CN114861038A - Live broadcast service data processing method and device, equipment and medium thereof - Google Patents

Abstract

The present application relates to the technical field of live broadcast, and discloses a method for processing data of live broadcast service and its device, equipment, and medium. data items of each data source and operation items for performing predefined operations on these data items, the data items include attribute data items of the anchor user, and the operation items include attribute tags; The corresponding data item in the source executes the predefined operation corresponding to the operation item, and determines the data set corresponding to the attribute label in the operation item, and the data set includes the data item sub-data corresponding to the multiple anchor users set; push the corresponding data set of the live broadcast service to the terminal device. The present application realizes the standardized operation of calling data items of multiple data sources by multiple live broadcast services through the data definition template, avoids the increase of data entropy, improves system efficiency, and saves system overhead.

Description

Live broadcast service data processing method and device, equipment and medium thereof

technical field

The present application relates to the technical field of live broadcast, and in particular, to a method for processing live broadcast service data and a corresponding device, computer equipment, and computer-readable storage medium.

Background technique

A webcasting platform that provides webcasting services usually embeds multiple livestreaming services in the access website or application of its platform. Each livestreaming service usually corresponds to an access page, which is used to provide users with data corresponding to the same type of service. set, so that the terminal device can parse out specific data according to the data set for interface display.

In the same platform, on the one hand, the data comes from multiple sources, and the data access protocols are also diversified. Different data originate from different databases or data tables and may need to be called through different protocols; on the other hand, the live broadcast business is also diversified. Yes, different live streaming services require different data items and data processing results. Therefore, with the development of the platform, according to the law of entropy increase, the platform data becomes more and more cluttered, which reduces the operating efficiency and data access efficiency of the platform and affects the user experience.

The more in-depth problem is that the complicated and disordered data will greatly increase the difficulty for new live broadcast services to obtain the data they need, which will further increase the difficulty of developing new services on the platform.

In view of the above aspects, it is necessary to improve the processing method of the data required by the live broadcast service in the live broadcast platform.

SUMMARY OF THE INVENTION

The primary purpose of the present application is to solve at least one of the above problems and provide a method for processing live broadcast service data and a corresponding device, computer equipment and computer-readable storage medium.

In order to meet the various purposes of the application, the application adopts the following technical solutions:

A kind of live broadcast service data processing method proposed to adapt to one of the purposes of this application, comprising the following steps:

Acquire a data definition template corresponding to the live broadcast service, where the data definition template includes data items distributed in multiple data sources that the live broadcast service needs to call and operation items for performing predefined operations on these data items, so The data item includes the attribute data item of the anchor user, and the operation item includes one or more attribute tags;

According to the operation items in the data definition template, the predefined operations corresponding to the operation items are performed on the corresponding data items in the multiple data sources, and the data sets corresponding to the attribute labels in the operation items are determined, The data set includes a subset of data items corresponding to multiple anchor users;

In response to the data calling instruction of the live broadcast service from the terminal device, the data set corresponding to the live broadcast service is pushed to the terminal device.

In some further embodiments, according to the operation items in the data definition template, the predefined operations corresponding to the operation items are performed on the corresponding data items in the multiple data sources, and the attributes corresponding to the operation items are determined. The data set corresponding to the label includes the following steps:

Parse the data definition template, and determine the data items and operation items in each data source;

Applying distributed lock calls to obtain the data items in the respective data sources, and determining the data item subsets of each anchor user according to the data items associated with the same anchor user;

Using an intelligent classification model pre-trained to a convergent state, according to each subset of data items, determine the attribute labels of the corresponding anchor users;

Filtering out a subset of the anchor user data items whose attribute tags match the attribute tags in the operation items;

Format the subset of data items of each anchor user, and construct a data set in a standardized format.

In some extended embodiments, the intelligent classification model is pre-iteratively trained to a convergent state, and the training process includes the following steps:

Invoke a single training sample from the training data set, the training sample includes attribute data corresponding to multiple attribute data items of an anchor user;

Vectorizing the attribute data in the training sample to obtain a sample vector;

Inputting the sample vector into an intelligent classification model to perform semantic extraction and classification mapping to obtain attribute labels predicted by classification;

Calculate the loss value of the attribute label predicted by the intelligent classification model according to the attribute label corresponding to the training sample;

According to the loss value, gradient update is performed on the intelligent classification model, or the iterative training is continued until the model reaches a convergent state.

In some extended embodiments, after the step of filtering out a subset of the anchor user data items whose attribute tags match the attribute tags in the operation items, the following steps are included:

Obtaining the popularity data of each attribute tag correspondingly determined by each anchor user, and the popularity data is statistically determined according to the user behavior data of the live broadcast room of the anchor user;

The data item subsets of each anchor user are sorted according to the popularity data, so that the subsequently generated data sets maintain the corresponding sorting.

In some extended embodiments, before the step of acquiring the popularity data of each attribute tag determined correspondingly by each anchor user, the following steps are included:

Obtaining user behavior data associated with each anchor user, the user behavior data corresponding to the description data generated by the corresponding access events of the anchor user's live broadcast room being followed by the user, gifted by the user, and entered by the user;

Perform statistics on the user behavior data to obtain user popularity corresponding to each anchor user, where the user popularity is a weighted statistical result of the number of various access events;

Corresponding to each attribute tag in the preset attribute tag system, the user popularity of the anchor user carrying the attribute tag is accumulated to obtain the accumulated popularity corresponding to the attribute tag;

Normalization is performed according to the accumulated heat of each attribute label in the attribute label system to obtain heat data corresponding to each attribute label.

In some extended embodiments, after the step of acquiring the data definition template corresponding to the live broadcast service, the following steps are included:

Scheduling the specified data item in the data definition template corresponding to the live broadcast service from its corresponding data source to the secondary cache for invocation according to the timing task trigger;

The data set obtained by performing the operation according to the data item of the data source in the second-level cache is stored in the first-level cache.

In some extended embodiments, before the step of acquiring the data definition template corresponding to the live broadcast service, the following steps are included:

The data source adaptation service is run, and the interface is opened to the external data source, so as to realize the access of the external data source and participate in providing the data of the data set for the live broadcast service.

A live broadcast service data processing device provided in accordance with one of the purposes of this application includes a template calling module, a data processing module, and a data push module, wherein: the template calling module is used to obtain the data definition corresponding to the live broadcast service. A template, the data definition template includes data items distributed in multiple data sources that need to be called by the live broadcast service and operation items for performing predefined operations on these data items, and the data items include the attributes of the anchor user A data item, the operation item includes one or more attribute tags; the data processing module is configured to, according to the operation item in the data definition template, execute the operation item relative to the corresponding data item in the multiple data sources. The corresponding predefined operation determines a data set corresponding to the attribute label in the operation item, and the data set includes a subset of data items corresponding to multiple anchor users; the data push module is used to respond to the terminal The data invocation instruction of the device for the live broadcast service pushes the corresponding data set of the live broadcast service to the terminal device.

In some further embodiments, the data processing module includes: a template parsing unit for parsing the data definition template to determine data items and operation items in each data source; a subset processing unit for Applying distributed lock calls to obtain the data items in the various data sources, and determining the data item subsets of each anchor user according to the data items associated with the same anchor user; the label determination unit is used for pre-training to The intelligent classification model of the convergence state determines the attribute labels of the corresponding anchor users according to the subsets of each data item; the anchor screening unit is used to filter out the anchor user data whose attribute tags match the attribute tags in the operation items. Item subset; the format unification unit is used to format the data item subset of each anchor user, and construct it into a standardized format data set.

In some extended embodiments, the intelligent classification model is placed in a training module to be pre-iteratively trained to a convergent state, and the training module includes: a sample calling unit for calling a single training sample from a training data set, the training sample Including attribute data corresponding to multiple attribute data items of an anchor user; a vector encoding unit for vectorizing the attribute data in the training sample to obtain a sample vector; a classification mapping unit for converting the sample vector Input the intelligent classification model to perform semantic extraction and classification mapping, and obtain the attribute labels predicted by the classification; the loss calculation unit is used to calculate the loss value of the attribute labels predicted by the intelligent classification model according to the attribute labels corresponding to the training samples; iteration A decision-making unit, configured to implement gradient update on the intelligent classification model according to the loss value, or continue iterative training until the model reaches a convergence state.

In some extended embodiments, the data processing module includes the following units that run after the host screening unit: a popularity calling unit, which is used to obtain the popularity data of each attribute tag determined correspondingly by each anchor user, and the popularity data Statistically determined according to the user behavior data of the live broadcast room of the host user; the sorting processing unit is configured to sort the data item subsets of each host user according to the popularity data, so that the subsequently generated data set maintains the corresponding sorting.

In some extended embodiments, the data processing module includes the following units that run before the popularity calling unit: a data description unit, configured to acquire user behavior data associated with each anchor user, the user behavior data corresponding to the The description data generated by the corresponding access events of the live broadcast room of the anchor user being followed by the user, gifted by the user, and entered by the user; the data statistics unit is used to perform statistics on the user behavior data, and obtain the users corresponding to each anchor user. Heat, the user heat is the weighted statistical result of the number of various access events; the heat accumulating unit is used to correspond to each attribute label in the preset attribute label system, and will carry the attribute label of the anchor user's The user's heat is accumulated to obtain the accumulated heat corresponding to the attribute label; the heat determination unit is configured to normalize the accumulated heat of each attribute label in the attribute label system to obtain the heat data corresponding to each attribute label.

In some extended embodiments, the live broadcast service data processing device of the present application further includes the following modules that are run after the template invocation module: a second-level cache module, configured to define data corresponding to the live broadcast service according to the timing task trigger. The specified data item in the template is dispatched from its corresponding data source to the second-level cache for invocation; the first-level cache module is used to store the data set obtained by performing the operation according to the data item of the data source in the second-level cache in the second-level cache. in the first level cache.

In some of the extended embodiments, the live broadcast service data processing apparatus of the present application further includes a data adaptation module that runs before the template invocation module, and is used to run a data source adaptation service and open an interface to an external data source to realize The external data source accesses and participates in providing the data of the data set for the live broadcast service.

A computer device provided in accordance with one of the purposes of this application, comprising a central processing unit and a memory, the central processing unit is used to call and run a computer program stored in the memory to execute the live broadcast service data processing described in this application. steps of the method.

A computer-readable storage medium provided for another purpose of the present application, which stores a computer program implemented according to the described live broadcast service data processing method in the form of computer-readable instructions, and the computer program is invoked by a computer to run , perform the steps included in the method.

A computer program product provided in accordance with another object of the present application includes a computer program/instruction, when the computer program/instruction is executed by a processor, the steps of the method described in any one of the embodiments of the present application are implemented.

Compared with the prior art, the present application has many technical advantages, including but not limited to the various aspects disclosed as follows:

First, the present application utilizes the correspondence between the live broadcast service and the data definition template, and realizes the standardized representation of the data to be invoked for different live broadcast services through the data items and operation items in the data definition template, wherein the data items include data items related to each anchor user. The corresponding attribute data item, and the operation item includes one or more attribute tags, thus, the operation item can play the role of a predefined operation, and then obtain the relevant data item corresponding to the live broadcast service according to the data definition template. The association relationship between the attribute data item and the attribute label is determined, and the data set containing the data item subsets of multiple anchor users is determined, that is, the anchor user that has a certain predefined operation corresponding relationship with the attribute label is determined. Corresponding subset of data items, so the subset of data items plays the role of associating description for the anchor user. When the terminal device needs to load the corresponding page of the live broadcast service and initiate a request to the server, it can The data set is pushed to the corresponding terminal device, so that the corresponding terminal device displays the key information of each anchor user according to the data item subset of the anchor user in the data set.

Secondly, in this application, on the one hand, the data definition template realizes the centralized definition of data items of multiple data sources, and on the other hand, it is associated with the live broadcast service in advance, thus realizing the mapping and sorting of the corresponding relationship between the data source and the live broadcast service. For the live broadcast platform, developers can maintain orderly call data items of multiple data sources by maintaining data definition templates, thereby serving the corresponding live broadcast business and processing data. The live broadcast business and its required data items It is always managed in an orderly manner, and there is no need to obtain data items separately, which can not only effectively avoid the situation where each data source repeatedly obtains the same data items required by different live broadcast services, and minimize the influence of the law of entropy increase, but also makes it easier for developers to control each live broadcast. The data logic between the data items to be called by the business is maintained.

In addition, because the application has sorted out the relationship between the live broadcast service and the data to be called, and optimized the resource occupation of data storage, access and other links, the storage and operation load of the service cluster can be saved, thereby saving the platform system overhead and the platform. deployment cost.

Description of drawings

The above and/or additional aspects and advantages of the present application will become apparent and readily understood from the following description of embodiments taken in conjunction with the accompanying drawings, wherein:

1 is a schematic flowchart of a typical embodiment of a method for processing live broadcast service data of the present application;

Figure 2 and Figure 3 are exemplary graphical user interfaces, respectively showing the interface effects of "guess you like" and "popular" corresponding live broadcast services;

4 is a schematic flowchart of a process of acquiring a data set corresponding to a live broadcast service according to a data definition template in an embodiment of the present application;

5 is a schematic flowchart of a training process of an exemplary intelligent classification model in an embodiment of the present application;

FIG. 6 is a schematic flowchart of a process of determining the popularity data of each attribute tag in an embodiment of the present application;

Fig. 7 is the principle block diagram of the live broadcast service data processing apparatus of the present application;

8 is a schematic structural diagram of a computer device used in the application;

Detailed ways

The following describes in detail the embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, but not to be construed as a limitation on the present application.

It will be understood by those skilled in the art that the singular forms "a", "an", "the" and "the" as used herein can include the plural forms as well, unless expressly stated otherwise. It should be further understood that the word "comprising" used in the specification of this application refers to the presence of stated features, integers, steps, operations, elements and/or components, but does not preclude the presence or addition of one or more other features, Integers, steps, operations, elements, components and/or groups thereof. It will be understood that when we refer to an element as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Furthermore, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. As used herein, the term "and/or" includes all or any element and all combination of one or more of the associated listed items.

It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It should also be understood that terms, such as those defined in a general dictionary, should be understood to have meanings consistent with their meanings in the context of the prior art and, unless specifically defined as herein, should not be interpreted in idealistic or overly formal meaning to explain.

Those skilled in the art can understand that the "client", "terminal" and "terminal device" used herein include both a wireless signal receiver device that only has a wireless signal receiver without transmission capability, and a wireless signal receiver device. A device with receive and transmit hardware that has receive and transmit hardware capable of two-way communication over a two-way communication link. Such devices may include: cellular or other communication devices such as personal computers, tablet computers, which have a single-line display or a multi-line display or a cellular or other communication device without a multi-line display; PCS (Personal Communications Service, Personal Communications System) ), which can combine voice, data processing, fax and/or data communication capabilities; PDA (Personal Digital Assistant), which can include radio frequency receivers, pagers, Internet/Intranet access, web browsers, notepads , calendar and/or GPS (Global Positioning System) receivers; conventional laptop and/or palmtop computers or other devices having and/or conventional laptop and/or palmtop radio frequency receivers computer or other device. As used herein, "client", "terminal", "terminal device" may be portable, transportable, mounted in a vehicle (air, marine and/or land), or adapted and/or configured to be locally operate, and/or in distributed form, operate at any other location on Earth and/or space. The "client", "terminal" and "terminal device" used here can also be a communication terminal, an Internet terminal, and a music/video playing terminal, such as a PDA, MID (Mobile Internet Device) and/or A mobile phone with music/video playback function, or a smart TV, set-top box, etc.

The hardware referred to by names such as "server", "client" and "service node" in this application is essentially an electronic device with the equivalent capability of a personal computer, which is a central processing unit (including an arithmetic unit and a controller). ), memory, input device and output device and other necessary components disclosed by the Von Neumann principle, the computer program is stored in its memory, and the central processing unit transfers the program stored in the external memory into the memory to run, and executes the program. The instructions in the interface interact with input and output devices to complete specific functions.

It should be pointed out that the concept of "server" referred to in this application can also be extended to the case of server clusters in the same way. According to the principles of network deployment understood by those skilled in the art, the servers should be logically divided. In physical space, these servers can be independent from each other but can be called through interfaces, or can be integrated into a physical server. A computer or a group of computers. Those skilled in the art should understand this modification, but should not limit the implementation of the network deployment manner of the present application.

Unless explicitly specified, one or more technical features of the present application can be deployed on the server and remotely invoked by the client to obtain the online service interface provided by the server to implement access, or can be directly deployed and run on the client to implement access.

The neural network model cited or possibly cited in this application, unless specified in plain text, can either be deployed on a remote server and invoked remotely on the client, or deployed on a client with competent device capabilities to directly invoke, in some embodiments , when it runs on the client, its corresponding intelligence can be obtained through transfer learning, so as to reduce the requirements on the client hardware running resources and avoid excessively occupying the client hardware running resources.

All kinds of data involved in this application, unless specified in plain text, can be stored in a server remotely or in a local terminal device, as long as it is suitable for being called by the technical solution of this application.

Those skilled in the art should know that: although the various methods of the present application are described based on the same concept to show commonality with each other, unless otherwise specified, these methods can be independently executed. Similarly, for the various embodiments disclosed in this application, they are all proposed based on the same inventive concept. Therefore, the concepts expressed in the same way, and the concepts that are appropriately transformed for convenience even though the concept expressions are different, should be regarded as equivalent. understand.

In the various embodiments to be disclosed in this application, unless the mutually exclusive relationship between each other is clearly indicated, the related technical features involved in the various embodiments can be cross-combined to flexibly construct new embodiments, as long as the combination does not deviate from the present invention. The creative spirit of the application can meet the needs in the prior art or solve a certain aspect of the deficiencies in the prior art. Variations on this will be known to those skilled in the art.

A method for processing live broadcast service data of the present application can be programmed into a computer program product, which can be implemented by being deployed in a computer device to run, whereby by accessing an interface opened after the computer program product runs, it can communicate with the computer through a graphical user interface. The process of the program product performs human-computer interaction to execute the method.

Please refer to FIG. 1 , in a typical embodiment of the method for processing live broadcast service data of the present application, the method includes the following steps:

Step S1100: Acquire a data definition template corresponding to the live broadcast service, where the data definition template includes data items distributed in multiple data sources that the live broadcast service needs to call and operations for performing predefined operations on these data items item, the data item includes the attribute data item of the host user, and the operation item includes one or more attribute tags:

In the web live broadcast platform, the collection, storage, and invocation of data items in various data sources inside and outside the platform are unified and displayed, and the corresponding data items are displayed through the corresponding page codes of the live broadcast service realized by the development team.

The data sources include various databases suitable for database engines, such as HBase, MySQL, and MongoDB. For large-scale webcast platforms, the databases are usually databases that support distributed storage. Each database can contain multiple pre-defined data tables, each data table is used to store multiple data items, and cross-database and cross-data tables can realize data connection through data item correspondence, such as through devices or users UID, or associated with the same attribute tag to correspond. Therefore, in the network platform, personal information related to an anchor user may be distributed in multiple data items of multiple data tables of multiple data sources, and may be a relatively complex tree or network relationship. Some data items relative to the personal information are attribute data items, which are used to store attribute data corresponding to the corresponding anchor user, for example, store the anchor user's age, gender, live broadcast theme description information, personal hobby keywords, etc. .

The live broadcast service usually corresponds to a page that can be accessed by users of the web live broadcast platform. For example, when the user accesses the "guess you like" page as shown in A corresponding live broadcast service is called. At this time, the live broadcast service is used to recommend video covers of multiple anchor users according to the user's personal preference, so that the user can quickly enter the live broadcast room of one of the anchor users. When the user accesses the "Popular" page as shown in Figure 3, the same is true, and correspondingly accesses a video cover that pushes the currently popular anchor user to the user, so that the user can enter one of the live broadcast rooms.

Online live broadcast platforms often provide multiple websites and multiple applications to open various services for platform users, and one website or one application often provides multiple live broadcast services, and the underlying data of all these live broadcast services are provided by the platform. Therefore, building a bridge between the live broadcast service and the data source to maintain the relationship between the live broadcast service and the complex distributed data items is helpful to improve the development efficiency and Data maintenance effect. Therefore, a data maintenance platform can be used to provide customized services for the development of each live broadcast business. Through this customized service, the development team can create and edit the data items required by a live broadcast business, form a corresponding data definition template, and implement it later. In the corresponding live broadcast service, the desired data set can be obtained by calling this data definition template.

The data definition template, in this application, not only includes the specification of the data items that the live broadcast service needs to use, but also includes the operation items used to instruct the implementation of the specified predefined operations on these data items. The data items may be distributed in a number of different data sources, and the operation items may specify the information they contain in various pre-protocol ways. In an example of this application, the operation item can be parsed by the application by including one or more attribute tags to predict the corresponding attribute tag for the data item associated with each anchor user, and then use The mapping relationship between the attribute labels of the operation items and the predicted attribute labels matches the subset of data items of some of the anchor users to construct the data set. According to this example, the predefined operation, which has been agreed in advance, can be parsed as a matching operation for data items.

As mentioned above, among the data items associated with each anchor user, some of them are attribute data items, which are particularly suitable for participating in the operations defined by the operation items, while the rest can be other types of data items, which can be Serving the needs of the background implementation of the live broadcast service, including participating in the operations defined by the operation items, is not limited by this application.

When the online live broadcast platform needs to prepare data sets for each live broadcast service, it will obtain the data definition template corresponding to each live broadcast service according to the predetermined trigger mechanism, and then perform pre-protocol analysis on the data definition template to determine the data items and Action items for later use.

Step S1200: According to the operation items in the data definition template, perform predefined operations corresponding to the operation items on the corresponding data items in the multiple data sources, and determine the operation items corresponding to the attribute labels. A data set, the data set includes a subset of data items corresponding to multiple anchor users:

As mentioned above, the operation item may include one or more attribute tags to indicate a predefined operation that is parsed as performing matching on the data item in the data definition template. Therefore, according to the data definition corresponding to the live broadcast service The operation items in the template, that is, one or more attribute tags, analyze whether the data items matching each anchor user are related to the one or more attribute tags. The data item subsets formed by the data items corresponding to the user are filtered out, and finally each filtered data item subset is formed into a data set.

The matching relationship between the operation item and the data item of the anchor user associated with the attribute tag can be confirmed by analyzing whether the data item of the anchor user contains the same attribute tag specified by the operation item, or pre-training can be used to confirm the matching relationship between the two. The neural network model in the convergent state performs classification mapping on the data items of each anchor user to predict the attribute labels, and then analyzes whether the predicted attribute labels correspond to the attribute labels in the operation items to confirm. The manner of realizing this matching operation is relatively flexible, and those skilled in the art can flexibly choose other means to replace them according to the principles and examples disclosed in the present application, which will not affect the embodiment of the inventive spirit of the present application.

After the data set is determined, the data set includes a subset of data items corresponding to multiple anchor users, and the subset of data items includes one or more attribute data items, and the attribute data items generally It is a data item suitable for being directly displayed on the page corresponding to the live broadcast service.

In order to facilitate the parsing and calling of the data set using a unified code, the data set may be formatted according to a preset unified format.

Step S1300: In response to the terminal device's data calling instruction for the live broadcast service, push the corresponding data set of the live broadcast service to the terminal device:

After the data sets of each live broadcast service are prepared through the foregoing steps, the data sets are in a callable state, and steps S1100 and S1200 can be executed cyclically subsequently to realize dynamic update of the data sets.

When a large number of platform users log in to the website or application of the web live broadcast platform on their respective terminal devices and need to load or preload the page corresponding to the live broadcast service, they can initiate a live broadcast service loading request to the server. In response to the request, the server Then according to the corresponding relationship between the live broadcast service and the data set, the corresponding data set is called and pushed to the terminal device that triggered the request, so that the data set can be displayed in the graphical user interface of the terminal device. The anchor user information described by each data item subset of , including the personal information of the anchor user, etc., thus, as shown in Figures 2 and 3, the video covers corresponding to multiple anchor users can be displayed.

In the case where the data set is encapsulated in a unified format, as long as the background code corresponding to the page in the terminal device parses the data set according to the unified format, the same subset of anchor user data items can be obtained, and the same subset of data items of the host can be obtained. The application category and website category used to load the page are irrelevant. Therefore, for multiple networks and multiple applications developed by the webcast platform, as long as they comply with the same format protocol, they can use the described information according to their business logic. Data sets in a unified format can make personalized data presentation, so that in the process of code realization of the live service page corresponding to the terminal device, there is no need to directly face the underlying data, and there is no need to directly operate the underlying data source. Data access and call operations.

According to the present exemplary embodiment, it is not difficult to understand that, compared with the prior art, the present application has many technical advantages, including but not limited to the various aspects disclosed as follows:

First, the present application utilizes the correspondence between the live broadcast service and the data definition template, and realizes the standardized representation of the data to be invoked for different live broadcast services through the data items and operation items in the data definition template, wherein the data items include data items related to each anchor user. The corresponding attribute data item, and the operation item includes one or more attribute tags, thus, the operation item can play the role of a predefined operation, and then obtain the relevant data item corresponding to the live broadcast service according to the data definition template. The association relationship between the attribute data item and the attribute label is determined, and the data set containing the data item subsets of multiple anchor users is determined, that is, the anchor user that has a certain predefined operation corresponding relationship with the attribute label is determined. Corresponding subset of data items, so the subset of data items plays the role of associating description for the anchor user. When the terminal device needs to load the corresponding page of the live broadcast service and initiate a request to the server, it can The data set is pushed to the corresponding terminal device, so that the corresponding terminal device displays the key information of each anchor user according to the data item subset of the anchor user in the data set.

Secondly, in this application, on the one hand, the data definition template realizes the centralized definition of data items of multiple data sources, and on the other hand, it is associated with the live broadcast service in advance, thus realizing the mapping and sorting of the corresponding relationship between the data source and the live broadcast service. For the live broadcast platform, developers can maintain orderly call data items of multiple data sources by maintaining data definition templates, thereby serving the corresponding live broadcast business and processing data. The live broadcast business and its required data items It is always managed in an orderly manner, and there is no need to obtain data items separately, which can not only effectively avoid the situation where each data source repeatedly obtains the same data items required by different live broadcast services, and minimize the influence of the law of entropy increase, but also makes it easier for developers to control each live broadcast. The data logic between the data items to be called by the business is maintained.

In addition, because the application has sorted out the relationship between the live broadcast service and the data to be called, and optimized the resource occupation of data storage, access and other links, the storage and operation load of the service cluster can be saved, thereby saving the platform system overhead and the platform. deployment cost.

Referring to FIG. 4, in some further embodiments, in step S1200, according to the operation items in the data definition template, perform predefined operations corresponding to the operation items on the corresponding data items in the multiple data sources, Determining the data set corresponding to the attribute label in the operation item includes the following steps:

Step S1210, parse the data definition template, and determine the data items and operation items in each data source:

For the data definition template obtained according to the live service call, it can be parsed correspondingly according to the pre-protocol, so as to determine the operation items and data items distributed in each data source. For the server, the data definition template Each data item parsed in the data source, the corresponding data table, data field, etc. in the data source are all predefined and clear.

Step S1220, applying a distributed lock call to obtain the data items in the respective data sources, and determining the data item subsets of each anchor user according to the data items associated with the same anchor user:

In order to realize the invocation of specific data items, after the data items required by the live broadcast service are determined, the corresponding data items can be invoked from each data source, that is, the corresponding data can be obtained. As mentioned above, since the data stored in the data items are usually stored in association with the anchor user, the anchor user can be the clustering unit to construct a data item subset for each anchor user, and this subset corresponds to the anchor user The data composition in the data item. When implementing a call to a data item in a data source, taking into account the risk of data access conflict, a distributed lock can be applied to the accessed data item to perform atomic operations, thereby ensuring that the data call process will not cause data disorder.

Step S1230, adopting the intelligent classification model pre-trained to the convergence state, according to each data item subset, determine the attribute label of each corresponding anchor user:

In this embodiment, an intelligent classification model pre-trained to a convergent state is used to classify and map the data item subsets of each anchor user, so as to predict the attribute label corresponding to each anchor user. The training process of the intelligent classification model will be disclosed in the subsequent embodiments of this application, and will not be shown here for the time being.

The intelligent classification model, after being trained to a convergent state, acquires the representation learning ability of accurately extracting deep semantic features from the data item subset of the anchor user. On this basis, the deep semantic features are classified and mapped to Obtain each attribute tag to which the anchor user belongs. For example, in a scenario example, the attribute tag may be "popular anchor", "dance anchor", "sports anchor", "knowledge anchor" and so on. One or more attribute labels predicted by the intelligent classification model can be confirmed. For example, if a host user is allowed to be both a "popular host" and a "knowledge host" at the same time, whether the attribute labels are multi-level The system can be implemented flexibly.

The input of the intelligent classification model is determined according to the data items used as model training samples when it is pre-trained. properties, and only the attribute data items corresponding to the host user can be used to construct its input. Of course, those skilled in the art may also use other data items of the host user to participate in constructing the input of the intelligent classification model.

Step S1240, filtering out a subset of anchor user data items whose attribute tags match the attribute tags in the operation items:

For an anchor user, when the attribute label of the operation item is in a one-to-one relationship with the attribute label predicted by the intelligent classification model and is exactly the same, it can be regarded as realizing the predefined matching relationship of the operation item. A subset of data items is filtered out to construct the dataset.

For an anchor user, when there are multiple attribute labels in the operation item and the intelligent classification model only predicts a single attribute label, and the predicted attribute label is one of the multiple attribute labels in the operation item, it can be regarded as the operation. The predefined matching relationship of the item can be used to filter the data item subset of the host user to construct the data set.

For a host user, when there is a single attribute label in the operation item and the intelligent classification model predicts multiple attribute labels, if one of the multiple attribute labels is the same as the single attribute label in the operation item, it can be regarded as the realization of the The matching relationship defined by the operation item can be used to filter the data item subset of the host user to construct the data set.

For an anchor user, when the operation item contains multiple attribute tags and the intelligent classification model also predicts multiple attribute tags, the set relationship can be used to confirm whether the two match. A specific example is to examine multiple attributes in the operation item. Whether the tag constitutes a subset of the predicted multiple attribute tags, if so, it can be regarded as the predefined matching relationship between the two implementation operation items, and the data item subset of the anchor user can be filtered out to construct the data set.

It can be seen that whether the subset of data items of the anchor user achieves the matching specified by the matching relationship predefined by the operation item, the implementation is flexible and diverse, and those skilled in the art can flexibly implement according to the various examples here, It does not affect the embodiment of the inventive spirit of the present application.

Step S1250, formatting the data item subsets of each anchor user, and constructing a data set in a standardized format:

After the foregoing steps are processed to match the data item subsets of each anchor user required by the live broadcast service, these data item subsets can be constructed into a standardized format data set. In order to achieve the unification of standardized formats, the subset of data items corresponding to each anchor user can be encapsulated in a unified format according to the preset format protocol, so that subsequent unified analysis can be performed according to the format protocol, and then the encapsulation can be performed in a unified format. A subset of good data items constructs the data set.

In this embodiment, it is disclosed that the data definition template implements predefined operations by encapsulating attribute labels in its operation items, thereby instructing to perform intelligent classification and mapping on the subset of data items called from the data source to predict the corresponding attribute labels, and then according to the operation The matching operation indicated by the predefined operation of the item, according to whether the predicted attribute label matches the attribute label of the operation item, the data item subset of the anchor user required by the live broadcast service is selected and constructed as a unified format data set. In this way, the functions of the data definition template are enriched, so that the data definition template can not only realize the sorting of data items from multiple sources, but also reflect more complex business logic, instructing the server to intelligently carry out the data item subsets of each anchor user. Through analysis and matching, a data set that accurately corresponds to the data requirements of the live broadcast business can be obtained, which avoids the tediousness of developers repeatedly developing complex business logic, but also provides richer technical support, which greatly improves the efficiency of software engineering development.

Referring to FIG. 5, in some extended embodiments, the intelligent classification model is pre-iteratively trained to a convergent state, and the training process includes the following steps:

Step S2100, calling a single training sample from the training data set, where the training sample includes attribute data corresponding to multiple attribute data items of an anchor user:

Before training the intelligent classification model of the present application, prepare a training data set, the training data set includes a large number of training samples, and the number of training samples is limited to be suitable for training the model to a convergent state.

The training samples are composed of attribute data collected from a plurality of attribute data items of each anchor user, and the specific data items can be flexibly determined by those skilled in the art according to the prediction capability required by the model. Each training sample corresponds to an attribute label in advance so that it can be used to supervise the output of the model.

Step S2200, vectorize the attribute data in the training sample to obtain a sample vector:

For each training sample that is about to be input into the intelligent classification model to implement the training of the model, each relevant attribute data is vectorized by querying the vocabulary, so as to obtain an embedding vector as a sample vector.

Step S2300, inputting the sample vector into the intelligent classification model to perform semantic extraction and classification mapping, and obtain the attribute label predicted by classification:

The sample vector is input into the intelligent classification model for processing. The intelligent classification model has a built-in text feature extraction model, which is used to extract deep semantic feature information from sample vectors and realize the representation learning of training samples. The text feature extraction model can be implemented using neural network models such as Bert, TextCNN, and LSTM. After the deep semantic feature information is extracted from the sample vector by the text feature extraction model, the corresponding comprehensive vector is obtained, and the comprehensive vector is further input into the fully connected layer in the model for full connection, so as to map to the preset classification space, and the classifier performs Classification and predict the corresponding attribute labels. The classification space includes a plurality of classifications corresponding to the attribute labels given in advance. When the classification probability of each classification is predicted for a comprehensive vector, the attribute label corresponding to the classification with the largest classification probability is the model for the described classification. The attribute labels predicted from the training samples.

Step S2400: Calculate the loss value of the attribute label predicted by the intelligent classification model according to the attribute label corresponding to the training sample:

For the attribute label predicted by the model, the attribute label corresponding to the training sample can be called to calculate the model loss value. Since the model is implemented by a classifier, the cross entropy function can be used as the loss function when calculating the loss value. Calculate the stated loss value.

Step S2500, implement gradient update to the intelligent classification model according to the loss value, or continue iterative training until the model reaches a convergence state:

After determining the loss value, the loss value can be compared with a preset threshold for identifying whether the model has reached a convergence state, and when the loss value reaches the preset threshold, it is considered that the model has been trained to converge. state, so that the training of the model can be terminated, and the model can be put into the attribute label corresponding to the subset of data items used to predict the anchor user. When the loss value does not reach the preset threshold, the model has not converged at this time. Therefore, the gradient update is performed on the model according to the loss value, and the weight of the model is corrected by backpropagation, so that the model is further approached and converged. Then, from The next training sample is called in the training data set, and the steps of this embodiment are repeated to continue to perform iterative training on the model until the model is trained to a convergent state.

In this embodiment, the training process of the intelligent classification model of the present application is provided. According to the training process, the data items originally distributed in multiple data sources are called centrally through the data definition model, and then used to train the intelligent classification model, so that the The intelligent classification model obtains the ability to determine the attribute label corresponding to the anchor user according to the data item subset of the anchor user, so that it can serve the prediction of the attribute label of each data item subset, and form a data item specified by the data definition template. Data cleaning mechanism, therefore, through the data definition template, the selection of data item subsets can be realized, thereby improving the acquisition efficiency of the data set corresponding to the live broadcast business, and due to the existence of the artificial intelligence factor of the intelligent classification model, it can be upgraded to live broadcast The accuracy with which the service matches the data item subset of the anchor user.

In some extended embodiments, the step S1240, after the step of filtering out a subset of the anchor user data items whose attribute tags match the attribute tags in the operation items, includes the following steps:

Step S1245: Acquire the popularity data of each attribute tag correspondingly determined by each anchor user, and the popularity data is statistically determined according to the user behavior data of the live broadcast room of the anchor user:

For each attribute tag in the attribute tag system adopted in this application, popularity data representing the visited popularity of each attribute tag can be pre-formed, and the popularity of the corresponding attribute tag can be reflected through the popularity data. The popularity data of each attribute tag can be statistically determined according to the number of visits of the anchor user carrying the attribute tag, and the number of times the anchor user has been visited by the user can be determined by accessing the user behavior data of the anchor user. This will be further exemplified by the embodiment of , which will be temporarily clicked here.

On the basis of the popularity data of each attribute tag, associate the popularity data corresponding to the attribute tag for each anchor user for sorting.

Step S1246, sorting the data item subsets of each anchor user according to the popularity data, so that the subsequently generated data sets maintain the corresponding sorting:

Then, according to the popularity data, the data item subsets of each anchor user are sorted from high to low, so that the higher the popularity, the higher the sorting, so that the data set constructed according to these data item subsets is actually completed. The popularity of each corresponding anchor user is sorted.

This embodiment further improves the processing of the sorting relationship of each anchor user in the data set required by the live broadcast service, so that after the live broadcast service obtains the data set for analysis and display, it can be displayed in an orderly manner on the graphical user interface of the terminal device Each anchor user establishes a correlation between the arrangement order of the anchor users and the attribute labels specified by the live broadcast service, thereby highlighting the functional focus of the live broadcast service.

Referring to FIG. 6, in some extended embodiments, the step S1245, before the step of acquiring the popularity data of each attribute tag correspondingly determined by each anchor user, includes the following steps:

Step S1241, obtain the user behavior data associated with each anchor user, the user behavior data corresponds to the description data generated by the user's attention, the user's gift, and the user's entry into the corresponding access event in the anchor user's live broadcast room:

The live broadcast room of each anchor user will generate corresponding user behavior data when it is accessed by the users of the live broadcast platform. Corresponding access events will be triggered, and corresponding user behavior records will be generated in the background and uploaded to the server. Therefore, data mining can be performed on these user behavior data so as to obtain higher semantic information, that is, the popularity data serves to optimize the data set corresponding to the data definition template of the present application.

Step S1242: Counting the user behavior data to obtain user popularity corresponding to each anchor user, where the user popularity is a weighted statistical result of the number of various access events:

After obtaining the user behavior data, statistics can be performed for each anchor user one by one according to the number of access events corresponding to the user behavior data, so as to calculate the user popularity corresponding to each anchor user. In the recommended implementation, the preset weighting formula is used to count the number of different access events. For example, for the statistics of the three different access events of the anchor user being followed by the user, being given a gift by the user, and being entered by the user, the correlation is different. Weight, multiply the weight by the corresponding number of statistics and add them together to obtain the weighted statistical result, which is used as the popularity of an anchor user. An example of a weighted statistical formula is as follows:

S _i =α×Like+β×Present+γ×Enter

Among them, S _i represents the user popularity of a single anchor user, Like, Present, and Enter represent different access events, respectively, α, β, γ represent the respective weights corresponding to different access events, and their values can be determined by those skilled in the art as needed.

Step S1243: Corresponding to each attribute label in the preset attribute label system, accumulate the user popularity of the anchor user who carries the attribute label, and obtain the accumulated popularity corresponding to the attribute label:

Since each anchor user can be mapped to one or more attribute tags belonging to the preset attribute tag system, the user popularity of each anchor user is regarded as each attribute tag carried by it corresponds to the anchor user According to the obtained user popularity, for each attribute tag in the attribute tag system, the user popularity corresponding to each anchor user can be counted to obtain the cumulative popularity corresponding to each attribute tag. That is, for each attribute tag in the attribute tag system, its corresponding cumulative popularity Sum _j is the summation result of the user popularity obtained from each anchor user.

Step S1244: Normalize the accumulated heat of each attribute label in the attribute label system to obtain heat data corresponding to each attribute label:

After the above steps, the cumulative heating temperature Sum _j corresponding to each attribute label in the attribute label system has been determined. Accordingly, it can be normalized according to the following formula to provide a unified measure for the cumulative heating temperature of each attribute label, so as to determine each attribute label. Corresponding heat data Hot _j :

That is, for an attribute tag, its heat data is the product of its corresponding accumulated heat divided by the sum of the accumulated heat of all attribute tags in the attribute tag system. In this way, the popularity data of each attribute tag can be compared and sorted on the same metric benchmark.

In this embodiment, in order to optimize the data set determined according to the data definition template, the popularity data is determined in advance for each attribute tag in the attribute tag system according to the user behavior data corresponding to the anchor user. The determination of the popularity data refers to The popularity of the anchor users who carry the attribute tags. Therefore, in the data set optimized for the live broadcast service, the selected anchor users are sorted according to the attribute tags. The sorting result can better represent the traffic effect of the anchor users, which can improve the The matching accuracy of the data required by the live broadcast business.

In some extended embodiments, the step S1100, after the step of acquiring the data definition template corresponding to the live broadcast service, includes the following steps:

Step S1101, according to the timing task trigger, schedule the specified data item in the data definition template corresponding to the live broadcast service from its corresponding data source to the secondary cache for invocation:

The server may preset timing tasks, for example, using Sping's timer Schedule technology to regularly trigger update tasks, and in response to timing trigger tasks, change each data item specified in the data definition template corresponding to the live broadcast service from its corresponding multiple The data source is dispatched to the second-level cache, so that more efficient operations are performed on these data items in the second-level cache, for example, the matching operation in step S1200 is performed. In the recommended embodiment, the secondary cache can be implemented using Redis technology, and the data items of each anchor user are stored through key-value pairs. For example, the UID of the anchor user is used as the key field, and each data item is stored as the value field.

Step S1102: Store the data set obtained by performing the operation on the data item of the data source in the second-level cache in the first-level cache:

In order to further improve the data scheduling efficiency and shorten the data access response time, the data set obtained according to step S1200 can be stored in the first-level cache, which is more efficient than the second-level cache, and subsequent calls and updates to the data set can Both operate in the L1 cache. In the recommended alternative embodiment, the technology of Guava Cache can be used as the support to implement the first-level cache to achieve the effect of efficient access.

In this embodiment, considering that the distributed storage data items occupy more memory space and the data logic is lower level, the data items are placed in the second-level cache for operation, and the data sets constructed according to the data items are relatively It is called frequently, so it is placed in the first-level cache for operation, which comprehensively takes into account the balance of the realization efficiency of the entire business logic at different stages, which can improve the operating efficiency and robustness of the system, and ensure that each terminal device can efficiently access each live broadcast. The data set corresponding to the business can ensure the timely update of the data set, and realize the mutual decoupling of the underlying operations corresponding to the data items in various data sources.

In some extended embodiments, the step S1100, before the step of acquiring the data definition template corresponding to the live broadcast service, includes the following steps:

Step S1000, running a data source adaptation service, opening an interface to an external data source, so as to realize the access of the external data source and participate in providing the data of the data set for the live broadcast service:

In order to expand the compatibility of the webcast platform for data sources, a data source adaptation service can be run in the server, through which different external data source access protocols, such as PB, HTTP, YYP, etc., can be supported, and the corresponding data sources can be opened to external data sources. The mutual access interface, so that this data source realizes the support for mutual access of different database engines, so that each different data source can call these mutual access interfaces in advance to access the system that realizes the technical solution of the present application, so that the present application The data definition template for fetching its data items facilitates.

This embodiment further improves the data acquisition capability of the data definition template, so that the live broadcast service can further call external data through its data definition template, thereby enriching the information content of the live broadcast service and improving the accuracy of data matching through external data. The compatibility of the system implementing the technical solution of the present application is greatly enhanced.

Please refer to FIG. 7 , a live broadcast service data processing device provided for one of the purposes of this application includes a template calling module 1100, a data processing module 1200, and a data pushing module 1300, wherein: the template calling module 1100 uses in order to obtain a data definition template corresponding to the live broadcast service, the data definition template includes data items distributed in a plurality of data sources to be called by the live broadcast service and operation items for performing predefined operations on these data items, The data item includes the attribute data item of the host user, and the operation item includes one or more attribute tags; the data processing module 1200 is used for, according to the operation item in the data definition template, to perform corresponding operation items in multiple data sources. The data item performs a predefined operation corresponding to the operation item, and determines a data set corresponding to the attribute label in the operation item, and the data set includes a subset of data items corresponding to multiple anchor users; The data push module 1300 is configured to push a data set corresponding to the live broadcast service to the terminal device in response to the terminal device's data calling instruction for the live broadcast service.

In some further embodiments, the data processing module 1200 includes: a template parsing unit for parsing the data definition template to determine the data items and operation items in each data source; a subset processing unit for using The data item subsets of each anchor user are determined according to the data items associated with the same anchor user; the label determination unit is used for adopting pre-training The intelligent classification model to the convergent state determines the attribute labels of the corresponding anchor users according to the subsets of each data item; the anchor screening unit is used to filter out the anchor users whose attribute tags match the attribute tags in the operation items. Data item subset; the format unification unit is used to format the data item subset of each anchor user, and construct it into a standardized format data set.

In some extended embodiments, the intelligent classification model is placed in a training module to be pre-iteratively trained to a convergent state, and the training module includes: a sample calling unit for calling a single training sample from a training data set, the training sample Including attribute data corresponding to multiple attribute data items of an anchor user; a vector encoding unit for vectorizing the attribute data in the training sample to obtain a sample vector; a classification mapping unit for converting the sample vector Input the intelligent classification model to perform semantic extraction and classification mapping, and obtain the attribute labels predicted by the classification; the loss calculation unit is used to calculate the loss value of the attribute labels predicted by the intelligent classification model according to the attribute labels corresponding to the training samples; iteration A decision-making unit, configured to implement gradient update on the intelligent classification model according to the loss value, or continue iterative training until the model reaches a convergence state.

In some extended embodiments, the data processing module 1200 includes the following units that are run after the host screening unit: a popularity calling unit, which is used to obtain the popularity data of each attribute tag correspondingly determined by each anchor user, and the popularity The data is statistically determined according to the user behavior data in the live broadcast room of the anchor user; the sorting processing unit is used to sort the data item subsets of each anchor user according to the popularity data, so that the subsequently generated data sets maintain the corresponding sorting .

In some extended embodiments, the data processing module 1200 includes the following units that run before the popularity calling unit: a data description unit, configured to obtain user behavior data associated with each anchor user, the user behavior data pair The description data generated by the corresponding access events of the live broadcast room that should be followed by the user, gifted by the user, and entered by the user; the data statistics unit is used to perform statistics on the user behavior data, and obtain the corresponding data of each anchor user. User popularity, the user popularity is the weighted statistical result of the number of various access events; the popularity accumulating unit is used to correspond to each attribute tag in the preset attribute tag system, and will carry the anchor user of the attribute tag. Accumulate the user's heat of the attribute label to obtain the accumulated heat of the attribute label; the heat determination unit is used to normalize the accumulated heat of each attribute label in the attribute label system, and obtain the heat data corresponding to each attribute label. .

In some extended embodiments, the live broadcast service data processing apparatus of the present application further includes the following modules that are run after the template calling module 1100: a second-level cache module, which is used for triggering the data corresponding to the live broadcast service according to the timing task. The specified data item in the definition template is dispatched from its corresponding data source to the second-level cache for invocation; the first-level cache module is used to store the data set obtained by performing operations on the data items of the data source in the second-level cache in the first level cache.

In some extended embodiments, the live broadcast service data processing apparatus of the present application further includes a data adaptation module that runs before the template invocation module 1100, and is used to run the data source adaptation service and open an interface to an external data source to Participate in providing the data of the data set for the live broadcast service by realizing the access of the external data source.

To solve the above technical problems, the embodiments of the present application also provide computer equipment. As shown in FIG. 8 , a schematic diagram of the internal structure of the computer equipment. The computer device includes a processor, a computer-readable storage medium, a memory, and a network interface connected by a system bus. Wherein, the computer-readable storage medium of the computer device stores an operating system, a database and computer-readable instructions, the database may store a control information sequence, and when the computer-readable instructions are executed by the processor, the processor can be made to implement a A data processing method for live broadcast services. The processor of the computer device is used to provide computing and control capabilities and support the operation of the entire computer device. Computer-readable instructions may be stored in the memory of the computer device, and when the computer-readable instructions are executed by the processor, the processor may execute the method for processing live broadcast service data of the present application. The network interface of the computer equipment is used for communication with the terminal connection. Those skilled in the art can understand that the structure shown in FIG. 8 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer equipment to which the solution of the present application is applied. Include more or fewer components than shown in the figures, or combine certain components, or have a different arrangement of components.

In this embodiment, the processor is used to execute the specific functions of each module and its sub-modules in FIG. 7 , and the memory stores program codes and various types of data required to execute the above-mentioned modules or sub-modules. The network interface is used for data transmission between user terminals or servers. The memory in this embodiment stores the program codes and data required to execute all modules/sub-modules in the live broadcast service data processing apparatus of the present application, and the server can call the server's program codes and data to execute the functions of all sub-modules.

The present application further provides a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the method for processing live broadcast service data according to any embodiment of the present application. A step of.

The present application also provides a computer program product, including computer programs/instructions, when the computer program/instructions are executed by one or more processors, to implement the steps of the method described in any embodiment of the present application.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above-mentioned embodiments of the present application can be implemented by instructing relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the program is executed, it may include the flow of the embodiments of the above-mentioned methods. The aforementioned storage medium may be a computer-readable storage medium such as a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM), or a random access memory (Random Access Memory, RAM).

To sum up, the present application realizes the standardized operation of calling data items of multiple data sources by multiple live broadcast services through the data definition template, avoids the increase of data entropy, improves system efficiency, and saves system overhead.

Those skilled in the art can understand that various operations, methods, steps, measures, and solutions in the process discussed in this application may be alternated, modified, combined or deleted. Further, other steps, measures, and solutions in the various operations, methods, and processes that have been discussed in this application may also be alternated, modified, rearranged, decomposed, combined, or deleted. Further, steps, measures and solutions in the prior art with various operations, methods, and processes disclosed in this application may also be alternated, modified, rearranged, decomposed, combined or deleted.

The above are only part of the embodiments of the present application. It should be pointed out that for those skilled in the art, without departing from the principles of the present application, several improvements and modifications can also be made. It should be regarded as the protection scope of this application.

Claims (10)

1. a method for processing live broadcast service data, comprising the steps of: Acquire a data definition template corresponding to the live broadcast service, where the data definition template includes data items distributed in multiple data sources that the live broadcast service needs to call and operation items for performing predefined operations on these data items, so The data item includes the attribute data item of the anchor user, and the operation item includes one or more attribute tags; According to the operation items in the data definition template, the predefined operations corresponding to the operation items are performed on the corresponding data items in the multiple data sources, and the data sets corresponding to the attribute labels in the operation items are determined, The data set includes a subset of data items corresponding to multiple anchor users; In response to the data calling instruction of the live broadcast service from the terminal device, the data set corresponding to the live broadcast service is pushed to the terminal device. 2. The method for processing live broadcast service data according to claim 1, wherein, according to the operation items in the data definition template, the corresponding predefined data items of the operation items are executed to the corresponding data items in the multiple data sources. operation to determine the data set corresponding to the attribute label in the operation item, including the following steps: Parse the data definition template, and determine the data items and operation items in each data source; Applying distributed lock calls to obtain the data items in the respective data sources, and determining the data item subsets of each anchor user according to the data items associated with the same anchor user; Using an intelligent classification model pre-trained to a convergent state, according to each subset of data items, determine the attribute labels of the corresponding anchor users; Filtering out a subset of the anchor user data items whose attribute tags match the attribute tags in the operation items; Format the subset of data items of each anchor user, and construct a data set in a standardized format. 3. The method for processing live broadcast service data according to claim 2, wherein the intelligent classification model is iteratively trained to a convergence state in advance, and its training process comprises the steps: Invoke a single training sample from the training data set, the training sample includes attribute data corresponding to multiple attribute data items of an anchor user; Vectorizing the attribute data in the training sample to obtain a sample vector; Inputting the sample vector into an intelligent classification model to perform semantic extraction and classification mapping to obtain attribute labels predicted by classification; Calculate the loss value of the attribute label predicted by the intelligent classification model according to the attribute label corresponding to the training sample; According to the loss value, gradient update is performed on the intelligent classification model, or the iterative training is continued until the model reaches a convergent state. 4. The method for processing live broadcast service data according to claim 2, characterized in that, after filtering out the step of the anchor user data item subset whose attribute tag matches the attribute tag in the operation item, the method comprises the following steps: Obtaining the popularity data of each attribute tag correspondingly determined by each anchor user, and the popularity data is statistically determined according to the user behavior data of the live broadcast room of the anchor user; The data item subsets of each anchor user are sorted according to the popularity data, so that the subsequently generated data sets maintain the corresponding sorting. 5. The method for processing live broadcast service data according to claim 4, characterized in that: before the step of acquiring the heat data of each attribute label determined correspondingly by each anchor user, the method comprises the following steps: Obtaining user behavior data associated with each anchor user, the user behavior data corresponding to the description data generated by the corresponding access events of the anchor user's live broadcast room being followed by the user, gifted by the user, and entered by the user; Perform statistics on the user behavior data to obtain user popularity corresponding to each anchor user, where the user popularity is a weighted statistical result of the number of various access events; Corresponding to each attribute tag in the preset attribute tag system, the user popularity of the anchor user carrying the attribute tag is accumulated to obtain the accumulated popularity corresponding to the attribute tag; Normalization is performed according to the accumulated heat of each attribute label in the attribute label system to obtain heat data corresponding to each attribute label. 6. The method for processing live broadcast service data according to claim 1, wherein after the step of acquiring the corresponding data definition template of the live broadcast service, the method comprises the following steps: Scheduling the specified data item in the data definition template corresponding to the live broadcast service from its corresponding data source to the secondary cache for invocation according to the timing task trigger; The data set obtained by performing the operation according to the data item of the data source in the second-level cache is stored in the first-level cache. 7. The method for processing live broadcast service data according to claim 1, wherein before the step of acquiring the corresponding data definition template of the live broadcast service, the method comprises the following steps: The data source adaptation service is run, and the interface is opened to the external data source, so as to realize the access of the external data source and participate in providing the data of the data set for the live broadcast service. 8. A live broadcast service data processing device, characterized in that, comprising: The template calling module is used to obtain a data definition template corresponding to the live broadcast service, the data definition template includes the data items distributed in multiple data sources that the live broadcast service needs to call and is used to execute predefined data items for these data items The operation item of the operation, the data item includes the attribute data item of the anchor user, and the operation item includes one or more attribute tags; A data processing module, configured to perform predefined operations corresponding to the operation items on the corresponding data items in the multiple data sources according to the operation items in the data definition template, and determine the attribute labels in the operation items A corresponding data set, the data set includes a subset of data items corresponding to a plurality of anchor users; The data push module is used to push the data set corresponding to the live broadcast service to the terminal device in response to the data calling instruction of the live broadcast service from the terminal device. 9. A computer device comprising a central processing unit and a memory, wherein the central processing unit is used to call and run a computer program stored in the memory to execute the computer program according to any one of claims 1 to 7 steps of the method. 10. A computer-readable storage medium, characterized in that it stores a computer program implemented by the method according to any one of claims 1 to 7 in the form of computer-readable instructions, and the computer program is called by a computer At runtime, the steps included in the corresponding method are executed.

Images