CN116992872A

CN116992872A - Interactive information rendering method, device, equipment and storage medium

Info

Publication number: CN116992872A
Application number: CN202211305661.6A
Authority: CN
Inventors: 刘刚
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2022-10-24
Filing date: 2022-10-24
Publication date: 2023-11-03

Abstract

The application provides an interactive information rendering method, an interactive information rendering device, interactive information rendering equipment and a storage medium, which are applied to various scenes such as cloud technology, artificial intelligence, map field, intelligent traffic, auxiliary driving, vehicle-mounted and the like, wherein the method comprises the following steps: acquiring a hot word set and interaction information to be displayed aiming at target multimedia information release; the hot word set is generated based on at least two historical word recognition results and corresponding historical hot data, and the historical hot data of each historical word recognition result is obtained by correcting the initial hot data based on the average value of the initial hot data; word recognition processing is carried out on the interactive information to be displayed, and a target word recognition result is obtained; acquiring target hotwords matched with target word recognition results and corresponding target hotness data from a hotword library; and sending the target hotword and the target hotness data to the client for rendering. The application realizes accurate office digging of hot words in the interactive information to be displayed, and improves the display effect of the target multimedia information.

Description

Interactive information rendering method, device, equipment and storage medium

Technical Field

The application belongs to the technical field of computers, and particularly relates to an interactive information rendering method, an interactive information rendering device, interactive information rendering equipment and a storage medium.

Background

The barrage is an integral part of the video viewing experience. For the client account for publishing the barrage, the client account can obtain the sense of identity from the exposure and praise of the barrage. For a client account viewing a bullet screen, the bullet screen is an extension and supplement of video content, and is also a place where the client account seeks to watch a video accompany and harvest a co-emotion experience.

In the related art, when a video is played, a barrage is displayed in a playing page of the video in real time, and more interactive information is added by taking the barrage as a display window and a channel. However, the variety of the barrages is more, and the related technology does not accurately excavate and enhance the hot spot keywords in the barrages when playing and displaying the barrages, so that the hot spot barrage words corresponding to the video content and the current trend cannot be accurately displayed on the video, the watching experience and the interaction effect of the client account are affected, the interaction information between the client account and the video is reduced, and the display duration and the display effect of the video are further reduced.

Disclosure of Invention

In order to solve the technical problems, the application provides an interactive information rendering method, an interactive information rendering device, interactive information rendering equipment and a storage medium.

In one aspect, the application provides an interactive information rendering method, which comprises the following steps:

acquiring a hot word set and interaction information to be displayed aiming at target multimedia information release; the hot words in the hot word set are generated based on at least two historical word recognition results and historical heat data corresponding to the at least two historical word recognition results; the historical heat data of each historical word recognition result is obtained by correcting the initial heat data of each historical word recognition result by utilizing the average value of the initial heat data of each historical word recognition result;

performing word recognition processing on the interactive information to be displayed to obtain a target word recognition result corresponding to the interactive information to be displayed;

acquiring a target hotword matched with the target word recognition result and target hotness data corresponding to the target hotword from the hotword library;

and sending the target hotword and the target hotness data to a client so that the client renders the target hotword on the target multimedia information based on the target hotness data.

In another aspect, an embodiment of the present application provides an interactive information rendering apparatus, including:

The hot word interaction information acquisition module is used for acquiring a hot word set and interaction information to be displayed aiming at the target multimedia information release; the hot words in the hot word set are generated based on at least two historical word recognition results and historical heat data corresponding to the at least two historical word recognition results; the historical heat data of each historical word recognition result is obtained by correcting the initial heat data of each historical word recognition result by utilizing the average value of the initial heat data of each historical word recognition result;

the target word recognition result determining module is used for carrying out word recognition processing on the interactive information to be displayed to obtain a target word recognition result corresponding to the interactive information to be displayed;

the hot word data acquisition module is used for acquiring a target hot word matched with the target word recognition result and target heat data corresponding to the target hot word from the hot word library;

and the hotword data sending module is used for sending the target hotword and the target hotword data to a client so that the client renders the target hotword on the target multimedia information based on the target hotword data.

In another aspect, the present application proposes an electronic device for interactive information rendering, where the electronic device includes a processor and a memory, where at least one instruction or at least one program is stored in the memory, where the at least one instruction or at least one program is loaded and executed by the processor to implement an interactive information rendering method as described above.

In another aspect, the present application proposes a computer readable storage medium having stored therein at least one instruction or at least one program, the at least one instruction or the at least one program being loaded and executed by a processor to implement the interactive information rendering method as described above.

In another aspect, the application proposes a computer program product which, when executed by a processor, implements an interactive information rendering method as described above.

According to the interactive information rendering method, the interactive information rendering device, the electronic equipment and the storage medium, the average value of the initial heat data of each historical word recognition result is utilized to correct the initial heat data of each historical word recognition result, and the historical heat data of each historical word recognition result is obtained; based on at least two historical word recognition results and the historical heat data corresponding to the at least two historical word recognition results, a heat word set is generated, the initial heat data of each historical word recognition result is corrected based on the average value of the existing initial heat data, the determination accuracy of the historical heat data is improved, and therefore the construction accuracy of the heat word set is improved. After the hot word set is constructed, the interactive information to be displayed aiming at the target multimedia information is obtained, and word recognition processing is carried out on the interactive information to be displayed, so that a target word recognition result corresponding to the interactive information to be displayed is obtained; acquiring a target hotword matched with a target word recognition result and target hotness data corresponding to the target hotword from a hotword library; the target hotword and the target hotness data are sent to the client, so that the client renders the target hotword based on the target hotness data, accurate office digging of the target hotword in the interactive information to be displayed is realized, and the hot spot barrage words corresponding to the interactive information to be displayed and the current trend can be accurately displayed on the target multimedia information, so that the watching experience and the interactive effect of the client account and the interactive information between the client account and the video are improved, and further the display duration and the display effect of the target multimedia information are improved.

Drawings

In order to more clearly illustrate the embodiments of the application or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the application, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.

Fig. 1 is a schematic diagram illustrating an implementation environment of an interactive information rendering method according to an exemplary embodiment.

Fig. 2 is a flowchart illustrating a method of interactive information rendering according to an exemplary embodiment.

FIG. 3 is a schematic flow diagram illustrating one process of building a hot word set according to an exemplary embodiment.

FIG. 4 is a schematic diagram illustrating hot words in a client-side presentation hot word library, according to an example embodiment.

Fig. 5 is a second flowchart illustrating a method for interactive information rendering according to an exemplary embodiment.

Fig. 6 is a flow diagram illustrating a process for generating mask information according to an exemplary embodiment.

Fig. 7 is a flowchart illustrating a method of interactive information rendering according to an exemplary embodiment.

FIG. 8 is a system diagram illustrating a hotword mining and client enhanced presentation, according to an example embodiment.

Fig. 9 is a block diagram illustrating an interactive information rendering apparatus according to an exemplary embodiment.

Fig. 10 is a block diagram of a hardware architecture of a server for interactive information rendering according to an exemplary embodiment.

Detailed Description

Artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use the knowledge to obtain optimal results. In other words, artificial intelligence is an integrated technology of computer science that attempts to understand the essence of intelligence and to produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence, i.e. research on design principles and implementation methods of various intelligent machines, enables the machines to have functions of sensing, reasoning and decision.

The artificial intelligence technology is a comprehensive subject, and relates to the technology with wide fields, namely the technology with a hardware level and the technology with a software level. Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions.

Specifically, the construction process of the hot word set and the word recognition processing process of the interactive information to be displayed in the embodiment of the application relate to a text preprocessing technology in an AI natural language processing technology.

In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or server that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Fig. 1 is a schematic diagram illustrating an implementation environment of an interactive information rendering method according to an exemplary embodiment. As shown in fig. 1, the implementation environment may include at least a client 01 and a server 02, where the client 01 and the server 2 may be directly or indirectly connected through a wired or wireless communication manner, and the present application is not limited herein.

Specifically, the server 02 may be configured to generate a hot word set based on at least two historical word recognition results and historical heat data corresponding to the at least two historical word recognition results; the method comprises the steps of obtaining a hot word set and interaction information to be displayed aiming at target multimedia information; the method comprises the steps of carrying out word recognition processing on interactive information to be displayed to obtain target word recognition results corresponding to the interactive information to be displayed; the method comprises the steps of obtaining a target hotword matched with a target word recognition result and target hotness data corresponding to the target hotword from a hotword library; and the method is used for sending the target hotword and the target hotness data to the client. Optionally, the server 02 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDNs, and basic cloud computing services such as big data and artificial intelligence platforms.

Specifically, the client 01 may be configured to receive the target hotword and the target hotness data sent by the server; and rendering the target hotword based on the target hotness data. Alternatively, the client 01 may include, but is not limited to, a cell phone, a computer, an intelligent voice interaction device, an intelligent home appliance, a vehicle-mounted terminal, an aircraft, etc. The embodiment of the application can be applied to various scenes, including but not limited to rendering and displaying interaction information related to various scenes such as cloud technology, artificial intelligence, intelligent traffic, auxiliary driving and the like.

It should be noted that fig. 1 is only an example. In other scenarios, other implementation environments may also be included.

Fig. 2 is a flowchart illustrating a method of interactive information rendering according to an exemplary embodiment. The method may be used in the implementation environment of fig. 1. The present specification provides method operational steps as described above, for example, in the examples or flowcharts, but may include more or fewer operational steps based on conventional or non-inventive labor. The order of steps recited in the embodiments is merely one way of performing the order of steps and does not represent a unique order of execution. When implemented in a real system or server product, the methods illustrated in the embodiments or figures may be performed sequentially or in parallel (e.g., in a parallel processor or multithreaded environment). As shown in fig. 2, the method may include:

S101, acquiring a hot word set and interaction information to be displayed aiming at target multimedia information release; the hot words in the hot word set are generated based on at least two historical word recognition results and historical heat data corresponding to the at least two historical word recognition results; the historical heat data of each historical word recognition result is obtained by correcting the initial heat data of each historical word recognition result by using the average value of the initial heat data of each historical word recognition result.

Optionally, the target multimedia information may be professional production content (Professional Generated Content, PGC) and/or user production content (User Generated Content, UGC), which may include, but is not limited to: video, audio-visual, teletext content, etc. The target multimedia information can be presented to the client account in the form of Feeds, wherein Feeds are information streams which are continuously updated and presented to the client account content, and a plurality of message sources actively subscribed to by the client account are combined together to form a content aggregator to help the client account to continuously acquire the latest feed content.

Optionally, the interactive information to be displayed may be interactive information published by a client account viewing the target multimedia information for the target multimedia information during the playing process of the target multimedia information, and the interactive information may include, but is not limited to: bullet screens, comment data, etc.

The bullet screen refers to comment captions published by a client account at a specific time point on a time axis of target multimedia information when the client account views the target multimedia information. The comment data may refer to comments of the client account on the barrage content, i.e., the comment data refers to post-evaluation. The comment data can be regarded as a short text displayed below the bullet screen, and the comment data is also an expression of emotion and state of the client account, which is different from the presentation area of the bullet screen content.

In the embodiment of the present application, in the step S101, the server may generate a hotword set in advance, and when the client account issues the interactive information to be displayed (issues a barrage or issues a comment to a barrage displayed on a target multimedia information page) to the target multimedia information being watched, the server may acquire the hotword set generated in advance and the interactive information to be displayed issued by the client account.

Alternatively, the hot word set may be constructed in a variety of ways, not specifically limited herein. FIG. 3 is a schematic flow diagram illustrating one process of building a hot word set according to an exemplary embodiment. As shown in fig. 3, in one possible embodiment, the method for generating the hot word set may include:

S201, acquiring historical interaction information aiming at the release of the historical multimedia information in the historical time period.

In this embodiment, the historical time period may be flexibly set according to the service requirement, which is not specifically limited herein. Illustratively, the historical time period may be a certain time period before the playing time of the target multimedia information, for example, a week time, a month time, etc. before the playing time of the target multimedia information.

It should be noted that the source and type of the historical multimedia information are similar to those of the target multimedia information. That is, the historical multimedia information may be professional production content (Professional Generated Conte nt, PGC) and/or user production content (User Generated Content, UGC), which may include, but is not limited to: video, audio-visual, teletext content, etc. The historical multimedia information may be presented to the client account in the form of a Feeds stream.

S203, carrying out word recognition processing on the historical interaction information to obtain at least two historical word recognition results in the historical interaction information.

In this embodiment, the word recognition process may include, but is not limited to: preprocessing, named entity recognition (Name Entity Recognition, NER), word segmentation processing, part-of-speech tagging processing, and the like. Illustratively, the preprocessing may include, but is not limited to: removing stop words, removing Chinese characters, filtering barrage advertisements and the like. The NER refers to identifying a predefined entity type (name of person, organization, place name, etc.). The word segmentation process refers to splitting a text sequence into words with individual meanings, and may include, but is not limited to: dictionary-based, word library matching-based word segmentation methods, word frequency statistics-based word segmentation methods, knowledge understanding-based word segmentation methods and the like. The part-of-speech tagging refers to classifying words according to parts of speech, and can be completed by using a machine learning method.

In the step S203, after the server obtains the history interaction information, the history interaction information may be preprocessed to obtain the preprocessed history interaction information. In order to ensure that various types of phrases and words can be discovered, after the preprocessed historical interaction information is obtained, the server can perform NER, word segmentation processing, part-of-speech tagging processing and the like on the preprocessed historical interaction information, so that possible words in the interaction information to be displayed are recognized in a maximum range and accurately, namely at least two historical word recognition results in the interaction information to be displayed are recognized accurately.

Because the words mined by the server through NER, word segmentation, part-of-speech tagging and other methods may have partial overlapping, in order to ensure the recognition accuracy of at least two historical word recognition results and avoid the influence of the overlapping word recognition results on the subsequent mining process, the words obtained by various ways of recognition can be subjected to de-duplication processing after the recognition is completed, the de-duplicated words are obtained, and the de-duplicated words are used as the at least two historical word recognition results.

For example, to be presented with the interactive information "hoped for years, there is a prefix that causes the hot blood to boil: XX tip mathematics-! The results of the word segmentation process, the part-of-speech tagging process and the NER entity recognition of the interactive information to be displayed are respectively as follows:

Word segmentation: hope/years later/,/have/one/make/person/hot blood boiling/prefix/: XX/tip/math /)!

Part-of-speech tagging process: hope/V after many years/N,/W has/V one/M/Q vs/V person/N hot blood boiling/I/U prefix/N: WXX/NS tip/B math/N-! W/W

NER entity identification: XX.

Therefore, the hot blood boiling appears in the word segmentation processing result and the part of speech tagging processing, and the hot blood boiling appears in the word segmentation processing result and the part of speech tagging processing is de-duplicated once, namely, one hot blood boiling is reserved. It should be noted that if the repeated word recognition result originates from a different comment or bullet screen, the repeated word is reserved and no duplicate removal processing is performed.

S205, correcting the initial heat data of each historical word recognition result according to the average value of the initial heat data of each historical word recognition result to obtain the historical heat data of each historical word recognition result.

First, the generation process of initial heat data of each history word recognition result is introduced:

in an alternative embodiment, the method for generating the initial heat data of each historical word recognition result includes: determining a first word frequency of each historical word recognition result in a target time interval and a second word frequency in a historical time interval; and calculating the ratio of the first word frequency to the second word frequency to obtain initial heat data of each historical word recognition result.

Optionally, the first word frequency refers to the number or frequency of occurrence of each of the historical word recognition results within the target time interval, and the second word frequency refers to the number or frequency of occurrence of each of the historical word recognition results within the historical time period. Because the frequency of the occurrence of certain common words is high, the absolute value on the fluctuation is also high, words with few frequency bases are easy to surpass, and if the difference value between the word frequencies is used for measuring the heat data, the determination accuracy of the heat data is reduced. Based on the above, the embodiment of the application adopts the ratio of the first word frequency of each historical word recognition result in the target time interval to the second word frequency in the historical time interval to accurately measure the heat data of the word recognition result. The specific calculation formula can be as follows:

wherein w is _i Representing a certain historical word recognition result, T _j Represents a target time interval, F (w _i ，T _j ) Representing the historical word recognition result w _i In the target time interval T _j Is of the first word frequency, F (w _i ，T ₁ ，T ₂ ，...，T _j ) Representing the second word frequency, S (w _i ) Initial heat data representing the recognition result of a certain historical word.

Alternatively, the target time interval may be a time interval associated with the historical time period. As an example, the target time interval may be any time interval in the historical time period. The historical time period may be divided into a plurality of time intervals according to a time sequence, and the target time interval may be a last time interval in the time sequence among the plurality of time intervals. For example, the history period is one week, and the history period is divided into a plurality of time intervals in units of "days", and the target time interval is the last day in the history period. As another example, a historical time segment may be further determined, the historical time segment may be divided into a plurality of time intervals according to the time sequence, the target time interval may be a time interval of the time intervals, the time sequence is ordered to be the last time interval, and the historical time segment may be a time interval other than the target time interval in the historical time segment. For example, the history time period is one week, and the history time period is divided into a plurality of time intervals in units of "days", the target time interval being the last day in the history time period, and the history time period being the time other than the last day in the history time period.

Alternatively, the heat data may refer to the degree to which the historical word recognition results are of interest. Illustratively, the heat data may be a heat value, or a probability that it belongs to a certain heat value, or a "yes or no" result (e.g., the historical word recognition result is heat data, the historical word recognition result is not heat data, etc.).

In the embodiment of the application, the initial heat data of each historical word recognition result is determined through the ratio of the first word frequency of each historical word recognition result in the target time interval to the second word frequency in the historical time interval, so that certain historical word recognition results with larger word frequency every day can be eliminated, and the determination accuracy of the initial heat data can be improved.

In some embodiments, in order to further improve the accuracy of determining the initial heat data, some historical word recognition results with smaller frequency and higher ratio of the first word frequency to the second word frequency may be excluded from at least two historical word recognition results, so as to obtain updated historical word recognition results, and the subsequent hot word set construction process may construct based on the updated historical word recognition results. For example, for the historical word recognition result W1 and the historical word recognition result W2, 1000 times and 18 times (i.e., the first word frequency) respectively appear in the target time interval, and 1500 times and 20 times (i.e., the second word frequency) respectively appear in the historical time interval, the result of directly calculating the ratio of the first word frequency and the second word frequency is 0.67 and 0.9, and the ratio corresponding to W2 is greater than the ratio corresponding to W1, but it is obvious that the word W1 is more prone to be judged as a hot word. Therefore, W2 is a historical word recognition result with a smaller frequency and a higher ratio of the first word frequency to the second word frequency, and W2 can be deleted from at least two historical word recognition results to obtain an updated historical word recognition result, so that the determination accuracy of hot words and hot data is improved.

Alternatively, in the step S205, the embodiment of the present application may modify the initial heat data of each historical word recognition result in various manners, which is not limited herein. In one embodiment, continuing to fig. 3, in step S205, correcting the initial heat data of each historical word recognition result according to the average value of the initial heat data of each historical word recognition result to obtain the historical heat data of each historical word recognition result may include:

s2051, determining average word frequency corresponding to at least two historical word recognition results according to the second word frequency of each historical word recognition result in the historical time period, the number of at least two historical word recognition results and the number of time intervals contained in the historical time period.

S2053, correcting initial heat data corresponding to each historical word recognition result based on the average word frequency, the average value of the initial heat data of each historical word recognition result and the first word frequency of each historical word recognition result in a target time interval, and obtaining the historical heat data of each historical word recognition result.

In one possible embodiment, the step S2051-the step S2053 may be modified by using a bayesian averaging method. The bayesian averaging method can be used to correct the score calculation of the popularity problem. I.e. first estimating a value for the result and then continuously correcting with new information so that it more closely approximates an objective answer. For example, taking the client account voting ranking as an example, if the number of people for scoring the client account voting is small, the average score is calculated, and the situation that the average score is not objective enough is likely to occur. External information is introduced at this time, assuming that a part of the client accounts (C client accounts) are ticketed and all are given an average score (m minutes). The scores of the client account numbers are added to the scores of the existing client account numbers, and then the average is calculated, so that the average score can be corrected to increase the objectivity of the final score to a certain degree or angle. The formula for correcting the heat data by using the Bayesian averaging method can be as follows:

Wherein S' (w _i ) For the corrected historical heat data of a certain historical word recognition result, m is an a priori average score, and C is a constant, which is related to the overall situation of the sample. C is understood to mean the number of people given an average score, the larger C is, the smaller the overall distribution gap representing the desired score. In the hotword extraction scenario, m may be defined as the initial hotness data S (w _i ) Average (i.e. average of initial heat data of each historical word recognition result) of (a) using S _average A representation; c may be defined as the average word frequency of all historical word recognition results over a certain time interval (e.g., calculated according to the granularity of the day, C may be defined as the frequency of occurrence of a certain historical word recognition result over the day), using F _average I is the number of recognition results of the historical words, J is the number of time intervals (for example, total days) included in the historical time period, and the calculation formula of S2051 may be as follows:

wherein,,refers to a first sum of the second word frequency of each historical word recognition result over a historical period of time.

That is, in the above step S2051, a first sum of the second word frequency of each of the history word recognition results in the history period may be calculated, and a first ratio of the first sum to the number I of the history word recognition results may be calculated, followed by calculation of the first ratio and the time zone A second ratio between the numbers J, the second ratio being the average word frequency F _average 。

Alternatively, the correction procedure in the above step S2053 may be regarded as a correction to the above bayesian average formula S' (w _i ) The conversion process, the calculation formula corresponding to the step S2053 may be as follows:

that is, in the above-described step S2053, for each history word recognition result, the first word frequency F (w _i ,T _j ) And average word frequency F _average And a second sum of the first word frequency F (w _i ,T _j ) A third ratio to the second sum; calculating initial heat data S (w _i ) And the average value S of the initial heat data of the historical word recognition result _average A difference between the historical word recognition results, the difference being considered a personality score for the historical word recognition results; calculating a product between the difference and the third ratio; calculating the average S of the product and the initial heat data of the historical word recognition result _average And obtaining the historical heat data S' (w) of the historical word recognition result _i )。S′(w _i ) The larger the history word recognition result is, the higher the frequency of occurrence can be considered.

According to the embodiment of the application, the initial heat data corresponding to each historical word recognition result is corrected through the average word frequency, the average value of the initial heat data of each historical word recognition result and the first word frequency of each historical word recognition result in the target time interval, so that the personality score of the historical word recognition result, the average value of the initial heat data and the first word frequency of each historical word recognition result in the target time interval can be fully considered in determining the historical heat data corresponding to each historical word recognition result, and the accuracy and objectivity of the historical heat data determination are improved.

Optionally, in obtaining the historical heat data S' (w) of each historical word recognition result _i ) Thereafter, the historical heat data S' (w) may be based on the recognition result of each of the historical words _i ) The historical heat data S' (w) is deleted _i ) And (5) identifying the historical words which do not meet the preset conditions.

In one embodiment, the above S' (w) _i ) The correction formula of (a) can be regarded as that each historical word recognition result has a priori average score S _average And each history word recognition result has its own personality score (S (w _i )-S _average ) Is determined by the performance of each historical word recognition result. If the word frequency of the historical word recognition result is higher, namely the exposure is large, the weight with larger personality score is given. If the corrected historical heat data S' (w) of the historical word recognition result _i ) Equal to or less than S _average And considering the historical word recognition result as a non-hot word, and deleting the non-hot word from the at least two historical word recognition results.

In another embodiment, if the first word frequency of a certain historical word recognition result in the target time interval is smaller than F _average Word weight constraints for their personality scoresMake->The method is close to 0, and even if the initial heat data is higher, the corrected historical heat data can be close to average score, so that the historical word recognition result with less word frequency but larger quantity change ratio is screened out. If the word frequency of a certain historical word recognition result is large, the word weight is close to 1, but the word is often a common word, the initial heat data is close to average, the personality score (S (w _i )-S _average ) The historical heat data is also approximately equal, and the historical word recognition results can be screened out.

S207, determining a target historical word recognition result from at least two historical word recognition results according to the historical heat data of each historical word recognition result.

S209, generating a hot word set based on the target historical word recognition result.

The above steps S207 to S209 may be implemented in various ways, and are not particularly limited herein.

In one embodiment, in the step S207, each of the history word recognition results may be regarded as a target history word recognition result, and in the step S209, a hot word set is generated according to each of the history word recognition results, that is, the hot word in the hot word set is each of the history word recognition results.

In another embodiment, in the step S207, the individual score of each history word recognition result may be deleted as described above (S (w _i )-S _average ) And (3) screening out part of historical word recognition results from at least two historical word recognition results to obtain screened historical word recognition results. And in the step S209, a hot word set is generated according to the filtered historical word recognition result, that is, the hot word in the hot word set is the filtered historical word recognition result.

In the third embodiment, in the step S207, some historical word recognition results with smaller frequency may be excluded from at least two historical word recognition results, and the ratio of the first word frequency to the second word frequency is higher, so as to obtain updated historical word recognition results. And in the step S209, a hot word set is generated according to the updated historical word recognition result, that is, the hot word in the hot word set is the updated historical word recognition result.

In the fourth embodiment, a heat data threshold may be preset according to the service requirement, and in the step S207, a history word recognition result that the history heat data is greater than the heat data threshold is used as the target history word recognition result. And in the step S209, a hot word set is generated according to the target historical word recognition result, that is, the hot word in the hot word set is the target historical word recognition result.

Taking the interactive information to be displayed as a barrage as an example, the barrage hotword reflects the general attention problem and things of the client account in a certain time period. Whether the bullet screen hotword is extracted accurately or not directly influences the interactive experience of the client account and the playing effect of the target multimedia information. Table 1 shows the hot words obtained by hot word mining on the barrage using the method described above. From table 1, it can be seen that by mining bullet screen key hotwords and performing enhancement processing and rendering, emotion of the client account can be better expressed and bullet screen atmosphere can be improved.

TABLE 1 hotword obtained by hotword mining on barrages

According to the embodiment of the application, the initial heat data of each historical word recognition result is determined through the ratio of the first word frequency of each historical word recognition result in the target time interval to the second word frequency in the historical time interval, and certain historical word recognition results with larger word frequency per day can be eliminated, so that the determination accuracy of the initial heat data can be improved; by means of a Bayesian averaging method, based on average word frequency, the average value of initial heat data of each historical word recognition result and first word frequency of each historical word recognition result in a target time interval, the initial heat data corresponding to each historical word recognition result is corrected, so that the personality score of the historical word recognition result, the average value of the initial heat data and the first word frequency of each historical word recognition result in the target time interval can be fully considered in determining the historical heat data corresponding to each historical word recognition result, and the accuracy and objectivity of the historical heat data determination are improved; through the higher historical heat data of precision, can follow the accurate target historical word recognition result of determining in at least two historical word recognition results to the determination precision of the hot word in hot word set has been promoted, and then the construction precision of hot word set has been improved.

S103, carrying out word recognition processing on the interactive information to be displayed to obtain a target word recognition result corresponding to the interactive information to be displayed.

Optionally, the manner in which the server performs the word recognition processing on the interactive information to be displayed is similar to the manner in which the server performs the word recognition processing on the historical interactive information, and specifically, the step S203 may be referred to, which is not described herein.

S105, acquiring target hotwords matched with the target word recognition result and target hotness data corresponding to the target hotwords from the hotword library.

Because the hot words in the hot word library are generated based on at least two historical word recognition results and the historical heat data corresponding to the at least two historical word recognition results, the target word recognition results can be matched with the hot words in the hot word library, and if the target hot words matched with the target word recognition results exist in the hot word library, the target hot words and the target heat data corresponding to the target hot words are acquired.

Alternatively, matching may refer to: judging whether a hot word consistent with the content expressed by the target word recognition result exists in the hot word library, judging whether the hot word identical with the target word recognition result exists in the hot word library, judging whether the full name of the target word recognition result exists in the hot word library, judging whether the abbreviation of the target word recognition result exists in the hot word library, and the like. If any judgment result is yes, judging that the target hotword matched with the target word recognition result exists in the hotword library. For example, if the target word recognition result is xx medical university and there is a hot word "xx medical university" in the hot word library, then it is considered that "xx medical university" in the hot word library is the matched target hot word of the target word recognition result.

S107, the target hotword and the target hotness data are sent to the client side, so that the client side renders the target hotword on the target multimedia information based on the target hotness data.

In this embodiment, after obtaining the target hotword and the corresponding target hotness data, it may be sent to the client, and the client renders the target hotword on the target multimedia information based on the target hotness data. In one embodiment, the client may determine display information of the target hotword according to the target hotness data. For example, the client may determine a rendering size, a rendering location, a number of renderings, etc. of the target hotword from the target hotness data. And renders the target hotword on the target multimedia information according to the rendering size, rendering position, rendering number, etc.

In one possible embodiment, after the server constructs the hotword set, the hotword set may be further sent to the client, and the client performs an intuitive presentation on the hotword in the hotword set, thereby improving the interactive experience of the client. For example, the client may determine a hotword display size and a hotword display position of each hotword according to the historical hotness data corresponding to each hotword in the hotword library, and render each hotword according to the hotword display size of each hotword at the hotword display position of each hotword. For example, the larger the historical heat data, the larger the hotword display size, the more centered or closer the hotword display location is to a location that is easily viewable by the client account. Fig. 4 is a schematic diagram showing hot words in a hot word library at a client according to an exemplary embodiment, where, as shown in fig. 4, the hot words in the hot word library include AA, BB, CC, DD, EE, FF, GG, and AA corresponding to the highest historical heat data, and then the hot word display size of AA is the largest, and GG corresponding to the highest historical heat data, and then the hot word display size of GG is the largest.

Fig. 5 is a second flowchart illustrating a method for interactive information rendering according to an exemplary embodiment. In an alternative embodiment, as shown in fig. 5, after the step S105, the method may further include: s106, displaying and configuring the target hotword. Specifically, the step S106 may include:

s1061, acquiring a display strategy set; the display policy set includes display policies corresponding to different heat data.

In this embodiment, in order to render, enhance and display the target hotword, avoid that different hotwords are displayed in the same manner, so as to improve the interaction experience between the client account and the target multimedia information, after the server determines that the target hotword is obtained, a display policy set may be obtained as described in step S1061, where the display policy set includes display policies corresponding to different hotness data, that is, a mapping relationship between different hotness data and the corresponding display policies is established, so that different hotness data may correspond to different display policies.

As an example, the display policies corresponding to the different heat data may include, but are not limited to: the font attribute display strategy corresponding to the different heat data, the word position display strategy corresponding to the different heat data and the rendering frequency display strategy corresponding to the different heat data. The font attribute display strategies corresponding to the different heat data comprise font sizes, font colors, font types and the like corresponding to the different heat data. The word position display strategies corresponding to the different heat data comprise display areas corresponding to the different heat data and the like. The rendering frequency display strategy corresponding to the different heat data comprises the different heat data, the cycle display times during the playing of the target multimedia information and the like.

S1063, acquiring a target display strategy corresponding to the target heat data from the display strategies corresponding to the different heat data.

In this embodiment, the server may obtain, according to the mapping relationship, a target display policy corresponding to the target heat data from display policies corresponding to different heat data. The target display policy may include a target font property display policy corresponding to the target heat data, a target word position display policy corresponding to the target heat data, a target rendering frequency display policy corresponding to the target heat data, and the like.

S1065, performing display configuration processing on the target hotword according to the target display strategy to obtain the configured target hotword corresponding to the interaction information to be displayed.

In this embodiment, after the server obtains the target display policy, the display configuration process may be performed on the target hotword based on the target font attribute display policy, the target word position display policy, the target rendering frequency display policy, and the like corresponding to the target hotword, to obtain the configured target hotword corresponding to the interactive information to be displayed.

In an optional embodiment, as further shown in fig. 5, in step S1065, performing display configuration processing on the target hotword according to the target display policy to obtain a configured target hotword corresponding to the interaction information to be displayed may include:

S10651, configuring display attributes of fonts in the target hotword in the rendering process based on the target font attribute display strategy.

S10653, configuring the display position of the target hotword in the rendering process according to the target word position display strategy.

S10655, matching the rendering frequency of the target hotword in the rendering process based on the target rendering frequency display strategy.

S10657, generating the configured target hotword according to the target hotword, the display attribute, the display position and the rendering frequency.

In this embodiment, in the above steps S10651-S10657, the server may configure the display attribute of the fonts in the target hotword in the rendering process according to the target font attribute display policy, for example, configure the font size, the font color, the font type, and the like of the fonts in the target hotword in the rendering process. And configuring the display position of the target hotword in the rendering process according to the target word position display strategy. According to the target rendering frequency display strategy, the rendering frequency of the target hotword in the rendering process, namely the cycle display times in the rendering process, is configured. And finally, the server performs fusion processing on the target hotword, the display attribute, the display position and the rendering frequency to generate the configured target hotword. Because the display strategies of different hotwords are different, hotword texts with different callbacks and styles can be generated through the steps S10651-S10657.

It should be noted that, the steps S1063-S1065 and the steps S10651-S10657 may be performed by the client, that is, the server may send the target hotword and the corresponding target hotness data to the client, and the client performs the steps S1063-S1065 and the steps S10651-S10657.

Accordingly, continuing to refer to fig. 5, in the step S107, the sending the target hotword and the target hotword to the client, so that the client renders the target hotword on the target multimedia information based on the target hotword, may include:

s1071, sending the configured target hotword to the client so that the client renders the configured target hotword on the target multimedia information.

Optionally, after the server sends the configured target hotword to the client, the client may obtain a display attribute, a display position and a rendering frequency in the configured target hotword, and in a playing process of the target multimedia information, display a font of the target hotword according to the display attribute on an image frame corresponding to the target multimedia information, display the target hotword at a position corresponding to the display position in the client, and determine a cycle display number of times of the target hotword according to the rendering frequency, and perform cycle display on the target hotword according to the cycle display number.

In the embodiment of the application, the target display strategy corresponding to the target heat data is obtained from the display strategies corresponding to the different heat data, so that the configuration strategy of the target heat words according to the display strategy (namely, the configuration strategy of fonts, font sizes, font colors, circulation times, bullet screen display areas and the like corresponding to the heat) is realized, the heat word texts with different callbacks and styles are generated, different heat words can be displayed in different modes, namely, the rendering enhancement display of different styles is realized, bullet screen interactive content forms with different enhancement styles can be provided for the client account, the viewing experience of the client account and the interactive information between the client account and the target multimedia information are improved, and the display duration and the display effect of the target multimedia information are further improved.

In an alternative embodiment, as further shown in fig. 5, the number of the interaction information to be displayed is a plurality of, and the release time of each interaction information to be displayed is different, and the step S1071 may include:

s10711, acquiring release time corresponding to each of the interaction information to be displayed.

In this embodiment, there are usually a plurality of client accounts for viewing the target multimedia information, and different client accounts can issue corresponding interactive information to be displayed for the target multimedia information at different viewing time points, so that the issue time of each interactive information to be displayed is different. When the client account issues the interactive information to be displayed aiming at the target multimedia information, the server can acquire the issued interactive information to be displayed in real time, and store the issued interactive information to be displayed and the time line association of the target multimedia information in the database in real time for persistence.

S10713, sending the configured target hotwords corresponding to the interaction information to be displayed and the release time corresponding to the interaction information to be displayed to the client, so that the client sequentially renders the configured target hotwords corresponding to the interaction information to be displayed on the target multimedia information according to the release time corresponding to the interaction information to be displayed.

In this embodiment, since the server stores the published timeline association of the interaction information to be displayed and the target multimedia information in the database for persistence, when the server sends the configured target hotword to the client, the server may also send the respective release times of the interaction information to be displayed to the client. And the client sequentially renders the configured target hotwords corresponding to the interaction information to be displayed on the target multimedia information according to the release time corresponding to the interaction information to be displayed, so that the release time line of the interaction information to be displayed is consistent with the display time line of the configured target hotwords. For example, the earlier the release time corresponding to the interactive information to be displayed is, the earlier the display time of the corresponding configured target hotword is, whereas the later the release time of each interactive information to be displayed is, the later the display time of the corresponding configured target hotword is.

In the embodiment of the application, the server stores the issued time line association of the interaction information to be displayed and the target multimedia information in the database for persistence, so that the client can sequentially render the configured target hotwords corresponding to the interaction information to be displayed on the target multimedia information according to the respective release time of the interaction information to be displayed when rendering, the release time line of the interaction information to be displayed is consistent with the display time line of the configured target hotwords, the messy display of the configured target hotwords corresponding to the interaction information to be displayed is avoided, the rendering enhancement effect of the target hotwords is further improved, the bullet screen interaction content form of the enhancement style can be provided for the client account, the viewing experience of the client account and the interaction information between the client account and the target multimedia information are further improved, and the display duration and the display effect of the target multimedia information are further improved.

FIG. 6 is a schematic flow chart of generating mask information according to an exemplary embodiment, as shown in FIG. 6, in an alternative embodiment, the method may further include:

S301, acquiring target multimedia information from a preset multimedia library before displaying the target multimedia information.

Alternatively, the preset multimedia library may include a plurality of pieces of multimedia information, and the server may acquire the target multimedia information from the multimedia library in advance before the target multimedia information is played.

For example, the server may acquire the identification information of the target multimedia information, and search and acquire the target multimedia information from the preset multimedia library according to the identification information.

S303, performing foreground and background segmentation processing on the target multimedia information to obtain mask information corresponding to the target multimedia information; the mask information comprises a foreground area and a background area, wherein the foreground area is used for displaying an object main body in the target multimedia information, and the background area is used for displaying the configured target hotword.

S305, storing mask information to a content distribution network.

In this embodiment, in order to enable the configured target hotword to have a better display effect, before playing the target multimedia information, the server may perform main detection processing on the target multimedia information, that is, perform foreground and background segmentation processing on an image corresponding to the target multimedia information, obtain Mask information (i.e., mask file) corresponding to the target multimedia information, and store the Mask information in a content delivery network (Content Delivery Network, CDN).

Illustratively, in the above step S303, the server may perform foreground and background segmentation processing on the image corresponding to the target multimedia information using a semantic segmentation model (pointrand model), to obtain a Mask file, where the Mask file includes a foreground area and a background area, the foreground area is used to display an object body (including a person, an object, a text, etc.) in the target multimedia information, and the background area is used to display the configured target hotword. For example, the image corresponding to the target multimedia information is an RGB image of size h×w×3, the image is input to the poinrand model for foreground and background segmentation processing, the poinrand model returns h×w segmentation results (foreground segmentation result and background segmentation result), where H represents the width of the image and W represents the height of the image. The value range of each pixel in the segmentation result is 0-3 (4 types in total), and the segmentation result represents the background, the person, the object and the text respectively. Since the background class is needed in the embodiment of the present application, filtering processing may be performed on the segmentation result, where all pixels identified as 0 are set as a first value (i.e., the pixel value of the background segmentation result is set as a first value), and all pixels of other values are set as a second value (i.e., the pixel of the foreground segmentation result such as a person, an object, and a text is set as a second value), so as to obtain a binary Mask image, where a portion of the binary Mask image with a pixel value of 1 is used to display the configured target hotword, and a portion of the pixel of 0.5 is used to display the object body (such as the person, the object, and the text) in the target multimedia information. It should be noted that the first value and the second value may be set according to actual service requirements, which is not limited in particular, for example, the first value is 1, the second value is 0.5, and so on.

In the embodiment of the application, the foreground and background segmentation processing is carried out on the target multimedia information to obtain the Mask file corresponding to the target multimedia information, and the Mask file can be used for enabling the client to display the configured target hotword in the background area in the rendering process without displaying the configured target hotword on the main body in the target multimedia information, so that the main body in the target multimedia information is prevented from being blocked by the target hotword, namely, the Mask file is used for realizing the hotword Mask filtering effect, the rendering effect of the client is effectively improved, and the rendering enhancement effect of the target hotword is further improved, thereby improving the interactive information and interactive experience between the client account and the target multimedia information, and further improving the display duration and the display effect of the target multimedia information.

In an optional embodiment, in S1071, the sending the configured target hotword to the client so that the client renders the configured target hotword on the target multimedia information may include:

mask information is obtained from a content distribution network.

And sending the configured target hotword and the mask information to the client so that the client displays the object main body in the foreground area and displays the configured target hotword in the background area.

In this embodiment, since the preset multimedia library may include a plurality of pieces of multimedia information, the server may generate mask information corresponding to each piece of multimedia information in advance, and establish an association relationship between the identification information of each piece of multimedia information and the corresponding mask information. When the server sends the configured target hotword to the client, the server can acquire mask information of the target multimedia information from mask information corresponding to the pre-generated multimedia information according to the pre-established association relation and the identification information of the target multimedia information. And sending the configured target hotword and the mask information to the client. The client displays the object main body in the foreground area, and displays the configured target hotword in the background area, so that the main body in the target multimedia information is prevented from being blocked by the target hotword, namely, the Mask file is utilized to realize the hotword Mask filtering effect, the rendering effect of the client is effectively improved, the rendering enhancement effect of the target hotword is further improved, the interactive information and interactive experience between the client account and the target multimedia information are further improved, and the display duration and the display effect of the target multimedia information are further improved.

Fig. 7 is a flowchart third illustrating a method for rendering interactive information according to an exemplary embodiment, where, as shown in fig. 7, the interactive information to be displayed is a barrage by taking target multimedia information as a target video, and the method for rendering interactive information may include:

the configured target hotword and release time acquisition process comprises the following steps:

s401, acquiring a target video.

S403, decoding the target video to obtain a plurality of video frames corresponding to the target video.

S405, in the process of playing a plurality of video frames, the server stores the time line association of a plurality of barrages issued by the client account aiming at the target video and the target video in a barrage database in real time for persistence. The specific process is referred to the above step S10711, and will not be described herein.

S407, the client acquires the configured target hotword corresponding to each of the multiple barrages and the release time corresponding to each of the multiple barrages, which are sent by the server. The specific process is referred to the above step S10713, and will not be described herein.

S409, the client sends a Mask file acquisition request to the server, and acquires the Mask file from the server.

S4011, the client displays the object main body in a foreground area and displays the configured target hotword in a background area.

Mask diagram generation:

s501, before playing the target video, the server acquires the target video from a preset multimedia library. In the process of obtaining the target video, please refer to the above step S301, which is not described herein.

S503, decoding the target video by the server to obtain a plurality of video frames corresponding to the target video.

S505, the server performs foreground and background segmentation processing on the video frames to obtain Mask files corresponding to the target video. The specific process is referred to the above step S303, and will not be described herein.

S507, converting the format of the Mask file by the server, and storing the Mask file into the CDN.

FIG. 8 is a schematic diagram of a system for hotword mining and client-side enhanced presentation of interactive information to be presented as a bullet screen, such as shown in FIG. 8, according to an exemplary embodiment, the system may include:

PGC and UGC content production and consumption end

(1) The method comprises the steps of firstly acquiring an interface address of an uploading server through communication with an uplink content interface service and a downlink content interface service, then uploading a local file, wherein local video content can be selected to match with music, a filter template, a beautifying function of a video and the like in the shooting process;

(2) As a consumption end, the method communicates with a content distribution outlet server to acquire index information of corresponding content, and can download corresponding streaming media files for video and play and watch the streaming media files through a local player; for graphics context, CDN service communication with direct edge deployment can be realized;

(3) Simultaneously, operation data, reading speed, completion rate, reading time, clamping, loading time, playing click and the like of the client account browsing in the uploading and downloading processes are reported to a server;

(4) The consumer usually browses the consumption data through Feeds stream mode, provides the entrance for directly reporting and feeding back uncomfortable content of the consumer, directly interfaces with the manual auditing system, and confirms and rechecks.

(5) The bullet screen display module displays the bullet screen after enhancement in the video content playing process, the Mask is used for avoiding shielding the picture main body, and the bullet screen interaction process with the user is completed;

uplink and downlink content interface server

(1) The method comprises the steps of directly communicating with a content production end, and storing contents submitted from a client end, usually titles, publishers, abstracts, cover charts and release time of the contents, or directly entering a server through the server by shot videos, and storing the files into a video content storage service;

(2) Writing meta information of video content, such as video file size, cover map link, code rate, file format, title, release time, author, etc. into a content database;

(3) And submitting the uploaded file and content meta-information to a dispatching center service for subsequent content processing and circulation.

Third, content database

(1) The core database of content, in which the meta-information of the content released by all producers is stored, is focused on the meta-information of the content itself, such as file size, cover map link, code rate, file format, title, release time, author, video file size, video format, mark of whether original or not, or classification of the content in the manual review process (including one, second, third-level classification and label information, such as a video content of an explanation mobile phone, wherein the first-level classification is science and technology, the second-level classification is a smart mobile phone, and the third-level classification is a domestic mobile phone);

(2) The information in the content database is read in the manual auditing process, and meanwhile, the result and the state of the manual auditing are returned to the content database;

(3) The scheduling center mainly comprises machine processing and manual auditing processing, and the machine processing mainly carries out various quality judgments, such as low-quality filtering, content label classification, content weight removal and the like. The result of machine processing is written into a content database, and the content processed by the machine is not subjected to repeated two processes by a worker;

fourth, dispatch center service

(1) The whole dispatching process of video content circulation is responsible for receiving the content through an uplink content interface server and a downlink content interface server, and then acquiring meta information of the content from a content meta information database;

(2) As an actual scheduling controller for the operation of the video link, the scheduling center service can schedule the multi-mode uncomfortable content identification service system to process the picture content in the link according to the type of the content, for example, filter the picture content and mark the image content correspondingly;

(3) The scheduling manual auditing system and the machine processing system control the scheduling sequence and priority;

(4) Content is enabled through a manual auditing system, and then is provided to content consumers of the terminal, namely content index information obtained by a consumer end through a content outlet distribution service (usually a recommendation engine or a search engine or operation) directly showing pages;

fifth, manual auditing service

(1) The method is characterized in that the method is generally a World Wide Web (WEB) system, a machine filtering result is received on a link, the result is manually confirmed and checked, the checked result is written into a content information metadata base and recorded, and meanwhile, the actual effect of a machine processing and filtering model can be evaluated on line through the manually checked result;

(2) Reporting the detailed stream of auditing such as the source of a task, auditing results, auditing starting time and ending time and the like in the manual auditing process to a statistics server;

sixth, content storage service

(1) The content storage server is a group of storage servers with a wide distribution range and a relatively close distance from a client account, the CDN acceleration server is arranged at the periphery of the content storage service to carry out distributed cache acceleration, and video and picture contents uploaded by a content producer are stored through the uplink and downlink content interface servers;

(2) After acquiring the content index information, the terminal consumer can directly access the video content storage server to download the corresponding content;

(3) Besides being used as a data source for external service unexpectedly, the system is also used as a data source for internal service for the downloaded file system to acquire the original video data for relevant processing, and the paths of the internal and external data sources are usually arranged separately, so that the mutual influence is avoided, for example, when the cloud end detects the video file or the picture main body Mask, the original video data needs to be read from the paths. That is, the preset multimedia library in the above-described step S301 may be regarded as a content storage service from which the server acquires the target multimedia information.

Seventh, bullet screen template library

(1) The bullet screen template library is saved to obtain a template strategy, and the template strategy is generally configuration strategies such as fonts, font sizes, font colors, circulation times, bullet screen display areas and the like corresponding to a certain heat degree of the bullet screen;

(2) The bullet screen is usually matched by operators of bullet screens, and the client can also have a strategy actively set by a user, and if the client is set, the priority of the strategy can be configured;

eighth, mining hot spot barrage word stock

(1) According to the detailed procedures and steps described above (i.e., steps S201-S2013 above), including pre-processing of barrage content, keyword popularity calculation, popularity correction, outputting popularity dictionary, the popularity dictionary being stored in hot barrage word stock;

(2) The bullet screen enhancement and cloud detection service communication is adopted, inquiry and cooperation output of bullet screen hot spot information are provided, and meanwhile, a hotness word stock is updated regularly;

nine, bullet screen enhancement and cloud detection service

(1) Obtaining a Mask file of the picture body according to the enhancement method and the picture body detection mode (namely the steps S106, S107 and S301-S305);

(2) Acquiring strategy configuration enhanced by the barrage and Mask files of a main body through interactive information service communication and a barrage template library, and then issuing corresponding strategies to a client for rendering, displaying and playing the barrage by the client;

Tenth interactive information service

(1) According to the flow described above, the client and the bullet screen enhancement are communicated with the cloud detection service, and the bullet screen data persistent storage and the bullet screen enhancement information and configuration acquisition and distribution are completed.

Eleven barrage database

(1) The method comprises the steps of storing historical barrage information and running water records, providing barrage data sources for hot barrage word stock mining, and leading historical comment data into the database in an offline mode to serve as the mined historical data sources; i.e. the history interaction information in step S201 may be obtained from the barrage database.

(2) And communicating with the interactive information service, and storing bullet screen data issued by the client.

Twelve-step statistics server

(1) Receiving reports of consumption running water of a content auditing end and a content consumption end;

(2) Carrying out statistics mining and analysis on the reported running water, filtering and reading by a machine algorithm model, and simultaneously providing monitoring and analysis of a scheduling effect on content enabling rate and content auditing backlog time delay;

(3) According to the log information pushed by the statistical reporting interface server, performing statistical analysis to calculate basic result data such as click rate, reaching rate, conversion rate, client account related data and operation characteristics of the content;

(4) And the recommendation engine server is communicated with the recommendation engine server to provide content recommendation service based on the operation statistical information of the client account and the associated data information of the client account.

Fig. 9 is a block diagram illustrating an interactive information rendering apparatus according to an exemplary embodiment. As shown in fig. 9, the apparatus may include at least:

the hot word interaction information acquisition module 601 is configured to acquire a hot word set and interaction information to be displayed for target multimedia information release; the hot words in the hot word set are generated based on at least two historical word recognition results and historical heat data corresponding to the at least two historical word recognition results, and the at least two historical word recognition results are obtained by carrying out word recognition processing on historical interaction information published in a historical time period; the historical heat data of each historical word recognition result is obtained by correcting the initial heat data of each historical word recognition result by utilizing the average value of the initial heat data of each historical word recognition result; the initial heat data of each historical word recognition result is obtained by determining a ratio of a first word frequency in a target time interval to a second word frequency in a historical time interval based on each historical word recognition result; the target time interval is a time interval associated with a historical time period.

The target word recognition result determining module 603 is configured to perform word recognition processing on the interactive information to be displayed, so as to obtain a target word recognition result corresponding to the interactive information to be displayed.

The hotword data obtaining module 605 is configured to obtain, from a hotword library, a target hotword that matches the target word recognition result and target hotness data corresponding to the target hotword.

And the hotword data sending module 607 is configured to send the target hotword and the target hotword data to the client, so that the client renders the target hotword on the target multimedia information based on the target hotword data.

In an alternative embodiment, the apparatus may further include a hot word set construction module, where the hot word set construction module may include:

the history interaction information acquisition unit is used for acquiring history interaction information aiming at the release of the history multimedia information in a history time period.

The historical word recognition result generation unit is used for carrying out word recognition processing on the historical interaction information to obtain at least two historical word recognition results in the historical interaction information.

And the correction unit is used for correcting the initial heat data of each historical word recognition result according to the average value of the initial heat data of each historical word recognition result to obtain the historical heat data of each historical word recognition result.

The target historical word recognition result generation unit is used for determining a target historical word recognition result from at least two historical word recognition results according to the historical heat data of each historical word recognition result.

And the hot word set generating unit is used for generating a hot word set based on the target historical word recognition result.

In an alternative embodiment, the hot vocabulary construction module may further include:

and the word frequency determining unit is used for determining a first word frequency of each historical word recognition result in the target time interval and a second word frequency in the historical time interval.

The initial heat data generating unit is used for calculating the ratio of the first word frequency to the second word frequency to obtain initial heat data of each historical word recognition result.

In an alternative embodiment, the correction unit may include:

the average word frequency determining subunit is configured to determine an average word frequency corresponding to the at least two historical word recognition results according to the second word frequency of each historical word recognition result in the historical time period, the number of the at least two historical word recognition results, and the number of time intervals included in the historical time period.

And the heat data correction subunit is used for correcting the initial heat data corresponding to each historical word recognition result based on the average word frequency, the average value of the initial heat data of each historical word recognition result and the first word frequency of each historical word recognition result in the target time interval, and obtaining the historical heat data of each historical word recognition result.

In an alternative embodiment, the apparatus may further include:

the display strategy set acquisition module is used for acquiring a display strategy set; the set of display policies includes display policies corresponding to different heat data.

The target display strategy acquisition module is used for acquiring the target display strategy corresponding to the target heat data from the display strategies corresponding to the different heat data.

And the configuration processing module is used for carrying out display configuration processing on the target hotword according to the target display strategy to obtain the configured target hotword corresponding to the interactive information to be displayed.

Accordingly, the hotword data sending module 607 includes:

and the configured target hotword sending unit is used for sending the configured target hotword to the client so that the client renders the configured target hotword on the target multimedia information.

In an alternative embodiment, the target display policy includes a target font property display policy, a target word position display policy, and a target rendering frequency display policy, and the configuration processing module includes:

and the display attribute configuration unit is used for configuring the display attribute of the fonts in the target hotword in the rendering process based on the target font attribute display strategy.

The display position configuration unit is used for configuring the display position of the target hotword in the rendering process according to the target word position display strategy.

The rendering frequency configuration unit is used for configuring the rendering frequency of the target hotword in the rendering process based on the target rendering frequency display strategy.

The configured target hotword generation unit is used for generating the configured target hotword according to the target hotword, the display attribute, the display position and the rendering frequency.

In an optional embodiment, the number of the interactive information to be displayed is a plurality of, and the release time of each interactive information to be displayed is different, and the configured target hotword sending unit includes:

the release time acquisition subunit is used for acquiring release times corresponding to the interaction information to be displayed respectively.

The hot word time sending subunit is used for sending the configured target hot words corresponding to the interaction information to be displayed and the release time corresponding to the interaction information to be displayed to the client, so that the client sequentially renders the configured target hot words corresponding to the interaction information to be displayed on the target multimedia information according to the release time corresponding to the interaction information to be displayed.

In an alternative embodiment, the apparatus further comprises:

the target multimedia information acquisition module is used for acquiring target multimedia information from a preset multimedia library before the target multimedia information is displayed.

The segmentation module is used for carrying out foreground and background segmentation processing on the target multimedia information to obtain mask information corresponding to the target multimedia information; the mask information comprises a foreground area and a background area, wherein the foreground area is used for displaying an object main body in the target multimedia information, and the background area is used for displaying the configured target hotword.

And the storage module is used for storing the mask information to the content distribution network.

In an alternative embodiment, the segmentation module includes:

the voice segmentation unit is used for inputting the target multimedia information into the semantic segmentation model to carry out foreground and background segmentation processing, and obtaining a foreground segmentation result and a background segmentation result.

The pixel setting unit is used for setting the pixel value of the background segmentation result as a first numerical value to obtain a background area; setting the value of the foreground segmentation result as a second value to obtain a foreground region; the first value is greater than the second value.

And the mask information generating unit is used for generating mask information according to the background area and the foreground area.

In an optional embodiment, the configured target hotword sending unit includes:

and the mask information acquisition subunit is used for acquiring the mask information from the content distribution network.

And the hotword mask sending subunit is used for sending the configured target hotword and mask information to the client so that the client displays the object main body in the foreground area and the configured target hotword in the background area.

In an alternative embodiment, the method further comprises:

the hot word set sending module is used for sending the hot word set to the client so that the client can determine the hot word display size and the hot word display position of each hot word according to the historical heat data corresponding to each hot word in the hot word set; and rendering each hotword by the client according to the hotword display size and the hotword display position of each hotword.

It should be noted that, the device embodiment provided by the embodiment of the present application and the method embodiment described above are based on the same inventive concept.

The embodiment of the application also provides an electronic device for rendering the interactive information, which comprises a processor and a memory, wherein at least one instruction or at least one section of program is stored in the memory, and the at least one instruction or the at least one section of program is loaded and executed by the processor so as to realize the interactive information rendering method provided by any embodiment.

Embodiments of the present application also provide a computer readable storage medium that may be provided in a terminal to store at least one instruction or at least one program for implementing an interactive information rendering method in a method embodiment, the at least one instruction or the at least one program being loaded and executed by a processor to implement the interactive information rendering method as provided in the method embodiment described above.

Alternatively, in the present description embodiment, the storage medium may be located in at least one network server among a plurality of network servers of the computer network. Alternatively, in the present embodiment, the storage medium may include, but is not limited to: a U-disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing program codes.

The memory of the embodiments of the present specification may be used for storing software programs and modules, and the processor executes various functional applications and data processing by executing the software programs and modules stored in the memory. The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, application programs required for functions, and the like; the storage data area may store data created according to the use of the device, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory may also include a memory controller to provide access to the memory by the processor.

Embodiments of the present application also provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the interactive information rendering method provided by the above method embodiment.

The embodiment of the interactive information rendering method provided by the embodiment of the application can be executed in a terminal, a computer terminal, a server or similar computing devices. Taking the example of running on a server, fig. 10 is a block diagram of a hardware architecture of a server for interactive information rendering according to an exemplary embodiment. As shown in fig. 10, the server 700 may vary considerably in configuration or performance and may include one or more central processing units (Central Processing Units, CPU) 710 (the central processing unit 710 may include, but is not limited to, a microprocessor MCU or a processing device such as a programmable logic device FPGA), a memory 730 for storing data, one or more storage mediums 720 (e.g., one or more mass storage devices) for storing applications 723 or data 722. Wherein memory 730 and storage medium 720 may be transitory or persistent. The program stored in the storage medium 720 may include one or more modules, each of which may include a series of instruction operations on the server. Still further, the central processor 710 may be configured to communicate with the storage medium 720 and execute a series of instruction operations in the storage medium 720 on the server 700. The server 700 may also include one or more power supplies 760, one or more wired or wireless network interfaces 750, one or more input/output interfaces 740, and/or one or more operating systems 721, such as Windows ServerTM, mac OS XTM, unixTM, linuxTM, freeBSDTM, and the like.

Input-output interface 740 may be used to receive or transmit data via a network. The specific example of the network described above may include a wireless network provided by a communication provider of the server 700. In one example, the input-output interface 740 includes a network adapter (Network Interface Controller, NIC) that can connect to other network devices through a base station to communicate with the internet. In one example, the input/output interface 740 may be a Radio Frequency (RF) module for communicating with the internet wirelessly.

It will be appreciated by those of ordinary skill in the art that the configuration shown in fig. 10 is merely illustrative and is not intended to limit the configuration of the electronic device described above. For example, server 700 may also include more or fewer components than shown in fig. 10, or have a different configuration than shown in fig. 10.

It should be noted that: the sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for the device and server embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and references to the parts of the description of the method embodiments are only required.

It will be appreciated by those of ordinary skill in the art that all or part of the steps of implementing the above embodiments may be implemented by hardware, or may be implemented by a program to instruct related hardware, and the program may be stored in a computer readable storage medium, where the storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.

The foregoing is only illustrative of the present application and is not to be construed as limiting thereof, but rather as various modifications, equivalent arrangements, improvements, etc., within the spirit and principles of the present application.

Claims

1. An interactive information rendering method, characterized in that the method comprises:

2. The interactive information rendering method according to claim 1, wherein the hot word set generating method comprises:

acquiring historical interaction information aiming at the release of the historical multimedia information in a historical time period;

performing word recognition processing on the historical interaction information to obtain at least two historical word recognition results in the historical interaction information;

correcting the initial heat data of each historical word recognition result according to the average value of the initial heat data of each historical word recognition result to obtain the historical heat data of each historical word recognition result;

according to the historical heat data of each historical word recognition result, determining a target historical word recognition result from the at least two historical word recognition results;

And generating the hot word set based on the target historical word recognition result.

3. The interactive information rendering method according to claim 2, wherein the generating method of the initial heat data of each of the history word recognition results comprises:

determining a first word frequency of each historical word recognition result in a target time interval and a second word frequency in the historical time interval; the target time interval is a time interval associated with the historical time period;

and calculating the ratio of the first word frequency to the second word frequency to obtain initial heat data of each historical word recognition result.

4. The interactive information rendering method according to claim 3, wherein said correcting the initial heat data of each of the historical word recognition results according to the average value of the initial heat data of each of the historical word recognition results to obtain the historical heat data of each of the historical word recognition results comprises:

determining average word frequencies corresponding to the at least two historical word recognition results according to the second word frequency of each historical word recognition result in the historical time period, the number of the at least two historical word recognition results and the number of time intervals contained in the historical time period;

And correcting the initial heat data of each historical word recognition result based on the average word frequency, the average value of the initial heat data of each historical word recognition result and the first word frequency of each historical word recognition result in the target time interval to obtain the historical heat data of each historical word recognition result.

5. The interactive information rendering method according to any one of claims 1 to 4, wherein after acquiring a target hotword that matches the target word recognition result and target hotword corresponding target hotword data, the method further comprises:

acquiring a display strategy set; the display strategy set comprises display strategies corresponding to different heat data;

acquiring a target display strategy corresponding to the target heat data from the display strategies corresponding to the different heat data;

performing display configuration processing on the target hotword according to the target display strategy to obtain a configured target hotword corresponding to the interaction information to be displayed;

the sending the target hotword and the target hotword to a client, so that the client renders the target hotword on the target multimedia information based on the target hotword, including:

And sending the configured target hotword to the client so that the client renders the configured target hotword on the target multimedia information.

6. The method for rendering interactive information according to claim 5, wherein the target display policy includes a target font property display policy, a target word position display policy, and a target rendering frequency display policy, and the performing display configuration processing on the target hotword according to the target display policy to obtain a configured target hotword corresponding to the interactive information to be displayed includes:

configuring the display attribute of the fonts in the target hotword in the rendering process based on the target font attribute display strategy;

according to the target word position display strategy, configuring the display position of the target hotword in the rendering process;

configuring the rendering frequency of the target hotword in the rendering process based on the target rendering frequency display strategy;

and generating the configured target hotword according to the target hotword, the display attribute, the display position and the rendering frequency.

7. The method for rendering interactive information according to claim 5, wherein the number of interactive information to be displayed is plural, and the release time of each interactive information to be displayed is different, and the sending the configured target hotword to the client so that the client renders the configured target hotword on the target multimedia information includes:

Acquiring release time corresponding to each of a plurality of interactive information to be displayed;

and sending the configured target hotwords corresponding to the interaction information to be displayed and the release time corresponding to the interaction information to be displayed to the client, so that the client sequentially renders the configured target hotwords corresponding to the interaction information to be displayed on the target multimedia information according to the release time corresponding to the interaction information to be displayed.

8. The interactive information rendering method according to claim 5, further comprising:

before the target multimedia information is displayed, acquiring the target multimedia information from a preset multimedia library;

performing foreground and background segmentation processing on the target multimedia information to obtain mask information corresponding to the target multimedia information; the mask information comprises a foreground area and a background area, the foreground area is used for displaying an object main body in the target multimedia information, and the background area is used for displaying the configured target hotword;

storing the mask information to a content distribution network.

9. The interactive information rendering method according to claim 8, wherein the performing foreground and background segmentation on the target multimedia information to obtain mask information corresponding to the target multimedia information includes:

Inputting the target multimedia information into a semantic segmentation model to carry out foreground and background segmentation processing to obtain a foreground segmentation result and a background segmentation result;

setting a pixel value of the background segmentation result as a first numerical value to obtain the background region; setting the value of the foreground segmentation result as a second value to obtain the foreground region; the first value is greater than the second value;

and generating the mask information according to the background area and the foreground area.

10. The interactive information rendering method of claim 8, wherein the sending the configured target hotword to the client to cause the client to render the configured target hotword on the target multimedia information comprises:

acquiring the mask information from the content distribution network;

and sending the configured target hotword and the mask information to the client so that the client displays the object body in the foreground area and displays the configured target hotword in the background area.

11. The interactive information rendering method according to any one of claims 1 to 4, characterized in that the method further comprises:

The hot word set is sent to the client, so that the client determines the hot word display size and the hot word display position of each hot word according to the historical hot data corresponding to each hot word in the hot word set; and rendering each hotword by the client according to the hotword display size and the hotword display position of each hotword.

12. An interactive information rendering apparatus, characterized in that the apparatus comprises:

13. An electronic device for interactive information rendering, characterized in that it comprises a processor and a memory in which at least one instruction or at least one program is stored, the at least one instruction or the at least one program being loaded and executed by the processor to implement the interactive information rendering method according to any one of claims 1 to 11.

14. A computer-readable storage medium having stored therein at least one instruction or at least one program that is loaded and executed by a processor to implement the interactive information rendering method of any one of claims 1 to 11.

15. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the interactive information rendering method of any one of claims 1 to 11.