Commodity pushing method and system based on consumer address

Abstract

The invention discloses a commodity pushing method and system based on a consumer address, and particularly relates to the technical field of positioning and message pushing, and the system comprises a data collection module which is used for monitoring and collecting consumer personal information and mobile phone historical record information data; according to the invention, based on the consumer outdoor information, the environmental noise information and the network connection information generated in the daily activities of the consumer, the demand degree of the consumer for commodity pushing is evaluated, the pushing frequency is preliminarily adjusted according to the evaluation result, and then the machine learning algorithm model is constructed to optimize the pushing process, so that the recommendation accuracy is improved, and the user experience is improved. After a system sends commodity push to consumers according to an algorithm, a post-push emotion evaluation model is constructed through emotion abnormal information, software retention information and commodity evaluation information of consumers who do not purchase commodities after information acquisition, and the emotion tendency of the consumers to user software is judged; and the commodity pushing frequency of the consumer is adjusted for the second time according to the emotion tendency of the user.

Description

Technical Field

[0001] The present invention relates to the technical field of positioning and message push, and more specifically, to a commodity push method and system based on consumer address. Background Art

[0002] With the rapid development of e-commerce platforms, online shopping has become an important way for people to obtain goods. In order to improve consumer experience and platform profitability, how to push goods based on consumers' geographic locations has become an important research direction.

[0003] Modern consumers' demand for goods is personalized and regionalized, and factors such as living environment and regional differences affect consumption habits. In big cities, consumers may pay more attention to the quality and brand of goods, while in rural areas, price and practicality may be the main considerations. Therefore, accurate push based on consumer addresses can increase purchase conversion rates. In addition, the consumer's geographical location directly determines the efficiency of logistics distribution. Pushing goods based on consumer addresses can help the platform optimize based on factors such as logistics resources and inventory distribution, reducing delivery time and costs.

[0004] However, for individual consumers, frequent and homogeneous push content can easily cause negative emotions. Product push that is too close to the address may sometimes cause some people to be suspicious and wary, which in turn has an adverse effect on product sales. Therefore, a more advanced product push method and system based on consumer addresses is needed. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a product push method and system based on consumer address, which evaluates the consumer's demand for product push based on the consumer's outdoor information, environmental noise information and network connection information generated in the consumer's daily activities, and preliminarily adjusts the push frequency according to the evaluation results, and then constructs a machine learning algorithm model to optimize the push process to improve the recommendation accuracy. After the system sends product push to the consumer according to the algorithm, it constructs a post-push emotion evaluation model by obtaining the abnormal emotion information, software retention information and product evaluation information of the consumer who did not purchase the product after obtaining the information, and judges the consumer's emotional tendency towards the user's software, and makes a second adjustment to the consumer's product push frequency based on the user's emotional tendency.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method and system for pushing goods based on consumer addresses, comprising the following steps:

[0008] Step S1, monitoring and collecting consumer personal information and mobile phone history record information data, including consumer outdoor information, environmental noise information and network connection information, obtaining abnormal emotion information, software retention information and product evaluation information of consumers who did not purchase the product after receiving the information;

[0009] Step S2, constructing an address push demand evaluation model, obtaining an address push demand evaluation index, and evaluating consumers' demand for product push;

[0010] Step S3, for the consumer push information after the push frequency is initially adjusted, a machine learning algorithm model is constructed to optimize the push process;

[0011] Step S4, construct a post-push emotion evaluation model, judge the consumer's emotional tendency towards the user's software through the abnormal emotion information, software retention information and product evaluation information in the model, and make a second adjustment to the consumer's product push frequency based on the user's emotional tendency.

[0012] Specifically, in step S1, by reducing the frequency of product push and improving the accuracy and practicality of product push, consumers' trust in the user platform is improved. For consumer groups with a large base, the system can build a big data processing framework-Apache Spark to process data. For special consumers obtained after screening, only Python combined with Scikit-learn or TensorFlow can achieve machine learning optimization goals. In actual production, due to factors such as equipment cost and the necessity of real-time monitoring, the present invention is based on the premise of continuously monitoring consumer activity information when the user product push system works for a long time and the operating environment is relatively stable, which may lead to abnormal privacy data protection at a certain monitoring moment.

[0013] In step S2, the consumer's demand for product push is determined by evaluating the collected consumer personal information and mobile phone history records, so as to preliminarily decide the frequency of pushing products to consumers. The specific process is to collect the consumer's outdoor information, environmental noise information and network connection information, where the consumer's outdoor information includes the consumer's outdoor coefficient, which is calibrated as , the environmental noise information includes the environmental noise coefficient, calibrated as , the network connection information includes the network connection coefficient, calibrated as ;

[0014] Consumer outdoor coefficient The camera is used to identify the light and color environment around the consumer, and the light intensity change value collected by the camera within half an hour is calibrated as , obtain the real-time external light intensity in the consumer area, calibrated as , and calibrate the light intensity collected by the camera as , then the consumer outdoor coefficient ;

[0015] Environmental noise coefficient By collecting microphone environmental noise, the type of the collected microphone environmental noise is calibrated as , obtain the proportion of wind noise in the microphone ambient noise, calibrated as , then the environmental noise coefficient ;

[0016] Network connection coefficient By collecting the number of Wi-Fi signals and Bluetooth devices detected by consumer devices, it is calibrated as , obtain the number and range of Wi-Fi signal strength fluctuations within half an hour, and mark them as and , then the network connection coefficient ;

[0017] The address push demand evaluation model is constructed by weighting the consumer's outdoor information, environmental noise information, and network connection information to generate the address push demand evaluation index. , the corresponding coefficients are consumer outdoor coefficient , Environmental noise coefficient and network connectivity coefficient , the formula is ;

[0018] at the same time, It is set according to the actual situation. For example, the expert empowerment method is adopted, that is, experts in related fields are invited to determine the weights of various indicators through professional opinion surveys and comprehensive evaluations to ensure that the weight coefficients can accurately reflect the importance of various indicators in the address push demand assessment. In addition, the use of multiple methods such as hierarchical analysis method and fuzzy comprehensive evaluation method to determine the weight coefficients can ensure the objectivity and scientificity of the weight coefficients. I will not elaborate on this here.

[0019] In step S2, the address push demand evaluation index obtained in the address push demand evaluation model is used to reflect the degree of demand for product push at the consumer's address. The larger its value is, the more likely it is that the consumer is outdoors, the higher the environmental noise is, the more unstable the network connection is, the less time the consumer spends on mobile shopping, and the smaller the demand for product push.

[0020] In step S2, when the address push demand evaluation index When it is greater than the set high-frequency threshold, it means that the target consumers have little demand for product push, so the product push frequency should be reduced and the subsequent steps should be taken;

[0021] When the address push demand evaluation index When it is less than or equal to the set high-frequency threshold, it means that the target consumers have a high demand for product push, so the product push frequency should be increased and subsequent steps should be taken.

[0022] In step S3, a machine learning algorithm model is constructed to optimize the product push method based on the consumer address by using a collaborative filtering method based on geographic information. The main purpose is to use geographic location information to improve the accuracy of recommendations and make the recommended products more relevant to the consumer's location.

[0023] Furthermore, the construction method of the machine learning algorithm model of the present invention is as follows:

[0024] Step S3.1, data collection and preprocessing, collects consumer location information, product location information, consumer behavior data, converts the geographic location information of consumers and stores into coordinate form, and uses the longitude and latitude information to calculate the geographic distance between consumers and stores, providing similarity calculation for recommendation;

[0025] Step S3.2, extract geographic information features, recommend nearby stores or products based on distance. Use the geographic distance between consumers and stores as an additional feature of collaborative filtering, and calculate the distance between the two places using the spherical distance formula, which is: , where d is the distance between two places, R is the radius of the earth, are the latitudes of the two points, is the longitude;

[0026] Step S3.3, collaborative filtering modeling, using collaborative filtering based on consumer-store or consumer-product, combined with geographic similarity for optimization, using the geographic location and purchase record similarity between consumers to recommend products. For example, if consumers i and j have similar geographic locations and have purchased similar products, it can be inferred that consumer i likes consumer j's products, and similarity is calculated based on this: ,in, is the geographical similarity, is the behavioral similarity between consumers, is a hyperparameter to control geographic weight;

[0027] Step S3.4, based on the similarity between the consumer and the product, predict the consumer's interest score for the product and make a recommendation. Taking the predicted score of consumer u for product s as an example, the calculation formula is: , Score the predictions, is the neighbor set of consumer u, is the similarity between consumer u and its neighbor consumer j, is the rating of the neighbor consumer j on the product s. Based on the consumer's rating prediction results, the product with the highest rating is recommended to each consumer.

[0028] In step S4, a post-push emotion evaluation model is constructed. The specific process is that after the system sends a product push to the consumer according to the algorithm, the abnormal emotion information, software retention information and product evaluation information of the consumer who did not purchase the product after obtaining the information are obtained, where the abnormal emotion information includes the touch screen force coefficient, which is calibrated as , the software retention information includes the software uninstallation coefficient, which is calibrated as , the product evaluation information includes the malicious evaluation coefficient, which is calibrated as ;

[0029] Touch screen force coefficient By collecting the changes in the strength of the consumer's finger touching the screen after receiving the product push, the pressure caused by the consumer's finger touching the screen 1000 times before receiving the product push is obtained and calibrated as , x is the number of times the finger touches the screen, x=1, 2, 3, ..., x is a positive integer, and the pressure of the finger touching the screen after the consumer receives the product push is counted and calibrated as , y is the number of times the finger touches the screen up to the time of statistics, y=0, 1, 2, 3, ..., y<x and is a positive integer, then the touch screen force coefficient ;

[0030] Software uninstall coefficient The number of consumers who did not purchase the product within one day after receiving the product push is collected and marked as , count the number of consumers who uninstalled the user software after receiving the product push but did not purchase the product within one day, and mark it as , then the software uninstall coefficient ;

[0031] Malicious evaluation coefficient By collecting the number of malicious reviews posted by consumers on various platforms in the past fifteen days, we calibrate it as , and at the same time obtain the number of malicious reviews posted by consumers on the user software platform in the past fifteen days, marked as , then the malicious evaluation coefficient ;

[0032] The post-push emotion evaluation model is constructed by weighting the abnormal emotion information of consumers who did not purchase the product after obtaining the information, the software retention information, and the product evaluation information to generate the post-push emotion evaluation index. , and the corresponding coefficients are touch screen force coefficient , Software Uninstallation Coefficient and malicious evaluation coefficient , the formula is ;

[0033] In step S4, the post-push emotion evaluation index obtained in the post-push emotion evaluation model is used to reflect the emotional tendency of consumers who did not purchase the product after obtaining the information towards the user software. The larger the post-push emotion evaluation index, the greater the screen touching action of the consumer after receiving the push and the more intense the emotion, the higher the probability of deleting the user software, the more serious the negative emotion, and the worse the emotional tendency towards the user software.

[0034] In step S4, when the sentiment evaluation index is pushed When it is greater than the set negative threshold, it indicates that the consumer has a negative emotional tendency towards the user's software, and the number of subsequent push notifications is reduced, and the consumer is included in the special consumer ranks. Only for the special consumer samples, the system process of steps S2-S4 is repeated in the ranks, and for consumers whose second emotional evaluation index is still less than the negative threshold, the number of push notifications is greatly reduced;

[0035] Emotional evaluation index after push When it is less than or equal to the set bad threshold, it means that the consumer's emotional tendency towards the user software is in a good attitude. Continue to collect consumer information, and increase or keep the number of product pushes unchanged based on the consumer's software browsing situation.

[0036] A commodity push system based on consumer address, used in the above-mentioned commodity push method based on consumer address, comprising:

[0037] The data collection module monitors and collects consumer personal information and mobile phone history information, including consumer outdoor information, environmental noise information, and network connection information, and obtains abnormal emotional information, software retention information, and product evaluation information of consumers who did not purchase the product after receiving the information;

[0038] The address push demand evaluation module builds an address push demand evaluation model, obtains the address push demand evaluation index, and evaluates consumers' demand for product push;

[0039] The push optimization module pushes information to consumers after the initial adjustment of the push frequency, and builds a machine learning algorithm model to optimize the push process;

[0040] The post-push emotion evaluation module builds a post-push emotion evaluation model. Through the abnormal emotion information, software retention information and product evaluation information in the model, it judges the consumer's emotional tendency towards the user software, and makes a second adjustment to the consumer's product push frequency based on the user's emotional tendency.

[0041] Technical effects and advantages of the present invention:

[0042] The present invention evaluates the degree of consumer demand for product push based on the consumer's outdoor information, environmental noise information and network connection information generated in the consumer's daily activities, and preliminarily adjusts the push frequency according to the evaluation results, and then constructs a machine learning algorithm model to optimize the push process and improve the recommendation accuracy. After the system sends product push to consumers according to the algorithm, it constructs a post-push emotion evaluation model by obtaining the abnormal emotion information, software retention information and product evaluation information of consumers who did not purchase the product after obtaining the information, and judges the consumer's emotional tendency towards user software, and makes a second adjustment to the consumer's product push frequency based on the user's emotional tendency. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 The present invention is a flow chart of a method for pushing goods based on consumer addresses.

[0044] Figure 2 This is a flow chart of a commodity push system based on consumer address of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] The present invention is a method and system for pushing goods based on consumer addresses. It is based on the consumer's outdoor information, environmental noise information and network connection information generated in the consumer's daily activities, evaluates the consumer's demand for product push, and preliminarily adjusts the push frequency according to the evaluation result, and then builds a machine learning algorithm model to optimize the push process and improve the recommendation accuracy. After the system sends the product push to the consumer according to the algorithm, it builds a post-push emotion evaluation model by obtaining the abnormal emotion information, software retention information and product evaluation information of the consumer who did not purchase the product after obtaining the information, and judges the consumer's emotional tendency towards the user's software, and makes a second adjustment to the consumer's product push frequency based on the user's emotional tendency.

[0047] Example 1

[0048] like Figure 1 As shown, a method and system for pushing goods based on consumer addresses are as follows:

[0049] A method and system for pushing goods based on consumer addresses, comprising the following steps:

[0050] Step S1, monitoring and collecting consumer personal information and mobile phone history record information data, including consumer outdoor information, environmental noise information and network connection information, obtaining abnormal emotion information, software retention information and product evaluation information of consumers who did not purchase the product after receiving the information;

[0051] Step S2, constructing an address push demand evaluation model, obtaining an address push demand evaluation index, and evaluating consumers' demand for product push;

[0052] Step S3, for the consumer push information after the push frequency is initially adjusted, a machine learning algorithm model is constructed to optimize the push process;

[0053] Step S4, construct a post-push emotion evaluation model, judge the consumer's emotional tendency towards the user's software through the abnormal emotion information, software retention information and product evaluation information in the model, and make a second adjustment to the consumer's product push frequency based on the user's emotional tendency.

[0054] Specifically, in step S1, by reducing the frequency of product push and improving the accuracy and practicality of product push, consumers' trust in the user platform is improved. For consumer groups with a large base, the system can build a big data processing framework-Apache Spark to process data. For special consumers obtained after screening, only Python combined with Scikit-learn or TensorFlow can achieve machine learning optimization goals. In actual production, due to factors such as equipment cost and the necessity of real-time monitoring, the present invention is based on the premise of continuously monitoring consumer activity information when the user product push system works for a long time and the operating environment is relatively stable, which may lead to abnormal privacy data protection at a certain monitoring moment.

[0055] In step S2, the consumer's demand for product push is determined by evaluating the collected consumer personal information and mobile phone history records, so as to preliminarily decide the frequency of pushing products to consumers. The specific process is to collect the consumer's outdoor information, environmental noise information and network connection information, where the consumer's outdoor information includes the consumer's outdoor coefficient, which is calibrated as , the environmental noise information includes the environmental noise coefficient, calibrated as , the network connection information includes the network connection coefficient, calibrated as ;

[0056] Consumer outdoor coefficient The camera is used to identify the light and color environment around the consumer, and the light intensity change value collected by the camera within half an hour is calibrated as , obtain the real-time light intensity outside the consumer area, calibrated as , and calibrate the light intensity collected by the camera as , then the consumer outdoor coefficient ;

[0057] Environmental noise coefficient By collecting microphone environmental noise, the type of the collected microphone environmental noise is calibrated as , obtain the proportion of wind noise in the microphone ambient noise, calibrated as , then the environmental noise coefficient ;

[0058] Network connection coefficient By collecting the number of Wi-Fi signals and Bluetooth devices detected by consumer devices, it is calibrated as , obtain the number and range of Wi-Fi signal strength fluctuations within half an hour, and mark them as and , then the network connection coefficient ;

[0059] The address push demand evaluation model is constructed by weighting the consumer's outdoor information, environmental noise information, and network connection information to generate the address push demand evaluation index. , the corresponding coefficients are consumer outdoor coefficient , Environmental noise coefficient and network connectivity coefficient , the formula is ;

[0060] at the same time, It is set according to the actual situation. For example, the expert empowerment method is adopted, that is, experts in related fields are invited to determine the weights of various indicators through professional opinion surveys and comprehensive evaluations to ensure that the weight coefficients can accurately reflect the importance of various indicators in the address push demand assessment. In addition, the use of multiple methods such as hierarchical analysis method and fuzzy comprehensive evaluation method to determine the weight coefficients can ensure the objectivity and scientificity of the weight coefficients. I will not elaborate on this here.

[0061] In step S2, the address push demand evaluation index obtained in the address push demand evaluation model is used to reflect the degree of demand for product push at the consumer's address. The larger its value is, the more likely it is that the consumer is outdoors, the higher the environmental noise is, the more unstable the network connection is, the less time the consumer spends on mobile shopping, and the smaller the demand for product push.

[0062] In step S2, when the address push demand evaluation index When it is greater than the set high-frequency threshold, it means that the target consumers have little demand for product push, so the product push frequency should be reduced and the subsequent steps should be taken;

[0063] When the address push demand evaluation index When it is less than or equal to the set high-frequency threshold, it means that the target consumers have a high demand for product push, so the product push frequency should be increased and subsequent steps should be taken.

[0064] In step S3, a machine learning algorithm model is constructed to optimize the product push method based on the consumer address by using a collaborative filtering method based on geographic information. The main purpose is to use geographic location information to improve the accuracy of recommendations and make the recommended products more relevant to the consumer's location.

[0065] Furthermore, the construction method of the machine learning algorithm model of the present invention is as follows:

[0066] Step S3.1, data collection and preprocessing, collects consumer location information, product location information, consumer behavior data, converts the geographic location information of consumers and stores into coordinate form, and uses the longitude and latitude information to calculate the geographic distance between consumers and stores, providing similarity calculation for recommendation;

[0067] Step S3.2, extract geographic information features, recommend nearby stores or products based on distance. Use the geographic distance between consumers and stores as an additional feature of collaborative filtering, and calculate the distance between the two places using the spherical distance formula, which is: , where d is the distance between two places, R is the radius of the earth, are the latitudes of the two points, is the longitude;

[0068] Step S3.3, collaborative filtering modeling, using collaborative filtering based on consumer-store or consumer-product, combined with geographic similarity for optimization, using the geographic location and purchase record similarity between consumers to recommend products. For example, if consumers i and j have similar geographic locations and have purchased similar products, it can be inferred that consumer i likes consumer j's products, and similarity is calculated based on this: ,in, is the geographical similarity, is the behavioral similarity between consumers, is a hyperparameter to control geographic weight;

[0069] Step S3.4, based on the similarity between the consumer and the product, predict the consumer's interest score for the product and make a recommendation. Taking the predicted score of consumer u for product s as an example, the calculation formula is: , Score the predictions, is the neighbor set of consumer u, is the similarity between consumer u and its neighbor consumer j, is the rating of the neighbor consumer j on the product s. Based on the consumer's rating prediction results, the product with the highest rating is recommended to each consumer.

[0070] In step S4, a post-push emotion evaluation model is constructed. The specific process is that after the system sends a product push to the consumer according to the algorithm, the abnormal emotion information, software retention information and product evaluation information of the consumer who did not purchase the product after obtaining the information are obtained, where the abnormal emotion information includes the touch screen force coefficient, which is calibrated as , the software retention information includes the software uninstallation coefficient, which is calibrated as , the product evaluation information includes the malicious evaluation coefficient, which is calibrated as ;

[0071] Touch screen force coefficient By collecting the changes in the strength of the consumer's finger touching the screen after receiving the product push, the pressure caused by the consumer's finger touching the screen 1000 times before receiving the product push is obtained and calibrated as , x is the number of times the finger touches the screen, x=1, 2, 3, ..., x is a positive integer, and the pressure of the finger touching the screen after the consumer receives the product push is counted and calibrated as , y is the number of times the finger touches the screen up to the time of statistics, y=0, 1, 2, 3, ..., y<x and is a positive integer, then the touch screen force coefficient ;

[0072] Software uninstall coefficient The number of consumers who did not purchase the product within one day after receiving the product push is collected and marked as , count the number of consumers who uninstalled the user software after receiving the product push but did not purchase the product within one day, and mark it as , then the software uninstall coefficient ;

[0073] Malicious evaluation coefficient By collecting the number of malicious reviews posted by consumers on various platforms in the past fifteen days, we calibrate it as , and at the same time obtain the number of malicious reviews posted by consumers on the user software platform in the past fifteen days, marked as , then the malicious evaluation coefficient ;

[0074] The post-push emotion evaluation model is constructed by weighting the abnormal emotion information of consumers who did not purchase the product after obtaining the information, the software retention information, and the product evaluation information to generate the post-push emotion evaluation index. , and the corresponding coefficients are touch screen force coefficient , Software Uninstallation Coefficient and malicious evaluation coefficient , the formula is ;

[0075] In step S4, the post-push emotion evaluation index obtained in the post-push emotion evaluation model is used to reflect the emotional tendency of consumers who did not purchase the product after obtaining the information towards the user software. The larger the post-push emotion evaluation index, the greater the screen touching action of the consumer after receiving the push and the more intense the emotion, the higher the probability of deleting the user software, the more serious the negative emotion, and the worse the emotional tendency towards the user software.

[0076] In step S4, when the sentiment evaluation index is pushed When it is greater than the set negative threshold, it indicates that the consumer has a negative emotional tendency towards the user's software, and the number of subsequent push notifications is reduced, and the consumer is included in the special consumer ranks. Only for the special consumer samples, the system process of steps S2-S4 is repeated in the ranks, and for consumers whose second emotional evaluation index is still less than the negative threshold, the number of push notifications is greatly reduced;

[0077] Emotional evaluation index after push When it is less than or equal to the set bad threshold, it means that the consumer's emotional tendency towards the user software is in a good attitude. Continue to collect consumer information, and increase or keep the number of product pushes unchanged based on the consumer's software browsing situation.

[0078] Embodiment 2: This embodiment is a system embodiment of embodiment 1, and is used to implement a method for pushing goods based on a consumer address introduced in embodiment 1, and specifically includes:

[0079] The data collection module monitors and collects consumer personal information and mobile phone history information, including consumer outdoor information, environmental noise information, and network connection information, and obtains abnormal emotional information, software retention information, and product evaluation information of consumers who did not purchase the product after receiving the information;

[0080] The address push demand evaluation module builds an address push demand evaluation model, obtains the address push demand evaluation index, and evaluates consumers' demand for product push;

[0081] The push optimization module pushes information to consumers after the initial adjustment of the push frequency, and builds a machine learning algorithm model to optimize the push process;

[0082] The post-push emotion evaluation module builds a post-push emotion evaluation model. Through the abnormal emotion information, software retention information and product evaluation information in the model, it judges the consumer's emotional tendency towards the user software, and makes a second adjustment to the consumer's product push frequency based on the user's emotional tendency.

[0083] The above formulas are all dimensionless and calculated numerically. Specific dimension removal can be achieved by various means such as standardization, which will not be elaborated here. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0084] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, an ATA hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state ATA hard disk.

[0085] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0086] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0087] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0088] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0089] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0090] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile ATA hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0091] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for pushing goods based on consumer addresses, characterized in that: The steps include: Step S1, monitoring and collecting consumers' personal address information and mobile phone history record information data, including consumers' outdoor information, environmental noise information and network connection information, obtaining consumers' abnormal emotions information, software retention information and product evaluation information after receiving information but not purchasing products; Step S2, constructing an address push demand evaluation model, obtaining an address push demand evaluation index through the consumer's outdoor information, environmental noise information, and network connection information in the model, and evaluating the consumer's demand for product push based on the consumer's personal address; Step S3, for the consumer push information after the push frequency is initially adjusted, a machine learning algorithm model is constructed to extract the consumer's geographic location information, improve the accuracy of the recommendation, make the recommended products more relevant to the consumer's location, and optimize the push process; Step S4, constructing a post-push emotion evaluation model to determine the consumer's emotional inclination towards the user's software, and making a second adjustment to the consumer's product push frequency based on the user's emotional inclination.

2. A method for pushing goods based on consumer addresses according to claim 1, characterized in that: In step S2, the collected consumer personal information and mobile phone history records are evaluated to determine the consumer's demand for product push, thereby preliminarily determining the frequency of pushing products to consumers. The specific process is to collect consumer outdoor information, environmental noise information, and network connection information, where the consumer outdoor information includes the consumer outdoor coefficient, which is calibrated as , the environmental noise information includes the environmental noise coefficient, calibrated as , the network connection information includes the network connection coefficient, calibrated as ; Consumer outdoor coefficient The camera is used to identify the light and color environment around the consumer, and the light intensity change value collected by the camera within half an hour is calibrated as , obtain the real-time external light intensity in the consumer area, calibrated as , and calibrate the light intensity collected by the camera as , then the consumer outdoor coefficient ; Environmental noise factor By collecting microphone environmental noise, the type of the collected microphone environmental noise is calibrated as , obtain the proportion of wind noise in the microphone ambient noise, calibrated as , then the environmental noise coefficient ; Network connection coefficient By collecting the number of Wi-Fi signals and Bluetooth devices detected by consumer devices, it is calibrated as , obtain the number and range of Wi-Fi signal strength fluctuations within half an hour, and mark them as and , then the network connection coefficient .

3. A method for pushing goods based on consumer addresses according to claim 2, characterized in that: The address push demand evaluation model is constructed by weighting the consumer's outdoor information, environmental noise information, and network connection information to generate the address push demand evaluation index. , the corresponding coefficients are consumer outdoor coefficient , Environmental noise coefficient and network connectivity coefficient , the formula is , where is the weight coefficient of the corresponding indicator.

4. A method for pushing goods based on consumer addresses according to claim 3, characterized in that: In step S2, when the address push demand evaluation index When the frequency is greater than the set high-frequency threshold, the product push frequency is reduced and the subsequent steps are performed; When the address push demand evaluation index When the frequency is less than or equal to the set high-frequency threshold, increase the product push frequency and proceed to the next step.

5. A method for pushing goods based on consumer addresses according to claim 1, characterized in that: In step S3, a machine learning algorithm model is constructed to optimize the product push method based on the consumer address by using a collaborative filtering method based on geographic information. The main purpose is to use geographic location information to improve the accuracy of recommendations and make the recommended products more relevant to the consumer's location; Furthermore, the method for constructing the machine learning algorithm model is as follows: Step S3.1, data collection and preprocessing, collects consumer location information, product location information, consumer behavior data, converts the geographic location information of consumers and stores into coordinate form, and uses the longitude and latitude information to calculate the geographic distance between consumers and stores, providing similarity calculation for recommendation; Step S3.2, extracting geographic information features, recommending stores or products that are closer based on distance, and using the geographic distance between consumers and stores as an additional feature of collaborative filtering; Step S3.3, collaborative filtering modeling, using collaborative filtering based on consumer-store or consumer-product, combined with geographic similarity for optimization, using the geographic location and purchase record similarity between consumers to recommend products; Step S3.4, based on the similarity between the consumer and the product, predict the consumer's interest score for the product and make a recommendation. Based on the consumer's score prediction result, recommend the highest-scoring product to each consumer.

6. A method for pushing goods based on consumer addresses according to claim 5, characterized in that: In step S4, a post-push emotion evaluation model is constructed. The specific process is that after the system sends a product push to the consumer according to the algorithm, the abnormal emotion information, software retention information and product evaluation information of the consumer who did not purchase the product after obtaining the information are obtained, where the abnormal emotion information includes the touch screen force coefficient, which is calibrated as , the software retention information includes the software uninstallation coefficient, which is calibrated as , the product evaluation information includes the malicious evaluation coefficient, which is calibrated as ; Touch screen force coefficient By collecting the changes in the strength of the consumer's finger touching the screen after receiving the product push, the pressure caused by the consumer's finger touching the screen 1000 times before receiving the product push is obtained and calibrated as , x is the number of times the finger touches the screen, x=1, 2, 3, ..., x is a positive integer, and the pressure of the finger touching the screen after the consumer receives the product push is counted and calibrated as , y is the number of times the finger touches the screen up to the time of statistics, y=0, 1, 2, 3, ..., y<x and is a positive integer, then the touch screen force coefficient ; Software uninstall coefficient The number of consumers who did not purchase the product within one day after receiving the product push is collected and marked as , count the number of consumers who uninstalled the user software after receiving the product push but did not purchase the product within one day, and mark it as , then the software uninstall coefficient ; Malicious evaluation coefficient By collecting the number of malicious reviews posted by consumers on various platforms in the past fifteen days, we calibrate it as , and at the same time obtain the number of malicious reviews posted by consumers on the user software platform in the past fifteen days, marked as , then the malicious evaluation coefficient .

7. A method for pushing goods based on consumer addresses according to claim 6, characterized in that: The post-push emotion evaluation model is constructed by weighting the abnormal emotion information of consumers who did not purchase the product after obtaining the information, the software retention information, and the product evaluation information to generate the post-push emotion evaluation index. , and the corresponding coefficients are touch screen force coefficient , Software Uninstallation Coefficient and malicious evaluation coefficient , the formula is .

8. A method for pushing goods based on consumer addresses according to claim 7, characterized in that: In step S4, when the sentiment evaluation index is pushed When the sentiment evaluation index is greater than the set bad threshold, the number of subsequent push notifications is reduced, and the consumer is included in the special consumer category. Only for the special consumer sample, the system process of steps S2-S4 is repeated within the category. For consumers whose sentiment evaluation index is still less than the bad threshold for the second time, the number of push notifications is greatly reduced. Emotional evaluation index after push When it is less than or equal to the set bad threshold, the consumer information will continue to be collected, and the number of product pushes will increase or remain unchanged based on the consumer's software browsing situation.

9. A commodity push system based on consumer address, used to implement a commodity push method based on consumer address as claimed in any one of claims 1 to 8, characterized in that: include: The data collection module monitors and collects consumer personal information and mobile phone history information, including consumer outdoor information, environmental noise information, and network connection information, and obtains abnormal emotional information, software retention information, and product evaluation information of consumers who did not purchase the product after receiving the information; The address push demand evaluation module builds an address push demand evaluation model, obtains the address push demand evaluation index, and evaluates consumers' demand for product push; The push optimization module pushes information to consumers after the initial adjustment of the push frequency, and builds a machine learning algorithm model to optimize the push process; The post-push emotion evaluation module builds a post-push emotion evaluation model. Through the abnormal emotion information, software retention information and product evaluation information in the model, it judges the consumer's emotional tendency towards the user software, and makes a second adjustment to the consumer's product push frequency based on the user's emotional tendency.

Images