CN110472991B - Data processing method, device, server and storage medium - Google Patents
Data processing method, device, server and storage medium Download PDFInfo
- Publication number
- CN110472991B CN110472991B CN201910570263.9A CN201910570263A CN110472991B CN 110472991 B CN110472991 B CN 110472991B CN 201910570263 A CN201910570263 A CN 201910570263A CN 110472991 B CN110472991 B CN 110472991B
- Authority
- CN
- China
- Prior art keywords
- commodity operation
- historical
- commodity
- operation frequency
- group
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0203—Market surveys; Market polls
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0224—Discounts or incentives, e.g. coupons or rebates based on user history
Landscapes
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention relates to the field of information technology processing, and discloses a data processing method, a data processing device, a server and a storage medium. A data processing method, comprising: acquiring historical commodity operation records and real-time characteristic information of a target user; wherein the characteristic information includes one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to the historical commodity operation records of the user and the historical characteristic information. By adopting the implementation mode of the invention, the accuracy of predicting the commodity operation frequency is effectively improved.
Description
Technical Field
The embodiment of the invention relates to the field of information technology processing, in particular to a data processing method, a data processing device, a server and a storage medium.
Background
With the development of internet technology, applications for providing commodity services on the market are on the rise. In order to keep the share of the application to the user market and control the user not to run off, the operator needs to allocate commodity service resources to different users reasonably according to the commodity operation behaviors of the different users. However, the inventors found that the following problems exist in the related art: in the related technology, each user is generally modeled to predict the commodity operation behavior of the user, the modeling operation workload is huge, and the time consumption and the consumption in the prediction process are huge; and the ordering behavior of the user in a future period of time is predicted only through the commodity operation behavior of each user in a historical period of time, and the accuracy of the prediction result is low.
Disclosure of Invention
The embodiment of the invention aims to provide a data processing method, a data processing device, a server and a storage medium, which effectively improve the accuracy of predicting the commodity operation frequency.
In order to solve the above technical problem, an embodiment of the present invention provides a data processing method, including: acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of the user and historical characteristic information.
The embodiment of the invention also provides a data processing method, which comprises the following steps: acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of a user and historical characteristic information; and acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
An embodiment of the present invention further provides a data processing apparatus, including: the acquisition module is used for acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; the prediction module predicts the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to the historical commodity operation records of the user and the historical characteristic information.
An embodiment of the present invention further provides a data processing apparatus, including: the data acquisition module is used for acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; the frequency prediction module is used for predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of a user and historical characteristic information; and the stage prediction module is used for acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
Embodiments of the present invention also provide a server, including: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to cause the at least one processor to perform: acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to the historical commodity operation records of the user and the historical characteristic information.
Embodiments of the present invention also provide a server, including: acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of a user and historical characteristic information; and acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
Embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which when executed by a processor implements the data processing method described above.
Embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which when executed by a processor implements the data processing method described above.
Compared with the prior art, the method and the device for processing the commodity operation record have the advantages that the historical commodity operation record and the real-time characteristic information of the target user are obtained; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; that is to say, in the embodiment, in addition to predicting the commodity operation frequency through the historical commodity operation record, a plurality of kinds of characteristic information (such as commodity operation strategy information, weather information, holiday sign information, and the like, which all affect whether a user performs a commodity operation behavior) which affect the commodity operation in the actual life are also used as the input of the predicted commodity operation frequency, so that the predicted commodity operation frequency is closer to the commodity operation behavior in the actual life, and the accuracy of the predicted commodity operation frequency is effectively improved. In addition, the preset prediction model for predicting the commodity operation frequency is obtained by training according to the historical commodity operation record of the user and the historical characteristic information, namely, the model training is carried out by referring to the characteristic information of multiple dimensions, so that the reference data of the model training is rich, the prediction model obtained by training can predict the commodity operation frequency more truly and accurately, and the reference value of the predicted commodity operation frequency is higher.
In addition, after the obtaining of the historical commodity operation record of the target user, the method further includes: acquiring characteristic data representing the commodity operation frequency according to the historical commodity operation record of the target user; the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information comprises the following steps: predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the feature data and the real-time feature information; the prediction model is obtained by training according to characteristic data of a user, which represents commodity operation frequency, and historical characteristic information; in the embodiment, the historical commodity operation records are subjected to data processing, and the characteristic data obtained after the processing is used as the input of the prediction commodity operation frequency instead of directly using one historical commodity operation record as the input of the prediction model, so that the operation load of the prediction model in work is effectively reduced, and the work efficiency of the prediction model is also improved.
In addition, the characteristic data representing the commodity operation frequency comprises historical commodity operation trends; the historical commodity operation trend is obtained through the following modes: grouping dates in the historical time period; wherein, the dates in the same group have the same week identification; acquiring the operation frequency of commodities on each date in each group; calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group; taking the set of the commodity operation trends of each group as the historical commodity operation trend; the embodiment provides a mode for processing historical commodity operation records, namely, a commodity operation trend in a week unit is calculated according to the historical commodity operation records, so that the change trend of commodity operation behaviors of a user in a week can be reflected; the method has the advantages that the commodity operation behavior of the user is considered to have the periodic characteristic, the commodity operation behavior characteristics of the user are extracted, and the problem of inaccurate prediction results caused by accumulated errors when the commodity operation time is used for prediction is solved.
In addition, the calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group comprises: sequentially counting the commodity operation frequency in each group according to a preset sliding window, wherein the commodity operation frequency of each group is sequenced according to the date sequence corresponding to the commodity operation frequency; calculating the index value of the commodity operation frequency sequentially taken by the sliding window until the sliding window finishes taking the commodity operation frequency in each group; and taking the set of the index values of the commodity operation frequency in each group as the commodity operation trend of each group, wherein the index values of the commodity operation frequency in each group are sorted and set according to the sequence of the commodity operation frequency taken by the sliding window.
In addition, the characteristic data representing the commodity operation frequency comprises a historical commodity operation mean value; the historical commodity operation mean value is obtained through the following method: grouping dates within the historical time period; wherein, the dates in the same group have the same week identification; acquiring the operation times of commodities on each date in each group; calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group; taking the set of the commodity operation mean values of each group as the historical commodity operation mean values; the embodiment provides a mode for processing historical commodity operation records, namely, a commodity operation mean value in a week unit is calculated according to the historical commodity operation records, so that the change condition of commodity operation behaviors of a user in a week can be reflected; the method has the advantages that the commodity operation behavior of the user is considered to have the periodic characteristic, the commodity operation behavior characteristics of the user are extracted, and the problem of inaccurate prediction results caused by accumulated errors when the commodity operation time is used for prediction is solved.
In addition, the commodity operation strategy information includes one or any combination of the following: voucher information, money information, point card information and red packet information; the commodity operation strategy information in this embodiment specifically refers to marketing information, and since the marketing strategy and the consumption amount generated by the commodity operation have a large influence on whether the user performs the commodity operation, the commodity operation strategy information includes voucher information, point card information, red packet information, and other information related to the marketing strategy and the commodity operation amount, and the accuracy of commodity operation frequency prediction can be higher.
Additionally, the weather information includes one or a combination of: temperature information and climate information; the holiday identification information is used for identifying whether the date is a holiday or not; since the temperature information, the weather information, and the holiday identification information have a large influence on whether the user performs the commodity operation, the above information is used as one of the feature information, and the accuracy of the commodity operation frequency prediction can be made higher.
In addition, the feature information specifically includes: characteristic information coded by a one-hot coding mode; one-hot coding (one-hot coding) is one-bit effective coding, the characteristic information is coded in a one-hot coding mode, discrete data can be conveniently processed, the problem of classifying characteristic information samples in data mining can be solved better, and the characteristics contained in the characteristic information are expanded.
In addition, the prediction model is composed of a plurality of machine learning models; the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency comprises the following steps: acquiring commodity operation frequency output by each machine learning model according to the historical commodity operation record and the real-time characteristic information of the target user; determining the commodity operation frequency of the target user according to the commodity operation frequency output by each machine learning model; that is to say, the commodity operation frequency output in the embodiment is obtained through the integrated and fused machine learning model, the application of the integrated and fused model is beneficial to subsequent study of the model learning curve and the model weight parameters, and the application of the integrated and fused model is beneficial to exponentially reducing the error rate of the model output, i.e. the accuracy of the commodity operation frequency output is effectively improved.
Drawings
Fig. 1 is a flowchart of a data processing method according to a first embodiment of the present invention;
FIG. 2 is a flowchart of a data processing method according to a second embodiment of the present invention;
FIG. 3 is a flowchart for obtaining historical commodity operation trends according to a second embodiment of the present invention;
FIG. 4 is a flowchart of obtaining the average value of historical commodity operations according to the second embodiment of the present invention;
FIG. 5 is a flowchart of a data processing method according to a third embodiment of the present invention;
FIG. 6 is a flowchart of a data processing method according to a fourth embodiment of the present invention;
fig. 7 is a block diagram showing the construction of a data processing apparatus according to a fifth embodiment of the present invention;
fig. 8 is a block diagram showing the construction of a data processing apparatus according to a sixth embodiment of the present invention;
fig. 9 is a block diagram showing the construction of a server according to a seventh embodiment of the present invention;
fig. 10 is a block diagram showing the construction of a server according to an eighth embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, it will be appreciated by those of ordinary skill in the art that numerous technical details are set forth in order to provide a better understanding of the present application in various embodiments of the present invention. However, the technical solution claimed in the present application can be implemented without these technical details and various changes and modifications based on the following embodiments. The following embodiments are divided for convenience of description, and should not constitute any limitation to the specific implementation manner of the present invention, and the embodiments may be mutually incorporated and referred to without contradiction.
The first embodiment of the present invention relates to a data processing method, and the commodity operation in the present embodiment can be understood as ordering operation by shopping application software, takeaway application software, or the like; a specific flow of the data processing method in this embodiment is shown in fig. 1, and specifically includes:
and 102, predicting the commodity operation frequency of a target user through a preset prediction model according to the historical commodity operation record and the real-time characteristic information.
In the embodiment, historical commodity operation records and real-time characteristic information of a target user are obtained; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; that is to say, in the embodiment, in addition to predicting the commodity operation frequency through the historical commodity operation record, a plurality of kinds of characteristic information (such as commodity operation strategy information, weather information, holiday sign information, and the like, which all affect whether a user performs a commodity operation behavior) which affect the commodity operation in the actual life are also used as the input of the predicted commodity operation frequency, so that the predicted commodity operation frequency is closer to the commodity operation behavior in the actual life, and the accuracy of the predicted commodity operation frequency is effectively improved. In addition, the preset prediction model for predicting the commodity operation frequency is obtained by training according to the historical commodity operation record of the user and the historical characteristic information, namely, the model training is carried out by referring to the characteristic information of multiple dimensions, so that the reference data of the model training is rich, the prediction model obtained by training can predict the commodity operation frequency more truly and accurately, and the reference value of the predicted commodity operation frequency is higher.
The following description specifically describes implementation details of the data processing method according to the present embodiment, taking a commodity operation as an ordering operation, and the following description is only provided for facilitating understanding of the implementation details and is not necessary for implementing the present embodiment.
In step 101, historical commodity operation records and real-time characteristic information of a target user are acquired. In this embodiment, the target user specifically refers to a user who is expected to be single-frequency in a future period of time, and the historical commodity operation record of the target user specifically refers to a commodity operation record of the target user in a previous period of time taking the current time as a starting point; for example, for the target user a, the commodity operation record of the target user a in the past three months is acquired as the historical commodity operation record of the target user a. Since the user usually has a certain consumption behavior habit, the referenceable value of predicting the future commodity operation frequency through the historical commodity operation records is high.
In the embodiment, the real-time characteristic information is acquired, which comprises real-time commodity operation strategy information, real-time weather information and real-time holiday identification information of a target user; the real-time commodity operation strategy information of the target user specifically refers to discount information used for deducting consumption amount when the user places an order or discount information used for giving other discount when the user places an order, for example, the commodity operation strategy information can comprise one or any combination of the following information, namely voucher information, money information, point collecting card information and red packet information; the voucher information may specifically refer to the remaining number of vouchers owned by the current user account, and the voucher information may specifically refer to the remaining number of vouchers owned by the current user account. Due to the fact that the various types of the preferential information have great influence on whether the user performs order placing operation or not (for example, when the preferential strength of the preferential information of the user is large, the user tends to perform order placing operation, but when the preferential strength of the preferential information of the user is small, the user may not perform order placing operation due to low order placing cost performance), the commodity operation strategy information is obtained to serve as input of the commodity operation frequency prediction, and the accuracy degree of the commodity operation frequency prediction is higher.
In addition, the weather information and the holiday identification information in the acquired real-time characteristic information are related to a time period in which the predicted commodity operation frequency is expected; specifically, if it is expected that the commodity operation frequency of the target user in the future week is predicted, acquiring weather information predicted in the future week and holiday identification information of each date in the future week; the weather information in the future week can be acquired through known weather forecast conditions, and can include one or any combination of the following information: temperature information and climate information (e.g., specific temperature values, sunny climate, rainy and snowy climate, windy climate, etc.); the holiday identification information of each date in the next week is used to identify whether each date is a holiday (for example, holiday identification information of a weekday in the next week is 0, and holiday identification information of a weekend is 1). The weather information and the holiday identification information have a large influence on whether the order placing operation is carried out by the user or not (for example, when the weather is fine and suitable for going out, the order placing operation is less likely to be carried out by the user, when the weather is bad and not suitable for going out, the order placing operation is more likely to be carried out by the user, when the holiday day is a holiday day, the time for the user to carry out off-line shopping or autonomous cooking is more abundant, the order placing operation is less likely to be carried out, and when the holiday day is not a holiday day, the user is busy in working, the order placing operation is more likely to be carried out), so that the weather information and the holiday day information are obtained as the input for predicting the commodity operation frequency, and the accuracy degree for predicting the commodity operation frequency is higher.
In the present embodiment, the prediction model for predicting the frequency of product operations is a machine learning model obtained by training based on the history of product operations by the user and the history of characteristic information. It should be noted that the historical commodity operation records of the users used in training the prediction model are historical commodity operation records of a plurality of users (it should be noted that the more the number of the users, the more abundant the historical commodity operation records of the users used for training, the more the model can be trained in a rich data environment to improve the accuracy of the output of the trained prediction model, and the more the prediction model obtained by training can be applied to a wider user group. In addition, historical characteristic information used when training the prediction model is related to historical time periods corresponding to historical commodity operation records of a plurality of users; specifically, if historical commodity operation records of a plurality of users in past history within three months are used in training the prediction model, the acquired historical feature information is feature information in past history within three months, for example: acquiring the condition that the plurality of users use the cash voucher when carrying out commodity operation recording in past three months (such as whether the users use the cash voucher to carry out commodity operation and the amount of the cash voucher), the condition that the users use the cash voucher (such as whether the users use the cash voucher to carry out commodity operation and the amount of the cash voucher), the condition that the users use a red package (such as whether the users use the red package to carry out commodity operation and the amount of the red package), the condition of a point collecting card (such as whether the users own the point collecting card and use the point collecting card), and the like; acquiring weather information (which can comprise one or any combination of the following information: temperature information and climate information) of each day in past three months; and acquiring the holiday identification information of each day in three months of past history. That is to say, when the prediction model is trained, the historical commodity operation behaviors of the user are reproduced as much as possible through the characteristic information of multiple dimensions, so that the prediction effect of the trained prediction model is close to the actual situation.
Compared with the prior art, the embodiment acquires real-time characteristic information, including real-time commodity operation strategy information, real-time weather information and real-time holiday identification information of a target user; the real-time commodity operation strategy information of the target user specifically refers to discount information used for deducting consumption amount when the user places an order or discount information used for giving other discount when the user places an order, for example, the commodity operation strategy information can comprise one or any combination of the following information, namely voucher information, money information, point collecting card information and red packet information; the voucher information may specifically refer to the remaining number of vouchers owned by the current user account, and the voucher information may specifically refer to the remaining number of vouchers owned by the current user account. Due to the fact that the various types of the preferential information have great influence on whether the user performs order placing operation or not (for example, when the preferential strength of the preferential information of the user is large, the user tends to perform order placing operation, but when the preferential strength of the preferential information of the user is small, the user may not perform order placing operation due to low order placing cost performance), the commodity operation strategy information is obtained to serve as input of the commodity operation frequency prediction, and the accuracy degree of the commodity operation frequency prediction is higher.
The second embodiment of the invention relates to a data processing method, and provides two modes for processing data of historical commodity operation records. In the present embodiment, the commodity operation is described as the ordering operation, and the data processing method in the present embodiment is shown in fig. 2, and the flow of fig. 2 will be specifically described below:
specifically, in the step, the acquired historical commodity operation records of the target user are subjected to data processing to obtain characteristic data for representing the commodity operation frequency, and the processed characteristic data is used as input for predicting the commodity operation frequency instead of directly using one historical commodity operation record as input of a prediction model, so that the operation load of the prediction model in working is effectively reduced; meanwhile, the preset training model is obtained through training the processed historical commodity operation records of the multiple users. In this step, two data processing modes for historical commodity operation records are provided, and the two data processing modes are specifically described below with reference to a flowchart.
In a first mode, the obtained feature data used for characterizing the operation frequency of the commodity may refer to a historical commodity operation trend, and a flowchart for acquiring the historical commodity operation trend is shown in fig. 3, and includes:
sub-step 301, grouping the dates of the historical time period.
Specifically, the historical time periods corresponding to the obtained historical commodity operation records of the target user are grouped according to dates, the dates in the same group have the same week identification, and the week identification can be understood as Monday and Tuesday 8230; total 7 week identifications of Sunday. For example, a historical product operation record of a target user in a past month is acquired, the past month has 30 days (numbers 1 to 30), the numbers 1 to 30 are divided into 7 groups (Monday group, tuesday group, 8230; \8230; sunday group), the Monday group includes 4 dates (numbers 3, 10, 17, and 24), the Tuesday group includes 4 dates (numbers 4, 11, 18, and 25) \8230;, and the Sunday group includes 5 dates (numbers 2, 9, 16, 23, and 30).
In substep 302, the product operation frequency on each date in each group is obtained.
Specifically, the commodity operation frequency of each date in this step is equal to the total number of ordering operations within each date divided by 1 (i.e., equal to the total number of ordering operations within each date), for example, in a monday group, if the user has 1 ordering operation in total for number 3, the commodity operation frequency of number 3 is 1; the user has 2 ordering operations in number 10, the commodity operation frequency of number 10 is 2, and the like. The commodity operation frequency of each date in each group is sorted in the order of the date corresponding to the commodity operation frequency, for example, the commodity operation frequency of 4 dates (No. 3, no. 10, no. 17, no. 24) in the weekly group is also sorted in the order of No. 3, no. 10, no. 17, no. 24.
And a substep 303, sequentially counting the commodity operation frequency in each group according to a preset sliding window, and calculating an index value of the commodity operation frequency sequentially taken by the sliding window until the sliding window finishes taking the commodity operation frequency in each group.
Specifically, a sliding window with a certain length is preset, the number of the commodity operation frequency ordered according to the date sequence in each group is taken, and the index value of the commodity operation frequency taken by each sliding window is calculated until all the commodity operation frequencies in the group are taken out by the sliding window; the length of the sliding window in the step can be set according to actual requirements, and is not limited specifically; the calculation of the index value in this step may be understood as calculating an average value or a sum, and the like, and is not particularly limited. For example, the length of the preset sliding window is 3, and the operation frequency of the commodities in the weekly group is counted through the sliding window, and the operation frequency of the commodities in the weekly group is arranged as the following table:
the result of sequentially taking the commodity operation frequency from the preset sliding window is as follows:
(1,2,1)；(2,1,1)；(1,1,2)；(1,2,0)；(2,0,1)
calculating the average value of the commodity operation frequency sequentially taken by the sliding window as follows:
(1+2+1)/3；(2+1+1)/3；(1+1+2)/3；(1+2+0)/3；(2+0+1)/3
namely, the index values of the operation frequency of the commodities in the weekly group are sequentially as follows:
4/3，4/3,4/3,1,1
in substep 304, a set of index values of the commodity operation frequency in each group is defined as a commodity operation trend in each group.
Specifically, the calculated index values of the commodity operation frequency in each group are sorted according to the sequence of the commodity operation frequency taken by the sliding window, and the set of the index values sorted according to the sequence is the commodity operation trend of each group. For example, the index values of the commodity operation frequency in the weekly group obtained in step 303 are sorted and collected in order, and the commodity operation trend of the weekly group is obtained as:
4/3→4/3→4/3→1→1
as can be seen from the operation trend of the commodities in the monday group, the ordering operation behavior of the user in monday tends to gradually decrease.
And a substep 305 of using the set of the product operation trends of each group as historical product operation trends.
Specifically, since the dates in the historical period are divided into 7 groups according to the week identification, the set of 7 groups of product operation trends is taken as the historical product operation trend, that is, the historical product operation trend includes the product operation trend of the weekday group to the product operation trend of the weekday group, and the historical product operation trend is characteristic data for characterizing the frequency of product operation.
The above is the first way of processing data of historical commodity operation records; in a second manner, the obtained feature data for characterizing the commodity operation frequency may be referred to as a historical commodity operation mean value, and a flowchart for obtaining the historical commodity operation mean value is shown in fig. 4, and includes:
sub-step 401, grouping dates of the historical time period. In this step, the dates in the same group have the same week identification, and this step is substantially the same as step 301, and will not be described herein again.
In substep 402, the number of commodity operations for each date in each group is obtained.
Specifically, the total number of ordering operations in each date is directly obtained in the step, and other calculation is not needed. For example, within a week group, the total number of orders placed by the user at number 3 is 1, and the total number of orders placed at number 10 is 2 \8230; \8230, and so on.
In substep 403, the average value of the commodity operation in each group is calculated according to the commodity operation times of each date in each group.
Specifically, the commodity operation average value of each group is calculated from the sum of the commodity operation times of each date in each group and the total number of days in each group. For example, the number of commodity operations in a weekly group is arranged as follows:
the sum of the number of times of commodity operation on each date in the monday group is:
(ii) (1, 2,1, 0) =9 times
The total days in the weekly cohort were: 9 days
The mean of the commodity operation in the weekly group is: 9/9=1
In the substep 404, the set of commodity operation averages in each group is used as historical commodity operation averages.
Specifically, since the dates in the historical time period are divided into 7 groups according to the day of the week identification, a set of 7 groups of commodity operation means is used as the historical commodity operation means, that is, the historical commodity operation means includes the commodity operation means of the weekday group to the commodity operation means of the weekday group, and the historical commodity operation means is characteristic data for representing the frequency of commodity operation. For example, the historical commodity operation average includes commodity operation average values of monday group to sunday group as follows: 1,8/7,1,2,1/3,2/3, and the like, and can also roughly reflect the change trend of the ordering behavior of the user in one week.
It should be noted that, in the second embodiment, the commodity operation and value of each group may also be calculated according to the commodity operation times of each date in each group, and the characteristic data used subsequently is not particularly limited.
In the embodiment, the characteristic data representing the operation frequency of the commodity is obtained in a grouping manner through the week identification, and actually, the ordering behavior characteristics of the user are extracted, and the characteristic that the ordering behavior of the user has periodicity is also considered; because errors are accumulated continuously along with the time when the ordering time and the ordering behavior are adopted for prediction, the failure of the prediction result is caused, and the problem of inaccurate prediction result caused by the accumulated errors of the time span is solved by adopting the two modes.
After the historical commodity operation records are subjected to data processing to obtain characteristic data for representing the commodity operation frequency, step 203 is executed:
and 203, predicting the commodity operation frequency of the target user through a preset prediction model according to the characteristic data and the real-time characteristic information.
Specifically, the characteristic data and the real-time characteristic information are used as the input of the prediction model, so that the prediction model can directly output the predicted commodity operation trend in a period of time in the future according to the input commodity operation trend without performing data processing on a single historical commodity operation record, the operation load of the prediction model in work is effectively reduced, and the work efficiency of the prediction model is also improved. This step is substantially the same as step 102 and will not be described herein.
Compared with the prior art, the embodiment takes the characteristic data obtained after the historical commodity operation records are processed as the input of the prediction model, so that the prediction model does not need to perform data processing on the historical commodity operation records which are independently formed, the running load of the prediction model during working is effectively reduced, and the working efficiency of the prediction model is also improved. In addition, historical commodity operation trend or historical commodity operation mean value is calculated according to the week identification groups to approximately reflect the change situation of ordering operation behaviors of the user in a week, actually, ordering behavior characteristics of the user are extracted, and the characteristic that the ordering behaviors of the user have periodicity is also considered; because errors are accumulated continuously along with the lapse of time when the ordering time and the ordering behavior are simply adopted for prediction, the failure of the prediction result is caused, and the problem of inaccurate prediction result caused by the accumulated errors of the time span is avoided by adopting the two modes.
A third embodiment of the present invention relates to a data processing method, and is substantially the same as the first or second embodiment, wherein the characteristic information is embodied in an encoded form; the prediction model is composed of a plurality of machine learning models. In the present embodiment, the commodity operation is described as the ordering operation, and the data processing method in the third embodiment of the present invention is shown in fig. 5, and the flow of fig. 5 is specifically described below:
specifically, the characteristic information in the present embodiment may be embodied in a coding manner, for example, a coding manner by a one-hot coding manner; the feature information described here refers to both feature information input to the prediction model for predicting the next single frequency and feature information input when the prediction model is trained. The one-hot coding, namely one-hot coding and one-bit effective coding, mainly adopts an N-bit state register to code N states, each state has an independent register bit, and only one bit is effective at any time, namely, only 1 bit of a single characteristic in each sample is ensured to be in a state 1, and the rest bits are in a state 0; the problem that discrete data are not well processed is solved through the single-hot coding, the problem that characteristic information samples are classified through data mining can be better solved, and the function of expanding the characteristics is achieved to a certain extent.
In one example, the commodity operation strategy information included in the characteristic information can be coded in the following way:
when in the prediction stage of the prediction model, the commodity operation strategy information comprises voucher information and coin information; the voucher information specifically refers to the remaining number of vouchers owned by the current user account, and the remaining number of vouchers can be encoded as shown in the following table:
or, the voucher information specifically indicates whether the voucher in the validity period is owned under the current user account, and if the voucher in the validity period is owned under the current user account, the code is 1; if the voucher in the valid period is not owned under the current user account, the code is 0. In the prediction stage of the prediction model, the coding rule and mode of the token information can be set by referring to the convenience of the token information as above.
In the prediction model training stage, the commodity operation strategy information comprises voucher information, money information, point card information and red packet information; the voucher information specifically indicates whether a user uses a voucher to place an order, and if the user uses the voucher to place an order, the code is 1; if the user does not use the voucher to place an order, the code is 0; the voucher information further includes the amount of the voucher used by the user when placing an order, and the amount of the voucher used can be encoded as shown in the following table:
| amount of voucher used | One-hot encoding |
| 1 | 0001 |
| 2 | 0010 |
| 3 | 0100 |
| 4 | 1000 |
Or, the code is 1 when the use amount of the voucher is more than 3, and the code is 0 when the use amount of the voucher is less than 3. In the stage of forecasting model training, the coding rules and modes of the token information, the point card information and the red packet information can be set by referring to the convenience of the token information.
In one example, for weather information included in the feature information, the weather information may be encoded as follows:
the weather information comprises temperature information and climate information; the temperature information and the climate information can be subdivided into the following dimensional characteristics: temperature, weather phenomena, wind-up level, suitable egress level, etc., wherein the temperature characteristics may be encoded as shown in the following table:
| temperature of | One-hot encoding |
| The temperature is more than or equal to 30 DEG C | 1000 |
| The temperature is higher than 30 ℃ and higher than 15 DEG C | 0100 |
| The temperature is more than 0 ℃ at the temperature of more than or equal to 15 DEG C | 0010 |
| At 0 ℃ or higher | 0001 |
The weather phenomenon characteristics may be encoded as shown in the following table:
the wind rise characteristics can be encoded as shown in the following table:
| degree of wind | One-hot encoding |
| Degree of strong wind | 100 |
| Extent of breeze | 010 |
| Degree of absence of |
001 |
The appropriate egress level characteristics may be encoded as shown in the following table:
| suitable for the degree of going out | One-hot encoding |
| Is suitable for going out | 1000 |
| Is generally suitable for going out | 0100 |
| Is not suitable for outdoor use | 0010 |
| Is very unsuitable for going out | 0001 |
Combining the above-mentioned unique hot codes of several dimensions, the specific weather information code at a certain date can be obtained as shown in the following table:
| weather information | One-hot encoding |
| Sunny day/20 deg.C/breeze/suitable for going out | 100 0100 010 1000 |
| Cloudy day/10 ℃/gale/unsuitable for outdoor use | 010 0010 100 0010 |
| Rain and snow/0 ℃/strong wind/very unsuitable for going out | 001 0001 100 0001 |
In one example, the holiday identification information included in the feature information may be encoded as follows: when a specific date is a holiday, the code is 1; when a specific date is not a holiday, the code is 0.
Note that this embodiment mode can also be implemented in combination with the second embodiment mode; in connection with the data processing of the historical product operation records in the second embodiment, the historical product operation records and the historical feature information of the user at a certain date can be shown in the following table:
in other words, the digitized historical commodity operation records and the characteristic information are used as input to the prediction model, so that the machine learning algorithm of the machine learning model can conveniently process and calculate data.
The other parts in this step are substantially the same as step 101, and are not described herein again.
502, acquiring commodity operation frequency output by each machine learning model according to historical commodity operation records and real-time characteristic information;
Specifically, the prediction model in the present embodiment is a combination of several machine learning models, each of which is obtained by training a user's historical commodity operation record and historical feature information, and the output result is a predicted value of the commodity operation frequency of the target user. Respectively inputting the acquired historical commodity operation records and real-time characteristic information of the target user into each machine learning model forming a prediction model to respectively obtain predicted values of commodity operation frequency of the target user, which are output by each machine learning model; in the present embodiment, the predicted value of the commodity operation frequency output by each machine learning model is averaged, and the obtained average commodity operation frequency is determined as the predicted commodity operation frequency of the target user. By means of integrating and fusing a plurality of models, the learning curve of each model and the model weight parameter of each model can be researched subsequently; and the model integrated and fused is applied to prediction, so that the error rate of model output is exponentially reduced, and the accuracy of the operation frequency of the output commodity is effectively improved. In practical applications, the machine learning models may include an LR model, an xgboost model, a randomforest model, an SVR model, and the like.
Compared with the prior art, the embodiment performs digital processing such as coding on the characteristic information, and then takes the characteristic information as input of the prediction model, so that the data processing and calculation of the machine learning algorithm of the machine learning model are facilitated; the prediction model is integrated and fused by a plurality of machine learning models, so that the subsequent study on the learning curve of each model and the model weight parameter of each model is facilitated; and the model integrated and fused is applied for prediction, so that the error rate of model output is exponentially reduced, and the accuracy of the operation frequency of the output commodity is effectively improved.
A fourth embodiment of the present invention relates to a data processing method. A specific flow of the data processing method in the present embodiment is shown in fig. 6, and specifically includes:
And 603, acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
In this embodiment, the determination of the life cycle stage of the target user is performed according to the predicted commodity operation frequency of the target user in the first, second, or third embodiment; the life cycle in the embodiment can be understood as the whole process from the beginning of using the commodity service provided in the application to the end of using the commodity service; according to the user using behaviors, the life cycle of the user can be divided into four stages: an entry phase, a stabilization phase, a decline phase and a loss phase; aiming at users in different life cycle stages, different commodity service strategies need to be provided so as to increase the viscosity of the users; therefore, in the embodiment, the life cycle stage of the user is determined by predicting the commodity operation frequency of the user in a future period of time, so as to perform subsequent operation service work. The following describes implementation details of the data processing method of the present embodiment in detail, and the following description is only provided for facilitating understanding of the implementation details and is not necessary for implementing the present embodiment.
The first mode is specifically to determine the life cycle stage of the predicted target user by comparing the predicted commodity operation frequency of the target user with a preset commodity operation frequency threshold value. Since the user's life cycle is generally divided into four phases according to the user's usage behavior: in the entering period, the stabilizing period, the declining period and the loss period, the commodity operation frequency threshold values are respectively set for the four stages; the threshold value can be determined according to the actual operation experience and requirements of the operator. When the predicted commodity operation frequency of the target user is compared with the preset commodity operation frequency threshold value, an index value of the predicted commodity operation frequency of the target user can be obtained and compared with the preset commodity operation frequency threshold value, for example, after 7 commodity operation frequency values of the target user A in the future week are obtained, an average value of the 7 commodity operation frequency values can be taken as the total commodity operation frequency of the target user A in the future week and compared with the preset commodity operation frequency threshold value; since the commodity operation frequency of the user in the entry period is in an ascending trend, the commodity operation frequency threshold value in the entry period can be preset to be a numerical value capable of representing the lowest commodity operation frequency of the user in the entry period, and when the total commodity operation frequency of the target user A in a future week is greater than the commodity operation frequency threshold value in the entry period, the target user A can be roughly judged to be in the entry period; the judgment rules of other life cycle stages are similar. The method is easy to realize in practical application, and users in different life cycle stages can be quickly divided through the preset threshold value.
The second way is to determine the life cycle stage of the predicted target user by comparing the predicted commodity operation frequency of the target user with the historical commodity operation frequency. For example, after 7 commodity operation frequency values of the target user a in the next week are obtained, an average value of the 7 commodity operation frequency values may be taken as a total commodity operation frequency of the target user a in the next week; meanwhile, the total commodity operation frequency of the target user A in the past week is obtained and compared with the total commodity operation frequency of the target user A in the future week, and the change trend of the commodity operation frequency of the target user A can be obtained according to the comparison result, so that the life cycle stage of the target user A is judged; since the commodity operation frequency of the user in the entry period is in an ascending trend, when the total commodity operation frequency of the target user a in the future week is greater than the total commodity operation frequency in the past week, it can be roughly determined that the target user a is currently in the entry period; since the commodity operation frequency of the user in the decline period is in a descending trend, when the total commodity operation frequency of the target user a in the future week is less than the total commodity operation frequency in the past week, the target user a can be roughly judged to be in the decline period at present; the judgment rules of other life cycle stages are similar. In the method, the two parties for comparison are the commodity operation frequency of the same target user, and the change trend of the commodity operation frequency of the target user can be reflected on the comparison and judgment result, so that the life cycle stage obtained by judgment is more accurate.
Compared with the prior art, the embodiment judges the life cycle stage of the target user according to the commodity operation frequency of the target user predicted in the first embodiment, the second embodiment or the third embodiment, so that an operator can plan and implement a targeted service strategy for the target user in advance according to the life cycle stage of the target user, the viscosity and the cognition degree of the target user for the commodity service provided by the operator are improved, the commodity operation behavior habit of the target user is developed, and the control of the operation cost is facilitated.
A fifth embodiment of the present invention relates to a data processing apparatus, as shown in fig. 7, including: an acquisition module 701 and a prediction module 702.
An obtaining module 701, configured to obtain a historical commodity operation record and real-time feature information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
the prediction module 702 predicts the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of the user and historical characteristic information.
In an example, the obtaining module 701 is further configured to, after obtaining the historical article operation record of the target user, further include: acquiring characteristic data representing the commodity operation frequency according to the historical commodity operation record of the target user; the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information comprises the following steps: predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the feature data and the real-time feature information; the prediction model is obtained by training according to the characteristic data of the user representing the commodity operation frequency and the historical characteristic information.
In one example, the characteristic data representing the operation frequency of the commodity, which is obtained by the obtaining module 701, includes a historical commodity operation trend; the historical commodity operation trend is obtained by the following method: grouping dates within the historical time period; wherein, the dates in the same group have the same week identification; acquiring the operation frequency of commodities on each date in each group; calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group; and taking the set of the commodity operation trends of each group as the historical commodity operation trends.
In one example, the obtaining module 701 calculates the operation trend of the commodities in each group according to the operation frequency of the commodities on each date in each group, including: sequentially counting the commodity operation frequency in each group according to a preset sliding window, wherein the commodity operation frequency of each group is sequenced according to the date sequence corresponding to the commodity operation frequency; calculating the index values of the commodity operation frequency sequentially taken by the sliding window until the sliding window finishes taking the commodity operation frequency in each group; and taking the set of the index values of the commodity operation frequency in each group as the commodity operation trend of each group, wherein the index values of the commodity operation frequency in each group are sorted and set according to the sequence of the commodity operation frequency taken by the sliding window.
In one example, the characteristic data representing the commodity operation frequency obtained by the obtaining module 701 includes a historical commodity operation mean value; the historical commodity operation mean value is obtained through the following method: grouping dates within the historical time period; wherein, the dates in the same group have the same week identification; acquiring the operation times of commodities on each date in each group; calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group; and taking the set of the commodity operation mean values of each group as the historical commodity operation mean values.
In one example, the article operation policy information acquired by the acquiring module 701 includes one or any combination of the following: voucher information, money information, point card information, and red envelope information.
In one example, the weather information acquired by the acquiring module 701 includes one or a combination of the following: temperature information and climate information; the holiday identification information is used for identifying whether the date is a holiday. In an example, the feature information acquired by the acquiring module 701 specifically includes: and (4) encoding the characteristic information by a one-hot encoding mode.
In one example, the predictive model used by prediction module 702 consists of several machine learning models; the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency comprises the following steps: acquiring commodity operation frequency output by each machine learning model according to the historical commodity operation record and the real-time characteristic information of the target user; and determining the commodity operation frequency of the target user according to the commodity operation frequency output by each machine learning model.
It should be understood that this embodiment is an example of the apparatus corresponding to the first, second, or third embodiment, and may be implemented in cooperation with the second or third embodiment of the first embodiment. Related technical details mentioned in the second embodiment or the third embodiment of the first embodiment are still valid in this embodiment, and are not described herein again in order to reduce repetition. Accordingly, the related technical details mentioned in the present embodiment can also be applied to the second embodiment or the third embodiment of the first embodiment.
It should be noted that, in practical applications, one logical unit may be one physical unit, may be a part of one physical unit, and may also be implemented by a combination of multiple physical units. In addition, in order to highlight the innovative part of the present invention, elements that are not so closely related to solving the technical problems proposed by the present invention are not introduced in the present embodiment, but this does not indicate that other elements are not present in the present embodiment.
A sixth embodiment of the present invention relates to a data processing apparatus, as shown in fig. 8, including: a data acquisition module 801, a frequency prediction module 802 and a phase prediction module 803.
A data acquisition module 801, configured to acquire historical commodity operation records and real-time feature information of a target user; wherein the characteristic information includes one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
a frequency prediction module 802, configured to predict, according to the historical commodity operation record and the real-time characteristic information, the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency; the prediction model is obtained by training according to historical commodity operation records of a user and historical characteristic information;
and the phase prediction module 803 is configured to obtain a life cycle phase of the target user according to the predicted commodity operation frequency of the target user.
In one example, the phase prediction module 803 obtains the life cycle phase of the target user according to the predicted commodity operation frequency of the target user, including: acquiring the index value of the predicted commodity operation frequency of the target user; and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and a preset life cycle stage threshold value.
In one example, the phase prediction module 803 obtains the life cycle phase of the target user according to the predicted commodity operation frequency of the target user, including: acquiring the predicted index value of the commodity operation frequency of the target user and the historical index value of the commodity operation frequency of the target user; and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and the historical index value of the commodity operation frequency of the target user.
The data acquisition module 801 in the present embodiment has substantially the same function as the acquisition module 701 in the fifth embodiment; the frequency prediction module 802 in this embodiment has substantially the same function as the prediction module 702 in the fifth embodiment, and is not described herein again.
It should be understood that this embodiment is an example of an apparatus corresponding to the fourth embodiment, and that this embodiment can be implemented in cooperation with the fourth embodiment. The related technical details mentioned in the fourth embodiment are still valid in this embodiment, and are not described herein again in order to reduce repetition. Accordingly, the related-art details mentioned in the present embodiment can also be applied to the fourth embodiment.
It should be noted that, in practical applications, one logical unit may be one physical unit, may be a part of one physical unit, and may also be implemented by a combination of multiple physical units. In addition, in order to highlight the innovative part of the present invention, a unit which is less closely related to solving the technical problem proposed by the present invention is not introduced in the present embodiment, but it does not indicate that no other unit exists in the present embodiment.
A seventh embodiment of the present invention relates to a server, as shown in fig. 9, including at least one processor 901; and memory 902 communicatively connected to the at least one processor 901; and a communication component 903 communicatively coupled to the data processing apparatus, the communication component 903 receiving and transmitting data under the control of the processor 901; wherein the memory 902 stores instructions executable by the at least one processor 901, the instructions being executable by the at least one processor 901 to implement:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to the historical commodity operation records of the user and the historical characteristic information.
Specifically, the data processing apparatus includes: one or more processors 901 and a memory 902, and fig. 9 illustrates one processor 901 as an example. The processor 901 and the memory 902 may be connected by a bus or by other means, and fig. 9 illustrates the connection by the bus as an example. Memory 902, which is a computer-readable storage medium, may be used to store computer software programs, computer-executable programs, and modules. The processor 901 executes various functional applications of the device and data processing, i.e., implements the above-described data processing method, by executing computer software programs, instructions, and modules stored in the memory 902.
The memory 902 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store a list of options, etc. Further, the storage 902 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, the memory 902 may optionally include memory located remotely from the processor 901, which may be connected to an external device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
One or more modules are stored in the memory 902 and when executed by the one or more processors 901 perform the data processing method of any of the method embodiments described above.
The product can execute the method provided by the embodiment of the application, has corresponding functional modules and beneficial effects of the execution method, and can refer to the method provided by the embodiment of the application without detailed technical details in the embodiment.
In the embodiment, the real-time characteristic information is acquired, which comprises real-time commodity operation strategy information, real-time weather information and real-time holiday identification information of a target user; the real-time commodity operation strategy information of the target user specifically refers to discount information used for deducting consumption amount when the user places an order or discount information used for giving other discount when the user places an order, for example, the commodity operation strategy information can comprise one or any combination of the following information, namely voucher information, money information, point collecting card information and red packet information; the voucher information may specifically refer to the remaining number of vouchers owned under the current user account, and the token information may specifically refer to the remaining number of tokens owned under the current user account, and the like. Due to the fact that the various types of the preferential information have great influence on whether the user performs order placing operation or not (for example, when the preferential strength of the preferential information of the user is large, the user tends to perform order placing operation, but when the preferential strength of the preferential information of the user is small, the user may not perform order placing operation due to low order placing cost performance), the commodity operation strategy information is obtained to serve as input of the commodity operation frequency prediction, and the accuracy degree of the commodity operation frequency prediction is higher.
An eighth embodiment of the present invention is directed to a server, as shown in fig. 10, including at least one processor 001; and, memory 002 communicatively coupled to the at least one processor 001; and a communication component 003 in communication connection with the data processing apparatus, the communication component 003 receiving and transmitting data under the control of the processor 001; wherein the memory 002 stores instructions executable by the at least one processor 001, the instructions being executable by the at least one processor 001 to implement:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of a user and historical characteristic information;
and acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
Specifically, the data processing apparatus includes: one or more processors 001 and a memory 002, and one processor 001 is illustrated in fig. 10. The processor 001 and the memory 002 may be connected by a bus or by other means, and fig. 10 illustrates the connection by a bus as an example. The memory 002 serves as a computer-readable storage medium for storing computer software programs, computer-executable programs, and modules. The processor 001 executes various functional applications of the device and data processing, i.e., implements the above-described data processing method, by running computer software programs, instructions, and modules stored in the memory 002.
The memory 002 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store a list of options, etc. Further, the storage 002 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, the memory 002 may optionally include memory located remotely from the processor 001, which may be connected to an external device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The one or more modules are stored in the memory 002 and, when executed by the one or more processors 001, perform the data processing method of any of the method embodiments described above.
The product can execute the method provided by the embodiment of the application, has corresponding functional modules and beneficial effects of the execution method, and can refer to the method provided by the embodiment of the application without detailed technical details in the embodiment.
In the embodiment, the life cycle stage of the target user is judged according to the predicted commodity operation frequency of the target user, so that the operator can plan and implement a targeted service strategy for the target user in advance according to the life cycle stage of the target user, the viscosity and the cognition degree of the target for commodity service provided by the operator are improved, the commodity operation behavior habit of the target user is cultivated, and the operation cost control is facilitated.
A ninth embodiment of the present invention relates to an example of a computer-readable storage medium, that is, as can be understood by those skilled in the art, all or part of the steps in the above data processing method embodiment can be implemented by a program to instruct related hardware, where the program is stored in a storage medium and includes several instructions to enable a device (which may be a single chip, a chip, or the like) or a processor (processor) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
It will be understood by those of ordinary skill in the art that the foregoing embodiments are specific examples for carrying out the invention, and that various changes in form and details may be made therein without departing from the spirit and scope of the invention in practice.
The embodiment of the application discloses an A1 data processing method, which comprises the following steps:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to the historical commodity operation records of the user and the historical characteristic information.
A2. The data processing method according to A1, after the obtaining of the historical commodity operation record of the target user, further comprising:
acquiring characteristic data representing the commodity operation frequency according to the historical commodity operation record of the target user;
the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information comprises the following steps:
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the feature data and the real-time feature information; the prediction model is obtained by training according to characteristic data of a user, which represents commodity operation frequency, and historical characteristic information.
A3. The data processing method of A2, wherein the characteristic data representing the commodity operation frequency comprises historical commodity operation trends;
the historical commodity operation trend is obtained by the following method:
grouping dates in the historical time period; wherein, the dates in the same group have the same week identification;
acquiring the operation frequency of commodities on each date in each group;
calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group;
and taking the set of the commodity operation trends of each group as the historical commodity operation trends.
A4. The data processing method according to A3, wherein the calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group includes:
sequentially counting the commodity operation frequency in each group according to a preset sliding window, wherein the commodity operation frequency of each group is sorted according to the date sequence corresponding to the commodity operation frequency;
calculating the index values of the commodity operation frequency sequentially taken by the sliding window until the sliding window finishes taking the commodity operation frequency in each group;
and taking the set of the index values of the commodity operation frequency in each group as the commodity operation trend of each group, wherein the index values of the commodity operation frequency in each group are sorted and collected according to the sequence of the commodity operation frequency taken by the sliding window.
A5. The data processing method as described in A2, wherein the characteristic data representing the commodity operation frequency includes a historical commodity operation mean value;
the historical commodity operation mean value is obtained through the following method:
grouping dates within the historical time period; wherein, the dates in the same group have the same week identification;
acquiring the operation times of commodities on each date in each group;
calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group;
and taking the set of the commodity operation mean values of each group as the historical commodity operation mean values.
A6. The data processing method according to A1, wherein the commodity operation policy information includes one or any combination of the following:
voucher information, money information, point card information, and red envelope information.
A7. The data processing method according to A1, wherein the weather information includes one or a combination of the following: temperature information and climate information;
the holiday identification information is used for identifying whether the date is a holiday.
A8. The data processing method according to any one of claims A1 to A7, wherein the feature information specifically includes: and (4) encoding the characteristic information by a one-hot encoding mode.
A9. The data processing method according to A1, wherein the prediction model is composed of a plurality of machine learning models;
the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency comprises the following steps:
according to the historical commodity operation records and the real-time characteristic information of the target user, acquiring the commodity operation frequency output by each machine learning model;
and determining the commodity operation frequency of the target user according to the commodity operation frequency output by each machine learning model.
The embodiment of the application discloses a B1 data processing method, which comprises the following steps:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein the characteristic information includes one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of a user and historical characteristic information;
and acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
B2. The data processing method according to B1, wherein the obtaining a life cycle stage of the target user according to the predicted commodity operation frequency of the target user includes:
obtaining the index value of the predicted commodity operation frequency of the target user;
and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and a preset life cycle stage threshold value.
B3. The data processing method according to B1, wherein the obtaining a life cycle stage of the target user according to the predicted commodity operation frequency of the target user includes:
acquiring the predicted index value of the commodity operation frequency of the target user and the historical index value of the commodity operation frequency of the target user;
and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and the historical index value of the commodity operation frequency of the target user.
The embodiment of the application discloses C1. A data processing device, includes:
the acquisition module is used for acquiring historical commodity operation records and real-time characteristic information of a target user; wherein the characteristic information includes one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
the prediction module predicts the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of the user and historical characteristic information.
C2. The data processing apparatus according to C1, further comprising, after the obtaining of the historical commodity operation record of the target user:
acquiring characteristic data representing the commodity operation frequency according to the historical commodity operation record of the target user;
the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information comprises the following steps:
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the feature data and the real-time feature information; the prediction model is obtained by training according to the characteristic data of the user representing the commodity operation frequency and the historical characteristic information.
C3. The data processing device according to C2, wherein the characteristic data representing the commodity operation frequency includes historical commodity operation trends;
the historical commodity operation trend is obtained through the following modes:
grouping dates within the historical time period; wherein, the dates in the same group have the same week identification;
acquiring the operation frequency of commodities on each date in each group;
calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group;
and taking the set of the commodity operation trends of each group as the historical commodity operation trend.
C4. The data processing apparatus according to C3, wherein the calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group includes:
sequentially counting the commodity operation frequency in each group according to a preset sliding window, wherein the commodity operation frequency of each group is sequenced according to the date sequence corresponding to the commodity operation frequency;
calculating the index value of the commodity operation frequency sequentially taken by the sliding window until the sliding window finishes taking the commodity operation frequency in each group;
and taking the set of the index values of the commodity operation frequency in each group as the commodity operation trend of each group, wherein the index values of the commodity operation frequency in each group are sorted and set according to the sequence of the commodity operation frequency taken by the sliding window.
C5. The data processing device as described in C2, wherein the characteristic data representing the commodity operation frequency includes a historical commodity operation mean value;
the historical commodity operation mean value is obtained through the following method:
grouping dates in the historical time period; wherein, the dates in the same group have the same week identification;
acquiring the operation times of commodities on each date in each group;
calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group;
and taking the set of the commodity operation mean values of each group as the historical commodity operation mean values.
C6. The data processing apparatus according to C1, wherein the product operation policy information includes one or any combination of the following:
voucher information, money information, point card information, and red envelope information.
C7. The data processing apparatus of C1, the weather information comprising one or a combination of: temperature information and climate information;
the holiday identification information is used for identifying whether the date is a holiday.
C8. The data processing apparatus according to any one of claims C1 to C7, wherein the feature information is specifically: characteristic information coded by a one-hot coding mode.
C9. The data processing apparatus according to C1, wherein the predictive model is composed of a plurality of machine learning models;
the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency comprises the following steps:
acquiring commodity operation frequency output by each machine learning model according to the historical commodity operation record and the real-time characteristic information of the target user;
and determining the commodity operation frequency of the target user according to the commodity operation frequency output by each machine learning model.
The embodiment of the application discloses D1. A data processing device, including:
the data acquisition module is used for acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
the frequency prediction module is used for predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of a user and historical characteristic information;
and the stage prediction module is used for acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
D2. The data processing apparatus according to D1, wherein the obtaining a life cycle stage of the target user according to the predicted commodity operation frequency of the target user includes:
obtaining the index value of the predicted commodity operation frequency of the target user;
and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and a preset life cycle stage threshold value.
D3. The data processing apparatus according to D1, wherein the obtaining a life cycle stage of the target user according to the predicted commodity operation frequency of the target user includes:
acquiring the predicted index value of the commodity operation frequency of the target user and the historical index value of the commodity operation frequency of the target user;
and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and the historical index value of the commodity operation frequency of the target user.
The embodiment of the application discloses E1. A server includes:
at least one processor; and (c) a second step of,
a memory communicatively coupled to the at least one processor; wherein, the first and the second end of the pipe are connected with each other,
the memory stores instructions executable by the at least one processor to cause the at least one processor to perform:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein the characteristic information includes one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of the user and historical characteristic information.
E2. The server of E1, the at least one processor capable of performing: the data processing method according to any one of A2 to A9.
The embodiment of the application discloses a server, including:
at least one processor; and (c) a second step of,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to cause the at least one processor to perform:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to historical commodity operation records of a user and historical characteristic information;
and acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
F2. The server of F1, the at least one processor capable of performing: the data processing method according to any one of B2 to B3.
The embodiment of the application discloses G1. A computer-readable storage medium, which stores a computer program, wherein the computer program realizes the data processing method of any one of A1-A9 when being executed by a processor.
An embodiment of the application discloses h1. A computer-readable storage medium storing a computer program which, when executed by a processor, implements the data processing method of any one of B1-B3.
Claims (24)
1. A method of data processing, comprising:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein the characteristic information includes one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
acquiring characteristic data representing the commodity operation frequency according to the historical commodity operation record of the target user, wherein the characteristic data representing the commodity operation frequency refers to a historical commodity operation trend or a historical commodity operation mean value; the historical commodity operation trend and the historical commodity operation mean value are obtained through the following modes: grouping dates in the historical time period, wherein the dates in the same group have the same week identification, and acquiring the commodity operation times of each date in each group; calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group, and taking the set of the commodity operation trends of each group as the historical commodity operation trend; calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group, and taking the set of the commodity operation mean values of each group as the historical commodity operation mean value;
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to characteristic data of multiple users representing commodity operation frequency and historical characteristic information.
2. The data processing method according to claim 1, wherein the calculating of the operation trend of the commodities for each group according to the operation frequency of the commodities on each date in each group comprises:
sequentially counting the commodity operation frequency in each group according to a preset sliding window, wherein the commodity operation frequency of each group is sorted according to the date sequence corresponding to the commodity operation frequency;
calculating the index value of the commodity operation frequency sequentially taken by the sliding window until the sliding window finishes taking the commodity operation frequency in each group;
and taking the set of the index values of the commodity operation frequency in each group as the commodity operation trend of each group, wherein the index values of the commodity operation frequency in each group are sorted and set according to the sequence of the commodity operation frequency taken by the sliding window.
3. The data processing method of claim 1, wherein the commodity operation policy information comprises one or any combination of the following:
voucher information, money information, point card information, and red envelope information.
4. The data processing method of claim 1, wherein the weather information comprises one or a combination of: temperature information and climate information;
the holiday identification information is used for identifying whether the date is a holiday.
5. The data processing method according to any one of claims 1 to 4, wherein the feature information is specifically: and (4) encoding the characteristic information by a one-hot encoding mode.
6. The data processing method of claim 1, wherein the predictive model consists of a number of machine learning models;
the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency comprises the following steps:
according to the historical commodity operation records and the real-time characteristic information of the target user, acquiring the commodity operation frequency output by each machine learning model;
and determining the commodity operation frequency of the target user according to the commodity operation frequency output by each machine learning model.
7. A method of data processing, comprising:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein the characteristic information includes one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
acquiring characteristic data representing the commodity operation frequency according to the historical commodity operation record of the target user, wherein the characteristic data representing the commodity operation frequency refers to a historical commodity operation trend or a historical commodity operation mean value; the historical commodity operation trend and the historical commodity operation mean value are obtained through the following modes: grouping dates in the historical time period, wherein the dates in the same group have the same week identification, and acquiring the commodity operation times of each date in each group; calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group, and taking the set of the commodity operation trends of each group as the historical commodity operation trend; calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group, and taking the set of the commodity operation mean values of each group as the historical commodity operation mean value; predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to characteristic data representing commodity operation frequency of a plurality of users and historical characteristic information;
and acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
8. The data processing method of claim 7, wherein the obtaining the life cycle stage of the target user according to the predicted commodity operation frequency of the target user comprises:
obtaining the index value of the predicted commodity operation frequency of the target user;
and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and a preset life cycle stage threshold value.
9. The data processing method of claim 7, wherein the obtaining the life cycle stage of the target user according to the predicted commodity operation frequency of the target user comprises:
acquiring the predicted index value of the commodity operation frequency of the target user and the historical index value of the commodity operation frequency of the target user;
and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and the historical index value of the commodity operation frequency of the target user.
10. A data processing apparatus, comprising:
the acquisition module is used for acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
acquiring characteristic data representing the commodity operation frequency according to the historical commodity operation record of the target user, wherein the characteristic data representing the commodity operation frequency refers to a historical commodity operation trend or a historical commodity operation mean value; the historical commodity operation trend and the historical commodity operation mean value are obtained through the following modes: grouping dates in the historical time period, wherein the dates in the same group have the same week identification, and acquiring the commodity operation times of each date in each group; calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group, and taking the set of the commodity operation trends of each group as the historical commodity operation trend; calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group, and taking the set of the commodity operation mean values of each group as the historical commodity operation mean value; the prediction module predicts the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to characteristic data of multiple users representing commodity operation frequency and historical characteristic information.
11. The data processing apparatus according to claim 10, wherein said calculating the operation tendency of the commodities for each group based on the operation frequency of the commodities on each date in each group comprises:
sequentially counting the commodity operation frequency in each group according to a preset sliding window, wherein the commodity operation frequency of each group is sorted according to the date sequence corresponding to the commodity operation frequency;
calculating the index values of the commodity operation frequency sequentially taken by the sliding window until the sliding window finishes taking the commodity operation frequency in each group;
and taking the set of the index values of the commodity operation frequency in each group as the commodity operation trend of each group, wherein the index values of the commodity operation frequency in each group are sorted and set according to the sequence of the commodity operation frequency taken by the sliding window.
12. The data processing apparatus according to claim 10, wherein the commodity operation policy information includes one or any combination of the following:
voucher information, money information, point card information, red envelope information.
13. The data processing apparatus of claim 10, wherein the weather information comprises one or a combination of: temperature information and climate information;
the holiday identification information is used for identifying whether the date is a holiday.
14. The data processing apparatus according to any one of claims 10 to 13, wherein the feature information is specifically: and (4) encoding the characteristic information by a one-hot encoding mode.
15. The data processing apparatus of claim 10, wherein the predictive model consists of a number of machine learning models;
the predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency comprises the following steps:
acquiring commodity operation frequency output by each machine learning model according to the historical commodity operation record and the real-time characteristic information of the target user;
and determining the commodity operation frequency of the target user according to the commodity operation frequency output by each machine learning model.
16. A data processing apparatus, comprising:
the data acquisition module is used for acquiring historical commodity operation records and real-time characteristic information of a target user; wherein the characteristic information includes one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information; obtaining characteristic data representing commodity operation frequency according to the historical commodity operation record of the target user, wherein the characteristic data representing the commodity operation frequency refers to a historical commodity operation trend or a historical commodity operation average value; the historical commodity operation trend and the historical commodity operation mean value are obtained through the following modes: grouping dates in the historical time period, wherein the dates in the same group have the same week identification, and acquiring the commodity operation times of each date in each group; calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group, and taking the set of the commodity operation trends of each group as the historical commodity operation trend; calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group, and taking the set of the commodity operation mean values of each group as the historical commodity operation mean value; the frequency prediction module is used for predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to characteristic data representing commodity operation frequency of a plurality of users and historical characteristic information;
and the stage prediction module is used for acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
17. The data processing device of claim 16, wherein the obtaining of the lifecycle stage of the target user according to the predicted commodity operation frequency of the target user comprises:
acquiring the index value of the predicted commodity operation frequency of the target user;
and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and a preset life cycle stage threshold value.
18. The data processing device of claim 16, wherein the obtaining of the lifecycle stage of the target user according to the predicted commodity operation frequency of the target user comprises:
acquiring the predicted index value of the commodity operation frequency of the target user and the historical index value of the commodity operation frequency of the target user;
and judging the life cycle stage of the target user according to the comparison result of the index value of the commodity operation frequency and the historical index value of the commodity operation frequency of the target user.
19. A server, comprising:
at least one processor; and (c) a second step of,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to cause the at least one processor to perform:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
acquiring characteristic data representing the commodity operation frequency according to the historical commodity operation record of the target user, wherein the characteristic data representing the commodity operation frequency refers to a historical commodity operation trend or a historical commodity operation mean value; the historical commodity operation trend and the historical commodity operation mean value are obtained through the following modes: grouping dates in the historical time period, wherein the dates in the same group have the same week identification, and acquiring the commodity operation times of each date in each group; calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group, and taking the set of the commodity operation trends of each group as the historical commodity operation trend; calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group, and taking the set of the commodity operation mean values of each group as the historical commodity operation mean value;
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to characteristic data of a plurality of users representing commodity operation frequency and historical characteristic information.
20. The server according to claim 19, wherein said at least one processor is capable of performing: a data processing method as claimed in any one of claims 2 to 6.
21. A server, comprising:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to cause the at least one processor to perform:
acquiring historical commodity operation records and real-time characteristic information of a target user; wherein, the characteristic information comprises one of the following or any combination thereof: commodity operation strategy information, weather information and holiday identification information;
acquiring characteristic data representing the commodity operation frequency according to the historical commodity operation record of the target user, wherein the characteristic data representing the commodity operation frequency refers to a historical commodity operation trend or a historical commodity operation mean value; the historical commodity operation trend and the historical commodity operation mean value are obtained through the following modes: grouping dates in the historical time period, wherein the dates in the same group have the same week identification, and acquiring the commodity operation times of each date in each group; calculating the commodity operation trend of each group according to the commodity operation frequency of each date in each group, and taking the set of the commodity operation trends of each group as the historical commodity operation trend; calculating the commodity operation mean value of each group according to the commodity operation times of each date in each group, and taking the set of the commodity operation mean values of each group as the historical commodity operation mean value;
predicting the commodity operation frequency of the target user through a preset prediction model for predicting the commodity operation frequency according to the historical commodity operation record and the real-time characteristic information; the prediction model is obtained by training according to characteristic data representing commodity operation frequency of a plurality of users and historical characteristic information;
and acquiring the life cycle stage of the target user according to the predicted commodity operation frequency of the target user.
22. The server according to claim 21, wherein said at least one processor is capable of performing: a data processing method as claimed in any one of claims 8 to 9.
23. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the data processing method of any one of claims 1 to 6.
24. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the data processing method of any one of claims 7 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910570263.9A CN110472991B (en) | 2019-06-27 | 2019-06-27 | Data processing method, device, server and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910570263.9A CN110472991B (en) | 2019-06-27 | 2019-06-27 | Data processing method, device, server and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110472991A CN110472991A (en) | 2019-11-19 |
| CN110472991B true CN110472991B (en) | 2022-10-28 |
Family
ID=68507074
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910570263.9A Active CN110472991B (en) | 2019-06-27 | 2019-06-27 | Data processing method, device, server and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110472991B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112132622B (en) * | 2020-09-25 | 2021-07-16 | 北京达佳互联信息技术有限公司 | Data estimation method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559758A (en) * | 2013-11-06 | 2014-02-05 | 上海煦荣信息技术有限公司 | Intelligent vending system and intelligent vending method |
| CN105260803A (en) * | 2015-11-06 | 2016-01-20 | 国家电网公司 | Power consumption prediction method for system |
| CN107818478A (en) * | 2017-10-31 | 2018-03-20 | 携程计算机技术（上海）有限公司 | Overseas reward voucher distribution method and system based on provisional profit |
-
2019
- 2019-06-27 CN CN201910570263.9A patent/CN110472991B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559758A (en) * | 2013-11-06 | 2014-02-05 | 上海煦荣信息技术有限公司 | Intelligent vending system and intelligent vending method |
| CN105260803A (en) * | 2015-11-06 | 2016-01-20 | 国家电网公司 | Power consumption prediction method for system |
| CN107818478A (en) * | 2017-10-31 | 2018-03-20 | 携程计算机技术（上海）有限公司 | Overseas reward voucher distribution method and system based on provisional profit |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110472991A (en) | 2019-11-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109741177A (en) | Appraisal procedure, device and the intelligent terminal of user credit | |
| JP6330901B2 (en) | Hierarchical hidden variable model estimation device, hierarchical hidden variable model estimation method, payout amount prediction device, payout amount prediction method, and recording medium | |
| CN110333991B (en) | Method for predicting maximum resource utilization rate of cloud platform tasks | |
| CN110147389A (en) | Account number treating method and apparatus, storage medium and electronic device | |
| US8290880B2 (en) | System and method for estimating a parameter that represents data describing a physical system | |
| CN108154311A (en) | Top-tier customer recognition methods and device based on random forest and decision tree | |
| CN111143685A (en) | Recommendation system construction method and device | |
| CN103617146B (en) | A kind of machine learning method and device based on hardware resource consumption | |
| KR101860457B1 (en) | Method for analyzing weather affect and apparatus for executing the method | |
| CN110472991B (en) | Data processing method, device, server and storage medium | |
| CN110490635B (en) | Commercial tenant dish transaction prediction and meal preparation method and device | |
| CN107784511A (en) | A kind of customer loss Forecasting Methodology and device | |
| CN113838303B (en) | Parking lot recommendation method and device, electronic equipment and storage medium | |
| CN103646670A (en) | Method and device for evaluating performances of storage system | |
| CN110134839A (en) | Time series data characteristic processing method, apparatus and computer readable storage medium | |
| CN116740946B (en) | Highway traffic flow big data analysis processing method | |
| CN111464337B (en) | Resource allocation method and device and electronic equipment | |
| CN111127114A (en) | Method and device for determining power load declaration data | |
| CN110069708A (en) | A kind of promotion effect predictor method, device, medium and equipment promoted across medium | |
| CN115630733A (en) | Order delivery time estimation method, system, medium and electronic device | |
| CN112396236B (en) | Traffic flow prediction method, system, server and storage medium | |
| CN103020081A (en) | Offsite keyword throwing-in method and system | |
| CN112489422A (en) | Traffic flow prediction method based on algorithm model analysis | |
| CN111784071A (en) | Stacking integration-based license occupation and prediction method and system | |
| CN110222842A (en) | A kind of network model training method, device and storage medium |
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 |