CN111127074B - Data recommendation method - Google Patents
Data recommendation method Download PDFInfo
- Publication number
- CN111127074B CN111127074B CN201911172534.1A CN201911172534A CN111127074B CN 111127074 B CN111127074 B CN 111127074B CN 201911172534 A CN201911172534 A CN 201911172534A CN 111127074 B CN111127074 B CN 111127074B
- Authority
- CN
- China
- Prior art keywords
- user
- record
- data
- main body
- receiving
- 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
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Landscapes
- Business, Economics & Management (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a data recommendation method, which is characterized by comprising the following steps: acquiring an identity of a user; acquiring a user superior record; inputting the user superior record into a potential demand generation model obtained by pre-training, and outputting a potential demand prediction result of a behavior main body; wherein generating the model according to the potential demand comprises at least one of: receiving a determination and analysis instruction, dividing a tag into a determination region, wherein the determination region takes a defined tag value range as a judgment basis, and determining the behavior main body when the user superior record belongs to one of the tag division regions, and matching the user superior record with the subordinate record of the same behavior main body in a sample library; and receiving a relative analysis instruction, dividing the labels into relative areas, wherein the relative areas take the size of an interval label value range as a judgment standard, and determining the behavior main body when the user superior record meets the relative areas, and matching the user superior record with the lower records of the same behavior main body in a sample library.
Description
Technical Field
The invention relates to the field of AI data analysis and deduction, in particular to a data recommendation method.
Background
With the development of science and technology, more and more analysis prediction based on big data appears; however, analysis and statistics of big data are tedious, manual operation is difficult, operation cost is high, economic cost is not saved, all industries are not communicated, statistical prediction modes are disordered, and big data cannot be effectively utilized. Some private enterprises or small industries still rely on subjective consciousness judgment of people due to insufficient funds, so that the defects of poor accuracy, slow update consciousness, uneven personnel level, incapability of copying and small effective range are caused.
At present, the continuous accumulation of big data forms a statistical problem, and the analysis and prediction deduction layer of the big data is continuous, most of the data is extracted from a plurality of business systems and contains historical data, and the big data such as some error data or the collision among the data cannot be effectively utilized, and the big data is a factor influencing the result when the big data is related to a target individual. The existing big data analysis algorithm is basically based on simple summary statistics, and some parts can use a point relative value. The adoption of the relative value algorithm is very likely to cause the potential demands of some users to be submerged, and finally, the later analysis result is misjudged.
Aiming at the technical problem that the prediction result of big data is inaccurate because the potential demands of users are submerged when the big data is analyzed by adopting a relative value algorithm in the prior art, no effective solution is proposed at present.
Disclosure of Invention
The invention aims to solve the technical problem that the prediction result of big data is inaccurate due to the fact that the potential demands of users are submerged, and provides a data recommendation method.
The invention adopts the following technical scheme for solving the technical problems: a data recommendation method, comprising: acquiring an identity of a user; acquiring a user superior record; inputting the user superior record into a potential demand generation model obtained by pre-training, and outputting a potential demand prediction result of a behavior main body; the step of inputting the user superior record into a pre-trained potential demand generation model comprises the step of generating the model according to the potential demand, and the method comprises at least one of the following steps: receiving a determination and analysis instruction, dividing a tag into a determination region, wherein the determination region takes a defined tag value range as a judgment basis, and determining the behavior main body when the user superior record belongs to one of the tag division regions, and matching the user superior record with the subordinate record of the same behavior main body in a sample library; and receiving a relative analysis instruction, dividing the labels into relative areas, wherein the relative areas take the size of an interval label value range as a judgment standard, and determining the behavior main body when the user superior record meets the relative areas, and matching the user superior record with the lower records of the same behavior main body in a sample library.
Preferably, before the step of generating the model for potential demand, the method further comprises the steps of: receiving a plurality of user histories, wherein the plurality of user histories respectively correspond to a plurality of users; receiving a tag instruction, screening according to the plurality of user histories, and deriving type parameters; an instruction to calculate intermediate data is received.
Preferably, after the potential demand generating model, the method further comprises: and receiving user analysis instructions one by one, and outputting a potential demand prediction result of the behavior main body for a service scene of data recommendation.
Preferably, receiving a tag instruction, filtering according to a plurality of user histories, and deriving a type parameter, including: s2021, setting a label according to preset information; s2022, performing label setting on the plurality of user histories according to labels; s2023, generating the type parameter after screening the plurality of user histories; s2024, storing the type parameter; s2025, reading the type parameter.
Preferably, after receiving the tag instruction and filtering according to the plurality of user histories, the method further comprises: and receiving a member circulation instruction, and grouping users, wherein the users are classified into a virginee purchasing user, an active user and a sleeping user.
Preferably, the step of receiving the instruction to calculate the intermediate data includes: receiving a label weight setting instruction; according to the respective duty ratio of the tag in the user history record, data coupling is carried out to obtain type parameters in the user history record; according to the type parameters in the user history record, determining a behavior main body corresponding to the user history record: from a plurality of user histories, acquiring histories belonging to the same behavior main body as reference information, and determining a lower record for optimizing the corresponding behavior main body according to type parameters; and determining a lower record forming sample library for optimizing the corresponding behavior main body according to the type parameter.
Preferably, determining the history record for optimizing the corresponding behavior subject according to the type parameter includes: according to a preset data screening algorithm, removing the unreliable data from the reference information; and analyzing and obtaining a lower record in the behavior main body according to the trusted data except the untrusted data in the reference information.
Preferably, receiving an instruction for setting tag weight includes: and setting a weight value for the tag according to preset information adjustment conditions.
The embodiment of the invention also provides computer equipment, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, and is characterized in that the data recommendation method of any one of the above is realized when the processor executes the computer program.
The embodiment of the invention also provides a computer readable storage medium, which is characterized in that the computer readable storage medium stores a computer program of any one of the data recommendation methods.
The beneficial effects of the invention are as follows:
1) The method solves the technical problem that the prediction result of big data is inaccurate because the potential demands of users are submerged;
2) The method and the system realize the great improvement of the prediction and recommendation accuracy of the potential demand data and reduce the popularization cost;
3) According to the method and the system, the marketing schemes are generated one by one according to the habit 1V1 of the user, so that accurate operation of potential demands of the user is realized.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. In the drawings:
FIG. 1 is a flow chart of processing data using a data recommendation method provided in accordance with an embodiment of the present invention;
FIG. 2 is a schematic diagram of various prediction schemes in a model generation process according to potential requirements, provided in accordance with an embodiment of the present invention;
FIG. 3 is a flow chart of a data recommendation method provided according to an embodiment of the present invention;
FIG. 4 is a detailed flowchart of a data recommendation method according to an embodiment of the present invention;
fig. 5 is a flowchart of sample library formation in a data recommendation method according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention will be described in further detail with reference to the accompanying drawings. The exemplary embodiments of the present invention and their descriptions herein are for the purpose of explaining the present invention, but are not to be construed as limiting the invention.
To clearly define the concepts of "model" and "sample library", the following description will now be made for these two words:
the formation of models corresponds to the "experience" in a computer, usually in the form of "data", so that the main content of the study of machine learning is the algorithm on the computer that generates "models" from the data (also called "learning algorithm"). With learning algorithms, we input empirical data and the computer can build a model based on the existing data.
The sample library corresponds to a training set, i.e., a set of empirical data used in a machine learning training process.
Example 1
In the embodiment of the present invention, a data recommendation method is provided, and fig. 1 is a flowchart of a method for recommending data according to an embodiment of the present invention, where as shown in fig. 1, a data recommendation method includes:
s101, acquiring an identity of a user;
s102, acquiring a user superior record;
s103, inputting the user superior record into a potential demand generation model obtained by pre-training, and outputting a potential demand prediction result of the behavior main body;
specifically, in combination with a SaaS model, a store performs subscription and then logs in to the system, acquires a buyer ID identifier through a seller_nick code, acquires an authorized binding API, synchronizes client back-end data with RDS data, acquires combined order data of last shopping of the buyer from an RDS version of an alicloud database, performs data cleaning, and then, guides the combined order data into the model to find a matched behavior main body, and if the last shopping is the 5 th shopping, performs data analysis on a lower record (namely the 6 th shopping record) of the behavior main body and then recommends the data as a prediction result to a relevant user.
FIG. 2 is a schematic diagram of various prediction schemes in a model generation process according to potential requirements, provided in accordance with an embodiment of the present invention;
the step of inputting the user superior record into the pre-trained potential demand generation model includes S204, generating the model according to the potential demand, including at least one of the following steps:
receiving a determination area analysis instruction, dividing a tag into a determination area, wherein the determination area takes a defined tag value range as a judgment basis, determining a behavior main body when a user upper record belongs to one of the tag division areas, and matching lower records of the same behavior main body in a sample library;
specifically, the user superior record is input into a pre-trained item_record sample library potential demand generation model, and potential demand prediction results of the output behavior main body are written into an Ariycloud database RDS version.
According to the method, after a definite area is defined, the number of times of next shopping of main labels of the infants A and B is respectively 6 times and 3 times according to the limit of a definite area, the lower records of the same behavior main body are searched in a item_recommend sample library, recommendation information obtained by the infants A is two-section milk powder, recommendation obtained by the infants B is one-section milk powder, accurate recommendation of users one by one is realized, user information pushing accuracy is improved, periodic updating of the item_recommend sample library is realized, and a high-accuracy recommendation method is greatly weakened by a large margin.
And receiving a relative analysis instruction, dividing the labels into relative areas, taking the size of the range of the interval label values as a judgment standard in the relative areas, determining a behavior main body when the upper-level records of the users meet the relative areas, and matching the lower-level records of the same behavior main body in a sample library.
Specifically, continuing to use the purchased milk powder for children as an actual application scene, and supposing that the item_assembled sample library is recorded and analyzed to obtain the first section of milk powder for infants, wherein the first section of milk powder needs to be purchased 5 times and the second section of milk powder needs to be purchased 6 times; assuming that the item_recommended sample library records that the relative analysis results in the most milk powder to be purchased next time for infants, the number of people shopping for many times is usually much smaller than the number of people shopping for few times. The prior infants A and B display that the 5 th and 4 th times of the purchasing of a section of milk powder is displayed in the superior record of the user, and the infants A and B recommend a section of milk powder according to the relative analysis, so that the pushing is not accurate for the infant A; if the analysis is adopted to analyze that the infant A recommends the second section of milk powder for the 6 th time and the infant B recommends the first section of milk powder for the 5 th time, the pushing obtained by the two persons is accurate. As can be seen from the above, the conventional data recommendation method using the relative analysis is very easy to submerge the original data record of the 6 th shopping.
It should be noted that the relative analysis is to effectively reduce unnecessary recommendation cost and improve generalization capability of the model under the condition of ensuring accuracy by adopting a technical means in the case that the same shopping record, namely the lower-level record, cannot be obtained for various reasons.
It should be noted that, in step S204, the lower level record may be recommended according to the optimal sub-scheme arrangement and combination mode according to the needs of the customer store, and the above mentioned "analysis instruction" and "relative analysis instruction" may be used alone or in combination according to the needs of the customer.
Fig. 3 is a flowchart of a data recommendation method according to an embodiment of the present invention, as shown in fig. 3,
before the step of generating the model according to the potential requirements, the method further comprises the following steps:
s201, receiving a plurality of user histories, wherein the user histories correspond to a plurality of users respectively;
s202, receiving a tag instruction, screening according to a plurality of user histories, and deriving type parameters;
s203, receiving an instruction for calculating intermediate data.
Specifically, the plurality of user histories includes: and (3) calling MaxCompute big data calculation API by using intermediate data such as sales related data, shopping interval data, logistics aging data, guest unit price tags and the like, and analyzing and processing mass data economically and efficiently by adopting distributed calculation.
After generating the model from the potential requirements, the method further comprises:
s205, receiving analysis instructions of users one by one, and outputting a potential demand prediction result of the behavior main body for a service scene of data recommendation.
As can be seen from the foregoing, fig. 4 is a detailed flowchart of a data recommendation method according to an embodiment of the present invention.
Preferably, receiving a tag instruction, filtering according to a plurality of user histories, and deriving a type parameter, including: s2021, setting a label according to preset information;
s2022, performing label setting on the plurality of user histories according to labels;
s2023, generating the type parameter after screening the plurality of user histories;
s2024, storing the type parameter;
s2025, reading the type parameter.
Preferably, after receiving the tag instruction and filtering according to the plurality of user histories, the method further comprises: and receiving a member circulation instruction, and grouping users, wherein the users are classified into a virginee purchasing user, an active user and a sleeping user.
Specifically, the user is classified according to the comparison of the latest shopping time and the current day data, so that the next record of the similar behavior main body is recommended conveniently, the multidimensional data are integrated, and the most needed articles are recommended.
Fig. 5 is a flowchart of sample library formation in a data recommendation method according to an embodiment of the present application, and as shown in fig. 5, the step of receiving an instruction for calculating intermediate data includes: receiving a label weight setting instruction; according to the respective duty ratio of the tag in the user history record, data coupling is carried out to obtain type parameters in the user history record; according to the type parameters in the user history record, determining a behavior main body corresponding to the user history record: from a plurality of user histories, acquiring histories belonging to the same behavior main body as reference information, and determining a lower record for optimizing the corresponding behavior main body according to type parameters; and determining a lower record forming sample library for optimizing the corresponding behavior main body according to the type parameter.
It should be noted that the tag weight is set according to the customer store requirements, and the requirements of each store for tag data are different.
Determining a history record for optimizing a corresponding behavior subject according to the type parameter, wherein the history record comprises: according to a preset data screening algorithm, removing the unreliable data from the reference information; and analyzing and obtaining a lower record in the behavior main body according to the trusted data except the untrusted data in the reference information.
Receiving a label weight setting instruction, comprising: and setting a weight value for the tag according to preset information adjustment conditions.
Specifically, taking an e-commerce as an example, recommending an article with highest purchase possibility of a customer as a final purpose, setting weights on related labels according to customer purchase demands in an actual scene, avoiding the occurrence of long tail theory, submerging some real potential demand data, customizing the settings according to the customer demands, and improving the accuracy of the data.
Example two
The embodiment of the invention provides an application scene of a data recommendation method, wherein the new baby is an actual scene, and the new baby means no history message, so that a scheme of using relative analysis is needed to operate related content. By searching similar baby or the history record of similar baby, the found potential demands are subjected to baby demand priority arrangement, preferential selection and continuous data accumulation to generate a more comprehensive model.
A data recommendation method, comprising: acquiring an identity of a user; acquiring a user superior record; inputting the user superior record into a potential demand generation model obtained by pre-training, and outputting a potential demand prediction result of a behavior main body; the step of inputting the user superior record into the pre-trained potential demand generation model comprises the step of generating the model according to the potential demand, wherein the step of inputting the user superior record into the pre-trained potential demand generation model comprises the following steps of: and receiving a relative analysis instruction, dividing the labels into relative areas, wherein the relative areas take the size of an interval label value range as a judgment standard, and determining the behavior main body when the user superior record meets the relative areas, and matching the user superior record with the lower records of the same behavior main body in a sample library.
Specifically, in the case of no matching history data, the store needs to set a relative area by adopting a relative analysis manner, mobilize an approximate relative area, search a matched behavior body for a item_record sample library formed according to a similar baby of a last new baby, a similar baby history record and the like, perform data analysis according to a lower record in the item_record sample library, and recommend the data analysis to the client.
Preferably, before the step of generating the model for potential demand, the method further comprises the steps of: receiving a plurality of user histories, wherein the plurality of user histories respectively correspond to a plurality of users; receiving a tag instruction, screening according to the plurality of user histories, and deriving type parameters; an instruction to calculate intermediate data is received.
Preferably, after the potential demand generating model, the method further comprises: and receiving user analysis instructions one by one, and outputting a potential demand prediction result of the behavior main body for a service scene of data recommendation.
Preferably, receiving a tag instruction, filtering according to a plurality of user histories, and deriving a type parameter, including:
s2021, setting a label according to preset information;
s2022, performing label setting on the plurality of user histories according to labels;
s2023, generating the type parameter after screening the plurality of user histories;
s2024, storing the type parameter;
s2025, reading the type parameter.
Preferably, after receiving the tag instruction and filtering according to the plurality of user histories, the method further comprises: and receiving a member circulation instruction, and grouping users, wherein the users are classified into a virginee purchasing user, an active user and a sleeping user.
Preferably, the step of receiving the instruction to calculate the intermediate data includes: receiving a label weight setting instruction; according to the respective duty ratio of the tag in the user history record, data coupling is carried out to obtain type parameters in the user history record; according to the type parameters in the user history record, determining a behavior main body corresponding to the user history record: from a plurality of user histories, acquiring histories belonging to the same behavior main body as reference information, and determining a lower record for optimizing the corresponding behavior main body according to type parameters; and determining a lower record forming sample library for optimizing the corresponding behavior main body according to the type parameter.
Specifically, the relative area is set, the range of the relative area is gradually narrowed each time the data iterates, the setting of the range of the relative area aims at various labels, when the range of the relative area is met by the superior records of the user, the retrieved data are reasonable data after multi-dimensional estimation, after the relative area is gradually narrowed, the recommended multiple babies have priority, the influence factors of the babies on the recommendation have priority, and then the data accuracy meeting the potential requirements of the user is higher.
Preferably, determining the history record for optimizing the corresponding behavior subject according to the type parameter includes: according to a preset data screening algorithm, removing the unreliable data from the reference information; and analyzing and obtaining a lower record in the behavior main body according to the trusted data except the untrusted data in the reference information.
Preferably, receiving an instruction for setting tag weight includes: and setting a weight value for the tag according to preset information adjustment conditions.
Example III
In the embodiment of the invention, by taking double 11 ladder marketing as an example, the system combines the ranking requirement of double 11 preheating meeting places, and the preheating and purchasing collection number can be enabled to show the trend of increasing or decreasing or other waveforms by distributing marketing numbers according to the percentage of days, so that the double 11 preheating purpose is better achieved.
According to statistics of the past year big data, 11 days of 11 months are burst days, 11 months of 10 days are consolidation days, the preheating period is 11 months of 01 days-11 months of 10 days, the first weekend after 11 months of 01 days begins, and the last weekend before the preheating period of 11 months of 01 days is the leading day.
In the whole hot marketing period, data recommendation needs to carry out data analysis on the historical records of users according to the setting of shops, and recommending proper articles to corresponding users, and on the basis of meeting the users, the whole sales performance of the shops is strongly optimized, so that the 'burst' is not recommended in a large quantity only, but the recommendation is accurately realized according to the demands of clients, and the recommendation time, the recommendation of babies, the recommendation mode, the consumption behavior and the like of the clients are predicted by combining the historical data, so that the optimal sales state of the shops is achieved.
A data recommendation method, comprising: acquiring an identity of a user; acquiring a user superior record; inputting the user superior record into a potential demand generation model obtained by pre-training, and outputting a potential demand prediction result of a behavior main body; the step of inputting the user superior record into the pre-trained potential demand generation model comprises the step of generating the model according to the potential demand, wherein the step comprises at least one of the following steps: receiving a determination and analysis instruction, dividing a tag into a determination region, wherein the determination region takes a defined tag value range as a judgment basis, and determining the behavior main body when the user superior record belongs to one of the tag division regions, and matching the user superior record with the subordinate record of the same behavior main body in a sample library; and receiving a relative analysis instruction, dividing the labels into relative areas, wherein the relative areas take the size of an interval label value range as a judgment standard, and determining the behavior main body when the user superior record meets the relative areas, and matching the user superior record with the lower records of the same behavior main body in a sample library.
Preferably, before the step of generating the model for potential demand, the method further comprises the steps of: receiving a plurality of user histories, wherein the plurality of user histories respectively correspond to a plurality of users; receiving a tag instruction, screening according to the plurality of user histories, and deriving type parameters; an instruction to calculate intermediate data is received.
Preferably, after the potential demand generating model, the method further comprises: and receiving user analysis instructions one by one, and outputting a potential demand prediction result of the behavior main body for a service scene of data recommendation.
Preferably, receiving a tag instruction, filtering according to a plurality of user histories, and deriving a type parameter, including:
s2021, setting a label according to preset information;
s2022, performing label setting on the plurality of user histories according to labels;
s2023, generating the type parameter after screening the plurality of user histories;
s2024, storing the type parameter;
s2025, reading the type parameter.
Preferably, after receiving the tag instruction and filtering according to the plurality of user histories, the method further comprises: and receiving a member circulation instruction, and grouping users, wherein the users are classified into a virginee purchasing user, an active user and a sleeping user.
Preferably, the step of receiving the instruction to calculate the intermediate data includes: receiving a label weight setting instruction; according to the respective duty ratio of the tag in the user history record, data coupling is carried out to obtain type parameters in the user history record; according to the type parameters in the user history record, determining a behavior main body corresponding to the user history record: from a plurality of user histories, acquiring histories belonging to the same behavior main body as reference information, and determining a lower record for optimizing the corresponding behavior main body according to type parameters; and determining a lower record forming sample library for optimizing the corresponding behavior main body according to the type parameter.
Preferably, determining the history record for optimizing the corresponding behavior subject according to the type parameter includes: according to a preset data screening algorithm, removing the unreliable data from the reference information; and analyzing and obtaining a lower record in the behavior main body according to the trusted data except the untrusted data in the reference information.
Specifically, taking the recommending time of recommending the baby as an example, reading historical data of a corresponding client, analyzing aiming at user habits, ordering and purchasing at 7-9 points in the middle of working, 13-14 points on the noon break, 18-20 points on the noon break, and 22-23 points before sleeping on some users in the recommending time, analyzing the data of the historical ordering time of the user by taking the time as a main label, importing user data to be predicted after the same single behavior main body generates type parameters, predicting the optimal potential recommending time, and recommending to set the optimal potential recommending baby according to the optimal potential recommending time.
Preferably, receiving an instruction for setting tag weight includes: and setting a weight value for the tag according to preset information adjustment conditions.
The embodiment of the invention also provides a computer device which is used for solving the technical problem of inaccurate big data prediction results caused by submerged potential demands of users, and the computer device comprises a memory, a processor and a computer program which is stored in the memory and can be run on the processor, wherein the processor realizes the data recommendation method when executing the computer program.
The embodiment of the invention also provides a computer readable storage medium for solving the technical problem of inaccurate big data prediction results caused by submerged potential demands of users, wherein the computer readable storage medium stores a computer program for executing the data recommendation method.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (8)
1. A data recommendation method, comprising:
s101, acquiring an identity of a user;
s102, acquiring a user superior record;
s103, inputting the user superior record into a potential demand generation model obtained by pre-training, and outputting a potential demand prediction result of the behavior main body;
the step of inputting the user superior record into the pre-trained potential demand generation model comprises the following steps: s204, generating the model according to the potential demand, wherein the model comprises at least one of the following steps:
receiving a determination and analysis instruction, dividing a tag into a determination region, wherein the determination region takes a defined tag value range as a judgment basis, and determining the behavior main body when the user superior record belongs to one of the tag division regions, and matching the user superior record with the subordinate record of the same behavior main body in a sample library;
receiving a relative analysis instruction, dividing the label into relative areas, wherein the relative areas take the size of an interval label value range as a judgment standard, and determining the behavior main body when the user superior record meets the relative areas, and matching the user superior record with the lower records of the same behavior main body in a sample library;
before the step of generating the model according to the potential requirements, the method further comprises the steps of:
s201, receiving a plurality of user histories, wherein the user histories respectively correspond to a plurality of users;
s202, receiving a tag instruction, screening according to the plurality of user histories, and deriving type parameters;
s203, receiving an instruction for calculating intermediate data;
the step of receiving the instruction for calculating the intermediate data comprises the following steps:
receiving a label weight setting instruction;
according to the respective duty ratio of the tag in the user history record, data coupling is carried out to obtain type parameters in the user history record;
determining a behavior main body corresponding to the user history record according to the type parameter in the user history record:
acquiring histories belonging to the same behavior main body from the plurality of user histories as reference information, and determining the lower-level records for optimizing the corresponding behavior main body according to the type parameters;
and determining the lower record forming sample library for optimizing the corresponding behavior main body according to the type parameter.
2. The data recommendation method of claim 1, wherein after generating said model based on said potential demand, said method further comprises: s205, receiving analysis instructions of users one by one, and outputting a potential demand prediction result of the behavior main body for a service scene of data recommendation.
3. The data recommendation method of claim 1, wherein receiving a tag instruction, filtering according to the plurality of user histories, and deriving a type parameter comprises:
s2021, setting a label according to preset information;
s2022, performing label setting on the plurality of user histories according to labels;
s2023, generating the type parameter after screening the plurality of user histories;
s2024, storing the type parameter;
s2025, reading the type parameter.
4. The data recommendation method according to claim 1, wherein after receiving the tag instruction, filtering according to the plurality of user histories, and deriving the type parameter, further comprising: and receiving a member circulation instruction, and grouping the users, wherein the users are classified into a virginee purchasing user, an active user and a sleeping user.
5. A data recommendation method according to claim 1, wherein determining the history for optimizing the corresponding behavior body based on the type parameter comprises:
removing the unreliable data from the reference information according to a preset data screening algorithm;
and analyzing and obtaining the lower record in the behavior main body according to the trusted data except the untrusted data in the reference information.
6. The data recommendation method of claim 5, receiving a set tag weight instruction, comprising: and setting a weight value for the tag according to preset information adjustment conditions.
7. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the data recommendation method according to any of claims 1 to 6 when executing the computer program.
8. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program for executing the data recommendation method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911172534.1A CN111127074B (en) | 2019-11-26 | 2019-11-26 | Data recommendation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911172534.1A CN111127074B (en) | 2019-11-26 | 2019-11-26 | Data recommendation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111127074A CN111127074A (en) | 2020-05-08 |
| CN111127074B true CN111127074B (en) | 2023-04-25 |
Family
ID=70496691
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911172534.1A Active CN111127074B (en) | 2019-11-26 | 2019-11-26 | Data recommendation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111127074B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112227007B (en) * | 2020-09-18 | 2023-08-25 | 张国政 | Control method for purchasing clothes treating agent for washing machine |
| CN112435111B (en) * | 2020-12-30 | 2023-12-19 | 广东金赋科技股份有限公司 | Tax hall guiding and handling method and device based on hall and user data |
| CN113378122A (en) * | 2021-08-02 | 2021-09-10 | 成都联帮医疗科技股份有限公司 | Data processing system and method based on SaaS service cloud platform |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016203652A1 (en) * | 2015-06-19 | 2016-12-22 | 株式会社Ｕｂｉｃ | System related to data analysis, control method, control program, and recording medium therefor |
| CN107203899A (en) * | 2017-05-03 | 2017-09-26 | 北京小米移动软件有限公司 | Information-pushing method and equipment |
| CN107844548A (en) * | 2017-10-30 | 2018-03-27 | 北京锐安科技有限公司 | A kind of data label method and apparatus |
| WO2019153518A1 (en) * | 2018-02-08 | 2019-08-15 | 平安科技（深圳）有限公司 | Information pushing method and device, computer device and storage medium |
| CN110222272A (en) * | 2019-04-18 | 2019-09-10 | 广东工业大学 | A kind of potential customers excavate and recommended method |
| WO2019174395A1 (en) * | 2018-03-13 | 2019-09-19 | 阿里巴巴集团控股有限公司 | Method and apparatus for information recommendation, and device |
| WO2019192261A1 (en) * | 2018-04-02 | 2019-10-10 | 阿里巴巴集团控股有限公司 | Payment mode recommendation method and device and equipment |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107368488A (en) * | 2016-05-12 | 2017-11-21 | 阿里巴巴集团控股有限公司 | A kind of method for determining user behavior preference, the methods of exhibiting and device of recommendation information |
| US10846617B2 (en) * | 2017-05-12 | 2020-11-24 | Adobe Inc. | Context-aware recommendation system for analysts |
-
2019
- 2019-11-26 CN CN201911172534.1A patent/CN111127074B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016203652A1 (en) * | 2015-06-19 | 2016-12-22 | 株式会社Ｕｂｉｃ | System related to data analysis, control method, control program, and recording medium therefor |
| CN107203899A (en) * | 2017-05-03 | 2017-09-26 | 北京小米移动软件有限公司 | Information-pushing method and equipment |
| CN107844548A (en) * | 2017-10-30 | 2018-03-27 | 北京锐安科技有限公司 | A kind of data label method and apparatus |
| WO2019153518A1 (en) * | 2018-02-08 | 2019-08-15 | 平安科技（深圳）有限公司 | Information pushing method and device, computer device and storage medium |
| WO2019174395A1 (en) * | 2018-03-13 | 2019-09-19 | 阿里巴巴集团控股有限公司 | Method and apparatus for information recommendation, and device |
| WO2019192261A1 (en) * | 2018-04-02 | 2019-10-10 | 阿里巴巴集团控股有限公司 | Payment mode recommendation method and device and equipment |
| CN110222272A (en) * | 2019-04-18 | 2019-09-10 | 广东工业大学 | A kind of potential customers excavate and recommended method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111127074A (en) | 2020-05-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Martínez et al. | A machine learning framework for customer purchase prediction in the non-contractual setting | |
| Dogan et al. | Customer segmentation by using RFM model and clustering methods: a case study in retail industry | |
| US11769194B2 (en) | Method and system for presenting items in online environment based on previous item selections | |
| Beladev et al. | Recommender systems for product bundling | |
| CN102629360B (en) | A kind of effective dynamic commodity recommend method and commercial product recommending system | |
| CN111127074B (en) | Data recommendation method | |
| CN104463637A (en) | Commodity recommendation method and device based on electronic business platform and server | |
| CN104239338A (en) | Information recommendation method and information recommendation device | |
| US11854022B2 (en) | Proactively predicting transaction dates based on sparse transaction data | |
| KR20140056731A (en) | Purchase recommendation service system and method | |
| CN111460316B (en) | Knowledge system-oriented personalized recommendation method and computer storage medium | |
| Hemalatha | Market basket analysis–a data mining application in Indian retailing | |
| EP3400571A1 (en) | Consumer decision tree generation system | |
| KR101639656B1 (en) | Method and server apparatus for advertising | |
| US20140372178A1 (en) | Correlating product sales to store segmentation | |
| JP2016212792A (en) | Commodity recommendation device, commodity recommendation method and program | |
| JP4386973B2 (en) | Hierarchical prediction model construction apparatus and method | |
| US10318984B1 (en) | Predictive recommendation system using tiered feature data | |
| GÜR ALI | Driver moderator method for retail sales prediction | |
| CN111708945A (en) | Product recommendation method and device, electronic equipment and computer storage medium | |
| Bernat et al. | Modelling customer lifetime value in a continuous, non-contractual time setting | |
| CN110020918A (en) | A kind of recommendation information generation method and system | |
| CN114817741A (en) | Financial product accurate recommendation method and device | |
| US20210216907A1 (en) | Selecting optimum primary and secondary parameters to calibrate and generate an unbiased forecasting model | |
| JP2014232346A (en) | Information recommendation device, information recommendation method, and information recommendation program |
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 |