, and the specific execution process can refer to the embodiment shown in fig. 2, which is not described herein again.
Further, the server may broadcast the target analysis policy and the recommendation policy, so that the nodes in the blockchain perform consensus check on the target analysis policy and the recommendation policy; and if the received consensus verification result indicates that the verification is passed, the server packs the target analysis strategy and the recommendation strategy into a block and issues the block to a block chain. The block chain is a set of decentralized infrastructure with distributed storage characteristics, and particularly is a data structure formed by data blocks in a linked list-like manner according to a time sequence, so that data which are in a sequential relationship and can be verified in a system can be safely stored, and the data cannot be tampered and counterfeited in a cryptographic manner. By the method, after the server obtains the structural characteristics of the target behavior list, the server can obtain the target analysis strategy for at least one user behavior record from the block chain, so that the server can calculate at least one user behavior record based on the target analysis strategy to obtain the interest score for the target object; after the server calculates the interest score for the target object, the server can acquire the recommendation strategy for the target object from the block chain, so that the server can recommend the target object to the client based on the recommendation strategy, the object recommendation efficiency is improved, the object recommendation effect is improved, and the traceability and the irreparability of the strategies (the target analysis strategy and the recommendation strategy) are ensured through the block chain.
In the embodiment of the invention, a server collects user behavior data from a client, wherein the user behavior data is generated by browsing behaviors of a user operating the client on each object in an object set; the server carries out statistical processing on the user behavior data, and stores at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, wherein each user behavior record comprises a browsing statistic value aiming at one object in an object set; the server acquires the structural characteristics of the target behavior list and determines a target analysis strategy aiming at least one user behavior record based on the structural characteristics; the server calculates at least one user behavior record based on a target analysis strategy to obtain an interest score aiming at a target object, wherein the target object is any object in an object set; and the server determines a recommendation strategy aiming at the target object according to the interest score and recommends the target object to the client based on the recommendation strategy. By implementing the method, the interest score aiming at the target object can be determined from the user behavior data, the recommendation strategy aiming at the target object is determined based on the interest score, and the target object is recommended based on the recommendation strategy, so that the object recommendation process has pertinence, and the object recommendation effect is improved.
The object recommending apparatus according to the embodiment of the present invention will be described in detail with reference to fig. 4. It should be noted that the object recommendation apparatus shown in fig. 4 is used for executing the method according to the embodiment of the present invention shown in fig. 2-3, for convenience of description, only the portion related to the embodiment of the present invention is shown, and details of the specific technology are not disclosed, and reference is made to the embodiment of the present invention shown in fig. 2-3.
Referring to fig. 4, which is a schematic structural diagram of an object recommendation device according to the present invention, the object recommendation device 40 may include: the system comprises an acquisition module 401, a processing module 402, a storage module 403, a determination module 404, a calculation module 405 and a recommendation module 406.
An obtaining module 401, configured to collect user behavior data from a client, where the user behavior data is generated by browsing behaviors of a user operating the client on each object in an object set;
a processing module 402, which performs statistical processing on the user behavior data;
a storage module 403, configured to store at least one user behavior record obtained based on user behavior data statistics in a preset behavior list to obtain a target behavior list, where each user behavior record includes a browsing statistic value for an object in an object set;
the obtaining module 401 is further configured to obtain structural features of the target behavior list;
a determining module 404, further configured to determine a target analysis policy for at least one user behavior record based on the structural features;
a calculating module 404, configured to calculate at least one user behavior record based on a target analysis policy to obtain an interest score for a target object, where the target object is any object in an object set;
a determination module 405, configured to determine a recommendation policy for the target object according to the interest score;
and a recommending module 406, configured to recommend the target object to the client based on the recommending policy.
In one implementation, the structural features include row and column information of the target behavior list, the row and column information including a number of rows or columns; the determining module 404 is specifically configured to:
acquiring an initial weight corresponding to the row and column information;
determining a weighting mode for each browsing statistic in at least one user behavior record based on the initial weight;
and formulating a target analysis strategy aiming at least one user behavior record based on a weighting mode.
In one implementation, the calculating module 405 is specifically configured to:
acquiring a target user behavior record aiming at a target object from the target behavior list, and determining a browsing statistic value aiming at the target object in the target user behavior record;
acquiring a target weighting mode of the browsing statistic value for the target object from a target analysis strategy, and performing weighting processing on the browsing statistic value of the target object based on the target weighting mode to obtain a weighted browsing statistic value for the target object;
determining a browsing statistic sum for each object in the object set from the target behavior list;
an interest score for the target object is determined based on a ratio between the weighted browsing statistics and the sum of the browsing statistics.
In one implementation, the target user behavior record is any one of the user behavior records in the target behavior list, and the target user behavior record includes browsing statistics for the target object; the processing module 402 is specifically configured to:
determining browsing times aiming at a target object and browsing duration aiming at the target object based on user behavior data, and acquiring a first weight corresponding to the browsing times and a second weight corresponding to the browsing duration;
weighting the browsing times of the target object based on the first weight to obtain the weighted browsing times of the target object;
weighting the browsing duration of the target object based on the second weight to obtain the weighted browsing duration of the target object;
and summing the weighted browsing duration and the weighted browsing times to obtain a browsing statistic value of the target object.
In one implementation, the target analysis strategy indicates a first scoring model and a second scoring model for calculating at least one user behavior record, and a first model weight corresponding to the first scoring model and a second model weight corresponding to the second scoring model; the calculating module 405 is specifically configured to:
calling a first scoring model to calculate at least one user behavior record to obtain a first interest score aiming at a target object;
calling a second scoring model to calculate at least one user behavior record to obtain a second interest score aiming at the target object;
if the first interest score and the second interest score both meet the preset screening condition, performing weighting processing on the first interest score by adopting a first model weight to obtain a first weighted interest score, and performing weighting processing on the second interest score by adopting a second model weight to obtain a second weighted interest score;
and summing the first weighted interest score and the second interest score to obtain an interest score for the target object.
In one implementation, the recommendation policy for the target object includes a recommendation frequency and a recommendation path for the target object; the recommending module 406 is specifically configured to:
if the interest score is larger than the score threshold value, determining that the recommendation frequency for the target object is a first recommendation frequency, and determining that the recommendation path for the target object is a first transmission path;
if the interest score is smaller than or equal to the score threshold, determining that the recommendation frequency for the target object is a second recommendation frequency and the recommendation path for the target object is a second transmission path; the first recommended frequency is higher than the second recommended frequency, and the transmission efficiency of the first transmission path is higher than that of the second transmission path.
In one implementation, the recommending module 406 is further configured to:
broadcasting the target analysis strategy and the recommendation strategy so that nodes in the block chain carry out consensus verification on the target analysis strategy and the recommendation strategy;
if the received consensus verification result indicates that the verification is passed, packaging the target analysis strategy and the recommendation strategy into a block;
the blocks are issued into a blockchain.
In the embodiment of the present invention, the obtaining module 401 collects user behavior data from the client, where the user behavior data is generated by browsing behaviors of a user operating the client on each object in an object set; the processing module 402 performs statistical processing on the user behavior data; the storage module 403 stores at least one user behavior record obtained based on user behavior data statistics in a preset behavior list to obtain a target behavior list, where each user behavior record includes a browsing statistics value for one object in an object set; the obtaining module 401 obtains the structural features of the target behavior list; the determination module 404 determines a target analysis policy for at least one user behavior record based on the structural features; the calculation module 405 calculates at least one user behavior record based on a target analysis policy to obtain an interest score for a target object, where the target object is any one object in an object set; the determination module 404 determines a recommendation policy for the target object according to the interest score; the recommendation module 406 recommends the target object to the client based on the recommendation policy. By implementing the method, the interest score aiming at the target object can be determined from the user behavior data, the recommendation strategy aiming at the target object is determined based on the interest score, and the target object is recommended based on the recommendation strategy, so that the object recommendation process has pertinence, and the object recommendation effect is improved.
Referring to fig. 5, a schematic structural diagram of an object recommendation device according to an embodiment of the present invention is provided. As shown in fig. 5, the object recommending apparatus may be the server 102 in the object recommending system shown in fig. 1, and includes: at least one processor 501, an input device 503, an output device 504, a memory 505, at least one communication bus 502. Wherein a communication bus 502 is used to enable connective communication between these components. The input device 503 may be a control panel, a microphone, or the like, and the output device 504 may be a display screen, or the like. The memory 505 may be a high-speed RAM memory or a non-volatile memory (non-volatile memory), such as at least one disk memory. The memory 505 may alternatively be at least one memory device located remotely from the processor 501. Wherein the processor 501 may be combined with the apparatus described in fig. 4, the memory 505 stores a set of program codes, and the processor 501, the input device 503, and the output device 504 call the program codes stored in the memory 505 to perform the following operations:
a processor 501, configured to collect user behavior data from a client, where the user behavior data is generated by browsing behaviors of a user operating the client on each object in an object set;
the processor 501 is configured to perform statistical processing on the user behavior data, and store at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, where each user behavior record includes a browsing statistic value for one object in the object set;
the processor 501 is further configured to obtain structural features of the target behavior list, and determine a target analysis policy for at least one user behavior record based on the structural features;
the processor 501 is configured to calculate at least one user behavior record based on a target analysis policy to obtain an interest score for a target object, where the target object is any object in an object set;
a processor 501 for determining a recommendation strategy for a target object according to the interest score;
and the processor 501 is configured to recommend the target object to the client based on the recommendation policy.
In one implementation, the structural features include row and column information of the target behavior list, the row and column information including a number of rows or columns; the processor 501 is specifically configured to:
acquiring an initial weight corresponding to the row and column information;
determining a weighting mode for each browsing statistic in at least one user behavior record based on the initial weight;
and formulating a target analysis strategy aiming at least one user behavior record based on a weighting mode.
In one implementation, the processor 501 is specifically configured to:
acquiring a target user behavior record aiming at a target object from the target behavior list, and determining a browsing statistic value aiming at the target object in the target user behavior record;
acquiring a target weighting mode of the browsing statistic value for the target object from a target analysis strategy, and performing weighting processing on the browsing statistic value of the target object based on the target weighting mode to obtain a weighted browsing statistic value for the target object;
determining a browsing statistic sum for each object in the object set from the target behavior list;
an interest score for the target object is determined based on a ratio between the weighted browsing statistics and the sum of the browsing statistics.
In one implementation, the target user behavior record is any one of the user behavior records in the target behavior list, and the target user behavior record includes browsing statistics for the target object; the processor 501 is specifically configured to:
determining browsing times aiming at a target object and browsing duration aiming at the target object based on user behavior data, and acquiring a first weight corresponding to the browsing times and a second weight corresponding to the browsing duration;
weighting the browsing times of the target object based on the first weight to obtain the weighted browsing times of the target object;
weighting the browsing duration of the target object based on the second weight to obtain the weighted browsing duration of the target object;
and summing the weighted browsing duration and the weighted browsing times to obtain a browsing statistic value of the target object.
In one implementation, the target analysis strategy indicates a first scoring model and a second scoring model for calculating at least one user behavior record, and a first model weight corresponding to the first scoring model and a second model weight corresponding to the second scoring model; the processor 501 is specifically configured to:
calling a first scoring model to calculate at least one user behavior record to obtain a first interest score aiming at a target object;
calling a second scoring model to calculate at least one user behavior record to obtain a second interest score aiming at the target object;
if the first interest score and the second interest score both meet the preset screening condition, performing weighting processing on the first interest score by adopting a first model weight to obtain a first weighted interest score, and performing weighting processing on the second interest score by adopting a second model weight to obtain a second weighted interest score;
and summing the first weighted interest score and the second interest score to obtain an interest score for the target object.
In one implementation, the recommendation policy for the target object includes a recommendation frequency and a recommendation path for the target object; the processor 501 is specifically configured to:
if the interest score is larger than the score threshold value, determining that the recommendation frequency for the target object is a first recommendation frequency, and determining that the recommendation path for the target object is a first transmission path;
if the interest score is smaller than or equal to the score threshold, determining that the recommendation frequency for the target object is a second recommendation frequency and the recommendation path for the target object is a second transmission path; the first recommended frequency is higher than the second recommended frequency, and the transmission efficiency of the first transmission path is higher than that of the second transmission path.
In one implementation, the processor 501 is further configured to:
broadcasting the target analysis strategy and the recommendation strategy so that nodes in the block chain carry out consensus verification on the target analysis strategy and the recommendation strategy;
if the received consensus verification result indicates that the verification is passed, packaging the target analysis strategy and the recommendation strategy into a block;
the blocks are issued into a blockchain.
In the embodiment of the present invention, the processor 501 collects user behavior data from the client, where the user behavior data is generated by browsing behaviors of a user operating the client on each object in the object set; the processor 501 performs statistical processing on the user behavior data, and stores at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, wherein each user behavior record comprises a browsing statistic value for one object in the object set; the processor 501 obtains structural features of the target behavior list and determines a target analysis strategy for at least one user behavior record based on the structural features; the processor 501 calculates at least one user behavior record based on a target analysis strategy to obtain an interest score for a target object, where the target object is any object in an object set; the processor 501 determines a recommendation policy for the target object according to the interest score, and recommends the target object to the client based on the recommendation policy. By implementing the method, the interest score aiming at the target object can be determined from the user behavior data, the recommendation strategy aiming at the target object is determined based on the interest score, and the target object is recommended based on the recommendation strategy, so that the object recommendation process has pertinence, and the object recommendation effect is improved.
The module in the embodiment of the present invention may be implemented by a general-purpose Integrated Circuit, such as a CPU (central Processing Unit), or an ASIC (application Specific Integrated Circuit).
It should be understood that, in the embodiment of the present invention, the Processor 501 may be a Central Processing Unit (CPU), and may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The bus 502 may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended ISA (EISA) bus, or the like, and the bus 402 may be divided into an address bus, a data bus, a control bus, or the like, and fig. 4 shows only one thick line for convenience of illustration, but does not show only one bus or one type of bus.
It will be understood by those skilled in the art that all or part of the processes of the methods of the above embodiments may be implemented by a computer program, which may be stored in a computer storage medium and may include the processes of the embodiments of the methods described above when executed. The computer storage medium may be a magnetic disk, an optical disk, a Read-only Memory (ROM), a Random Access Memory (RAM), or the like.
The above disclosure is only for the purpose of illustrating the preferred embodiments of the present invention, and it is therefore to be understood that the invention is not limited by the scope of the appended claims.

Claims (10)

1. An object recommendation method, characterized in that the method comprises:
collecting user behavior data from a client, wherein the user behavior data is generated by browsing behaviors of a user operating the client on each object in an object set;
performing statistical processing on the user behavior data, and storing at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, wherein each user behavior record comprises a browsing statistic value for one object in the object set;
acquiring structural characteristics of the target behavior list, and determining a target analysis strategy aiming at the at least one user behavior record based on the structural characteristics;
calculating the at least one user behavior record based on the target analysis strategy to obtain an interest score for a target object, wherein the target object is any one object in the object set;
and determining a recommendation strategy aiming at the target object according to the interest score, and recommending the target object to the client based on the recommendation strategy.
2. The method of claim 1, wherein the structural feature comprises row and column information of the target behavior list, the row and column information comprises a number of rows or columns, and the determining the target analysis policy for the at least one user behavior record based on the structural feature comprises:
acquiring an initial weight corresponding to the row and column information;
determining a weighting manner for each browsing statistic in the at least one user behavior record based on the initial weight;
and formulating a target analysis strategy aiming at the at least one user behavior record based on the weighting mode.
3. The method of claim 2, wherein the calculating the at least one user behavior record based on the target analysis strategy to obtain an interest score for a target object comprises:
acquiring a target user behavior record aiming at the target object from the target behavior list, and determining a browsing statistic value aiming at the target object in the target user behavior record;
acquiring a target weighting mode of the browsing statistic value for the target object from the target analysis strategy, and performing weighting processing on the browsing statistic value of the target object based on the target weighting mode to obtain a weighted browsing statistic value for the target object;
determining a total of browsing statistics for each object in the set of objects from the list of target behaviors;
determining an interest score for the target object based on a ratio between the weighted view statistics and the sum of view statistics.
4. The method of claim 1, wherein a target user behavior record is any one of the user behavior records in the target behavior list, the target user behavior record comprises browsing statistics for the target object, and the statistically processing the user behavior data comprises:
determining browsing times for the target object and browsing duration for the target object based on the user behavior data, and acquiring a first weight corresponding to the browsing times and a second weight corresponding to the browsing duration;
weighting the browsing times of the target object based on the first weight to obtain the weighted browsing times of the target object;
weighting the browsing duration of the target object based on the second weight to obtain the weighted browsing duration of the target object;
and summing the weighted browsing duration and the weighted browsing times to obtain a browsing statistic value of the target object.
5. The method of claim 1, wherein the target analysis strategy indicates a first scoring model and a second scoring model for computing the at least one user behavior record, and a first model weight corresponding to the first scoring model and a second model weight corresponding to the second scoring model, and wherein computing the at least one user behavior record based on the target analysis strategy results in an interest score for a target object, comprising:
calling the first scoring model to calculate the at least one user behavior record to obtain a first interest score aiming at the target object;
calling the second scoring model to calculate the at least one user behavior record to obtain a second interest score aiming at the target object;
if the first interest score and the second interest score both meet preset screening conditions, performing weighting processing on the first interest score by adopting the first model weight to obtain a first weighted interest score, and performing weighting processing on the second interest score by adopting the second model weight to obtain a second weighted interest score;
and summing the first weighted interest score and the second interest score to obtain an interest score for the target object.
6. The method of claim 1, wherein the recommendation policy for the target object comprises a recommendation frequency and a recommendation path for the target object, and wherein determining the recommendation policy for the target object according to the interest score comprises:
if the interest score is larger than a score threshold value, determining that the recommendation frequency for the target object is a first recommendation frequency, and determining that the recommendation path for the target object is a first transmission path;
if the interest score is smaller than or equal to the score threshold, determining that the recommended frequency for the target object is a second recommended frequency, and determining that the recommended path for the target object is a second transmission path; the first recommended frequency is higher than the second recommended frequency, and the transmission efficiency of the first transmission path is higher than that of the second transmission path.
7. The method of claim 1, wherein after determining the recommendation policy for the target object according to the interest score and recommending the target object to the client based on the recommendation policy, the method further comprises:
broadcasting the target analysis strategy and the recommendation strategy so that nodes in a block chain carry out consensus verification on the target analysis strategy and the recommendation strategy;
if the received consensus verification result indicates that the verification is passed, packaging the target analysis strategy and the recommendation strategy into a block;
issuing the block into the block chain.
8. An object recommendation apparatus, characterized in that the object recommendation apparatus comprises:
the system comprises an acquisition module, a processing module and a display module, wherein the acquisition module is used for acquiring user behavior data from a client, and the user behavior data is generated by browsing behaviors of a user operating the client on each object in an object set;
the processing module is used for carrying out statistical processing on the user behavior data;
a storage module, configured to store at least one user behavior record obtained based on the user behavior data statistics in a preset behavior list to obtain a target behavior list, where each user behavior record includes a browsing statistics value for an object in the object set;
the acquisition module is further used for acquiring the structural characteristics of the target behavior list;
a determination module for determining a target analysis strategy for the at least one user behavior record based on the structural features;
a calculation module, configured to calculate the at least one user behavior record based on the target analysis policy to obtain an interest score for a target object, where the target object is any object in the object set;
the determination module is further used for determining a recommendation strategy for the target object according to the interest score;
and the recommending module is used for recommending the target object to the client based on the recommending strategy.
9. An object recommendation device, characterized in that the object recommendation device comprises a processor, an input interface, an output interface and a memory, which are connected to each other, wherein the memory is used for storing a computer program, which comprises program instructions, and the processor is configured to call the program instructions to execute the object recommendation method according to any one of claims 1-7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program comprising program instructions that, when executed by a processor, cause the processor to perform the object recommendation method according to any one of claims 1 to 7.
CN202011523881.7A 2020-12-18 2020-12-18 Object recommendation method, device, equipment and storage medium Active CN112650921B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011523881.7A CN112650921B (en) 2020-12-18 2020-12-18 Object recommendation method, device, equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011523881.7A CN112650921B (en) 2020-12-18 2020-12-18 Object recommendation method, device, equipment and storage medium

Publications (2)

Publication Number Publication Date
CN112650921A true CN112650921A (en) 2021-04-13
CN112650921B CN112650921B (en) 2024-06-18

Family

ID=75358750

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011523881.7A Active CN112650921B (en) 2020-12-18 2020-12-18 Object recommendation method, device, equipment and storage medium

Country Status (1)

Country Link
CN (1) CN112650921B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113283348A (en) * 2021-05-28 2021-08-20 青岛海尔科技有限公司 Method and device for determining interest value, storage medium and electronic device
CN114004494A (en) * 2021-11-01 2022-02-01 工赋(青岛)科技有限公司 Policy recommendation method, device, equipment and storage medium
CN114119168A (en) * 2021-12-01 2022-03-01 中国建设银行股份有限公司 Information pushing method and device
CN114329187A (en) * 2021-12-08 2022-04-12 北京五八信息技术有限公司 Recommendation method and device of content object, electronic equipment and readable medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107194732A (en) * 2017-05-24 2017-09-22 努比亚技术有限公司 One kind application method for pushing, mobile terminal and computer-readable recording medium
US20190026816A1 (en) * 2016-03-25 2019-01-24 Alibaba Group Holding Limited Time-division Recommendation Method and Apparatus for Service Objects
CN109615487A (en) * 2019-01-04 2019-04-12 平安科技（深圳）有限公司 Products Show method, apparatus, equipment and storage medium based on user behavior
CN110348947A (en) * 2019-06-13 2019-10-18 阿里巴巴集团控股有限公司 Object recommendation method and device
CN111598644A (en) * 2020-04-01 2020-08-28 华瑞新智科技（北京）有限公司 Article recommendation method, device and medium

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190026816A1 (en) * 2016-03-25 2019-01-24 Alibaba Group Holding Limited Time-division Recommendation Method and Apparatus for Service Objects
CN107194732A (en) * 2017-05-24 2017-09-22 努比亚技术有限公司 One kind application method for pushing, mobile terminal and computer-readable recording medium
CN109615487A (en) * 2019-01-04 2019-04-12 平安科技（深圳）有限公司 Products Show method, apparatus, equipment and storage medium based on user behavior
CN110348947A (en) * 2019-06-13 2019-10-18 阿里巴巴集团控股有限公司 Object recommendation method and device
CN111598644A (en) * 2020-04-01 2020-08-28 华瑞新智科技（北京）有限公司 Article recommendation method, device and medium

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113283348A (en) * 2021-05-28 2021-08-20 青岛海尔科技有限公司 Method and device for determining interest value, storage medium and electronic device
CN114004494A (en) * 2021-11-01 2022-02-01 工赋(青岛)科技有限公司 Policy recommendation method, device, equipment and storage medium
CN114119168A (en) * 2021-12-01 2022-03-01 中国建设银行股份有限公司 Information pushing method and device
CN114329187A (en) * 2021-12-08 2022-04-12 北京五八信息技术有限公司 Recommendation method and device of content object, electronic equipment and readable medium

Also Published As

Publication number Publication date
CN112650921B (en) 2024-06-18

Similar Documents

Publication Publication Date Title
CN112650921B (en) Object recommendation method, device, equipment and storage medium
CN106326391B (en) Multimedia resource recommendation method and device
US20130238390A1 (en) Informing sales strategies using social network event detection-based analytics
CN108460627A (en) Marketing activity scheme method for pushing, device, computer equipment and storage medium
CN110825977A (en) Data recommendation method and related equipment
CN112561269A (en) Advisor recommendation method and device
CN113076416A (en) Information heat evaluation method and device and electronic equipment
CN110647547A (en) Consumption delay monitoring method and device, electronic equipment and computer readable storage medium
CN111209201B (en) Advertisement putting test method and device
CN110743169B (en) Anti-cheating method and system based on block chain
CN110535910B (en) Method and device for recalling breakpoint user and storage medium
CN103309885A (en) Method and device for identifying feature user in electronic trading platform, search method and device
CN105956061A (en) Method and device for determining similarity between users
CN106844504B (en) A kind of method and apparatus for sending song and singly identifying
CN113225580A (en) Live broadcast data processing method and device, electronic equipment and medium
CN108549674B (en) Recommendation method, recommendation device and storage medium
CN108595623A (en) A kind of game video method for pushing, device and computer storage media
CN110688582B (en) Application recommendation method, application recommendation device and terminal equipment
CN111193598B (en) Group chat session recommendation method and device
CN114329093A (en) Data processing method, device and equipment
CN114201696A (en) Message pushing method and device, storage medium and computer equipment
CN113312889A (en) Report processing method, device, terminal and storage medium
CN109214874B (en) IP product operation data processing method, device, equipment and readable storage medium
CN109413459B (en) User recommendation method and related equipment in live broadcast platform
CN108804462B (en) Advertisement recommendation method and device and server

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant