WO2021175021A1

WO2021175021A1 - Product push method and apparatus, computer device, and storage medium

Info

Publication number: WO2021175021A1
Application number: PCT/CN2021/071803
Authority: WO
Inventors: 刘继武
Original assignee: 深圳壹账通智能科技有限公司
Priority date: 2020-03-06
Filing date: 2021-01-14
Publication date: 2021-09-10
Also published as: CN111476595A

Abstract

A product push method and apparatus, a computer device, and a readable storage medium. The method comprises: obtaining preset protocol data, extracting node attributes of the preset protocol data (202); generating calculation factors according to the node attributes (204); obtaining protocol data and screening the protocol data by means of the calculation factors to obtain matching node data (206); mapping the matching node data to corresponding data nodes (208); generating product push information according to the data nodes (210); obtaining a specified data type and obtaining the matching node data from the data nodes according to the specified data type as object matching data (212); obtaining target push objects from an object database according to the object matching data and screening the product push information according to the specified data type to obtain target push information (214); pushing the target push information to the target push objects (216). The matching node data can be stored in a blockchain. The present method solves the problem that the types of product push in the prior art is too few.

Description

Product push method, device, computer equipment and storage medium

This application is based on the Chinese invention patent application filed on March 6, 2020 with the application number 202010151325.5 and titled "Product push method, device, computer equipment and storage medium", and claims its priority.

Technical field

This application relates to the field of data classification, in particular to a product push method, device, computer equipment and storage medium.

Background technique

At present, the push of a certain business product of the company is either initiated by the seller of the business or purchased related information services from the supplier. Either way, it will increase the cost. Traditional companies generally have a large customer resource database, and the degree of authenticity and data integrity are relatively high. If they buy products that are intelligently recommended by a third party, there will be certain information leakage problems. In the traditional technology, the user information is classified, and then the information is matched based on the classified information. In this way, the product information matched by the user or the object is too single. For example, a product data push method based on machine learning disclosed by the publication number CN109447685 is to classify the obtained user information and add category tags, and then obtain the corresponding product data according to the category tags, and finally follow the preset rules After generating the corresponding resource allocation result, pushing the information to the user indicates the corresponding user terminal. The inventor realizes that although this product push method can achieve targeted product push to specific users, the generated product push information is too single and cannot solve the singular information push in the prior art.

Summary of the invention

Based on this, it is necessary to address the above technical problems, and this application provides a product push method, device, computer equipment, and storage medium to solve the technical problem that the type of product push in the prior art is too single.

A product push method, the method includes:

Acquiring preset protocol data, and extracting node attributes of the preset protocol data;

Generating a calculation factor for screening protocol data according to the node attribute;

Obtain protocol data, and filter the protocol data by the calculation factor to obtain matching node data;

Mapping the matching node data to the corresponding data node;

Generating product push information according to the data node;

Acquiring a designated data type, and acquiring matching node data from the data node according to the designated data type as object matching data;

Acquiring the target push object from the object database according to the object matching data, and filtering the product push information according to the specified data type to obtain the target push information;

Push the target push information to the target push object

A product pushing device, the device comprising:

The attribute extraction module is used to obtain preset protocol data, and extract the node attributes of the preset protocol data;

A factor generating module, configured to generate a calculation factor for screening protocol data according to the node attribute;

The data screening module is used to obtain protocol data, and filter the protocol data by the calculation factor to obtain matching node data;

A data mapping module, configured to map the matching node data to the corresponding data node;

An information generation module, which is used to generate product push information according to the data node;

The object data acquisition module is configured to acquire a specified data type, and obtain matching node data from the data node according to the specified data type, as object matching data;

The data matching module is configured to obtain the target push object from the object database according to the object matching data, and filter the product push information according to the specified data type to obtain the target push information;

The information push module is used to push the target push information to the target push object.

A computer device, including a memory and a processor, and a computer program stored in the memory and capable of running on the processor, and when the processor executes the computer program, the product push method described below is implemented step:

Mapping the matching node data to the corresponding data node;

Generating product push information according to the data node;

Push the target push information to the target push object.

A computer-readable storage medium, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the product push method described below are realized:

Mapping the matching node data to the corresponding data node;

Generating product push information according to the data node;

Push the target push information to the target push object.

The above-mentioned product push methods, devices, computer equipment and storage media, through purposeful screening of the protocol data in the protocol data acquisition stage, and then store the filtered protocol data according to the data node, and finally obtain the specified type of protocol data The matching of the target push objects and the screening of the push information solve the technical problem of the single type of products pushed by the target push objects in the prior art.

Description of the drawings

In order to explain the technical solutions of the embodiments of the present application more clearly, the following will briefly introduce the drawings that need to be used in the description of the embodiments of the present application. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative labor.

Figure 1 is a schematic diagram of the application environment of the product push method;

Figure 2 is a schematic flow diagram of the product push method;

FIG. 3 is a schematic flowchart of step 204 in FIG. 2;

FIG. 4 is a schematic flowchart of step 206 in FIG. 2;

FIG. 5 is another schematic diagram of the process of step 206 in FIG. 2;

FIG. 6 is a schematic flowchart of step 208 in FIG. 2;

FIG. 7 is a schematic flowchart of step 212 in FIG. 2;

FIG. 8 is a schematic flowchart of step 214 in FIG. 2;

Figure 9 is a schematic diagram of a product pushing device;

Figure 10 is a schematic diagram of a computer device in an embodiment.

Detailed ways

Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the technical field of the application; the terms used in the specification of the application herein are only for describing specific embodiments. The purpose is not to limit the application; the terms "including" and "having" in the specification and claims of the application and the above-mentioned description of the drawings and any variations thereof are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific sequence.

The reference to "embodiments" herein means that a specific feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art clearly and implicitly understand that the embodiments described herein can be combined with other embodiments.

In order to make the objectives, technical solutions, and advantages of this application clearer, the following further describes the application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of this application.

The product push method provided in the embodiment of the present application can be applied to the application environment as shown in FIG. 1. Among them, the application environment may include a terminal 102, a network 106, and a server 104. The network 106 is used to provide a communication link medium between the terminal 102 and the server 104. The network 106 may include various connection types, such as wired and wireless communications. Link or fiber optic cable, etc.

The user can use the terminal 102 to interact with the server 104 through the network 106 to receive or send messages and so on. Various communication client applications, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc., may be installed on the terminal 102.

The terminal 102 may be various electronic devices with a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III, moving picture experts compress standard audio Level 3), MP4 (Moving Picture Experts Group Audio Layer IV, Motion Picture Experts compress standard audio level 4) Players, laptop portable computers and desktop computers, etc.

The server 104 may be a server that provides various services, for example, a background server that provides support for pages displayed on the terminal 102.

It should be noted that the product pushing method provided in the embodiments of the present application is generally executed by the server/terminal, and accordingly, the product pushing device is generally set in the server/terminal device.

It should be understood that the numbers of terminals, networks, and servers in FIG. 1 are merely illustrative. There can be any number of terminal devices, networks, and servers according to implementation needs.

Among them, the terminal 102 communicates with the server 104 through the network. The server 104 obtains the protocol data from the terminal 102, and filters the protocol data according to calculation factors to obtain matching node data, and then generates product push information based on the matching node data, matches the target push object, and then pushes the product push information to the target Push the object to realize the push of the product. Among them, the terminal 102 and the server 104 are connected through a network. The network can be a wired network or a wireless network. The terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablets, and portable wearable devices. , The server 104 can be implemented by an independent server or a cluster of multiple servers.

In one embodiment, as shown in FIG. 2, a product push method is provided. Taking the method applied to the server in FIG. 1 as an example for description, the method includes the following steps:

Step 202: Obtain preset protocol data, and extract node attributes of the preset protocol data.

The preset protocol data may be obtained according to actual business scenarios, and the actual business scenarios depend on the type of product pushed this time and the target of the product push. For example, it is known that the information of objects with the following three types of characteristics is used as the preset agreement data: age: 20-35; work location: remote area; whether or not a certain type of product has been purchased: yes.

Step 204: Generate calculation factors for filtering protocol data according to the node attributes.

Node attributes are dimensional features and values extracted from the structured data after structuring the preset protocol data. Specifically, it may be data information corresponding to the object, and the data information includes several dimensional features and data corresponding to the dimensional features. For example, Table 1 shows examples of some dimensional features in the preset protocol data and their corresponding data:

Table 1

The calculation factor is obtained based on the node data. A calculation factor includes an attribute generated based on the node attribute. For example, if you perform a cluster analysis on the obtained node attribute, and obtain 10 node attributes from the preset protocol data, then Performing cluster analysis on the node attributes of the three dimensions of Behavior_area: North China, Behavior_type: 3, and Item_category: 7350, the calculated factor can be that 7,350 products are purchased in North China.

Specifically, this embodiment can be applied to recommend insurance product information to users. The preset agreement data (preset insurance policy data) includes the geek users’ insurance areas: Henan, Hubei, and Hunan, and these provinces are in turn It belongs to Central China; for example, including Beijing, Tianjin, Shanxi Province, Hebei Province or Inner Mongolia Autonomous Region, these belong to North China; it can also extract the age of the user, such as 18-45 for young and middle-aged, 46-65 for Middle-aged people, over 66 belong to old age and other attributes, which can be the node attributes of the preset protocol data. Through the above method, the calculation factor can be obtained as: health insurance in North China.

Among them, clustering is a process of classifying data into different classes or clusters, so objects in the same cluster have great similarities, while objects between different clusters have great differences. Then it can be aggregated according to the similarity between different tags to generate a calculation factor. The number of calculation factors is not limited to one. Multiple calculation factors form a screening template, which is dedicated to the screening of agreement data of a specific business, and then the agreement data can be screened according to the screening template.

Step 206: Obtain protocol data, and filter the protocol data by calculation factors to obtain matching node data.

The protocol data includes historical protocol data and real-time protocol data. The historical agreement data is the agreement data of the previous purchase users, which can be obtained from other resource pools. Among them, the resource pool is a storage pool that stores data in different latitudes established according to specified rules. Its function is to make it easier for the server to extract data through Hadoop2. Then carry out data analysis, because the data structure will be extremely complicated with different types of insurance policies. Without classification and sorting, it is difficult to use the data effectively.

Historical protocol data are all structured data, which can be directly retrieved for analysis and use. Among them, both real-time agreement data and historical agreement data include characteristic data of various dimensions of the user object, such as age, occupation, income, location, family members, types of goods purchased, and so on.

For real-time protocol data, if the back-end server receives real-time protocol data, it will perform RSA data encryption processing on the real-time protocol data. RSA encryption is a kind of asymmetric encryption, which can complete the decryption without directly transferring the key. This can ensure the security of the information and avoid the risk of being cracked caused by the direct transfer of the key. The specific encryption process is current. There are technologies that will not be repeated here.

After the real-time protocol data is encrypted, the encrypted protocol data needs to be structured through Redis, and then the structured protocol data is saved in the resource pool for later use.

Redis is an open source high-performance key-value database developed in C language. It adapts to storage needs in different scenarios by providing a variety of key-value data types. So far, the key-value data types supported by Redis are as follows :

String type; hash type; list type; collection type; ordered collection type.

Application scenarios can be:

Cache (data query, short connection, news content, product content, etc.), session separation in distributed cluster architecture, online friend list in chat room, task queue (spike, snap-up, 12306, etc.), application ranking, website Access statistics, data expiration processing (can be accurate to milliseconds).

Step 208: Map the matching node data to the corresponding data node.

A data node is a node in a data node library set up separately, such as the product type (Enterprise Edition, Home Edition or Professional Edition) of the user object in the agreement data, purchase period (regular, lifetime), income, age, family members , Address, etc.

Each data node stores a unique identifier of the protocol data that meets the calculation factor, and the unique identifier is generated based on the order number and channel number of the protocol data. Among them, one protocol data can correspond to multiple data nodes, for example, it can be under the product type (Professional Edition) node, or under the data node that is regularly purchased. The advantage of mapping and saving the protocol data through different attributes of the protocol data is that the protocol data can be classified into categories, which is convenient for subsequent generation of product types that need to be pushed to the user based on the protocol data under the data node.

Furthermore, each data node can have multiple sub-nodes. For example, under the Professional Edition node, it can be divided into two nodes of the same level, the regular node and the lifetime node. Under the regular node, there can be different age groups of purchasing users. Sub-nodes, etc., the structure of the specific data node needs to be determined according to the specific situation, and it is not limited here.

Step 210: Generate product push information according to the data node.

Combining different data nodes can map multiple product push information, such as purchasing users who are older than 22 and lower than 28, address in first-tier cities, and income meets the preset value of these data nodes, and can push Professional Edition, etc. Product push information. Users who are older than 22 and lower than 28, live in a mountainous area, and whose income is not higher than a certain preset value are combined to get a free edition.

There is a many-to-many relationship between data nodes and product push information, and the details depend on specific application scenarios. For example, you can push product information to corresponding potential customers according to the following dimensions:

Age group, gender, occupation, income, etc.

Step 212: Obtain the specified data type, and obtain the protocol data from the data node according to the specified data type as the object matching data.

In this embodiment, protocol data can be obtained through Hadoop2, which improves the efficiency of data processing. Hadoop2 is a server cluster. The purpose of object matching data acquisition according to the specified data type obtained, and target push object acquisition according to the object matching data is that the amount of protocol data under the data node is large. If you want to achieve targeted push products for users, then It is possible to select a certain type or several types of protocol data from the protocol data under the data node by formulating the data type as the object matching data.

Optionally, if you need to change the product of the product, you only need to adjust the specified data type, and there is no need to filter and classify the data again, which greatly reduces the efficiency of product push.

Specifically, the server submits a job request through the client (for example, grabs all specified types of Free edition products), and then the server schedules specified resources from the resource pool, and the specified data type is obtained according to the specified product type .

Further, it receives the job request submitted by the client to capture the protocol data of all specified types of user objects, and then the Scheduler on the server is responsible for scheduling the specified resources (specify which product node data to capture) , Then, ApplicationsManager receives the data under the specific product node and executes the operation of fetching the data through a container (container).

Finally, the ApplicationMaster will be responsible for the application of the new container and the monitoring of the operation (subsequent data capture of this type will be operated and captured by the ApplicationMaster, and the monitoring of the capture situation).

Among them, the HDFS function in Hadoop2 can be used to collect protocol data scattered and stored under each data node. Among them, HDFS is a distributed file system with high reliability and high throughput.

Step 214: Obtain the target push object from the object database according to the object matching data, and filter the product push information according to the specified data type to obtain the target push information.

The object matching data includes some attribute data of the user object, such as age, income, date of birth, or identity information. Obtain objects that match the above attribute data from the object database as the target push objects, for example, objects that match the age of 20-30, income of 1W or more, unfixed occupation, and address in a certain area. Then you can also select product push information that meets the specified data type from the product push information obtained above, for example, products purchased by objects aged 20-30, income above 1W, unfixed occupation, and residential address in a certain area, and The product is pushed to the target push target as the target push information.

Step 216: Push the target push information to the target push object.

After entering the relevant order information for the user object, then the relevant order payment is made. After the payment is completed, if the user checks the product drift bottle check box, the background uses the activeMQ message middleware to asynchronously transmit the information to the background . Subsequently, the target push information generated for the user can be pushed to the user's terminal in the form of a drift bottle in the form of a product drift bottle.

In the above product push method, the calculation factor for user screening protocol data is obtained according to the protocol data of the specified user group, and the protocol data is filtered, and the appropriate protocol data is selected and inserted under the corresponding data node, which is targeted Obtain the node attributes of the preset protocol data of a certain type of object, and generate calculation factors for the node attributes to filter the protocol data, and also specify the target push information by specifying the data type and select the object from the protocol data to filter the data, which improves the The massive data processing capability of the drifting bottle is able to efficiently filter and sort the data for accurate push, and the data used during the period are all real order cases, and it is more persuasive to recommend to relevant users.

In one embodiment, as shown in FIG. 3, step 204 includes:

Step 302: Perform vectorization processing on the node attributes through the one-hot algorithm to obtain the vectorized attributes.

One-Hot expression is a structured way of text classification. It is the most intuitive and the most commonly used word expression so far. Specifically, a dictionary of node attributes is generated; for example: insurance type, insured user address, age, income, and the resulting dictionary is [insurance, land, household, category, age, year, income, income, investment, insurance, type, Use, address] (can be arranged in the order of pinyin, skipped here), 13 characters are obtained, and the three characters are expressed as ont-hot vectors in the following form:

"Guarantee": [1,0,0,0,0,0,0,0,0,0,0,0,0]

"Risk": [0,0,0,0,0,0,0,0,0,1,0,0,0]

...

"Address": [0,0,0,0,0,0,1,0,0,0,0,0,0]

At this time, the above node attributes are represented as follows:

Insurance type: 1,0,0,1,0,0,0,0,0,1,1,0,0

Insured user address: 1,1,1,0,0,0,0,0,1,0,0,1,1

...

Income: 0,0,0,0,0,1,1,0,0,0,0,0

Age and income are the node attributes and so on, so I won’t repeat them here.

Step 304: Calculate the Manhattan distance between the vectorized attributes.

Manhattan distance (Manhattan Di stance) is a kind of distance between vectors. For example, the Manhattan distance between two points a(x1,y1) and b(x2,y2) in a two-dimensional plane:

d ₁₂ =|x ₁ -x ₂ |+|y ₁ -y ₂ | Formula (1)

Among them, x1 is the X-axis coordinate of point a, y1 is the Y-axis coordinate; x2 is the X-axis coordinate of point b, and y2 is the Y-axis coordinate.

Representation of the Manhattan distance between two n-dimensional vectors a(x11,x12,...,x1n) and b(x21,x22,...,x2n):

Among them, k indicates that the point a or b is located in which dimension, and k is a positive integer.

In this embodiment, the Manhattan distance between two vector attributes can be calculated by formula (2) to obtain the similarity between the two vector attributes.

Step 306: If the number of vectorized attributes whose Manhattan distance is less than the preset distance is greater than the preset value, set the vectorized attributes whose Manhattan distance is less than the preset distance as the core attribute cluster.

In this embodiment, it is necessary to obtain a cluster whose Manhattan distance is less than a preset distance, and the number of vectorized attributes whose Manhattan distance is less than the preset distance is greater than a preset value, as the core attribute cluster. The smaller the Manhattan distance, the higher the similarity between the two vectors, and the greater the number, which indicates that the attribute cluster is more likely to be used as a calculation factor.

Step 308: Integrate the core attribute clusters to obtain calculation factors.

In this embodiment, if the purchase age is 25-35, the purchased product type is free edition, the user income is less than 5Kw, and the activity area is a critical area (mountain, ocean, desert, etc.) as a core attribute cluster, then The calculation factor of can be: a free version for young adults who live in remote areas for a long time.

Furthermore, it is known that it is necessary to push insurance products to geeks and adventurous people. For example, to recommend accident insurance products to some users who browse geeks and adventurous websites, a data screening template composed of calculation factors can be generated. This screening template is It can be obtained by combining multiple calculation factors. Among them, the calculation factor can be health insurance in North China, health insurance in Central China, accident insurance in North China, and so on. The selection of preset protocol data needs to be based on specific application scenarios.

Optionally, if an insurance product is recommended for a certain elderly user group, the insurance policy data of some elderly users who have purchased accident insurance can be obtained as the preset agreement data, and the final calculation factor generated according to the above method can be: age 65 or more Accident insurance, cancer insurance for age 65 and above, etc.

For example, if you want to recommend accident insurance products to a specific user, the generated calculation factor can be: the nodes in the calculation factor are age, insurance type, and occupation. Several nodes related to the accident insurance product comparison are used as calculation factors.

In this embodiment, the protocol data is filtered by the calculation factor, and the protocol data that also includes the nodes in the calculation factor are inserted under the corresponding data node. Different nodes can be included in different calculation factors to ensure the diversity of the selected protocol data. For example, in a certain protocol data, there are only node information such as protocol data type, user address, age, etc., but no node information such as income and family members, but the data of the income node is related to whether the user purchases the possibility and what type of purchase The possibility of the product, etc., may miss the protocol data because the data of a certain node is not considered.

This embodiment takes the push of a product as an example to illustrate the Hadoop2-based product push method. Specifically, the agreement data may be the agreement data generated after the user purchases the product, which includes the type of the purchased product of the user, and the user Identity information, address, income, home address and other data. After the user completes the payment, if the user selects the product recommendation drift bottle check box and clicks Finish, the background will use the activeMQ message middleware to asynchronously transmit the user’s protocol data to the background Perform storage and analysis. The general user completes this step and the purchase process has ended. The user's agreement data has certain reference significance for pushing suitable products to suitable user groups.

In one embodiment, as shown in FIG. 4, step 206 includes:

Step 402: Obtain historical protocol data and store it in HDFS.

The historical protocol data is the protocol data that has been obtained before a certain point in time, and the historical protocol data may be the protocol data that has been screened by the calculated factor, or it may be the protocol data that has not been screened by the calculated factor.

Step 404: Analyze historical protocol data through MapReduce to obtain historical analysis results.

MapReduce is a programming model for parallel operations on large-scale data sets (greater than 1TB). The reason for using MapReduce is its high fault tolerance. For example, if one of the machines is down, it can transfer the above computing tasks to another node to run, so that the task will not fail.

The analysis of historical protocol data through MapReduce is mainly to analyze whether the historical protocol data meets the calculation factor. For example, the calculation factor is "the free version of young adults living in remote areas for a long time", then it is necessary to compare whether the historical agreement data meets the three attributes of long-term active in dangerous areas, young adults and free version. Of course, there can also be other calculation factors combined with the calculation factor "Free version of young adults living in remote areas for a long time" to filter historical agreement data to ensure that the accuracy of the obtained agreement data is high.

Step 406: If the historical analysis result is that the historical protocol data meets the calculation factor, it is detected whether there is a key-value pair between the historical protocol data and the data node.

The key-value pair is a verification condition used to indicate whether the historical protocol data already exists under the data node (for example, the data node Free edition). If the historical protocol data has already appeared under the data node, a key-value pair will be generated to point to the data node.

Step 408: If the key-value pair does not exist, map the historical protocol data to the corresponding data node as screening data.

If there is no key-value pair pointing to the data node, insert the historical protocol data under the data node as the screening data. The screening data is the data used to screen the real-time protocol data. The screening data obtained in this way is used to screen the protocol data to obtain higher data similarity. Moreover, the historical agreement data is analyzed and inserted under the data node, and the purchase type, payment method, purchase frequency, etc. of the previous purchaser can be obtained, as a kind of screening data, and used as the basis for subsequent real-time agreement data processing.

Step 410: Screen the real-time protocol data sent through the activeMQ message middleware according to the screening data, and map the real-time protocol data that meets the screening data to the corresponding data node as matching node data.

Further, this embodiment can use the asynchronous message function of the activeMQ message queue to increase the continuity of the protocol data transfer to the back-end server during the purchase process, but the failure of this additional service will not affect the main purchase process. The matching node data is used to

In this embodiment, the historical protocol data is filtered in advance to obtain the screening data, and the screening data is used as the screening condition for the real-time protocol data, so that the subsequent received real-time protocol data can be calculated without calculation factors, but directly compared with The fit of the screening data, such as the similarity of income, age, address, family members, etc., has improved

The precision of data filtering.

In one embodiment, as shown in FIG. 5, step 206 further includes:

Step 502: If there is a key-value pair, obtain real-time protocol data sent through the activeMQ middleware.

If there is a key-value pair, it means that the protocol data has passed the filter of the calculation factor and is inserted under the corresponding data node.

Step 504: Filter the real-time protocol data according to the calculation factor to obtain matching node data.

If there is no newly inserted historical protocol data, the real-time protocol data can be directly filtered by calculation factors to obtain matching node data.

In this embodiment, the real-time protocol data is filtered by calculation factors to obtain matching node data, which ensures the real-time performance of the data and improves the real-time performance of the obtained data.

In one embodiment, as shown in FIG. 6, step 208 includes:

Step 602: Obtain the protocol code and channel code of the filtered protocol data.

The protocol code of the protocol data. In this embodiment, it can be the order number of the purchasing user. Taking the order number of a website integrating multiple e-commerce platforms as an example, it can be: GP0200000000310241. The channel number can be the channel source of the order data, such as Order data of e-commerce company A or order data of e-commerce company B.

Step 604: Generate an index code of the filtered protocol data according to the protocol code and the channel code.

The index code is the unique identification of the order data, and the reason for adding the channel code is that some small e-commerce companies use the same rules for the generated order numbers, there is a low probability of repetition, and the channel source of the protocol data can ensure the protocol data The uniqueness of the number.

Step 606: Map the filtered protocol data to the corresponding data node through index coding.

In this embodiment, only the index code of the protocol data is stored under the data node, so the same index code can be stored under multiple data nodes, and the index code is mapped to a specific protocol data on the protocol database, saving storage space .

In one embodiment, as shown in FIG. 7, step 212 includes:

Step 702: Receive a data capture request, where the data capture request includes a specified data type.

Step 704: Obtain matching node data from the data node according to the specified data type as object matching data.

Specifically, in this embodiment, the server receives a data extraction request submitted by a client (client). The data extraction request includes a specified data type. The specified data type may be: grab all specified types of Home Edition orders (such as specified types). Yes: Periodic, lifetime), and then the Scheduler on the server is responsible for scheduling the specified resources (specify which Home Edition node data is to be captured), and then obtain the specific Home Edition protocol data from the data node.

The operation in this embodiment can use the HDFS function (what algorithm in the HDFS function) in Hadoop2 to collect distributed stored protocol data. HDFS is a distributed file system with high reliability and high throughput. The data capture speed through Hadoop2 is high, which can improve the speed of data acquisition.

In one embodiment, as shown in FIG. 8, step 214 includes:

Step 802: Obtain at least one attribute value of the specified data type from the object matching data.

If the specified data types are age, income, and Home Edition types, the corresponding values under the same attribute obtained from the object matching data can be: age 30, income 2w/month, and Home Edition regular. Generally, at least one attribute value in the data type is specified.

Step 804: Obtain the user object from the object database, where the user object includes an attribute map having the same attribute as the attribute value. Based on the attribute value obtained above, an object with the same value or a similar value to the attribute value is obtained from the object database.

Step 806: Use the user corresponding to the attribute map that matches the attribute value as the target push object.

In this embodiment, the object in the specified direction is obtained from the object database by specifying the data type, and the object is pushed as the target, so that the obtained object is more targeted.

It should be understood that, although the steps in the flowcharts of FIGS. 2 to 8 are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless specifically stated in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least some of the steps in Figure 2-8 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. These sub-steps or The order of execution of the stages does not have to be carried out sequentially, but may be executed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

It should be emphasized that, in order to further ensure the privacy and security of the matching node data, the matching node data may also be stored in a node of a blockchain.

The blockchain referred to in this application is a new application mode of computer technology such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. Blockchain, essentially a decentralized database, is a series of data blocks associated with cryptographic methods. Each data block contains a batch of network transaction information for verification. The validity of the information (anti-counterfeiting) and the generation of the next block. The blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.

This application can be applied in the field of smart communities to push products to community members in smart communities, thereby promoting the construction of smart cities.

In one embodiment, as shown in FIG. 9, a product pushing device is provided, and the product pushing device corresponds to the product pushing method in the foregoing embodiment one-to-one. The product push device includes:

The attribute extraction module 902 is used to obtain preset protocol data and extract node attributes of the preset protocol data.

The factor generation module 904 is configured to generate calculation factors for screening protocol data according to the node attributes.

The data screening module 906 is used to obtain protocol data, and filter the protocol data by calculation factors to obtain matching node data.

The data mapping module 908 is used to map the matching node data to the corresponding data node.

The information generating module 910 is configured to generate product push information according to the data node.

The object data obtaining module 912 is configured to obtain a specified data type, and obtain matching node data from the data node according to the specified data type as object matching data.

The data matching module 914 is configured to obtain the target push object from the object database according to the object matching data, and filter the product push information according to the specified data type to obtain the target push information.

The information push module 916 is used to push the target push information to the target push object.

Further, the factor generation module 904 includes:

The vectorization sub-module is used to vectorize the node attributes through the one-hot algorithm to obtain the vectorized attributes.

The distance calculation sub-module is used to calculate the Manhattan distance between each vectorized attribute.

The core cluster determination sub-module is used to set the vectorized attribute with Manhattan distance less than the preset distance as the core attribute cluster if the number of vectorized attributes with Manhattan distance less than the preset distance is greater than the preset value.

The attribute cluster integration sub-module is used to integrate the core attribute clusters to obtain calculation factors.

Further, the data screening module 906 includes:

The historical data acquisition sub-module is used to acquire historical protocol data and store it in HDFS.

The historical analysis sub-module is used to analyze historical protocol data through MapReduce to obtain historical analysis results.

The mapping detection sub-module is used to detect whether there is a key-value pair between the historical protocol data and the data node if the historical analysis result is that the historical protocol data meets the calculation factor.

The filtering data acquisition sub-module is used to map the historical protocol data to the corresponding data node if there is no key-value pair, as the filtering data.

The first data mapping sub-module is used to screen the real-time protocol data sent through the activeMQ message middleware according to the screening data, and map the real-time protocol data that meets the screening data to the corresponding data node as matching node data.

Further, the data screening module 906 further includes:

The real-time data acquisition sub-module is used to acquire real-time protocol data sent through the activeMQ middleware if there is a key-value pair.

The second data mapping sub-module is used to filter the real-time protocol data according to the calculation factor to obtain matching node data.

Further, the data mapping module 908 includes:

The code obtaining sub-module is used to obtain the protocol code and channel code of the protocol data obtained by screening.

The index code generation sub-module is used to generate the index code of the filtered protocol data according to the protocol code and the channel code.

The coding mapping sub-module is used to map the filtered protocol data to the corresponding data node through index coding.

Further, the object data acquisition module 912 includes:

The request receiving sub-module is used to receive a data capture request, where the data capture request includes a specified data type.

The object data acquisition sub-module is used to obtain the matching node data from the data node according to the specified data type, as the object matching data.

Further, the data matching module 914 includes:

The attribute value obtaining sub-module is used to obtain at least one attribute value of the specified data type from the object matching data.

The object acquisition sub-module is used to acquire user objects from the object database, where the user objects include attribute mappings that have the same attributes as the attribute values.

The product push sub-module is used to target the user corresponding to the attribute mapping that matches the attribute value as the target push object.

The above-mentioned product push device obtains the calculation factor of user screening protocol data according to the protocol data of the specified user group, filters the protocol data, selects the appropriate protocol data and inserts it under the corresponding data node, in a targeted manner Obtain the node attributes of the preset protocol data of a certain type of object, and generate calculation factors for the node attributes to filter the protocol data, and also specify the target push information by specifying the data type and select the object from the protocol data to filter the data, which improves the mass Drift bottle data processing capabilities, so that the data can be efficiently filtered and sorted for accurate push, and the data used during the period are real purchase cases, and it is more convincing to recommend to relevant users.

In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. 10. The computer equipment includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor of the computer device is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer equipment is used to store user order data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a product push method is realized. The calculation factor of the user screening protocol data is obtained according to the protocol data of the specified user group, and the protocol data is screened, and the appropriate protocol data is selected and inserted Go to the corresponding data node, obtain the node attributes of the preset protocol data of a certain type of object in a targeted manner, and generate calculation factors for the node attributes to filter the protocol data, and specify the target push information and slave protocol by specifying the data type. Selecting objects in the data to filter data improves the ability to process massive amounts of drifting bottle data, so that the data can be efficiently filtered and sorted for accurate push, and the data used during the period are real purchase cases, and it is more persuasive to recommend to relevant users .

Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculation and/or information processing in accordance with pre-set or stored instructions. Its hardware includes, but is not limited to, a microprocessor, a dedicated Integrated Circuit (Application Specific Integrated Circuit, ASIC), Programmable Gate Array (Field-Programmable Gate Array, FPGA), Digital Processor (Digital Signal Processor, DSP), embedded equipment, etc.

Those skilled in the art can understand that the structure shown in FIG. 10 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may Including more or fewer parts than shown in the figure, or combining some parts, or having a different arrangement of parts.

In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored on the memory and running on the processor. The processor executes the computer program to implement the steps of the product push method in the above embodiment For example, step 202 to step 216 shown in FIG. 2, or when the processor executes the computer program, the function of each module/unit of the product pushing device in the above embodiment is realized, for example, the function of module 902 to module 916 shown in FIG. 9. To avoid repetition, I won’t repeat them here.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the product pushing method in the above-mentioned embodiment are implemented, for example, steps 202 to 2 shown in FIG. Step 216, or, when the processor executes the computer program, the function of each module/unit of the product pushing device in the above-mentioned embodiment is realized. For example, the functions of modules 902 to 916 shown in FIG. The user filters the calculation factors of the protocol data, and filters the protocol data, selects the appropriate protocol data and inserts it under the corresponding data node, and obtains the node attributes of the preset protocol data of a certain type of object in a targeted manner, and Generate calculation factors for node attributes to filter protocol data, and also specify data types to specify targets to push information and select objects from protocol data to filter data, which improves the ability to process massive amounts of drifting bottle data so that data can be efficiently filtered and sorted. Accurate push, and the data used during the period are real purchase cases, it is more convincing to recommend to relevant users.

A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage. In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database, or other media used in the embodiments provided in this application may include non-volatile and/or volatile memory. Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not a limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Channel (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

Those skilled in the art can clearly understand that, for the convenience and conciseness of description, only the division of the above functional units and modules is used as an example. In practical applications, the above functions can be allocated to different functional units and modules as needed. Module completion, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above.

The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be It is considered as the range described in this specification.

The above-mentioned embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it should not be understood as a limitation on the scope of the invention patent. It should be pointed out that for those of ordinary skill in the art, without departing from the concept of this application, several modifications, improvements, or equivalent substitutions of some technical features can be made, and these modifications or substitutions are not To make the essence of the same technical solution deviate from the spirit and scope of the technical solutions of the embodiments of this application belongs to the protection scope of this application. Therefore, the scope of protection of the patent of this application shall be subject to the appended claims.

Claims

A product push method, the method includes:

Acquiring preset protocol data, and extracting node attributes of the preset protocol data;

Generating a calculation factor for screening protocol data according to the node attribute;

Obtain protocol data, and filter the protocol data by the calculation factor to obtain matching node data;

Mapping the matching node data to the corresponding data node;

Generating product push information according to the data node;

Acquiring a designated data type, and acquiring matching node data from the data node according to the designated data type as object matching data;

Acquiring the target push object from the object database according to the object matching data, and filtering the product push information according to the specified data type to obtain the target push information;

Push the target push information to the target push object.
The method according to claim 1, wherein said generating a calculation factor for screening protocol data according to said node attribute comprises:

Performing vectorization processing on the node attributes by one-hot algorithm to obtain vectorized attributes;

Calculating the Manhattan distance between each of the vectorized attributes;

If the number of vectorized attributes whose Manhattan distance is less than the preset distance is greater than a preset value, set the vectorized attributes whose Manhattan distance is less than the preset distance as a core attribute cluster;

The core attribute clusters are integrated to obtain the calculation factor.
The method according to claim 1, wherein said acquiring protocol data and filtering said protocol data by said calculation factor to obtain matching node data comprises:

Obtain historical protocol data and store it in HDFS;

Analyze the historical protocol data through MapReduce to obtain historical analysis results;

If the historical analysis result is that the historical protocol data meets the calculation factor, detecting whether there is a key-value pair between the historical protocol data and the data node;

If the key-value pair does not exist, map the historical protocol data to the corresponding data node as screening data;

The real-time protocol data sent through the activeMQ message middleware is screened according to the screening data, and the real-time protocol data that meets the screening data is mapped to the corresponding data node as matching node data.
The method according to claim 3, wherein said acquiring protocol data and filtering said protocol data by said calculation factor to obtain matching node data further comprises:

If the key-value pair exists, obtain the real-time protocol data sent through the activeMQ middleware;

The real-time protocol data is screened according to the calculation factor to obtain the matching node data.
The method according to claim 1, wherein the mapping the filtered protocol data to the corresponding data node comprises:

Obtain the protocol code and channel code of the filtered protocol data;

Generating an index code of the filtered protocol data according to the protocol code and the channel code;

Map the filtered protocol data to the corresponding data node through the index coding.
The method according to claim 1, wherein said obtaining a specified data type, and obtaining matching node data from the data node according to the specified data type, as object matching data, comprises:

Receiving a data capture request, wherein the data capture request includes the specified data type;

Obtain the matching node data from the data node according to the specified data type as the object matching data.
The method according to claim 1, wherein the obtaining the target push object from the object database according to the object matching data comprises:

Obtaining at least one attribute value of the specified data type from the object matching data;

Obtaining a user object from the object database, wherein the user object includes an attribute map having the same attribute as the attribute value;

The user corresponding to the attribute mapping that matches the attribute value is used as the target push object.
A product push device, including:

The attribute extraction module is used to obtain preset protocol data, and extract the node attributes of the preset protocol data;

A factor generating module, configured to generate a calculation factor for screening protocol data according to the node attribute;

The data screening module is used to obtain protocol data, and filter the protocol data by the calculation factor to obtain matching node data;

A data mapping module, configured to map the matching node data to the corresponding data node;

An information generation module, which is used to generate product push information according to the data node;

The object data acquisition module is configured to acquire a specified data type, and obtain matching node data from the data node according to the specified data type, as object matching data;

The data matching module is configured to obtain the target push object from the object database according to the object matching data, and filter the product push information according to the specified data type to obtain the target push information;

The information push module is used to push the target push information to the target push object.
A computer device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the steps of the product push method described below are implemented:

Acquiring preset protocol data, and extracting node attributes of the preset protocol data;

Generating a calculation factor for screening protocol data according to the node attribute;

Obtain protocol data, and filter the protocol data by the calculation factor to obtain matching node data;

Mapping the matching node data to the corresponding data node;

Generating product push information according to the data node;

Acquiring a designated data type, and acquiring matching node data from the data node according to the designated data type as object matching data;

Acquiring the target push object from the object database according to the object matching data, and filtering the product push information according to the specified data type to obtain the target push information;

Push the target push information to the target push object.
The computer device according to claim 9, wherein said generating a calculation factor for filtering protocol data according to said node attribute comprises:

Performing vectorization processing on the node attributes by one-hot algorithm to obtain vectorized attributes;

Calculating the Manhattan distance between each of the vectorized attributes;

If the number of vectorized attributes whose Manhattan distance is less than the preset distance is greater than a preset value, set the vectorized attributes whose Manhattan distance is less than the preset distance as a core attribute cluster;

The core attribute clusters are integrated to obtain the calculation factor.
The computer device according to claim 9, wherein said acquiring protocol data and filtering said protocol data by said calculation factor to obtain matching node data comprises:

Obtain historical protocol data and store it in HDFS;

Analyze the historical protocol data through MapReduce to obtain historical analysis results;

If the historical analysis result is that the historical protocol data meets the calculation factor, detecting whether there is a key-value pair between the historical protocol data and the data node;

If the key-value pair does not exist, map the historical protocol data to the corresponding data node as screening data;

The real-time protocol data sent through the activeMQ message middleware is screened according to the screening data, and the real-time protocol data that meets the screening data is mapped to the corresponding data node as matching node data.
The computer device according to claim 11, wherein said acquiring protocol data and filtering said protocol data by said calculation factor to obtain matching node data further comprises:

If the key-value pair exists, obtain the real-time protocol data sent through the activeMQ middleware;

The real-time protocol data is screened according to the calculation factor to obtain the matching node data.
The computer device according to claim 9, wherein the mapping the filtered protocol data to the corresponding data node comprises:

Obtain the protocol code and channel code of the filtered protocol data;

Generating an index code of the filtered protocol data according to the protocol code and the channel code;

Map the filtered protocol data to the corresponding data node through the index coding.
8. The computer device according to claim 9, wherein said obtaining a specified data type and obtaining matching node data from the data node according to the specified data type as object matching data comprises:

Receiving a data capture request, wherein the data capture request includes the specified data type;

Obtain the matching node data from the data node according to the specified data type as the object matching data.
The computer device according to claim 9, wherein said obtaining the target push object from the object database according to the object matching data comprises:

Obtaining at least one attribute value of the specified data type from the object matching data;

Obtaining a user object from the object database, wherein the user object includes an attribute map having the same attribute as the attribute value;

The user corresponding to the attribute mapping that matches the attribute value is used as the target push object.
A computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the product pushing method described below are realized:

Acquiring preset protocol data, and extracting node attributes of the preset protocol data;

Generating a calculation factor for screening protocol data according to the node attribute;

Obtain protocol data, and filter the protocol data by the calculation factor to obtain matching node data;

Mapping the matching node data to the corresponding data node;

Generating product push information according to the data node;

Acquiring a designated data type, and acquiring matching node data from the data node according to the designated data type as object matching data;

Acquiring the target push object from the object database according to the object matching data, and filtering the product push information according to the specified data type to obtain the target push information;

Push the target push information to the target push object.
The computer-readable storage medium according to claim 16, wherein said generating a calculation factor for filtering protocol data according to said node attribute comprises:

Performing vectorization processing on the node attributes by one-hot algorithm to obtain vectorized attributes;

Calculating the Manhattan distance between each of the vectorized attributes;

If the number of vectorized attributes whose Manhattan distance is less than the preset distance is greater than a preset value, set the vectorized attributes whose Manhattan distance is less than the preset distance as a core attribute cluster;

The core attribute clusters are integrated to obtain the calculation factor.
The computer-readable storage medium according to claim 16, wherein said acquiring protocol data and filtering said protocol data by said calculation factor to obtain matching node data comprises:

Obtain historical protocol data and store it in HDFS;

Analyze the historical protocol data through MapReduce to obtain historical analysis results;

If the historical analysis result is that the historical protocol data meets the calculation factor, detecting whether there is a key-value pair between the historical protocol data and the data node;

If the key-value pair does not exist, map the historical protocol data to the corresponding data node as screening data;

The real-time protocol data sent through the activeMQ message middleware is screened according to the screening data, and the real-time protocol data that meets the screening data is mapped to the corresponding data node as matching node data.
18. The computer-readable storage medium according to claim 18, wherein said acquiring protocol data and filtering said protocol data by said calculation factor to obtain matching node data further comprises:

If the key-value pair exists, obtain the real-time protocol data sent through the activeMQ middleware;

The real-time protocol data is screened according to the calculation factor to obtain the matching node data.
The computer-readable storage medium according to claim 16, wherein the mapping the filtered protocol data to the corresponding data node comprises:

Obtain the protocol code and channel code of the filtered protocol data;

Generating an index code of the filtered protocol data according to the protocol code and the channel code;

Map the filtered protocol data to the corresponding data node through the index coding.