CN112116391A

CN112116391A - Multimedia resource delivery method and device, computer equipment and storage medium

Info

Publication number: CN112116391A
Application number: CN202010988546.8A
Authority: CN
Inventors: 何攀; 李俊; 高小平; 郑秋野; 秦烁; 黄冲; 王建明; 王宗; 张丹峰
Original assignee: Beijing Dajia Internet Information Technology Co Ltd
Current assignee: Beijing Dajia Internet Information Technology Co Ltd
Priority date: 2020-09-18
Filing date: 2020-09-18
Publication date: 2020-12-22

Abstract

The disclosure relates to a multimedia resource delivery method, a multimedia resource delivery device, computer equipment and a storage medium, and belongs to the technical field of computers. In the embodiment of the disclosure, from the comment data, the user account concerning the target object can be determined, or the user account touched by the target object can be reflected, and the tendency of the user accounts to the target object can be analyzed from the comment data, based on the learned user accounts, a user set can be generated for the target object according to the analyzed tendency, the user set generated according to the tendency can include the user account required by the target object to drop multimedia resources, the user account triggered or concerned by the target object can drop the associated multimedia resources in a targeted manner, and the dropping effect of the resources can be effectively improved.

Description

Multimedia resource delivery method and device, computer equipment and storage medium

Technical Field

The present disclosure relates to the field of computer technologies, and in particular, to a multimedia resource delivery method and apparatus, a computer device, and a storage medium.

Background

With the development of computer technology, feed streams such as short videos and information are in endlessly, people spend more and more time on information and short videos, and traditional marketing advertisements are inclined to information stream advertisements more and more. By putting the multimedia resources (such as advertisements and the like) related to the brand in various forms, the brand can be popularized to the user account so as to improve the brand image.

In the related technology, the multimedia resource delivery method generally delivers all user accounts indiscriminately, and some multimedia resource network deliveries can select user accounts of some genders, ages and regions for delivery. Obviously, the releasing mode cannot accurately touch the user account of the brand, or pay attention to the user account of the brand, and the releasing effect is poor.

Disclosure of Invention

The present disclosure provides a multimedia resource delivery method, an apparatus, a computer device and a storage medium, and provides a directional resource delivery method, which effectively improves the delivery effect of resources. The technical scheme of the disclosure is as follows:

according to an aspect of the embodiments of the present disclosure, a multimedia resource delivery method is provided, including:

acquiring at least one piece of comment data of a target object;

determining a classification label of at least one target user account according to the at least one piece of comment data, wherein the classification label is used for indicating the tendency of the at least one target user account to the target object, and the target user account is a user account issuing the at least one piece of comment data;

responding to a user set generation instruction of the target object, and generating a user set according to a user account corresponding to a target classification label in the at least one target user account;

and responding to a multimedia resource releasing instruction of the target object, releasing multimedia resources to the user accounts in the user set, wherein the multimedia resources are associated with the target object.

Optionally, the determining, according to the at least one piece of comment data, a category label of at least one target user account includes:

inputting words in the at least one piece of comment data into a classification label determination model, performing feature extraction on the words by the classification label determination model, and performing trend classification based on the extracted features to obtain a classification label of the at least one target user account.

inputting words in the at least one piece of comment data into a classification label determination model, performing feature extraction on the classification label determination model based on semantic relations between each word and preceding and following words in the words of each piece of comment data to obtain features of each word, and performing trend classification on the features of each word based on a full connection layer to obtain a classification label of the at least one target user account.

Optionally, the classification tags include positive classification tags, negative classification tags, and neutral classification tags;

generating a user set according to the user account corresponding to the target classification label in the at least one target user account, wherein the user set comprises any one of the following items:

combining user accounts corresponding to forward classification labels in the at least one target user account into a first user set;

combining user accounts corresponding to the negative direction classification labels in the at least one target user account into a second user set;

and combining the user accounts corresponding to the positive direction classification labels and the negative direction classification labels in the at least one target user account into a third user set.

Optionally, the user set generation instruction comprises a target time period;

generating a user set according to the user account corresponding to the target classification tag in the at least one target user account, including:

screening the at least one target user account according to the target time period to obtain at least one first user account, wherein the first user account is a user account for publishing comment data in the target time period;

and generating a user set according to the user account corresponding to the target classification label in the at least one first user account.

Optionally, the method further comprises:

obtaining comment data in real time;

matching the comment data with an object library, and determining an object corresponding to the comment data;

and determining a classification label of a second user account for the object according to the comment data, wherein the second user account is a user account for publishing the comment data.

Optionally, the obtaining at least one piece of comment data of the target object; determining a classification label of at least one target user account according to the at least one piece of comment data, including:

and obtaining at least one comment data and a corresponding classification label of the target object from the comment data and the corresponding classification label of at least one object.

Optionally, the method further comprises:

writing the comment data acquired in real time and the corresponding classification tags into a database to generate corresponding first index information;

and responding to a query instruction of any comment data, and querying to obtain corresponding comment data and corresponding classification labels according to the first index information.

Optionally, the method further comprises:

periodically writing the acquired comment data and the corresponding classification tags into a database, and generating second index information corresponding to each period;

and responding to an inquiry instruction of any period of comment data, and inquiring the period of comment data and the corresponding classification label according to the second index information.

Optionally, the method further comprises:

and acquiring public opinion indication information of the target object based on the classification label of the at least one target user account, wherein the public opinion indication information is used for indicating the tendency of part or all of the user accounts to the target object.

Optionally, the obtaining of the public opinion indication information of the target object based on the classification label of the at least one target user account includes at least one of:

acquiring positive public opinion indication information of the target object according to the number of the positive classification labels;

acquiring negative public opinion indication information of the target object according to the quantity of the negative classification labels;

and acquiring the target public opinion indication information of the target object according to the first number of the positive classification labels and the second number of the negative classification labels.

Optionally, the method further comprises at least one of:

responding to the negative public opinion indication information being larger than a first target threshold value, and sending first alarm information;

responding to the positive public opinion indication information being larger than a second target threshold value, and sending prompt information, wherein the prompt information is used for prompting that the public opinion of the target object is good;

and responding to the target public opinion indication information being smaller than a third target threshold value, and sending second alarm information.

Optionally, the obtaining public opinion indication information of the target object based on the classification label of the at least one target user account includes:

acquiring public opinion indication information of the target object in each period based on the classification label corresponding to the at least one piece of comment data acquired in each period;

the method further comprises the following steps:

and responding to a viewing instruction of the public opinion indication information, and sending the public opinion change trend information of the target object according to the public opinion indication information of each period of the target object.

According to a second aspect of the embodiments of the present disclosure, there is provided a multimedia resource delivering apparatus, including:

an acquisition unit configured to perform acquisition of at least one piece of comment data of a target object;

a determining unit configured to perform determining, according to the at least one piece of comment data, a classification label of at least one target user account, where the classification label is used to indicate a tendency of the at least one target user account to the target object, and the target user account is a user account that publishes the at least one piece of comment data;

the generating unit is configured to execute a user set generating instruction responding to the target object, and generate a user set according to a user account corresponding to a target classification label in the at least one target user account;

and the releasing unit is configured to execute a multimedia resource releasing instruction responding to the target object and release multimedia resources to the user accounts in the user set, wherein the multimedia resources are associated with the target object.

Optionally, the determining unit is configured to perform inputting a word in the at least one piece of comment data into a classification label determination model, performing feature extraction on the word by the classification label determination model, and performing trend classification based on the extracted feature to obtain a classification label of the at least one target user account.

the generation unit is configured to perform any one of:

Optionally, the user set generation instruction comprises a target time period;

the generation unit is configured to perform:

screening the at least one target user account according to the target time period to obtain at least one first target user account, wherein the first target user account is a user account for publishing comment data in the target time period;

and generating a user set according to the user account corresponding to the target classification label in the at least one first target user account.

Optionally, the obtaining unit is further configured to perform obtaining comment data in real time;

the determining unit is further configured to perform matching of the comment data with an object library, and determine an object corresponding to the comment data;

the determining unit is further configured to determine a classification label of the object by a first user account according to the comment data, where the first user account is a user account issuing the comment data.

Optionally, the obtaining unit and the determining unit are configured to perform obtaining at least one comment data and a corresponding classification tag of the target object from comment data and a corresponding classification tag of at least one object.

Optionally, the generating unit is further configured to perform writing the comment data acquired in real time and the corresponding classification tag into a database, and generate corresponding first index information;

the apparatus further comprises a first query unit;

the first query unit is configured to execute a query instruction responding to any comment data, and corresponding comment data and corresponding classification labels are obtained through query according to the first index information.

Optionally, the generating unit is further configured to periodically write the acquired comment data and the corresponding classification tag into a database, and generate second index information corresponding to each period;

the apparatus further comprises a second query unit;

the second query unit is configured to execute a query instruction responding to any period of comment data, and query the period of comment data and the corresponding classification label according to the second index information.

Optionally, the obtaining unit is further configured to perform obtaining public opinion indication information of the target object based on the classification label of the at least one target user account, where the public opinion indication information is used to indicate a tendency of some or all of the user accounts to the target object.

Optionally, the obtaining unit is configured to perform at least one of:

Optionally, the apparatus further comprises a first transmitting unit configured to perform at least one of:

Optionally, the obtaining unit is configured to perform obtaining public opinion indication information of the target object in each period based on a classification label corresponding to the at least one piece of comment data obtained in each period;

the apparatus further comprises a second transmitting unit;

the second transmitting unit is configured to perform transmitting public opinion change trend information of the target object according to public opinion indication information of each period of the target object in response to a viewing instruction of the public opinion indication information.

According to a third aspect of embodiments of the present disclosure, there is provided a computer device comprising:

one or more processors;

one or more memories for storing the one or more processor-executable program codes;

wherein the one or more processors are configured to execute the program code to implement any of the above methods for multimedia asset delivery.

According to a fourth aspect of the embodiments of the present disclosure, there is provided a storage medium, wherein program code of the storage medium, when executed by one or more processors of a computer device, enables the computer device to execute any one of the multimedia resource delivery methods described above.

According to a fifth aspect of embodiments of the present disclosure, there is provided a computer program product comprising one or more program codes, which when executed by one or more processors of a computer device, enable the computer device to perform any one of the above-mentioned multimedia resource delivery methods.

The technical scheme provided by the embodiment of the disclosure at least brings the following beneficial effects:

in the embodiment of the disclosure, from the comment data, the user account concerning the target object can be determined, or the user account touched by the target object can be reflected, and the tendency of the user accounts to the target object can be analyzed from the comment data, based on the learned user accounts, a user set can be generated for the target object according to the analyzed tendency, the user set generated according to the tendency can include the user account required by the target object to drop multimedia resources, the user account triggered or concerned by the target object can drop the associated multimedia resources in a targeted manner, and the dropping effect of the resources can be effectively improved.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the disclosure and are not to be construed as limiting the disclosure.

Fig. 1 is a schematic diagram of an implementation environment of a multimedia resource delivery method according to an exemplary embodiment;

FIG. 2 is a flow chart illustrating a method of multimedia resource delivery, according to an exemplary embodiment;

FIG. 3 is a flow chart illustrating a method for multimedia resource delivery, according to an exemplary embodiment;

FIG. 4 is a flow chart illustrating a method of multimedia resource delivery, according to an exemplary embodiment;

fig. 5 is a schematic structural diagram illustrating a multimedia resource delivery apparatus according to an exemplary embodiment;

FIG. 6 is a block diagram illustrating the structure of a terminal according to one exemplary embodiment;

fig. 7 is a schematic diagram illustrating a configuration of a server according to an example embodiment.

Detailed Description

In order to make the technical solutions of the present disclosure better understood by those of ordinary skill in the art, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.

It should be noted that the terms "first," "second," "third," and the like in the description and claims of the present disclosure and in the above-described drawings are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the disclosure described herein are capable of operation in sequences other than those illustrated or otherwise described herein. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

The user information to which the present disclosure relates may be information authorized by the user or sufficiently authorized by each party.

The following terms related to the present disclosure are explained as follows.

Public opinion: the term "public opinion" is short for "public opinion" and refers to the social attitude of the public as a subject to the orientation of social managers, enterprises, individuals and other organizations as objects, politics, society, morality and the like around the occurrence, development and change of social events of intermediaries in a certain social space. Public sentiment is the sum of the expressions of beliefs, attitudes, opinions, emotions, and the like expressed by more people about various phenomena, problems, and the like in the society.

Information stream (Feeds): has two definitions, broad and narrow. In a broad sense, information flow refers to the way people exchange information in various ways, from direct face-to-face conversation to the use of various modern transmission media, including channels and processes for information collection, transmission, processing, storage, retrieval, analysis, and so on. In particular, an information stream may refer to a set of information that is in the process of moving in the same direction, both spatially and temporally, that has a common information source and receiver of information, i.e. a collection of all information passed from one information source to another unit. Such flow can occur from person to person, person to institution, institution to institution, and institution to institution, including both tangible and intangible flow, the former being statements, drawings, books, and the like, and the latter being electrical, acoustic, optical, and the like.

In a narrow sense, from the viewpoint of modern information technology research, development and application, information flow refers to the flow of information in computer systems and communication networks during information processing.

Tendency: trends refer to a concept of the theory of utility, which refers to the attitude of decision makers to gain and risk. The tendency refers to a tendency of subjective internal happiness and dislike of a subject to a certain object, and belongs to internal evaluation. The tendency is both degree and targetability. For example, the classification tags include a positive classification tag, a negative classification tag, and a neutral classification tag. As another example, the trend may include very positive, neutral, negative, very negative.

The user's tendency can also be understood as the emotion of the user. The emotion is a part of the attitude in the whole, has coordination consistency with the inward feeling and the intention in the attitude, and is a physiological evaluation and experience of the attitude which is more complex and stable in physiology.

The following describes an environment in which the present disclosure may be implemented.

Fig. 1 is a schematic diagram of an implementation environment of a multimedia resource delivery method according to an exemplary embodiment, and as shown in fig. 1, the implementation environment includes at least one terminal 101 and a multimedia resource delivery platform 110. At least one terminal 101 is connected to the multimedia resource delivery platform 110 through a wireless network or a wired network.

The multimedia resource delivery platform 110 is, for example, at least one of a terminal, one or more servers, a cloud computing platform, and a virtualization center.

The terminal 101 is, for example, at least one of a smartphone, a game console, a desktop computer, a tablet computer, an electronic book reader, an MP3(Moving Picture Experts Group Audio Layer III, motion Picture Experts compression standard Audio Layer 3) player, an MP4(Moving Picture Experts Group Audio Layer IV, motion Picture Experts compression standard Audio Layer 4) player, and a laptop computer. The terminal is installed and operated with a multimedia resource application. The application program may be a client application or a browser application.

The multimedia resource delivery platform 110 is configured to provide a multimedia resource delivery service for the terminal 101. Illustratively, the terminal 101 can log in a user account to comment on any multimedia resource, and send comment data to the multimedia resource delivery platform 110. The multimedia resource delivery platform 110 can process the received comment data and determine a predicted classification label of the user account for an object, where the object is an object associated with the multimedia resource or an object related to the comment data. The multimedia resource delivery platform 110 obtains the comment data and the corresponding prediction classification label, and can analyze a certain object to determine a user set for the object, so as to subsequently deliver the multimedia resource of the object to the user account in the user set in a targeted manner.

The multimedia resource delivery platform 102 includes at least one of a server, a plurality of servers, a cloud computing platform, and a virtualization center. The multimedia resource delivery platform 102 is configured to provide a background service for an application supporting multimedia resource delivery. Optionally, the multimedia resource delivery platform 102 undertakes primary processing, and the terminal 101 undertakes secondary processing; or, the multimedia resource delivery platform 102 undertakes the secondary processing work, and the terminal 101 undertakes the primary processing work; or, the multimedia resource delivery platform 102 or the terminal 101 can respectively undertake processing separately. Or, the multimedia resource delivery platform 102 and the terminal 101 perform collaborative computing by using a distributed computing architecture.

Optionally, the multimedia resource delivery platform 102 includes at least one server 1021 and a database 1022, where the database 1022 is used to store data, and in the embodiment of the present disclosure, the database 1022 can store a sample image or a sample face image, so as to provide a data service for the at least one server 1021.

The server can be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, and a cloud server providing basic cloud computing services such as cloud service, a cloud database, cloud computing, cloud functions, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN, big data and artificial intelligence platform. The terminal can be, but is not limited to, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and the like.

Those skilled in the art will appreciate that the number of the terminals 101 and the servers 1021 can be greater or smaller. For example, the number of the terminals 101 and the servers 1021 may be only one, or the number of the terminals 101 and the servers 1021 may be several tens or several hundreds, or more, and the number of the terminals or the servers and the device types are not limited in the embodiment of the present disclosure.

The quality of public sentiment reflects the user's mood of an object such as a brand, a product or an event, and the brand is taken as an example, the user's reaction to the object such as the brand is concerned, and the brand image is improved and maintained. If negative direction of brand public opinion is found to be serious, certain measures can be taken to improve the brand image. For example, the brand may be advertised or the brand image may be enhanced by improving the quality of the brand. By considering the public opinion characteristics, the user concerned by the brand or the user reached by the brand is known in the resource release and the impression of the brand is given to the users, so that part of the users are selected to directionally release the multimedia resource, the image of the brand in the user centers can be effectively improved, the users know the brand, the multimedia resource release can attract the attention of the users, and the effectiveness and the effect of the resource release can be effectively improved. The following describes the flow of the multimedia resource delivery method through the embodiments shown in fig. 2 and fig. 3.

Fig. 2 is a flowchart illustrating a multimedia resource delivery method applied to a computer device, which is a terminal or a server, according to an exemplary embodiment, and referring to fig. 2, the method includes the following steps.

201. The computer device obtains at least one piece of comment data of the target object.

202. And the computer equipment determines the classification label of at least one target user account according to the at least one piece of comment data, wherein the target user account is the user account which issues the at least one piece of comment data.

203. And the computer equipment responds to the user set generation instruction of the target object and generates a user set according to the user account corresponding to the target classification label in the at least one target user account.

204. And the computer equipment responds to the multimedia resource releasing instruction of the target object and releases the multimedia resource to the user account in the user set, wherein the multimedia resource is associated with the target object.

Optionally, the determining a category label of at least one target user account according to the at least one piece of comment data includes:

and inputting the words in the at least one piece of comment data into a classification label determination model, performing feature extraction on the words by the classification label determination model, and performing trend classification based on the extracted features to obtain a classification label of the at least one target user account.

and inputting the words in the at least one piece of comment data into a classification label determination model, performing feature extraction on the classification label determination model based on the semantic relation between each word in each comment data and the preceding and following words to obtain the features of each word, and performing trend classification on the features of each word based on the full-link layer to obtain the classification label of the at least one target user account.

combining user accounts corresponding to the forward direction classification labels in the at least one target user account into a first user set;

combining the user accounts corresponding to the negative direction classification labels in the at least one target user account into a second user set;

and combining the user accounts corresponding to the positive classification label and the negative classification label in the at least one target user account into a third user set.

Optionally, the user set generation instruction comprises a target time period;

generating a user set according to the user account corresponding to the target classification label in the at least one target user account, including:

screening the at least one target user account according to the target time period to obtain at least one first user account, wherein the first user account is a user account for issuing comment data in the target time period;

Optionally, the method further comprises:

obtaining comment data in real time;

matching the comment data with an object library to determine an object corresponding to the comment data;

and determining a classification label of a second user account for the object according to the comment data, wherein the second user account is a user account for issuing the comment data.

Optionally, the obtaining at least one piece of comment data of the target object; determining a classification label of at least one target user account according to the at least one piece of comment data, wherein the classification label comprises:

Optionally, the method further comprises:

and responding to an inquiry instruction of the comment data in any period, and inquiring the comment data in the period and the corresponding classification label according to the second index information.

Optionally, the method further comprises:

Optionally, the method further comprises at least one of:

responding to the positive public opinion indication information being larger than a second target threshold value, sending prompt information, wherein the prompt information is used for prompting that the public opinion of the target object is good;

and sending second alarm information in response to the target public opinion indication information being smaller than a third target threshold value.

the method further comprises the following steps:

and responding to a viewing instruction of the public sentiment indication information, and sending the public sentiment change trend information of the target object according to the public sentiment indication information of each period of the target object.

Fig. 3 is a flowchart illustrating a multimedia resource delivery method according to an exemplary embodiment, and referring to fig. 3, the method includes the following steps.

301. And the computer equipment acquires the comment data in real time.

Comments are subjective or objective self-impression statements about things. The comment data can reflect the attitude, idea, viewpoint or tendency of the reviewer to the object to be commented on. Thus, the computer device can acquire comment data, and analyze the idea of the user account issuing the comment data from the comment data.

In the embodiment of the disclosure, a user can log in a user account on a terminal and operate on the terminal to publish comment data through the user account, and the computer device can acquire the comment data published by the user account in real time.

In a possible implementation manner, the computer device has a comment data publishing function, the terminal can send comment data to be published to the computer device, and the computer device receives the comment data, namely, the process of obtaining the comment data in real time is realized, and the comment data can be published to the corresponding network platform. The network platform refers to various network service support systems and network service activities based on the internet.

In another possible implementation manner, the computer device can monitor at least one website or at least one network platform, and when a piece of comment data is published by the at least one website or at least one network platform, the comment data can be obtained in real time.

Optionally, the process of the computer device obtaining the comment data from the website or the network platform may be: the computer equipment sends a comment data acquisition request to the equipment where the website or the network platform is located, and the equipment where the website or the network platform is located receives the data acquisition request and can respond to the comment data acquisition request and return the comment data to the computer equipment.

Optionally, the process of the computer device obtaining the comment data from the website or the network platform may be: and the computer equipment acquires the comment data published by the website or the network platform through the web crawler. Among them, web crawlers (also called web spiders, web robots, often called web chasers in FOAF communities) are programs or scripts that automatically capture web information according to certain rules.

The above provides several possible implementation manners of obtaining the comment data by the computer device, and of course, the obtaining process may also be implemented in other manners, and the embodiment of the present disclosure adopts any manner, which is not particularly limited.

302. And the computer equipment matches the comment data with an object library and determines an object corresponding to the comment data.

The object library may include a plurality of objects, the computer device needs to determine which object the comment data is directed to after acquiring the comment data, and the tendency of the object by the user account represented by the comment data may be analyzed after the determination. The object refers to a brand, product, commodity, event, or the like.

In a possible implementation manner, the object library may include names of a plurality of objects, the computer device matches the text content in the comment data with the names of a plurality of candidate objects in the object library, and the candidate object whose matching degree satisfies a condition is taken as the object corresponding to the comment data.

In a specific possible embodiment, the computer device can determine a matching degree between the text content of the comment data and the name of each candidate object, and take the candidate object with the matching degree greater than a threshold matching degree as the object corresponding to the comment data. Or, the computer device takes the candidate object with the highest matching degree in a plurality of matching degrees with the matching degree larger than the threshold value of the matching degree as the object corresponding to the comment data. Or the computer device takes the candidate object with the maximum matching degree as the object corresponding to the comment data. The matching degree satisfying condition in the embodiment of the present disclosure is not particularly limited.

303. And the computer equipment determines the classification label of a second user account for the object according to the comment data, wherein the second user account is the user account for issuing the comment data.

After the computer device determines the object corresponding to the comment data, the computer device may further determine a classification label of the user account for the object according to the comment data. The category label can express the user account's propensity for the object,

in one possible implementation, the category label includes three types: positive, negative and neutral. Through the determination of the classification label, whether the tendency of the user account to the object is positive, negative or neutral is analyzed, so that whether the tendency of the user account to the object is possibly improved if the multimedia resource of the object is released to the user account or whether the multimedia resource is easily accepted by the user account if the multimedia resource of the object is released to the user account can be determined.

In another possible implementation manner, the classification label may also adopt other classification manners, for example, the classification label includes five types: a very positive category label, a neutral category label, a negative category label, and a very negative category label. For another example, the category labels include two types: positive and negative classification tags, or positive and negative classification tags. The embodiment of the present disclosure does not limit the specific classification of the classification label.

In one possible implementation, the determination process of the classification label may be implemented by a classification label determination model, which is used to determine a classification label corresponding to the input comment data. Specifically, the computer device inputs words in the comment data into the classification label determination model, feature extraction is performed on the words by the classification label determination model, trend classification is performed based on the extracted features, and classification labels of the second user account are obtained. Optionally, the computer device may input a word in the at least one piece of comment data into the classification label determination model, perform feature extraction by the classification label determination model based on a semantic relationship between each word and preceding and following words in the word of each piece of comment data to obtain a feature of each word, perform trend classification on the feature of each word based on the full connection layer, and obtain a classification label of the at least one target user account.

The classification label determination model can automatically complete trend classification, and corresponding classification labels can be obtained by outputting words in the comment data into the classification label determination model through black box processing. The determination mode of the classification label is fast and efficient, reduces I/O (Input/Output) throughput of data, and improves data processing efficiency. For example, the class label determination model may be an LSTM (Long Short Term Memory) model, which may include a three-layer neural network. Of course, the classification label determination model may also adopt other structures, which is not limited in this disclosure.

In one possible implementation, the computer device may pre-process the comment data to obtain words in the comment data before inputting the words into the classification label determination model.

In a particular possible embodiment, the pre-processing process may include a word segmentation process. Specifically, the computer device may perform word segmentation processing on the comment data, and then input a word segmentation result into the classification label determination model. The word segmentation result may include words in the comment data.

Word segmentation refers to a process of decomposing a piece of content into a plurality of independent smaller text units such as words, phrases, and phrases. The computer equipment can be configured with a configuration file with a word segmentation function, and the terminal can perform word segmentation processing on the comment data content through the word segmentation function. Specifically, the computer device recombines the continuous word sequences in the comment data, if the combined phrase or phrase is a regular noun, the combined phrase or phrase is taken as a word segmentation result, and if the combined phrase or phrase is not a regular noun, the combination is ignored. Through this process, the computer device may tokenize the review data into at least one tokenization result.

Alternatively, the word segmentation process may be implemented by a word segmentation model trained based on a large amount of text and known word segmentation results. For example, the word segmentation Model may be a natural language processing Model such as a Hidden Markov Model (HMM) or a Conditional Random Field (CRF) Model. Of course, other models are also possible, and the embodiment of the disclosure does not limit this. Alternatively, the word segmentation process may also be implemented in other ways, for example, a word segmentation algorithm, or word segmentation software (such as a jieba word segmentation component), and the like. The embodiment of the present disclosure does not limit which way to perform word segmentation specifically.

In a possible implementation manner, after the word segmentation, the computer device may further remove some semantic-free words, and input the processed words into the classification label determination model. The semanteme-free word can refer to a word without semantics, for example, the semanteme-free word can be a help word, a stop word, etc. Stop Words refer to that in information retrieval, in order to save storage space and improve search efficiency, some characters or Words are automatically filtered before or after processing natural language data, and the characters or Words are called Stop Words. The stop words are manually input and are not automatically generated, and the generated stop words form a stop word list.

The process is a data cleaning process, and the data cleaning refers to a last procedure for finding and correcting recognizable errors in a data file, and comprises the steps of checking data consistency, processing invalid values and missing values and the like. In the embodiment of the disclosure, the computer device may perform data cleaning on the segmentation result, and remove redundant data or invalid data in the segmentation result. For example, the invalid data may include the above-described semantically-free word.

In a specific possible embodiment, besides the text content, other forms of content may exist in the comment data, such as expressions and the like. The computer equipment can also remove the non-text content before word segmentation to obtain the text content of the comment data, and then carry out word segmentation processing on the text content.

Words which have no effect on analysis tendency in all words contained in the comment data are removed through data cleaning, so that the calculation amount of the data can be reduced when classification labels are determined subsequently, the influence of the words is avoided, and the accuracy can be improved.

In a particular possible embodiment, the words of the comment data are in the form of word vectors. The computer device can convert the words of the processed comment data into word vectors, and the word vectors are input into the classification label determination model. The comment data are converted into the expression mode of the word vector, so that the comment data can be quickly processed, and the processing effect is improved. The obtaining mode of the word vector may be implemented by a natural language model, or may be implemented by other natural language processing modes, which is not limited in the embodiments of the present disclosure. For example, the computer device can translate words of the comment data into 256-dimensional word vectors via word2 vec. The word2vec is a group of correlation models used to generate word vectors.

The classification label determination model can be obtained by training based on the sample comment data and the target classification label corresponding to each comment data. Specifically, the training process of the classification label determination model is realized through the following steps one to four:

the method comprises the steps of firstly, obtaining sample comment data, wherein each sample comment data corresponds to a target classification label.

In the first step, the computer device may collect a large amount of comment data as sample comment data, label each sample comment data, and determine a target classification label corresponding to each sample comment data. The sample comment data may be comment data of the target object, or comment data of other objects. The sample comment data may include comment data for one or more objects, which is not limited by the disclosed embodiments.

Similarly to the above step 303, the computer device may also perform preprocessing such as word segmentation on the sample comment data, and then input the preprocessed sample comment data into the initial model for training.

For example, the computer device extracts comment data of a brand short video for nearly 30 days, marks it, and classifies the comment data as positive, negative, neutral.

And step two, inputting the words of the sample comment data into an initial model, and determining a prediction classification label of the sample user account on the object indicated by the sample comment data by the initial model.

In the second step, the model parameters of the initial model are initial values, the initial model can perform feature extraction on the words of the sample comment data, perform trend classification based on the extracted features, and determine the classification label corresponding to each sample comment data.

And step three, acquiring the error of the prediction classification label according to the prediction classification label and the target classification label.

By obtaining the error, the difference between the predicted classification label and the target classification label can be measured, and it can be understood that the smaller the difference, the smaller the error of the predicted classification label, and the higher the accuracy of the predicted classification label.

The error may be determined by a loss function, which may be a cross entropy function, for example, or by other objective functions.

And step four, according to the error, adjusting the model parameters of the initial model until the model parameters meet the target conditions, and obtaining a classification label determination model.

The first step to the third step are iterative processes, after each iterative process, the model parameters can be adjusted according to errors, so that the prediction accuracy of the model is improved, and the trained classification label determination model can accurately process comment data.

The above steps 301 to 303 are processes of acquiring, by a computer device, the comment data and the corresponding classification tags in real time, and the computer device can monitor the classification tags corresponding to the comment data in real time, grasp the classification tags of the user to the object in real time, and help to quickly make a corresponding policy for the classification tags or take corresponding measures, so that the impression of the user to the object can be quickly improved.

In a possible implementation manner, after determining the classification tag corresponding to the comment data, the computer device may write the classification tag into a database for storage and management, where the database is used for storage and management of the comment data and the corresponding classification tag. Specifically, the computer device may acquire the comment data and the corresponding classification tag in real time, that is, write the comment data and the corresponding classification tag into the database, or may write the comment data and the corresponding classification tag into the database periodically, that is, write the comment data and the corresponding classification tag into the database in batches. See, in particular, the following modes one and two.

In the first mode, the computer equipment writes the comment data acquired in real time and the corresponding classification tags into the database to generate corresponding first index information.

In the first mode, the computer device obtains the comment data in real time through the above steps 301 to 303, and after determining the classification tag, may directly write the comment data into the database.

In a possible implementation manner, the computer device can write the comment data into the database by importing the comment data into a database connection pool, the database connection pool comprises one or more database connections, and the database connection pool does not need to create a connection from the database every time, so that connection creation times are saved, and processing efficiency is improved.

For example, the database connection pool may be a pipeline, which is a solution in a big data scenario and is a BI (Business Intelligence)/OLAP (Online Analytical Processing) tool that needs to perform interactive real-time data presentation under complex massive data. The pipeline is a database link pool, is an efficient data query system, and is a real-time OLAP system. The computer equipment can directly import the comment data acquired in real time into the pipeline in real time.

After the relevant data of the comment data are imported into the database by the computer equipment, corresponding first index information can be generated for the comment data, and the relevant data of the comment data can be indexed through the first index information. Therefore, if the user wants to view the relevant data of the comment data, the user can quickly inquire the comment data through the first index information.

Specifically, the computer device responds to an inquiry instruction of any comment data, and inquires and obtains the corresponding comment data and the corresponding classification label according to the first index information. The first index information is used for indexing the comment data and the classification label corresponding to the comment data.

And in the second mode, the computer equipment periodically writes the acquired comment data and the corresponding classification tags into the database to generate second index information corresponding to each period.

The computer equipment can store the comment data acquired in real time into the database table, and periodically write the data in the database table into the database in batches. Therefore, the data of each period can generate second index information, and the data of one period can be indexed through the second index information.

For example, taking a cycle as one day as an example, the computer device can directly import the relevant data of the comment data acquired in real time into the offline hive, and then import the relevant data into the drive on a daily level. hive is a data warehouse tool based on Hadoop, which is used for data extraction, transformation and loading, and is a mechanism capable of storing, querying and analyzing large-scale data stored in Hadoop. The hive data warehouse tool can map the structured data file into a database table, provide SQL query function and convert SQL sentences into MapReduce tasks for execution.

For the query, in the second mode, the computer device responds to a query instruction for any period of comment data, and obtains the period of comment data and the corresponding classification tag according to the second index information.

304. The computer equipment obtains the comment data and the corresponding classification label of the target object from the comment data and the corresponding classification label of the at least one object. Based on the acquired data, the computer apparatus executes steps 305 to 306 described below, and executes steps 307 to 308 described below.

The target object may be an object that the user wants to analyze, for example, the user's own brand, an item on the user's own brand, for example, an event that the user wants to know, and the like.

In the above steps 301 to 303, the computer device determines the corresponding object and the classification tag for the comment data acquired in real time, and may obtain the comment data of at least one object and the corresponding classification tag by performing the above processes one or more times. When a certain object needs to be analyzed, the computer device may extract the comment data of at least one object obtained through the above process.

In the above implementation of writing the comment data and the corresponding classification tag into the database, in step 304, the computer device may extract at least one comment data and a corresponding classification tag of the target object from the database.

The steps 301 to 304 are a process of acquiring at least one piece of comment data of a target object, and determining a classification tag of at least one target user account according to the at least one piece of comment data, where the target user account is a user account issuing the at least one piece of comment data. In the above process, the comment data and the corresponding classification label are obtained in real time by the computer device for example. In another possible implementation manner, the computer device may also obtain at least one piece of comment data of the target object when the target object needs to be analyzed, and determine a classification label corresponding to the at least one piece of comment data. For example, the computer device may directly obtain at least one piece of comment data of the target object from the server corresponding to the target object, and perform the steps similar to step 303 described above.

305. And the computer equipment responds to the user set generation instruction of the target object and generates a user set according to the user account corresponding to the target classification label in the at least one target user account.

The user set generation instruction may be triggered by a user set generation operation of the user or periodically by the computer device. For example, a user wants to determine a user set according to the current public sentiment of the brand, so that when multimedia resources of the target object are released, the user accounts in the user set can be released, directional resource release is performed according to the public sentiment, and a better release effect can be achieved. Wherein the user set comprises a plurality of user accounts.

In an implementation manner in which the classification tags include positive classification tags, negative classification tags, and neutral classification tags, the computer device may generate different user sets according to different target classification tags, where the target classification tags may be set by related technicians as needed or selected by users, and the target classification tags selected by the users may be included in the user set generation instruction. This step 305 may include the following three cases, depending on the target classification label.

The first condition is as follows: and the computer equipment combines the user accounts corresponding to the forward direction classification labels in the at least one target user account into a first user set.

Case two: and the computer equipment combines the user accounts corresponding to the negative direction classification labels in the at least one target user account into a second user set.

Case three: and the computer equipment combines the user accounts corresponding to the positive classification label and the negative classification label in the at least one target user account into a third user set.

The above generation conditions of the three user sets are provided, any one of the conditions may be adopted in the embodiment of the present disclosure, any combination of the three conditions may also be adopted according to requirements, and the user or the technician may set or select the condition according to the requirements, which is not limited in the embodiment of the present disclosure.

In one possible implementation manner, the public sentiment has newness, and changes along with the change of time, so that when the public sentiment is analyzed to perform directional resource delivery, comment data in a certain time period can be acquired for analysis. The target time period may also be set when generating the user set so that the generated user set is adapted to the current public opinion.

In one possible implementation, the user set generation instruction includes a target time period, and the generation process of the user set may be: and the computer equipment screens the at least one target user account according to the target time period to obtain at least one first user account, wherein the first user account is a user account for publishing comment data in the target time period, and a user set is generated according to the user account corresponding to the target classification tag in the at least one first user account.

After the computer device determines the user accounts corresponding to the target classification tag in the at least one first user account, the user accounts may be combined into a user set. The process is the same as the process of generating the user set through combination in the above cases one to three, and details are not repeated here.

The generation process is described by taking an example that the computer device firstly screens the target user account numbers in the target time period and then further screens the target user account numbers based on the target classification tags, or the computer device firstly screens the target user account numbers based on the target classification tags and then further screens the target user account numbers based on the target time period, and combines the user account numbers obtained by the two steps of screening into a user set.

For example, a user may circle users who want certain category labels within a certain time period as targeted resource placement populations. Packaging process of the crowd management platform: the advertiser can select positive and negative crowds of some brands for a period of time to pack to generate a targeted crowd package, and the targeted crowd package can be put in a targeted mode after the targeted crowd package takes effect. Corresponding to the generated user set

In a possible implementation manner, the target time period obtained in the above implementation manner may be referred to as a first target time period, and the at least one comment data obtained in step 304 may be comment data in a second target time period. That is, the step 304 is: and acquiring at least one piece of comment data of the target object in a second target time period. Wherein the second target time period comprises the first target time period. That is, the first target time period may be the same as the second target time period, or may be a time period within the second target time period, and the length of the first target time period is smaller than that of the second target time period.

Through the above steps 301 to 305, the computer device can determine one or more user sets for the target object, and when multimedia resource delivery is required, can execute the following step 306. Of course, the computer device may also determine one or more sets of users for other objects.

306. And the computer equipment responds to the multimedia resource releasing instruction of the target object and releases the multimedia resource to the user account in the user set, wherein the multimedia resource is associated with the target object.

The multimedia resource release instruction may be triggered by a multimedia resource release operation of a user, or may be triggered by the computer device when receiving a multimedia resource to be released, which is not limited in the embodiment of the present disclosure. The multimedia resource can be a resource in a text form, a voice form, an image form, a video form or an audio form. The embodiment of the present disclosure does not limit the form of the multimedia resource.

When the target object multimedia resource needs to be released, the computer device may directly release the target object multimedia resource to the user account in the user set generated in step 305. Specifically, in the above process, the computer device may obtain an address of the user account of the user set, and release the multimedia resource to the address.

307. And the computer equipment acquires public opinion indication information of the target object based on the classification label of the at least one target user account, wherein the public opinion indication information is used for indicating the tendency of part or all of the user accounts to the target object.

In step 304, after the computer device obtains the comment data of the target object and the corresponding classification label, the computer device may further analyze the public sentiment of the target object, and the public sentiment indication information is used for representing the public sentiment. The tendency of the user account for publishing the comment data to the target object can be comprehensively known through the public opinion indication information.

In one possible implementation manner, the public opinion indication information may include one or more public opinion indication information, different classification tags may correspond to different public opinion indication information, and the process of acquiring the public opinion indication information by the computer device in the step 307 may include one, any two, or all of the following three cases, which is not limited by the embodiment of the disclosure.

In the first situation, the computer device obtains the forward public opinion indication information of the target object according to the number of the forward classification labels.

In one possible implementation manner, the public opinion indication information may be in the form of an index or a score, and when the number of forward classification tags is 300, the forward public opinion indication information is 300. The public opinion indication information can also be updated in real time when the comment data and the corresponding classification label are obtained in real time. When a forward classification label exists, one is added to the forward public opinion indication information.

And in the second situation, the computer equipment acquires negative public opinion indication information of the target object according to the quantity of the negative classification labels.

In one possible implementation, the public sentiment indicator may be in the form of an index or a score, and when the number of negative classification tags is 300, the negative public sentiment indicator is 300. The public opinion indication information can also be updated in real time when the comment data and the corresponding classification label are obtained in real time. When a negative classification label exists, one is added to the negative public opinion indication information.

And thirdly, the computer equipment acquires the target public opinion indication information of the target object according to the first number of the positive direction classification labels and the second number of the negative direction classification labels.

In one possible implementation manner, the public opinion indication information may be in the form of an index or a score, where a negative classification label corresponds to a negative score and a positive classification label corresponds to a positive score. For example, when the number of positive classification tags is 4000, the number of negative classification tags is 1000, and the public opinion indication information is 3000. The public opinion indication information can also be updated in real time when the comment data and the corresponding classification label are obtained in real time. When a forward classification label exists, a positive classification label is added to the public sentiment indication information. When a forward classification label exists, one is subtracted from the public sentiment indication.

308. And the computer equipment responds to the public opinion indication information meeting the target condition and sends alarm information or prompt information.

After the computer equipment acquires the public sentiment indication information, whether the current public sentiment of the target object needs to be alarmed or prompted to the user can be determined according to the public sentiment indication information. For example, if the public opinion indication information indicates that the public opinion of the target object is not good, an alarm message may be sent to alert the user or the related technical person, so that the user or the related technical person can take measures in time to improve the image of the target object. Therefore, the alarm or prompt message is automatically sent, the user or related technical personnel do not need to analyze the alarm or prompt message by themselves, and the real-time performance is better.

In implementations where the public opinion indication may include one or more different forms, the computer device may perform one, any two, or all of the following three scenarios. Specifically, which case or cases to use may be set by a user or a technician according to needs, which is not limited in the embodiment of the present disclosure.

In case one, the computer device sends first warning information in response to the negative public opinion indication information being greater than a first target threshold.

In this case one, if the public sentiment of the target object is poor by analyzing, the first warning information may be transmitted to warn that the public sentiment of the target object is poor. Therefore, the user can timely acquire the information with poor public sentiment of the target object by receiving the first warning information, so as to timely take measures. Such as advertising, etc.

And in the second situation, the computer equipment responds to the positive public opinion indication information being larger than the second target threshold value, and sends prompt information for prompting that the public opinion of the target object is good.

And in case III, the computer equipment responds to the target public opinion indication information being smaller than a third target threshold value, and sends second alarm information.

The target public opinion indication information is information of a comprehensive positive and negative classification label, and can represent the public opinion of the target object as a whole, so that if the target public opinion indication information is smaller, the public opinion of the target object is not good, and the second alarm information can be sent to alarm.

The first target threshold, the second target threshold and the third target threshold may be set by a relevant technician as required, for example, taking the first target threshold as an example for description, the first target threshold may be 3000, and when the negative public opinion indication information reaches 3000, an alarm may be sent.

It should be noted that, in the embodiment of the present disclosure, taking the example that the computer device performs both step 305 and step 306, and step 307, and performs step 308 when the public opinion indicating information meets the target condition, the step 307 and step 308 are optional steps, and in a possible embodiment, the computer device may perform only step 305 and step 306 after the step 304. In another possible embodiment, the computer device may also perform step 305 and step 306 after step 304, and perform step 307. In another possible embodiment, the computer device may also perform step 305 and step 306 after step 304, and perform step 307 and step 308. The above-described functions may be determined by a person skilled in the relevant art according to a need, and the embodiment of the present disclosure is not particularly limited thereto.

In a possible implementation manner, in the step 307, the computer device may periodically obtain the public opinion indication information, and further analyze the public opinion change trend of the target object according to a plurality of periodic public opinion indication information. Specifically, in step 307, the computer device obtains the public opinion indication information of the target object in each period based on the classification label corresponding to the at least one piece of comment data obtained in each period. If the user wants to check the public opinion indicating information of the target object, the user can check the public opinion indicating information, and when the computer equipment receives a checking instruction triggered by the checking operation, the computer equipment can respond to the checking instruction of the public opinion indicating information and send the public opinion change trend information of the target object according to the public opinion indicating information of each period of the target object. For example, taking a day as a cycle and a target brand as an example, the computer device may calculate daily public opinion indication information of the target brand, and when looking up the public opinion indication information of the target object, may convert the multiple days of public opinion indication information into public opinion change trend information. For example, the public opinion change trend information is a broken line, the broken line is formed by connecting a plurality of points, and each point is used for identifying daily public opinion indication information.

One specific example is provided below. As shown in fig. 4, in this specific example, the training process and the using process of the classification label determination model are introduced. In the model training process, comment data of brand short videos of nearly 30 days can be extracted and marked, and the comment data are divided into positive direction, negative direction and neutral. And then segmenting the comment text by utilizing the jieba (jieba) segmentation to remove some auxiliary words and stop words, and converting the words into 256-dimensional word vectors by utilizing word2 vec. Thus, sample data can be generated, and the sample is processed according to the following steps of 8: the ratio of 2 is divided into a training sample (train label) and a test sample (test label). Then a model of a 3-layer LSTM network structure can be defined, training samples are input, an LSTM public opinion model is generated after training is finished, finally, testing is carried out by using testing samples, when the final accuracy rate reaches more than 90%, model training can be considered to be finished, and the model is stored.

During the model using process, a user reviews the short video in real time, and sends review data to kafka in real time, wherein the kafka is an open source stream processing platform developed by an Apache software foundation. kafka is a high-throughput distributed publish-subscribe messaging system that can handle all the action flow data of a consumer in a web site. Therefore, the user publishes the comment data, and the comment data can be processed by subscribing the comment data of the user at kafka to determine public sentiment. Specifically, the comment data can be consumed (processed) in real time through the flink, a video brand library (namely, the object library, the object is taken as the video brand for example) is inquired, whether a certain brand is hit or not is judged, then the comment phrases are sequentially thrown into a Word2vec model and an LSTM prediction model, a positive and negative neutral public opinion result (namely, a classification label) is obtained, and then the brand public opinion result is sent to kafka. Among them, flink is an open source stream processing framework developed by the Apache software foundation. flink is capable of executing arbitrary stream data programs in a data parallel and pipelined manner, and flink's pipelined runtime system can execute batch and stream processing programs. After the brand consensus results are determined, the data may also be imported into the olap engine, where the data is consumed twice. Once, data was imported directly by kafka to offline hive, then daily to druid. And the other time is directly led into the pipeline in real time. The importing step also corresponds to the two ways of writing into the database. After the data is imported into the druid, an efficient index is generated, and when an advertiser (a management user of a target object) clicks a public opinion platform (namely a platform for processing comment data) to inquire the public opinion condition of a certain brand, the public opinion platform can acquire the public opinion index distribution trend (namely public opinion change trend information) of the brand in a near period of time through the druid api in real time. The public opinion platform can monitor the public opinions of the brands in real time, perform public opinion analysis, and determine brand indexes (namely public opinion indication information), wherein the brand indexes can comprise positive/negative indexes and the like. If the alarm is set, when the negative index (namely, the negative public opinion indication information) of the brand reaches 3000 points, short messages and mails are sent to inform the advertiser, namely, the first alarm information is sent. The advertiser can select some positive and negative crowds of the brand for a period of time to pack to generate a targeted crowd package (namely a user set), and the targeted crowd package can be targeted for delivery after the targeted crowd package takes effect. For the crowd package, a dmp (data Management platform) data Management platform may be responsible for crowd Management, and generate a public opinion crowd package (i.e., a user set). The advertisement engine can directionally deliver the advertisement materials (namely the multimedia resources of the target objects) of the brand advertisers to the part of people (namely the generated user set) defined in the public sentiment, so that accurate recommendation between the label crowd and the advertisement is realized. The label crowd is that the classification label is determined for the user through the classification label determining process, and the classification label is used as the label of the user, so that the user can be the label crowd.

In this specific example, the embodiment of the present disclosure may enable a brand advertiser to monitor the trend of positive and negative public sentiment indexes of its own brand in real time, and assist the advertiser in customs and advertisement placement decisions. And the brand advertisers can also monitor the public sentiment trend of some relevant competitive products, and the popular exposure short videos of the brand advertisers and the competitive products. This disclosed embodiment can also carry out the advertisement putting with the packing of brand positive and negative direction public opinion crowd to optimize the directional effect of label, realize the accurate recommendation between label crowd and the advertisement.

All the above optional technical solutions can be combined arbitrarily to form the optional embodiments of the present disclosure, and are not described herein again.

Fig. 5 is a schematic structural diagram illustrating a multimedia resource delivering apparatus according to an exemplary embodiment, and referring to fig. 5, the apparatus includes:

an acquisition unit 501 configured to perform acquisition of at least one piece of comment data of a target object;

a determining unit 502 configured to perform, according to the at least one piece of comment data, determining a classification label of at least one target user account, where the classification label is used to indicate a tendency of the at least one target user account to the target object, and the target user account is a user account issuing the at least one piece of comment data;

a generating unit 503 configured to execute a user set generating instruction in response to the target object, and generate a user set according to a user account corresponding to a target classification tag in the at least one target user account;

a delivering unit 504 configured to execute a multimedia resource delivering instruction in response to the target object, and deliver a multimedia resource to a user account in the user set, where the multimedia resource is associated with the target object.

Optionally, the determining unit 502 is configured to perform inputting a word in the at least one piece of comment data into a classification label determination model, performing feature extraction on the word by the classification label determination model, performing trend classification based on the extracted features, and obtaining a classification label of the at least one target user account.

Optionally, the determining unit 502 is configured to perform inputting a word in the at least one piece of comment data into a classification label determination model, performing feature extraction by the classification label determination model based on a semantic relationship between each word and preceding and following words in the word of each piece of comment data to obtain a feature of each word, and performing trend classification on the feature of each word based on a full connection layer to obtain a classification label of the at least one target user account.

the generating unit 503 is configured to perform any of:

Optionally, the user set generation instruction comprises a target time period;

the generating unit 503 is configured to perform:

Optionally, the obtaining unit 501 is further configured to perform obtaining comment data in real time;

the determining unit 502 is further configured to perform matching of the comment data with an object library, and determine an object corresponding to the comment data;

the determining unit 502 is further configured to perform determining, according to the comment data, a classification label of the object by a first user account, where the first user account is a user account issuing the comment data.

Optionally, the obtaining unit 501 and the determining unit 502 are configured to perform obtaining at least one comment data and a corresponding classification tag of the target object from the comment data and the corresponding classification tag of the at least one object.

Optionally, the generating unit 503 is further configured to perform writing the comment data acquired in real time and the corresponding classification tag into a database, and generate corresponding first index information;

the apparatus also includes a first query unit;

Optionally, the generating unit 503 is further configured to periodically write the obtained comment data and the corresponding classification tag into the database, and generate second index information corresponding to each period;

the apparatus also includes a second query unit;

Optionally, the obtaining unit 501 is further configured to perform obtaining public opinion indication information of the target object based on the classification label of the at least one target user account, where the public opinion indication information is used to indicate a tendency of some or all of the user accounts to the target object.

Optionally, the obtaining unit 501 is configured to perform at least one of the following:

Optionally, the obtaining unit 501 is configured to perform obtaining public opinion indication information of the target object in each period based on the classification label corresponding to the at least one piece of comment data obtained in each period;

the apparatus also includes a second transmitting unit;

the second transmitting unit is configured to perform transmitting public opinion change trend information of the target object according to the public opinion indication information of each period of the target object in response to a viewing instruction of the public opinion indication information.

It should be noted that: in the multimedia resource delivering device provided in the foregoing embodiment, when delivering the multimedia resource, only the division of the functional modules is exemplified, and in practical applications, the function distribution can be completed by different functional modules according to needs, that is, the internal structure of the multimedia resource delivering device is divided into different functional modules, so as to complete all or part of the functions described above. In addition, the multimedia resource delivery device provided by the above embodiment and the multimedia resource delivery method embodiment belong to the same concept, and specific implementation processes thereof are detailed in the method embodiment and are not described herein again.

The computer device in the above method embodiment can be implemented as a terminal. For example, fig. 6 is a block diagram illustrating a structure of a terminal according to an example embodiment. The terminal 600 may be a portable mobile terminal such as: a smart phone, a tablet computer, an MP3(Moving Picture Experts Group Audio Layer III, motion video Experts compression standard Audio Layer 3) player, an MP4(Moving Picture Experts Group Audio Layer IV, motion video Experts compression standard Audio Layer 4) player, a notebook computer or a desktop computer. The terminal 600 may also be referred to by other names such as user equipment, portable terminal, laptop terminal, desktop terminal, etc.

In general, the terminal 600 includes: a processor 601 and a memory 602.

The processor 601 may include one or more processing cores, such as a 4-core processor, an 8-core processor, and so on. The processor 601 may be implemented in at least one hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), and a PLA (Programmable Logic Array). The processor 601 may also include a main processor and a coprocessor, where the main processor is a processor for Processing data in an awake state, and is also called a Central Processing Unit (CPU); a coprocessor is a low power processor for processing data in a standby state. In some embodiments, the processor 601 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content that the display screen needs to display. In some embodiments, processor 601 may also include an AI (Artificial Intelligence) processor for processing computational operations related to machine learning.

The memory 602 may include one or more computer-readable storage media, which may be non-transitory. The memory 602 may also include high-speed random access memory, as well as non-volatile memory, such as one or more magnetic disk storage devices, flash memory storage devices. In some embodiments, a non-transitory computer readable storage medium in the memory 602 is used to store at least one instruction for execution by the processor 601 to implement the multimedia resource delivery method provided by the method embodiments in the present disclosure.

In some embodiments, the terminal 600 may further optionally include: a peripheral interface 603 and at least one peripheral. The processor 601, memory 602, and peripheral interface 603 may be connected by buses or signal lines. Various peripheral devices may be connected to the peripheral interface 603 via a bus, signal line, or circuit board. Specifically, the peripheral device includes: at least one of a radio frequency circuit 604, a display 605, a camera assembly 606, an audio circuit 607, a positioning component 608, and a power supply 609.

A peripheral interface 603 may be used to connect at least one I/O related peripheral to the processor 601 and memory 602. In some embodiments, the processor 601, memory 602, and peripheral interface 603 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 601, the memory 602, and the peripheral interface 603 may be implemented on a separate chip or circuit board, which is not limited in this embodiment.

The Radio Frequency circuit 604 is used for receiving and transmitting RF (Radio Frequency) signals, also called electromagnetic signals. The radio frequency circuitry 604 communicates with communication networks and other communication devices via electromagnetic signals. The rf circuit 604 converts an electrical signal into an electromagnetic signal to transmit, or converts a received electromagnetic signal into an electrical signal. Optionally, the radio frequency circuit 604 comprises: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a subscriber identity module card, and so forth. The radio frequency circuitry 604 may communicate with other terminals via at least one wireless communication protocol. The wireless communication protocols include, but are not limited to: the world wide web, metropolitan area networks, intranets, generations of mobile communication networks (2G, 3G, 4G, and 5G), Wireless local area networks, and/or WiFi (Wireless Fidelity) networks. In some embodiments, the radio frequency circuit 604 may also include NFC (Near Field Communication) related circuits, which are not limited by this disclosure.

The display 605 is used to display a UI (User Interface). The UI may include graphics, text, icons, video, and any combination thereof. When the display screen 605 is a touch display screen, the display screen 605 also has the ability to capture touch signals on or over the surface of the display screen 605. The touch signal may be input to the processor 601 as a control signal for processing. At this point, the display 605 may also be used to provide virtual buttons and/or a virtual keyboard, also referred to as soft buttons and/or a soft keyboard. In some embodiments, the display 605 may be one, disposed on the front panel of the terminal 600; in other embodiments, the display 605 may be at least two, respectively disposed on different surfaces of the terminal 600 or in a folded design; in other embodiments, the display 605 may be a flexible display disposed on a curved surface or a folded surface of the terminal 600. Even more, the display 605 may be arranged in a non-rectangular irregular pattern, i.e., a shaped screen. The Display 605 may be made of LCD (Liquid Crystal Display), OLED (Organic Light-Emitting Diode), and the like.

The camera assembly 606 is used to capture images or video. Optionally, camera assembly 606 includes a front camera and a rear camera. Generally, a front camera is disposed at a front panel of the terminal, and a rear camera is disposed at a rear surface of the terminal. In some embodiments, the number of the rear cameras is at least two, and each rear camera is any one of a main camera, a depth-of-field camera, a wide-angle camera and a telephoto camera, so that the main camera and the depth-of-field camera are fused to realize a background blurring function, and the main camera and the wide-angle camera are fused to realize panoramic shooting and VR (Virtual Reality) shooting functions or other fusion shooting functions. In some embodiments, camera assembly 606 may also include a flash. The flash lamp can be a monochrome temperature flash lamp or a bicolor temperature flash lamp. The double-color-temperature flash lamp is a combination of a warm-light flash lamp and a cold-light flash lamp, and can be used for light compensation at different color temperatures.

Audio circuitry 607 may include a microphone and a speaker. The microphone is used for collecting sound waves of a user and the environment, converting the sound waves into electric signals, and inputting the electric signals to the processor 601 for processing or inputting the electric signals to the radio frequency circuit 604 to realize voice communication. For the purpose of stereo sound collection or noise reduction, a plurality of microphones may be provided at different portions of the terminal 600. The microphone may also be an array microphone or an omni-directional pick-up microphone. The speaker is used to convert electrical signals from the processor 601 or the radio frequency circuit 604 into sound waves. The loudspeaker can be a traditional film loudspeaker or a piezoelectric ceramic loudspeaker. When the speaker is a piezoelectric ceramic speaker, the speaker can be used for purposes such as converting an electric signal into a sound wave audible to a human being, or converting an electric signal into a sound wave inaudible to a human being to measure a distance. In some embodiments, audio circuitry 607 may also include a headphone jack.

The positioning component 608 is used for positioning the current geographic Location of the terminal 600 to implement navigation or LBS (Location Based Service). The Positioning component 608 can be a Positioning component based on the Global Positioning System (GPS) in the united states, the beidou System in china, or the galileo System in russia.

Power supply 609 is used to provide power to the various components in terminal 600. The power supply 609 may be ac, dc, disposable or rechargeable. When the power supply 609 includes a rechargeable battery, the rechargeable battery may be a wired rechargeable battery or a wireless rechargeable battery. The wired rechargeable battery is a battery charged through a wired line, and the wireless rechargeable battery is a battery charged through a wireless coil. The rechargeable battery may also be used to support fast charge technology.

In some embodiments, the terminal 600 also includes one or more sensors 610. The one or more sensors 610 include, but are not limited to: acceleration sensor 611, gyro sensor 612, pressure sensor 613, fingerprint sensor 614, optical sensor 615, and proximity sensor 616.

The acceleration sensor 611 may detect the magnitude of acceleration in three coordinate axes of the coordinate system established with the terminal 600. For example, the acceleration sensor 611 may be used to detect components of the gravitational acceleration in three coordinate axes. The processor 601 may control the display screen 605 to display the user interface in a landscape view or a portrait view according to the gravitational acceleration signal collected by the acceleration sensor 611. The acceleration sensor 611 may also be used for acquisition of motion data of a game or a user.

The gyro sensor 612 may detect a body direction and a rotation angle of the terminal 600, and the gyro sensor 612 and the acceleration sensor 611 may cooperate to acquire a 3D motion of the user on the terminal 600. The processor 601 may implement the following functions according to the data collected by the gyro sensor 612: motion sensing (such as changing the UI according to a user's tilting operation), image stabilization at the time of photographing, game control, and inertial navigation.

Pressure sensors 613 may be disposed on the side bezel of terminal 600 and/or underneath display screen 605. When the pressure sensor 613 is disposed on the side frame of the terminal 600, a user's holding signal of the terminal 600 can be detected, and the processor 601 performs left-right hand recognition or shortcut operation according to the holding signal collected by the pressure sensor 613. When the pressure sensor 613 is disposed at the lower layer of the display screen 605, the processor 601 controls the operability control on the UI interface according to the pressure operation of the user on the display screen 605. The operability control comprises at least one of a button control, a scroll bar control, an icon control and a menu control.

The fingerprint sensor 614 is used for collecting a fingerprint of a user, and the processor 601 identifies the identity of the user according to the fingerprint collected by the fingerprint sensor 614, or the fingerprint sensor 614 identifies the identity of the user according to the collected fingerprint. Upon identifying that the user's identity is a trusted identity, the processor 601 authorizes the user to perform relevant sensitive operations including unlocking the screen, viewing encrypted information, downloading software, paying, and changing settings, etc. The fingerprint sensor 614 may be disposed on the front, back, or side of the terminal 600. When a physical button or vendor Logo is provided on the terminal 600, the fingerprint sensor 614 may be integrated with the physical button or vendor Logo.

The optical sensor 615 is used to collect the ambient light intensity. In one embodiment, processor 601 may control the display brightness of display screen 605 based on the ambient light intensity collected by optical sensor 615. Specifically, when the ambient light intensity is high, the display brightness of the display screen 605 is increased; when the ambient light intensity is low, the display brightness of the display screen 605 is adjusted down. In another embodiment, the processor 601 may also dynamically adjust the shooting parameters of the camera assembly 606 according to the ambient light intensity collected by the optical sensor 615.

A proximity sensor 616, also known as a distance sensor, is typically disposed on the front panel of the terminal 600. The proximity sensor 616 is used to collect the distance between the user and the front surface of the terminal 600. In one embodiment, when proximity sensor 616 detects that the distance between the user and the front face of terminal 600 gradually decreases, processor 601 controls display 605 to switch from the bright screen state to the dark screen state; when the proximity sensor 616 detects that the distance between the user and the front face of the terminal 600 is gradually increased, the processor 601 controls the display 605 to switch from the breath-screen state to the bright-screen state.

Those skilled in the art will appreciate that the configuration shown in fig. 6 is not intended to be limiting of terminal 600 and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components may be used.

The computer device in the above method embodiment can be implemented as a server. For example, fig. 7 is a schematic structural diagram illustrating a server 700 according to an exemplary embodiment, where the server 700 may have a relatively large difference due to different configurations or performances, and can include one or more processors (CPUs) 701 and one or more memories 702, where the memory 702 stores at least one program code, and the at least one program code is loaded and executed by the processors 701 to implement the multimedia resource delivery method provided by the above-described method embodiments. Certainly, the server can also have components such as a wired or wireless network interface and an input/output interface to facilitate input and output, and the server can also include other components for implementing the functions of the device, which is not described herein again.

In an exemplary embodiment, a computer readable storage medium, such as a memory, including at least one program code, which is executable by a processor to perform the multimedia resource delivery method in the above embodiments, is also provided. For example, the computer-readable storage medium can be a Read-Only Memory (ROM), a Random Access Memory (RAM), a Compact Disc Read-Only Memory (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, and the like.

In an exemplary embodiment, a computer program product or a computer program is also provided, which comprises one or more program codes, which are stored in a computer-readable storage medium. The one or more program codes can be read by one or more processors of the computer device from a computer-readable storage medium, and the one or more processors execute the one or more program codes, so that the computer device can execute the multimedia resource delivery method.

It should be understood that, in various embodiments of the present disclosure, the sequence numbers of the above-mentioned processes do not mean the execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure.

It should be understood that determining B from a does not mean determining B from a alone, but can also determine B from a and/or other information.

Those skilled in the art will appreciate that all or part of the steps for implementing the above embodiments can be implemented by hardware, or can be implemented by a program for instructing relevant hardware, and the program can be stored in a computer readable storage medium, and the above mentioned storage medium can be read only memory, magnetic or optical disk, etc.

The above description is intended only to illustrate the preferred embodiments of the present disclosure, and not to limit the present disclosure, so that any modification, equivalent replacement, or improvement made within the spirit and principle of the present disclosure should be included in the scope of the present disclosure.

Claims

1. A multimedia resource delivery method is characterized by comprising the following steps:

acquiring at least one piece of comment data of a target object;

2. The method for delivering multimedia resources according to claim 1, wherein the determining a category label of at least one target user account according to the at least one piece of comment data includes:

3. The multimedia resource delivery method of claim 1, wherein the classification labels comprise a positive classification label, a negative classification label, and a neutral classification label;

4. The method of multimedia resource delivery according to claim 1, wherein the user set generation instruction comprises a target time period;

5. The method of claim 1, further comprising:

6. The method for delivering multimedia resources according to claim 5, wherein the obtaining of the public opinion indication information of the target object based on the classification label of the at least one target user account comprises at least one of:

7. The method of multimedia resource delivery according to claim 6, further comprising at least one of:

8. A multimedia resource delivering device, comprising:

9. A computer device, comprising:

one or more processors;

wherein the one or more processors are configured to execute the program code to implement the multimedia resource delivery method of any of claims 1-7.

10. A storage medium, wherein program code in the storage medium, when executed by one or more processors of a computer device, enables the computer device to perform the multimedia resource delivery method of any of claims 1 to 7.