WO2020037917A1

WO2020037917A1 - User behavior data recommendation method, server and computer readable medium

Info

Publication number: WO2020037917A1
Application number: PCT/CN2018/123508
Authority: WO
Inventors: 王翼
Original assignee: 平安科技（深圳）有限公司
Priority date: 2018-08-22
Filing date: 2018-12-25
Publication date: 2020-02-27
Also published as: CN109242553A

Abstract

Disclosed in the embodiments of the present application are a user behavior data recommendation method, a server and a computer readable medium, relating to the analysis and sorting of user behavior data and implementing intelligent recommendation of user behavior data. The method comprises: reading user behavior information, and performing data cleaning on the user behavior information, so as to generate user behavior data; performing content recognition processing on the user behavior data to form a rule library, the rule library being configured to classify and store the user behavior data; extracting recommendation data from the rule library according to conditions of a front-end recommendation system, and pushing the recommendation data to the front-end recommendation system. The embodiments of the present application enable real-time and accurate recommendation when recommending user behavior data, so that no dirty data exists in the recommended user behavior data.

Description

User behavior data recommendation method, server and computer-readable medium

This application claims the priority of a Chinese patent application filed on August 22, 2018 with the Chinese Patent Office, application number 2018109655825, and application name "A Method for Recommending User Behavior Data, Servers, and Computer-readable Media", the entire contents of which Incorporated by reference in this application.

Technical field

The present application relates to the field of data analysis technology, and in particular, to a method for recommending user behavior data, a server, and a computer-readable medium.

Background technique

With the rapid development of information technology, information has shown an explosive expansion, and people's access to information has become more diverse and more convenient. At the same time, the timeliness of information has become higher and higher. For example, if a user bought a piece of clothing on Taobao yesterday and wants to buy a pair of swimming goggles to swim today, but finds that the system spares no effort to recommend pants and clothes to him, he simply ignores his behavior of looking for swimming goggles today. . This is because the system recommendation is based on the behavior track of the user yesterday to make recommendations to the user. That is to say, the existing recommendation algorithm is to capture the user's historical trajectory, and after algorithm analysis, the user's future trend can be analyzed after T + 1 days, resulting in inaccurate recommendation.

Summary of the Invention

The embodiment of the present application provides a method for recommending user behavior data, which can implement accurate real-time recommendation of user behavior data, and make the recommended user behavior data free of dirty data.

In a first aspect, an embodiment of the present application provides a method for recommending user behavior data. The method includes:

Reading user behavior information, and performing data cleaning on the user behavior information to generate user behavior data;

Performing content recognition processing on the user behavior data to form a rule base, where the rule base is used to classify and store the user behavior data;

Extract the qualified recommendation data from the rule base according to the conditions of the front-end recommendation system, and push the recommendation data to the front-end recommendation system.

In a second aspect, an embodiment of the present application provides a server, where the server includes:

A data cleaning unit, configured to read user behavior information, and perform data cleaning on the user behavior information to generate user behavior data;

A rule base generating unit, configured to perform content recognition processing on the user behavior data to form a rule base, where the rule base is used to classify and store the user behavior data;

An extraction unit, configured to extract recommendation data from the rule base according to conditions of a front-end recommendation system;

A recommendation unit, configured to push the recommendation data to a front-end recommendation system.

In a third aspect, an embodiment of the present application provides another server, including a processor, a memory, and a communication module. The memory is used to store program code, and the processor is used to call the program code to execute the first aspect. And any of its alternatives.

In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium. The computer storage medium stores a computer program, where the computer program includes program instructions, and the program instructions cause the processing when executed by a processor. The processor executes the method of the first aspect and any one of its optional ways.

In the embodiment of the present application, user behavior information is obtained in real time, and then the obtained user behavior information is subjected to data cleaning to obtain user behavior data without dirty data, and then content identification is performed on the user behavior data, and the classified To the rule base; finally, extract the recommendation data required by the front-end recommendation system from the above rule base according to the conditions of the front-end recommendation system, and push the recommendation data to the corresponding recommendation system, so that the front-end recommendation system obtains the recommendation system And there is no recommendation data of dirty data, so that the front-end recommendation system realizes accurate and real-time recommendation.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to explain the technical solution of the embodiment of the present application more clearly, the drawings used in the description of the embodiment will be briefly introduced below.

FIG. 1 is a schematic flowchart of a user behavior data recommendation method according to an embodiment of the present application; FIG.

2 is a schematic flowchart of another user behavior data recommendation method according to an embodiment of the present application;

3 is a schematic block diagram of a server according to an embodiment of the present application;

4 is a device provided by an embodiment of the present application;

FIG. 5 is a schematic structural diagram of a server according to an embodiment of the present application.

detailed description

In the following, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

It should be understood that when used in this specification and the appended claims, the terms "including" and "comprising" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude one or The presence or addition of a number of other features, wholes, steps, operations, elements, components, and / or sets thereof.

It should also be understood that the terminology used in the description of this application is for the purpose of describing particular embodiments only and is not intended to limit the application. As used in this specification and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

It should be further understood that the term "and / or" used in the specification of the application and the appended claims refers to any combination of one or more of the items listed in association and all possible combinations, and includes these combinations .

Referring to FIG. 1, FIG. 1 is a schematic flowchart of a method for recommending user behavior data according to an embodiment of the present application. As shown in the figure, the method may include:

101: Read user behavior information, and perform data cleaning on the user behavior information to generate user behavior data.

In the embodiment of the present application, the above-mentioned user behavior is an event consisting of time, place, person, interaction, and interactive content. For example, a user search is an event, at what time, on which platform, and which Internet protocol (Internet Protocol (IP) address, what was searched, and what was searched for. This is a complete event and a definition of user behavior; we can define tens of thousands of such events in a website or an application (Application, APP). With such events, you can connect user behavior and observe. A user is a new user after entering the website for the first time, and he may want to register, so the registration behavior is also an event. Registration requires filling in personal information, after which he may start searching for purchases, all of which is user behavior information.

When user behavior information needs to be analyzed and processed in real time, in order to obtain useful user behavior data, and make corresponding recommendations to users based on the user's behavior data. In the embodiment of the present application, the user behavior data is obtained by reading the user behavior information of the user in real time, and then performing data cleaning processing on the read user behavior information. The above-mentioned data cleaning processing on the read user behavior information may include operations such as error data cleaning, missing value data cleaning, repeated value data cleaning, or inconsistent data cleaning on the user behavior information.

Specifically, the data acquisition component spout of the storm framework, a real-time processing framework for data streams, is used to pull user behavior information data, and then spout distributes the pulled user behavior information data to the data processor bolt in storm according to preset rules. To perform various processing on user behavior data, such as filtering and cleaning user behavior information. Because there may be many dirty data in the user behavior data pulled by the above spout, for example, there may be some duplicate data, erroneous data, and incomplete data in the user behavior data. Therefore, after the user behavior information data is obtained, the user behavior information data is first subjected to a cleaning process. Specifically, the foregoing cleaning of user behavior information data includes cleaning of erroneous data, cleaning of missing values, cleaning of duplicate values, and cleaning of inconsistent data. After the user behavior information data is cleaned, the cleaned user behavior information data is passed to the next bolt component to continue processing.

Among them, the above storm is a free open source, distributed, highly fault-tolerant real-time processing framework. Storm supports creating topologies to transform data streams without end points. Storm is often used in the fields of real-time analysis, online machine learning, continuous computing, distributed remote calls, and so on. The above spout is a component that generates a source data stream in storm. Generally, spout reads data from an external data source and then converts it into internal source data. The above bolt is a component that performs data processing in Storm, and can perform any operation such as filtering, function operations, merging, and writing to a database.

As an optional implementation manner, before the user behavior information is read, the user behavior needs to be tracked, that is, the user behavior information of the user is collected. The collection of user behavior of the user may include collecting user behavior information based on a web server log or a client. Among them, the method of collecting user behavior information based on the Web server logs is relatively common. The log file is automatically generated by the web server, and the cost is small. It is relatively easy to develop a data analysis tool based on the log files. The technical method directly obtains the behavior data of the user's interaction with the website from the client. Collecting user behavior information data from the client can reduce human interference factors, and the obtained data is more true and accurate, which solves the shortcomings of server-side collection and reduces Consumption of servers. After the user behavior information is collected, the user behavior information is sent to a message queue, so that the real-time processing framework strom pulls the user behavior information from the message queue.

Tracker system is a set of professional tooling and intelligent database management system of tools, fixtures and measuring tools. It can manage the overall process of tooling, tools, fixtures and measuring tools in the production process of the enterprise, and track tooling and toolholder measuring tools in real time The process of purchasing, loading and unloading, repairing, scrapping, and calibration helps warehouse managers, craftsmen, manufacturing engineers, tooling and tool holder gauge supervisors, etc. to more effectively improve the tool management process and reduce production costs.

As an optional implementation manner, a Tracker system may be adopted to collect the foregoing user behavior information. When a website or app reaches a certain number of users, a tracker system is generally required to collect user behavior (such as user IP address, page source, city name, browser version, button location, etc.), page access performance, and abnormal errors. , And then report to the log server according to a certain policy. Search, recommendation, advertising center and other development teams analyze these logs to adjust and develop various functions; product managers, senior management, etc. use these logs to optimize operations and make correct decisions in a timely manner. The Tracker system plays an important role in a mature application. With the development of the business, the real-time requirements for it are getting higher and higher.

The Tracker system supports automatic dotted fields, automatic extension fields, and so on. The Tracker system's Application Programming Interface (API) is embedded in the event of each page of the website or application. A certain policy is set to send to the log server, and then synchronized. To the message queue as a data buffer. Pull messages from the message queue through the Storm framework, complete related filtering and calculations, and finally store them in the database (for example, Hbase, MYSQL, etc.). Among them, Hbase is a distributed, column-oriented open source database; MYSQL is a relational database.

User access will continuously generate data, which is either stored locally and sent to related applications when needed, or stored in a unified central storage area. The generated data will be captured, filtered, and processed by the Spout in Storm (such as protocol analysis, format analysis, data verification, etc.) between applications, and then sent to Bolt for data analysis. Finally, the available data is formed and stored in In a persistent medium (such as a DB) for other applications to obtain.

102: Perform content recognition processing on the user behavior data to form a rule base, where the rule base is used to classify and store the user behavior data.

In the embodiment of the present application, after the user behavior information is subjected to cleaning processing, content recognition processing is performed on the cleaned user behavior information, so as to classify the user behavior information according to the specific content of the user behavior information and a preset rule, and then The user behavior information is stored in a database into a rule base according to the classification of preset rules.

Specifically, the classifying the user behavior information according to the specific content and preset rules of the user behavior information may include: identifying the cleaned user behavior information, and classifying the cleaned user behavior information according to browsing behavior, click behavior, The input or search behavior is classified, and then sent to the corresponding bolt for processing (such as structured processing). After the corresponding bolt of the next layer above processes the user behavior information, the processed user behavior information is persistently stored in a database to form a rule base.

For example, there is a browsing behavior in the cleaned user behavior information, and the browsing behavior is that user A browses a movie-related webpage. After the content of the browsing user behavior information is identified, it is sent to the corresponding bolt that handles the browsing behavior. The bolt is structured according to the content of the browsing behavior. Specifically, the browsing behavior is described as follows: user, URL, The format of the theme, category, author, director, starring, release year is structured and then stored in the above rule base.

103: Extract recommendation data from the rule base according to the conditions of the front-end recommendation system, and push the recommendation data to the front-end recommendation system.

In the embodiment of the present application, after a large amount of user behavior information is stored in the rule base, a trajectory enhancement algorithm is used from the rule base to extract the eligible data from the rule base according to the conditions of different front-end recommendation systems, and then The extracted data is pushed to the corresponding front-end recommendation system.

Among them, because different front-end recommendation systems recommend different content to users, different front-end recommendation systems need to extract the recommendation data from the above-mentioned rule base; therefore, it is necessary to obtain from the above-mentioned rule base according to the conditions of different front-end recommendation systems. Extract recommendation data that satisfies the recommendation system. For example, the condition of a front-end recommendation system is the top 10 URLs in terms of total traffic, then the above recommendation data is the current top 10 URL information calculated from the rule base according to the algorithm.

Specifically, the function of the above-mentioned trajectory enhancement algorithm is to extract a large amount of data from the rule base, which includes the Uniform Resource Locator (URL), the access traffic, and the keyword information of the visit. Algorithm processing, and then root the front-end recommendation system conditions, such as merge the same URLs together, merge the upstream traffic and downstream traffic and sort them, and then sort the URLs based on the traffic summary, taking the first 80% of the total traffic, because The URL of 80% of the traffic is a URL that users often visit. The data in the rule base is further processed and saved to the database. The front-end system recommends these data to the user from the database.

Optionally, in the embodiment of the present application, after the qualified data is extracted from the above rule base and pushed to the corresponding front-end system, the already extracted data is stored in the historical behavior trajectory database, and At the same time, the extracted data in the rule base is deleted to save the storage space of the rule base.

It can be seen that, in the embodiment of the present application, user behavior information is obtained in real time, and then the obtained user behavior information is subjected to data cleaning to obtain user behavior data without dirty data, and then content identification is performed on the user behavior data. And sort the classification into the rule base; finally, according to the conditions of the front-end recommendation system, extract the recommendation data required by the front-end recommendation system from the above rule base, and push the recommendation data to the corresponding recommendation system, so that the front-end recommendation system obtains the Recommendation data required by the recommendation system without dirty data, so that the front-end recommendation system implements real-time accurate recommendations.

Referring to FIG. 2, FIG. 2 is a schematic flowchart of another user behavior data recommendation method according to an embodiment of the present application. As shown in the figure, the method may include:

201: Obtain user behavior information of a user from a web log file or from a user terminal.

202: Send the above user behavior information to a message queue.

In the embodiment of the present application, after the user behavior information is collected, the above user behavior information is sent to a message queue serving as a buffer area, for example, a message queue such as Kafka or MetaQ, so that the spout of the subsequent storm framework is removed from the message queue. Get the above user behavior information. Among them, Kafka is a high-throughput distributed publish and subscribe message system, which can process all action flow data in consumer-scale websites; MetaQ is a complete queue model message middleware, and the server is written in Java. Can be deployed on multiple software and hardware platforms.

203: The data acquisition component spout of the storm real-time processing framework reads the user behavior information from the message queue, and then distributes the user behavior information to the processing component bolt.

In the embodiment of the present application, after the obtaining component spout reads the user behavior information from the message queue, the user behavior information is distributed to different bolts for processing according to the type of the obtained user behavior information.

204: The processing component bolt of the real-time processing framework performs error data cleaning, missing value data cleaning, repeated value data cleaning, or inconsistent data cleaning on the user behavior information to obtain user behavior data.

In the embodiment of the present application, there may be a lot of dirty data in the user behavior data pulled by the spout, for example, there may be some duplicate data, erroneous data, and incomplete data in the user behavior data. Therefore, after the user behavior information data is obtained, the user behavior information data is first subjected to a cleaning process. Specifically, the foregoing cleaning of user behavior information data includes cleaning of erroneous data, cleaning of missing values, cleaning of duplicate values, and cleaning of inconsistent data. After the above user behavior information data is cleaned, the cleaned user behavior information data is passed to the next bolt to continue processing.

205: After the user behavior data is subjected to content recognition processing, the user behavior data is classified according to browsing behavior, click behavior, input behavior, or search behavior.

In the embodiment of the present application, after the user behavior information is subjected to cleaning processing, content recognition processing is performed on the cleaned user behavior information, so as to classify the user behavior information according to the specific content of the user behavior information and a preset rule, and then The user behavior information is stored in the database according to the classification of the preset rules to form a rule base.

Specifically, the classifying the user behavior information according to the specific content and preset rules of the user behavior information may include: identifying the cleaned user behavior information, and classifying the cleaned user behavior information according to browsing behavior, click behavior, The input behavior (or search behavior) is classified, and then sent to the corresponding bolt for processing (such as structured processing).

206: The classified user behavior data is structured and stored to form a rule base.

In the embodiment of the present application, after the corresponding bolt processes the user behavior information, the processed user behavior information is persistently stored in a database to form a rule base.

207: Use a trajectory enhancement algorithm to extract qualified recommendation data from the rule base according to the conditions of different front-end recommendation systems from the rule base, and push the recommendation data to the corresponding front-end recommendation system.

Specifically, the function of the above trajectory enhancement algorithm is to extract a large amount of data from the rule base, which includes the accessed URL and access traffic, the keyword information of the access, etc., and then processed by the trajectory enhancement algorithm, and then based on certain conditions, such as The same URLs are merged together, and the upstream traffic and downstream traffic that are accessed are merged and sorted, and then the URLs are sorted by the traffic summary, taking the first 80% of the total traffic, because the URLs of 80% of the traffic are URLs that users often visit, The data in the rule base is further processed and saved to the database, and the front-end system recommends these data to the user from the database.

208: Store the above recommendation data in a historical behavior trajectory database.

In the embodiment of the present application, the previously recommended user behavior data may be analyzed and processed on the user ’s historical behavior in the future. Therefore, after the user behavior data is recommended to the corresponding front-end recommendation system, the above recommendation is recommended. The data is stored in the historical behavior trajectory database, so that the above-mentioned user behavior data is subsequently obtained from the historical behavior trajectory database for analysis and processing.

209: Delete the above recommended data in the above rule base.

In the embodiment of the present application, in order to save the storage space of the rule base, after the user behavior data of the rule base is extracted, the extracted user behavior data is deleted from the above rule base.

It can be seen that the embodiment of the present application collects user behavior information of a user through a web log file or a user terminal, and then sends the user behavior information to a message queue, and then reads the user behavior information from the message queue, and analyzes the user behavior. The information is subjected to data cleaning to generate user behavior data, so as to delete duplicate data, erroneous data, and complete incomplete data in the user behavior information; then, perform content identification, classification and sorting on the above user behavior data to form a rule base; Extract the qualified recommendation data from the above rule base and push the above recommendation data to the front-end recommendation system. Through the embodiments of the present application, when recommending user behavior data, real-time and accurate recommendation can be achieved, and the recommended user behavior data does not contain dirty data.

An embodiment of the present application further provides a server, which is configured to execute a unit of any one of the foregoing methods. Specifically, referring to FIG. 3, FIG. 3 is a schematic block diagram of a server provided by an embodiment of the present application. The server in this embodiment includes a data cleaning unit 310, a rule base generation unit 320, an extraction unit 330, and a recommendation unit 340.

The data cleaning unit 310 is configured to read user behavior information, and perform data cleaning on the user behavior information to generate user behavior data.

A rule base generating unit 320 is configured to perform content recognition processing on the user behavior data to form a rule base, and the rule base is used to classify and store the user behavior data;

An extraction unit 330, configured to extract qualified recommendation data from the above rule base according to the conditions of the front-end recommendation system;

The recommendation unit 340 is configured to push the above recommendation data to a front-end recommendation system.

It can be seen that the embodiment of the present application generates user behavior data by reading the user behavior information and performing data cleaning on the above user behavior information, so as to delete duplicate data, erroneous data, and complete incomplete data in the user behavior information; then , Performing content recognition processing on the user behavior data to form a rule base; extracting qualified recommendation data from the rule base, and pushing the recommendation data to a front-end recommendation system. Through the embodiments of the present application, when recommending user behavior data, real-time and accurate recommendation can be achieved, and the recommended user behavior data does not contain dirty data.

Optionally, the server further includes:

An obtaining unit 350, configured to obtain user behavior information from a web log file, or directly obtain the user behavior information from a user terminal;

The sending unit 360 is configured to send the user behavior information to a message queue.

Optionally, the data cleaning includes: erroneous data cleaning, missing value data cleaning, repeated value data cleaning, or inconsistent data cleaning.

Optionally, the data cleaning unit 310 includes:

The reading unit 311 is configured to read the user behavior information from the message queue through the data acquisition component spout of the storm real-time processing framework, and then distribute the user behavior information to the processing component bolt;

A distribution unit 312, configured to distribute the user behavior information to a processing component of the real-time processing framework;

The cleaning unit 313 is configured to use the processing component bolt of the real-time processing framework to perform error data cleaning, missing value data cleaning, repeated value data cleaning, or inconsistent data cleaning on the user behavior information to obtain user behavior data.

Optionally, the reading unit is configured to pull the user behavior information from the message queue through a spout component of the Storm framework;

The distribution unit is configured to distribute the user behavior information to a Bolt component of a Storm framework.

Optionally, the rule base generating unit 320 includes:

A classification unit 321, configured to classify the user behavior data according to a browsing behavior, a click behavior, an input behavior, or a search behavior after the user behavior data performs content recognition processing;

A processing unit 322, configured to process the classified user behavior data in a structured manner;

The first storage unit 323 stores the user behavior data that is structured and processed to form the rule base.

Optionally, the above-mentioned extraction unit 330 is configured to use a trajectory enhancement algorithm to extract the recommendation data that meets the conditions from the rule base according to the conditions of different front-end recommendation systems from the rule base;

The recommendation unit 340 is configured to push the recommendation data to a corresponding front-end recommendation system.

Optionally, the conditions of the front-end recommendation system may include merging the same URLs together, merging and sorting the upstream and downstream traffic visited, and then sorting the URLs based on the traffic summary, and taking the previous preset of the total traffic Percent URL.

Optionally, the server further includes:

A second storage unit 370, configured to store the above-mentioned recommendation data in a historical behavior trajectory database;

The deleting unit 380 is configured to delete the recommended data in the rule base.

Referring to FIG. 4, FIG. 4 is a device provided by an embodiment of the present application. The device may be a server. As shown in FIG. 4, the device includes: one or more processors 401; one or more input devices 402, and one or more Output devices 403 and memory 404. The processor 401, the input device 402, the output device 403, and the memory 404 are connected through a bus 405. The memory 402 is configured to store instructions, and the processor 401 is configured to execute the instructions stored in the memory 402.

Where the device is used as a server, the processor 401 is configured to: read user behavior information, perform data cleaning on the user behavior information to generate user behavior data; perform content recognition processing on the user behavior data to form a rule base, The rule base is used to classify and store the user behavior data; extract recommendation data from the rule base according to the conditions of the front-end recommendation system, and push the recommendation data to the front-end recommendation system.

It should be understood that, in the embodiment of the present application, the processor 401 may be a central processing unit (CPU), and the processor may also be another general-purpose processor or a digital signal processor (DSP). Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The input device 402 may include a touchpad, a fingerprint sensor (for collecting user's fingerprint information and fingerprint orientation information), a microphone, etc., and the output device 403 may include a display (for example, a Liquid Crystal Display (LCD), etc.), Speakers, etc.

The memory 404 may include a read-only memory and a random access memory, and provide instructions and data to the processor 401. A portion of the memory 404 may also include non-volatile random access memory. For example, the memory 404 may also store device type information.

In specific implementation, the processor 401, the input device 402, and the output device 403 described in the embodiments of the present application may execute the implementation manner described in the first embodiment of a user behavior data recommendation method provided in the embodiments of the present application, The implementation manners in the second embodiment and the third embodiment may also implement the server implementation manners described in the embodiments of the present application, and details are not described herein again.

In another embodiment of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and the computer program is implemented when a processor executes: reading user behavior information, and reading the user behavior information. Perform data cleaning to generate user behavior data; perform content recognition processing on the user behavior data to form a rule base, the rule base is used to classify and store the user behavior data; and extract recommendation data from the rule base according to the conditions of the front-end recommendation system , Push the above recommendation data to the front-end recommendation system.

The computer-readable storage medium may be an internal storage unit of the terminal described in any one of the foregoing embodiments, such as a hard disk or a memory of the terminal. The computer-readable storage medium may also be an external storage device of the terminal, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, and a flash memory card provided on the terminal. (Flash Card), etc. Further, the computer-readable storage medium may further include both an internal storage unit of the terminal and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by the terminal. The computer-readable storage medium described above may also be used to temporarily store data that has been or will be output.

FIG. 5 is a schematic diagram of a server structure provided by an embodiment of the present application. The server 500 may have a large difference due to different configurations or performance, and may include one or more central processing units (CPUs) 522 (for example, , One or more processors) and a memory 532, one or more storage media 530 (eg, one or more storage devices) storing application programs 542 or data 544. The memory 532 and the storage medium 530 may be temporary storage or persistent storage. The program stored in the storage medium 530 may include one or more modules (not shown), and each module may include a series of instruction operations on the server. Furthermore, the central processing unit 522 may be configured to communicate with the storage medium 530, and execute a series of instruction operations in the storage medium 530 on the server 500.

The server 500 may also include one or more power sources 526, one or more wired or wireless network interfaces 550, one or more input-output interfaces 558, and / or, one or more operating systems 541, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM and more.

The steps performed by the server in the above embodiment may be based on the server structure shown in FIG. 5.

Those of ordinary skill in the art may realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the hardware and software, Interchangeability. In the above description, the composition and steps of each example have been described generally in terms of functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the system, server, terminal device, and unit described above can refer to the corresponding processes in the foregoing method embodiments, and are not repeated here .

In the several embodiments provided in this application, it should be understood that the disclosed systems, servers, and methods may be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the above units is only a logical function division. In actual implementation, there may be another division manner. For example, multiple units or components may be combined or may be combined. Integration into another system, or some features can be ignored or not implemented. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may also be electrical, mechanical or other forms of connection.

The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, which may be located in one place, or may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions in the embodiments of the present application.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist separately physically, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional unit.

When the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application is essentially a part that contributes to the existing technology, or all or part of the technical solution may be embodied in the form of a software product, which is stored in a storage medium Included are several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the above method in each embodiment of the present application. The foregoing storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks or optical disks and other media that can store program codes .

The above is only a specific implementation of this application, but the scope of protection of this application is not limited to this. Any person skilled in the art can easily think of various equivalents within the technical scope disclosed in this application. Modifications or replacements, and these modifications or replacements should be covered by the protection scope of this application. Therefore, the protection scope of this application shall be subject to the protection scope of the claims.

Claims

A method for recommending user behavior data includes:

Read user behavior information, and perform data cleaning on the user behavior information to generate user behavior data;

Performing content recognition processing on the user behavior data to form a rule base, where the rule base is used to classify and store the user behavior data;

Extract recommendation data from the rule base according to the conditions of the front-end recommendation system, and push the recommendation data to the front-end recommendation system.
The method according to claim 1, wherein before the reading user behavior information, the method further comprises:

Acquiring the user behavior information from a log file, or directly acquiring the user behavior information from a user terminal;

And sending the user behavior information to a message queue.
The method according to claim 2, wherein the data cleaning comprises: erroneous data cleaning, missing value data cleaning, repeated value data cleaning or inconsistent data cleaning.
The method according to claim 3, wherein the reading user behavior information comprises:

Reading the user behavior information from the message queue by a data acquisition component of the real-time processing framework;

The data cleaning of the user behavior information includes:

Distributing the user behavior information to a processing component of the real-time processing framework;

The processing component of the real-time processing framework performs data cleaning on the user behavior information to obtain user behavior data.
The method according to claim 3, wherein the real-time processing framework is a Storm framework;

The reading, by the data acquisition component of the real-time processing framework, the user behavior information from the message queue includes:

Pull the user behavior information from the message queue through the spout component of the Storm framework;

The processing component that distributes the user behavior information to the real-time processing framework includes:

Distribute the user behavior information to the Bolt component of the Storm framework.
The method according to any one of claims 1-5, wherein the performing a content recognition process on the user behavior data to form a rule base comprises:

After the user behavior data is subjected to content recognition processing, classifying the user behavior data according to browsing behavior, click behavior, input behavior, or search behavior;

Structure the classified user behavior data;

The structured processing of the user behavior data is stored to form the rule base.
The method according to any one of claims 1-5, wherein the extracting qualified recommendation data from the rule base and pushing the recommendation data to a front-end recommendation system comprises:

Using a trajectory enhancement algorithm to extract recommendation data from the rule base from the rule base according to the conditions of different front-end recommendation systems;

The recommendation data is pushed to a corresponding front-end recommendation system.
The method according to claim 7, wherein the conditions of the front-end recommendation system include merging the same URLs together, merging and sorting the upstream and downstream traffic accessed, and then sorting the URLs based on the traffic summary , Take the URL of the previous preset percentage of total traffic.
The method according to claim 8, wherein after the pushing the recommendation data to a corresponding front-end recommendation system, the method further comprises:

Storing the recommendation data in a historical behavior trajectory database;

Delete the recommendation data in the rule base.
A server is characterized in that it includes:

A data cleaning unit, configured to read user behavior information, and perform data cleaning on the user behavior information to generate user behavior data;

A rule base generating unit, configured to perform content recognition processing on the user behavior data to form a rule base, and the rule base is used to classify and store the user behavior data;

An extraction unit, configured to extract qualified recommendation data from the above rule base according to the conditions of the front-end recommendation system;

A recommendation unit is configured to push the above recommendation data to a front-end recommendation system.
The server according to claim 10, wherein the server further comprises:

An obtaining unit, configured to obtain user behavior information from a log file, or directly obtain the user behavior information from a user terminal;

A sending unit is configured to send the user behavior information to a message queue.
The server according to claim 11, wherein the data cleaning comprises: erroneous data cleaning, missing value data cleaning, repeated value data cleaning or inconsistent data cleaning.
The server according to claim 12, wherein the data cleaning unit comprises:

A reading unit, configured to read the user behavior information from the message queue through a data acquisition component of the real-time processing framework, and then distribute the user behavior information to the processing component;

A distribution unit, configured to distribute the user behavior information to a processing component of the real-time processing framework;

The cleaning unit is configured to process data of the user behavior information by a processing component of the real-time processing framework to obtain user behavior data.
The server according to claim 13, wherein the real-time processing framework is a Storm framework;

The reading unit is configured to pull the user behavior information from the message queue through the spout component of the Storm framework;

The distribution unit is configured to distribute the user behavior information to a Bolt component of a Storm framework.
The server according to any one of claims 10 to 14, wherein the rule base generating unit comprises:

A classification unit, configured to classify the user behavior data according to browsing behavior, click behavior, input behavior, or search behavior after performing content identification processing on the user behavior data;

A processing unit for structuredly processing the classified user behavior data;

The first storage unit stores the structured processing of the user behavior data to form the rule base.
The server according to any one of claims 10 to 14, wherein the extraction unit is configured to extract a recommendation from the rule base according to the conditions of different front-end recommendation systems from the rule base by using a trajectory enhancement algorithm. data.
The server according to claim 16, wherein the conditions of the front-end recommendation system include merging the same URLs together, merging and sorting the upstream and downstream traffic accessed, and then sorting the URLs based on the traffic summary , Take the URL of the previous preset percentage of total traffic.
The server according to claim 17, wherein the server further comprises:

A second storage unit 370, configured to store the above-mentioned recommendation data in a historical behavior trajectory database;

The deleting unit 380 is configured to delete the recommended data in the rule base.
A server, characterized in that the server includes a processor, a memory, and a communication module, wherein the memory is used to store program code, and the processor is used to call the program code to execute any one of claims 1-9 The method of one item.
A computer-readable storage medium, characterized in that the computer storage medium stores a computer program, wherein the computer program includes program instructions, and the program instructions, when executed by a processor, cause the processor to execute the program according to claim 1 The method according to any of -9.