CN110930187A - Store visitor crowd mining method, device, equipment and medium - Google Patents

Abstract

The present application discloses a method, device, device and medium for mining shop visitors, which relate to user visit prediction technology. The specific implementation scheme is as follows: according to the historical location information of the user and the location information of the store, determine the visitor crowd of the store in a historical specific time period; Set the online information and GPS check information within the time, and determine the convenience of visiting the store corresponding to each GPS check information; use the online information, GPS check information and its corresponding store of each user in the visiting crowd Visiting convenience constructs model features, and trains a visiting prediction model based on the model features; wherein, the visiting prediction model is used to predict the probability of a user visiting a store, so as to mine the crowd of store visitors. The embodiment of the present application can more accurately discover the possible visiting crowds when the store launches an activity, thereby improving the conversion rate and the effect of advertisement placement.

Description

Store visitor crowd mining method, device, equipment and medium

technical field

The present application relates to the field of Internet technologies, in particular to a user visit prediction technology, and in particular to a method, device, device and medium for mining crowds of store visitors.

Background technique

With the development of technology, traditional advertisers are no longer satisfied with the extensive advertising and other preferential activities. They hope to deliver advertisements to the people in need more accurately, thereby improving the ROI of advertising. Some more advanced advertisers will actually monitor the visitors brought by the advertisement to judge the effect of the advertisement.

At present, when traditional advertisers carry out advertising activities, they often circle a certain area for advertising offline. Correspondingly, online may add some crowd attributes for relatively more accurate advertising. However, these methods are still relatively extensive, and the accuracy of determining the possible visitor groups is not high, thus affecting the conversion rate and delivery effect of advertising.

SUMMARY OF THE INVENTION

Embodiments of the present application provide a method, device, device and medium for mining shop visitors, so as to improve the conversion rate and delivery effect of advertisements placed in stores.

In the first aspect, an embodiment of the present application provides a method for mining crowds of shoppers, including:

According to the historical location information of the user and the location information of the store, determine the visitor crowd of the store within a specific historical time period;

Obtain the online information and GPS check-in information of each user in the visiting crowd within the preset time period before the historical specific time period, and determine the convenience of visiting the store corresponding to each GPS check-in information;

Utilize the online information of each user in the visiting crowd, GPS check information and its corresponding store visit convenience to construct model features, and train a visit prediction model based on the model features;

Wherein, the visit prediction model is used to predict the probability of the user visiting the store, so as to mine the people who visit the store.

An embodiment in the above-mentioned application has the following advantages or beneficial effects: on the basis of online information, also in conjunction with GPS dosing information, and determine the convenience of store visits corresponding to each GPS dosing information, in combination with online information, GPS dosing information and its corresponding store visit convenience to build model features, train a visit prediction model, so that the model can predict the probability of user visits more accurately. Help advertisers achieve more accurate advertising and improve the conversion rate and delivery effect of advertising.

Optionally, the online information includes at least: user search information, portrait information, and APP installation information of the user terminal.

Optionally, the determining the convenience of visiting the store corresponding to each GPS check information includes:

According to the road network information and the location information of the store, the arrival time from different locations on the road network around the store to the store is calculated, and the isochronous information of the store arriving at the store is depicted according to the arrival time at the different locations, wherein different equal The arrival time from the position on the hour circle to the store is different;

According to the isochronous circle to which each GPS check information belongs, the store visit convenience corresponding to each GPS check information is determined.

An embodiment in the above application has the following advantages or beneficial effects: since whether the user visits the store is related to the convenience of visiting the store, the road network information is used to describe the isochronous circle of the store, and the isochronous circle is used to describe the store's isochronous circle. Time to measure the convenience of store visits, and let the model learn, so as to use the trained model to achieve accurate mining of store visitors.

Optionally, according to the isochronous circle to which each GPS check information belongs, determine the convenience of visiting the store corresponding to each GPS check information, including:

According to the isochronous circle to which each GPS check information belongs, and the holiday and weather information corresponding to each GPS check information, the convenience of visiting the store corresponding to each GPS check information is determined.

One of the embodiments in the above application has the following advantages or beneficial effects: in addition to store arrival time that affects whether a user visits or not, information on holidays and weather also affects information on holidays and weather. Therefore, the information on holidays and weather can also be used as store visits. The basis for determining the convenience, thereby further improving the accuracy of the model prediction.

Optionally, using the online information of each user in the visiting crowd, GPS check-in information and their corresponding store visits convenience to build model features, including:

Utilize the time dimension information that each GPS dotting information occurs to construct the time sequence feature of each GPS dotting information;

The model features are constructed by using the online information of each user in the visiting crowd, the GPS check-in information and their corresponding time series features and store visit convenience.

An embodiment in the above-mentioned application has the following advantages or beneficial effects: construct its time sequence feature according to the time dimension information that each GPS dotting information occurs, to distinguish the GPS dotting information in time, and further refine the construction of the feature, for accurate learning and prediction.

In a second aspect, an embodiment of the present application also provides a device for excavating crowds of people visiting a store, including:

The visiting crowd determination module is used to determine the visiting crowd of the store in the historical specific time period according to the historical location information of the user and the location information of the store;

The information processing module is used to obtain the online information and GPS check information of each user in the visiting crowd within the preset time before the historical specific time period, and to determine the convenience for visiting the store corresponding to each GPS check information Spend;

The feature building and model training module is used to construct model features using the online information, GPS check-in information and corresponding store visit convenience of each user in the visiting crowd, and train visiting prediction based on the model features Model;

Wherein, the visit prediction model is used to predict the probability of the user visiting the store, so as to mine the people who visit the store.

In a third aspect, an embodiment of the present application also provides an electronic device, including:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,

The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the store visit described in any embodiment of the present application Crowd mining methods.

In a fourth aspect, the embodiments of the present application further provide a non-transitory computer-readable storage medium storing computer instructions, the computer instructions being used to cause the computer to perform the store visitor crowd mining described in any embodiment of the present application method.

An embodiment in the above application has the following advantages or beneficial effects: on the basis of online information such as the user's search information, portrait information and APP installation information of the user terminal, it is also combined with GPS management information, road network information, store location, Offline information such as holidays and weather, use offline information to measure the convenience of users to the store, and refine the model features, so as to train a more accurate visit prediction model, and use this model to dig out the store’s launch event. It can help advertisers to achieve more accurate advertising and improve the conversion rate and delivery effect of advertising.

Other effects of the above-mentioned optional manners will be described below with reference to specific embodiments.

Description of drawings

The accompanying drawings are used for better understanding of the present solution, and do not constitute a limitation to the present application. in:

1 is a schematic flowchart of a method for mining crowds of shop visitors according to a first embodiment of the present application;

2 is a schematic flowchart of a method for mining crowds of shop visitors according to a second embodiment of the present application;

3 is a schematic structural diagram of a store visitor crowd mining device according to a third embodiment of the present application;

FIG. 4 is a block diagram of an electronic device used to implement the method for mining crowds of store visitors according to an embodiment of the present application.

Detailed ways

Exemplary embodiments of the present application are described below with reference to the accompanying drawings, which include various details of the embodiments of the present application to facilitate understanding, and should be considered as exemplary only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present application. Also, descriptions of well-known functions and constructions are omitted from the following description for clarity and conciseness.

FIG. 1 is a schematic flowchart of a method for mining shop visitors according to the first embodiment of the present application. This embodiment can be applied to mining potential visitors when a shop launches an event to improve the conversion rate of advertisement placement. The method can be performed by a shop visitor crowd mining apparatus, the apparatus is implemented by means of software and/or hardware, and is preferably configured in an electronic device, such as a server or a computer device. As shown in Figure 1, the method specifically includes the following:

S101. Determine, according to the historical location information of the user and the location information of the store, the visiting crowd of the store in a historical specific time period.

In order to mine store visitors, it is necessary to train a visit prediction model that can predict whether users will visit, and the training of the model needs to obtain sample data. In the embodiment of the present application, the visiting crowd who has visited the store is used as a sample, and the model features are constructed from the relevant data of the visiting crowd for training.

The historical specific time period may be any period of time, and the period of time may also be a period of any length, which may be configured according to the quantity of sample data and actual needs. As for which ones belong to the visiting users, it can be determined according to the historical location information of the users and the location information of the stores. For example, if a user whose historical location information matches the location information of the store within a specific time period in history belong to visiting users, all the visiting users constitute the visiting group.

S102: Acquire online information and GPS check information of each user in the visiting crowd within a preset time period before the historical specific time period, and determine the convenience of visiting the store corresponding to each GPS check information.

After determining the visiting population, it is also necessary to determine which information of the visiting population is used to construct the model features. As we all know, in the prior art, a certain area is usually delineated offline according to the location of the store, and advertisements are placed on the crowd in the area. Some existing technologies may also add some crowd attributes on this basis to improve the accuracy of advertisement placement. . However, it is very one-sided to determine the advertising crowd based on the location around the store. Even if the attribute information of the crowd is comprehensive, the accuracy can only be appropriately improved in a small range, which still cannot satisfy the current mining of store visitors. need. However, in the embodiment of the present application, various factors of online and offline information are combined, which can greatly improve the efficiency and the accuracy of visiting crowd mining.

Specifically, the online information includes at least: user search information, portrait information, and APP installation information of the user terminal. Among them, the user's search information refers to the query information that the user has searched on the Internet, the portrait information is the information obtained based on big data statistics to describe the characteristics of different users, and the APP installation information of the user terminal refers to the user's APP installation information on the user's terminal. What apps are installed on it. The above query information, portrait information and APP installation information are all information related to the prediction of the user's visit to the store, from which the content of the user's interest, the characteristics and attributes of the user, etc. can be obtained, and used as the basis for the visit prediction.

In addition, GPS dotting information, as a kind of offline information, also plays a role in the user's visit prediction. Wherein, the GPS dotting information refers to the real-time location information of the user obtained through the Internet, that is, the location information such as where the user has been or where he passed by, for example, it may be latitude and longitude information. When predicting whether a user will visit, the convenience of the user's visit can also be considered as a basis. For example, if the convenience of the visit is low, the possibility of the user's visit is lower, and vice versa. Therefore, the embodiments of the present application not only combine online information, but also combine offline GPS check information, and determine the convenience of store visit corresponding to each GPS check information, as a data basis for building model features.

In one embodiment, the determining the convenience of visiting the store corresponding to each GPS check information includes:

According to the road network information and the location information of the store, the arrival time from different locations on the road network around the store to the store is calculated, and the isochronous information of the store arriving at the store is depicted according to the arrival time at the different locations, wherein different equal The arrival time from the position on the hour circle to the store is different;

According to the isochronous circle to which each GPS check information belongs, the store visit convenience corresponding to each GPS check information is determined.

Among them, the traffic roads around the store can be obtained according to the road network information, radiating to each traffic road with the store as the center, to determine some random locations on each traffic road, and the arrival time from these locations to the store , the isochronous circle can be depicted according to the arrival time. The arrival time from the location point on the same isochronous circle to the store is the same or similar, and the arrival time from the location point on the different isochronous circle to the store is different or far away. GPS dotting on different isochronous circles has different arrival times to stores, so the corresponding stores have different access conveniences. Therefore, the corresponding store for each GPS dosing information can be determined according to the isochronous circles. Ease of access.

S103. Use the online information, GPS check-in information and the corresponding store visit convenience of each user in the visiting crowd to construct a model feature, and train a visit prediction model based on the model feature, wherein the The visit prediction model is used to predict the probability of a user visiting a store, so as to dig out the people who visit the store.

By combining online and offline information to construct model features, the problem of low accuracy of a single feature in the prior art is avoided, and the trained visit prediction model has higher accuracy.

The technical solutions of the embodiments of the present application, on the basis of online information, also combine offline GPS check information, and determine the convenience of store visit corresponding to each GPS check information, combine online information, GPS check information and their corresponding The convenience of store visits is based on the characteristics of the model, and the visit prediction model is trained, so that the model can predict the probability of user visits more accurately, and the model can be used to dig out the possible visiting groups when the store launches activities, so as to help advertisers. To achieve more accurate advertising, improve the conversion rate and delivery effect of advertising.

FIG. 2 is a schematic flowchart of a method for mining people visiting a store according to a second embodiment of the present application. This embodiment is further optimized on the basis of the above-mentioned embodiment. As shown in Figure 2, the method specifically includes the following:

S201 , according to the historical location information of the user and the location information of the store, determine the visiting crowd of the store in a historical specific time period.

S202: Acquire online information and GPS check information of each user in the visiting group within a preset time period before the historical specific time period.

Wherein, the online information at least includes: the user's search information, portrait information and APP installation information of the user terminal.

During model training, the portrait information and the APP installation information of the user terminal can also be embedded (embedding layer) first, and then used as the input of the model for training. As for the user's search information, it can be filtered first, that is, the similarity between the original search information and the store advertisement or activity content can be obtained through a specific interface. Divide the time dimension information to obtain search information with time series features, that is, search information in different time periods, let the model distinguish the search information according to the time series features, and learn according to different weights, so as to further refine the model features. to improve the accuracy of model predictions.

S203, according to the road network information and the location information of the store, calculate the arrival time from different locations on the road network around the store to the store, and describe the isochronous circle information arriving at the store according to the arrival time at the different locations, wherein, The arrival time to the store varies from location on the isochronous circle.

S204, according to the isochronous circle to which each GPS check information belongs, and the holiday and weather information corresponding to each GPS check information, determine the convenience of visiting the store corresponding to each GPS check information.

On the basis of the above-mentioned embodiment, in the process of determining the convenience of visiting the store in this embodiment, in addition to the isochronous circle information, the holiday and weather information are also combined. Specifically, whether it is a holiday or not, and the current weather conditions have an impact on the convenience of the user to the store. For example, during the holidays, people not only have time, but also are more willing to come to the store or store to buy the goods they need, so the convenience of visiting the store during the holidays is higher; while in bad weather such as wind or rain , the convenience of getting to the store is naturally lower, and vice versa. Therefore, in addition to considering the isochronous circle information, the convenience of the store to the store is also determined by combining the holiday and weather information, and then the model features are constructed for learning, so that the model can be based on the user's real-time GPS information, current holidays and Weather information, and other online information, comprehensively predict the probability of users coming to the store, so as to achieve accurate visitor user mining.

S205, using the time dimension information of the occurrence of each GPS dotting information to construct a time sequence feature of each GPS dotting information.

The time dimension information may refer to the time when each GPS dotting information is obtained, or the time when the GPS dotting information occurs. The time dimension information of each GPS check information is different, and the impact of GPS check information at different times on the user's visit prediction is different. For example, if a user's GPS check information was around the store one month ago, or the convenience of arriving at the store was high, it does not mean that the probability of the user coming to the store a month later is still high, which only indicates that the user a month ago The probability of arriving at the store is high.

Therefore, it is necessary to construct the time series feature of each GPS dotting information according to the time dimension information, that is, from the time dimension, according to the distance of time, different GPS dotting information is divided, and expressed by the way of time series features, and distinguish. Moreover, the weights of GPS dosing information of different time series features can be different, and the weight of the dotting information that occurred in the near future can be higher, and vice versa, so that the model has the ability to distinguish the GPS dotting information of different occurrence times, and Learn the features and patterns in it to make more accurate predictions.

S206. Use the online information and GPS check information of each user in the visiting crowd and their corresponding time series features and store visit convenience to construct model features, and train a visit prediction model based on the model features.

Wherein, the model may be, for example, a DCN (Deep&Cross Network, deep cross network) model.

The technical solutions of the embodiments of the present application are based on online information such as user search information, portrait information, and APP installation information of the user terminal, as well as offline information such as GPS management information, road network information, store location, holidays, and weather. information, use offline information to measure the convenience of users to the store, refine the model features, so as to train a more accurate visitor prediction model, use this model to dig out the possible visitors when the store launches activities, help Advertisers can achieve more accurate advertising and improve the conversion rate and delivery effect of advertising.

3 is a schematic structural diagram of a store visitor crowd mining device according to a third embodiment of the present application. This embodiment is applicable to mining possible visiting crowds when a store launches an event to improve the conversion rate of advertisement placement. The device can implement the method for mining crowds of shop visitors described in any embodiment of the present application. As shown in FIG. 3, the apparatus 300 specifically includes:

The visiting crowd determination module 301 is used to determine the visiting crowd of the store in a specific historical time period according to the historical location information of the user and the location information of the store;

The information processing module 302 is used to obtain the online information and GPS check-in information of each user in the visiting crowd within the preset time period before the historical specific time period, and determine the visit of the store corresponding to each GPS check-in information convenience;

The feature building and model training module 303 is used to construct model features using the online information, GPS check-in information and corresponding store visit convenience of each user in the visiting crowd, and train visiting based on the model features prediction model;

Wherein, the visit prediction model is used to predict the probability of the user visiting the store, so as to mine the people who visit the store.

Optionally, the online information includes at least: user search information, portrait information, and APP installation information of the user terminal.

Optionally, the information processing module 302 includes:

an information acquisition unit, used for acquiring online information and GPS check-in information of each user in the visiting crowd within a preset time period before the historical specific time period;

The isochronous circle characterization unit is used to calculate the arrival time from different locations on the road network around the store to the store according to the road network information and the location information of the store, and describe the arrival time of the store according to the arrival time of the different locations. Time lap information, wherein the arrival times from the positions on the different isochronous laps to the store are different;

The store visit convenience determination unit is configured to determine the store visit convenience corresponding to each GPS check information according to the isochronous circle to which the GPS check information belongs.

Optionally, the store visit convenience determining unit is specifically used for:

According to the isochronous circle to which each GPS check information belongs, and the holiday and weather information corresponding to each GPS check information, the convenience of visiting the store corresponding to each GPS check information is determined.

Optionally, the feature construction and model training module 303 is specifically used for:

The time sequence utilizes the time dimension information that each GPS dotting information occurs to construct the time sequence feature of each GPS dotting information;

The model features are constructed by using the online information of each user in the visiting crowd, the GPS check-in information and their corresponding time series features and store visit convenience.

The store visitor crowd mining apparatus 300 provided by the embodiment of the present application can execute the store visitor crowd mining method provided by any embodiment of the present application, and has functional modules and beneficial effects corresponding to the execution method. For the content not described in detail in this embodiment, reference may be made to the description in any method embodiment of this application.

According to the embodiments of the present application, the present application further provides an electronic device and a readable storage medium.

As shown in FIG. 4 , it is a block diagram of an electronic device of a method for mining crowds of shoppers according to an embodiment of the present application. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are by way of example only, and are not intended to limit implementations of the application described and/or claimed herein.

As shown in FIG. 4, the electronic device includes: one or more processors 401, a memory 402, and interfaces for connecting various components, including a high-speed interface and a low-speed interface. The various components are interconnected using different buses and may be mounted on a common motherboard or otherwise as desired. The processor may process instructions executed within the electronic device, including instructions stored in or on memory to display graphical information of the GUI on an external input/output device, such as a display device coupled to the interface. In other embodiments, multiple processors and/or multiple buses may be used with multiple memories and multiple memories, if desired. Likewise, multiple electronic devices may be connected, each providing some of the necessary operations (eg, as a server array, a group of blade servers, or a multiprocessor system). A processor 401 is taken as an example in FIG. 4 .

The memory 402 is the non-transitory computer-readable storage medium provided by the present application. Wherein, the memory stores instructions executable by at least one processor, so that the at least one processor executes the method for mining shop visitors provided by the present application. The non-transitory computer-readable storage medium of the present application stores computer instructions, and the computer instructions are used to cause the computer to execute the method for mining crowds of shop visitors provided by the present application.

As a non-transitory computer-readable storage medium, the memory 402 can be used to store non-transitory software programs, non-transitory computer-executable programs and modules, such as program instructions/modules ( For example, the visiting crowd determination module 301, the information processing module 302, and the feature construction and model training module 303 shown in FIG. 3). The processor 401 executes various functional applications and data processing of the server by running the non-transitory software programs, instructions and modules stored in the memory 402, ie, implements the method for mining shop visitors in the above method embodiments.

The memory 402 may include a stored program area and a stored data area, wherein the stored program area may store an operating system and an application program required by at least one function; Data created by the use of electronic equipment, etc. Additionally, memory 402 may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory 402 may optionally include a memory set remotely relative to the processor 401, and these remote memories may be connected to an electronic device implementing the method for mining crowds of shop visitors according to the embodiments of the present application. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

The electronic device for implementing the method for mining crowds of store visitors according to the embodiment of the present application may further include: an input device 403 and an output device 404 . The processor 401 , the memory 402 , the input device 403 and the output device 404 may be connected by a bus or in other ways, and the connection by a bus is taken as an example in FIG. 4 .

The input device 403 can receive input digital or character information, and generate key signal input related to user settings and function control of the electronic device implementing the method for mining crowds of shop visitors in the embodiment of the present application, such as touch screen, keypad, mouse, Input devices such as trackpads, touchpads, pointing sticks, one or more mouse buttons, trackballs, joysticks, etc. Output devices 404 may include display devices, auxiliary lighting devices (eg, LEDs), haptic feedback devices (eg, vibration motors), and the like. The display device may include, but is not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display, and a plasma display. In some implementations, the display device may be a touch screen.

Various implementations of the systems and techniques described herein can be implemented in digital electronic circuitry, integrated circuit systems, application specific ASICs (application specific integrated circuits), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include being implemented in one or more computer programs executable and/or interpretable on a programmable system including at least one programmable processor that The processor, which may be a special purpose or general-purpose programmable processor, may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device an output device.

These computational programs (also referred to as programs, software, software applications, or codes) include machine instructions for programmable processors, and may be implemented using high-level procedural and/or object-oriented programming languages, and/or assembly/machine languages calculation program. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and/or apparatus for providing machine instructions and/or data to a programmable processor ( For example, magnetic disks, optical disks, memories, programmable logic devices (PLDs), including machine-readable media that receive machine instructions as machine-readable signals. The term "machine-readable signal" refers to any signal used to provide machine instructions and/or data to a programmable processor.

To provide interaction with a user, the systems and techniques described herein may be implemented on a computer having a display device (eg, a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user ); and a keyboard and pointing device (eg, a mouse or trackball) through which a user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (eg, visual feedback, auditory feedback, or tactile feedback); and can be in any form (including acoustic input, voice input, or tactile input) to receive input from the user.

The systems and techniques described herein may be implemented on a computing system that includes back-end components (eg, as a data server), or a computing system that includes middleware components (eg, an application server), or a computing system that includes front-end components (eg, a user's computer having a graphical user interface or web browser through which a user may interact with implementations of the systems and techniques described herein), or including such backend components, middleware components, Or any combination of front-end components in a computing system. The components of the system may be interconnected by any form or medium of digital data communication (eg, a communication network). Examples of communication networks include: Local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.

A computer system can include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server arises by computer programs running on the respective computers and having a client-server relationship to each other.

According to the technical solutions of the embodiments of the present application, on the basis of online information such as the user's search information, portrait information, and APP installation information of the user terminal, GPS check-in information, road network information, store location, holidays, weather, etc. are also combined. offline information, use offline information to measure the convenience of users to the store, refine the model features, so as to train a more accurate visitor prediction model, and use this model to dig out the possible visitors to the store when the event is launched. Thereby helping advertisers to achieve more accurate advertising, improve the conversion rate and delivery effect of advertising.

It should be understood that steps may be reordered, added or deleted using the various forms of flow shown above. For example, the steps described in the present application can be performed in parallel, sequentially or in different orders, and as long as the desired results of the technical solutions disclosed in the present application can be achieved, no limitation is imposed herein.

The above-mentioned specific embodiments do not constitute a limitation on the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of this application shall be included within the protection scope of this application.

Claims (12)

1. a shop visits crowd mining method, is characterized in that, comprises: According to the historical location information of the user and the location information of the store, determine the visitor crowd of the store within a specific historical time period; Obtain the online information and GPS check-in information of each user in the visiting crowd within the preset time period before the historical specific time period, and determine the convenience of visiting the store corresponding to each GPS check-in information; Utilize the online information of each user in the visiting crowd, GPS check information and its corresponding store visit convenience to construct model features, and train a visit prediction model based on the model features; Wherein, the visit prediction model is used to predict the probability of the user visiting the store, so as to mine the people who visit the store. 2 . The method according to claim 1 , wherein the online information at least comprises: user search information, portrait information and APP installation information of the user terminal. 3 . 3. method according to claim 1, is characterized in that, described determining the shop visiting convenience corresponding to each GPS dotting information, comprises: According to the road network information and the location information of the store, the arrival time from different locations on the road network around the store to the store is calculated, and the isochronous information of the store arriving at the store is depicted according to the arrival time at the different locations, wherein different equal The arrival time from the position on the hour circle to the store is different; According to the isochronous circle to which each GPS check information belongs, the store visit convenience corresponding to each GPS check information is determined. 4. method according to claim 3, is characterized in that, described according to the isochronous circle that described each GPS dosing information belongs to, determine the store visiting convenience corresponding to described each GPS dosing information, comprising: According to the isochronous circle to which each GPS check information belongs, and the holiday and weather information corresponding to each GPS check information, the convenience of visiting the store corresponding to each GPS check information is determined. 5. The method according to any one of claims 1-4, wherein the online information, GPS check-in information and corresponding store visiting convenience of each user in the visiting crowd are utilized. Build model features, including: Utilize the time dimension information that each GPS dotting information occurs to construct the time sequence feature of each GPS dotting information; The model features are constructed by using the online information of each user in the visiting crowd, the GPS check-in information and their corresponding time series features and store visit convenience. 6. A shop visiting crowd excavation device, characterized in that, comprising: The visiting crowd determination module is used to determine the visiting crowd of the store in the historical specific time period according to the historical location information of the user and the location information of the store; The information processing module is used to obtain the online information and GPS check information of each user in the visiting crowd within the preset time before the historical specific time period, and to determine the convenience for visiting the store corresponding to each GPS check information Spend; The feature building and model training module is used to construct model features using the online information, GPS check-in information and corresponding store visit convenience of each user in the visiting crowd, and train visiting prediction based on the model features Model; Wherein, the visit prediction model is used to predict the probability of the user visiting the store, so as to mine the people who visit the store. 7 . The device according to claim 6 , wherein the online information at least comprises: user search information, portrait information, and APP installation information of the user terminal. 8 . 8. The apparatus according to claim 6, wherein the information processing module comprises: an information acquisition unit, used for acquiring online information and GPS check-in information of each user in the visiting crowd within a preset time period before the historical specific time period; The isochronous circle characterization unit is used to calculate the arrival time from different locations on the road network around the store to the store according to the road network information and the location information of the store, and describe the arrival time of the store according to the arrival time of the different locations. Time lap information, wherein the arrival times from the positions on the different isochronous laps to the store are different; The store visit convenience determination unit is configured to determine the store visit convenience corresponding to each GPS check information according to the isochronous circle to which each GPS check information belongs. 9. The device according to claim 8, wherein the store visit convenience determining unit is specifically used for: According to the isochronous circle to which each GPS check information belongs, and the holiday and weather information corresponding to each GPS check information, the convenience of visiting the store corresponding to each GPS check information is determined. 10. The device according to any one of claims 6-9, wherein the feature construction and model training module is specifically used for: The time sequence utilizes the time dimension information that each GPS dotting information occurs to construct the time sequence feature of each GPS dotting information; The model features are constructed by using the online information of each user in the visiting crowd, the GPS check-in information and their corresponding time series features and store visit convenience. 11. An electronic device, characterized in that, comprising: at least one processor; and a memory communicatively coupled to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the execution of any of claims 1-5 store visitor crowd mining method. 12. A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the method for mining shop visitors according to any one of claims 1-5 .

Images