CN117709311B

CN117709311B - Cloud-based lecture manuscript management method, device, equipment and storage medium

Info

Publication number: CN117709311B
Application number: CN202410165079.7A
Authority: CN
Inventors: 李翔; 吴云川; 詹歆; 赵璧
Original assignee: Xinlicheng Education Technology Co ltd
Current assignee: Xinlicheng Education Technology Co ltd
Priority date: 2024-02-05
Filing date: 2024-02-05
Publication date: 2024-05-03
Anticipated expiration: 2044-02-05
Also published as: CN117709311A

Abstract

The application provides a cloud-based lecture management method, a cloud-based lecture management device, equipment and a storage medium, wherein the cloud-based lecture management method is used for determining a talent dimension score of a talent dimension by acquiring an original lecture and carrying out talent analysis of at least one talent dimension on the original lecture, determining an overall quality score of the original lecture according to the talent dimension score, a dimension weight of the talent dimension and scene factor parameters, generating an optimization suggestion according to the overall quality score and the talent dimension score, and outputting the optimization suggestion based on the talent dimension score, so that high-quality optimization suggestion can be conveniently and efficiently provided for an object; and responding to an editing instruction of at least one object on the display page, editing the original lecture, generating a revised lecture, determining the comprehensive score of the revised lecture, generating a correction suggestion according to the comprehensive score, and automatically scoring and providing the correction suggestion when one or more objects are edited, thereby being beneficial to the co-collaboration of the lecture.

Description

Cloud-based lecture manuscript management method, device, equipment and storage medium

Technical Field

The application relates to the field of lectures, in particular to a lecture draft management method, device and equipment based on a cloud, and a storage medium.

Background

The conventional lecture creation and management method is generally based on a single-machine editing software, each lecturer edits the lecture locally and then merges the manuscripts of each version manually, so the following problems exist: 1. decentralized editing and collaboration difficulties: different lecturers scatter in different places to edit lectures, and the collaboration is difficult, so that version confusion and conflict are easy to cause; 2. the expression quality of the talents is uneven, the lecture quality can be modified only by experienced people, the complexity and subjectivity are high, and the high-quality expression is difficult to ensure; 3. the efficiency is low: the manual merging of the multi-version manuscripts is time-consuming, labor-consuming, error-prone and not suitable for large-scale collaboration and urgent lecture production.

Disclosure of Invention

The embodiment of the application provides a cloud-based lecture draft management method, device, equipment and storage medium, which are used for solving at least one problem existing in the related technology, and the technical scheme is as follows:

in a first aspect, an embodiment of the present application provides a method for lecture management based on cloud, including:

acquiring an original lecture, carrying out at least one talent dimension talent analysis on the original lecture, and determining a talent dimension score of the talent dimension;

Determining the overall quality score of the original lecture according to the talent dimension score, the dimension weight of the talent dimension and the scene factor parameter;

Generating an optimization suggestion according to the overall quality score and the spoken dimension score, and displaying the optimization suggestion on a display page of a cloud server;

Responding to an editing instruction of at least one object on the display page, editing the original lecture manuscript, and generating a revised lecture manuscript;

and determining the comprehensive score of the revised lecture, and generating a revised suggestion according to the comprehensive score.

In one embodiment, the performing of the spoken analysis of the original lecture for at least one spoken dimension, determining a spoken dimension score for the spoken dimension includes at least one of:

Calculating emotion intensity scores of all sentences and emotion continuity scores of all sentences in the original lecture, and determining a first talent dimension score of emotion transmission force dimension according to the number of sentences, the emotion intensity scores and the emotion continuity scores;

calculating the effectiveness score of the repair methods applied to each sentence in the original lecture, and determining a second spoken dimension score of the spoken dimension of the repair effect according to the effectiveness score, the total word number of the original lecture and the variety number of the repair methods;

Calculating the structure score of each paragraph in the original lecture, and determining a third talent dimension score of the structure rationality dimension according to the structure score and the number of the paragraphs;

Calculating the consistency score of each statement in the original lecture, and determining a fourth talent dimension score of the consistency dimension of the statement according to the consistency score and the number of the statement;

Calculating the definition score of each argument in the original lecture, and determining a fifth talent dimension score of the argument definition dimension according to the definition score and the number of the arguments;

calculating the score of each visual expression in the original lecture, and determining a sixth talent dimension score of the visual dimension according to the score of the visual expression and the number of the visual expressions;

calculating the score of each metaphor in the original lecture, and determining a seventh talent dimension score of a metaphor using dimension according to the score of the metaphor and the number of the metaphors;

And calculating the score of each emotion guiding point in the original lecture, and determining an eighth talent dimension score of the emotion guiding dimension according to the score of the emotion guiding point and the number of the emotion guiding points.

In one embodiment, the determining the overall quality score of the original lecture according to the spoken dimension score, the dimension weight of the spoken dimension, and the scene factor parameter includes:

According to at least one of the first talent dimension score, the second talent dimension score, the third talent dimension score, the fourth talent dimension score, the fifth talent dimension score, the sixth talent dimension score, the seventh talent dimension score and the eighth talent dimension score, carrying out weighted calculation according to the dimension weight and the dimension weight coefficient of the talent dimension, and obtaining a talent dimension quality score;

determining an overall quality score of the original lecture according to the sum value of the talent dimension quality score and the scene factor parameter;

Wherein the scene factor parameter characterizes the score of the remaining influencing factors except for the spoken dimension.

In one embodiment, the determining the composite score for the revised lecture includes:

determining a logical dimension score for a logical dimension of the revised lecture;

determining a language activity dimension score for the language activity dimension of the revised lecture;

Determining an emotion communication force dimension score for an emotion communication force dimension of the revised lecture;

Determining a workout effect dimension score for a workout effect dimension of the revised lecture;

And carrying out weighted calculation according to the logical dimension score, the language mobility dimension score, the emotion conveying force dimension score, the pedigree effect dimension score and a preset weight to obtain the comprehensive score of the revised lecture.

In one embodiment, the generating correction advice based on the composite score includes:

When the composite score of the revised lecture is less than a first scoring threshold:

determining a first target score of the lowest score among the logical dimension score, the language mobility dimension score, the emotion conveying force dimension score and the pedigree effect dimension score, determining a first target dimension corresponding to the first target score, and generating a first correction sub-suggestion according to the first target dimension;

Or alternatively

Determining a second target score lower than a second score threshold in the logic dimension score, the language mobility dimension score, the emotion conveying force dimension score and the pedigree effect dimension score, determining a second target dimension corresponding to the second target score, and generating a second modifier suggestion according to the second target dimension.

In one embodiment, the method further comprises:

When the number of the objects is a plurality of, determining corresponding revision lectures of a plurality of different versions;

Determining a first score of a structural rationality dimension, a second score of a discourse consistency dimension, a third score of an image dimension, a fourth score of a metaphor use dimension, and a fifth score of an emotion guiding dimension of different versions of the revised lecture;

determining a logical dimension merging score according to a first preset weight parameter, the first score and the second score, determining a language mobility dimension merging score according to a second preset weight parameter, the third score and the fourth score, and determining an emotion conveying force dimension merging score according to a third preset weight parameter and the fifth score;

determining the total merging score of the merged lectures obtained after the revised lectures of different versions are merged according to the logical dimension merging score, the language mobility dimension merging score and the emotion conveying force dimension merging score;

And synchronizing the combined lectures with the highest total combined score on the cloud server.

In one embodiment, the method further comprises:

When the number of the objects is a plurality of and the editing contents of different objects have conflicts, determining editing scores or score differences of different editing contents in at least one talent dimension;

determining a target conflict resolution strategy according to the score difference or the editing score so as to determine target editing content;

And combining the target editing content with the original lecture manuscript to obtain a target combined lecture manuscript.

In a second aspect, an embodiment of the present application provides a lecture management apparatus based on a cloud, including:

the acquisition module is used for acquiring an original lecture, carrying out at least one talent dimension talent analysis on the original lecture, and determining a talent dimension score of the talent dimension;

The determining module is used for determining the overall quality score of the original lecture according to the talent dimension score, the dimension weight of the talent dimension and the scene factor parameter;

The display module is used for generating an optimization suggestion according to the overall quality score and the talent dimension score and displaying the optimization suggestion on a display page of a cloud server;

The response module is used for responding to an editing instruction of at least one object on the display page, editing the original lecture and generating a revised lecture;

And the generation module is used for determining the comprehensive score of the revised lecture and generating a revised suggestion according to the comprehensive score.

In one embodiment, the generating module is further configured to:

When the number of the objects is a plurality of and the editing contents of different objects have conflicts, determining the score difference of different editing contents in at least one talent dimension;

Determining target editing content through a preset conflict resolution strategy according to the score difference;

In a third aspect, an embodiment of the present application provides an electronic device, including: a processor and a memory in which instructions are stored, the instructions being loaded and executed by the processor to implement the method of any of the embodiments of the above aspects.

In a fourth aspect, embodiments of the present application provide a computer readable storage medium storing a computer program, which when executed implements a method in any one of the embodiments of the above aspects.

The beneficial effects in the technical scheme at least comprise:

The method comprises the steps of obtaining an original lecture, carrying out at least one talent dimension talent analysis on the original lecture, determining a talent dimension score of the talent dimension, determining an overall quality score of the original lecture according to the talent dimension score, a dimension weight of the talent dimension and scene factor parameters, generating an optimization suggestion according to the overall quality score and the talent dimension score, displaying the optimization suggestion on a display page of a cloud server, and outputting the optimization suggestion based on the talent dimension score, so that high-quality optimization suggestion can be conveniently and efficiently provided for an object; and responding to an editing instruction of at least one object on the display page, editing the original lecture, generating a revised lecture, determining the comprehensive score of the revised lecture, generating a correction suggestion according to the comprehensive score, and automatically scoring when one or more objects are edited, so as to provide the correction suggestion for the revision, thereby being beneficial to the co-collaboration of the lecture.

The foregoing summary is for the purpose of the specification only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the present application will become apparent by reference to the drawings and the following detailed description.

Drawings

In the drawings, the same reference numerals refer to the same or similar parts or elements throughout the several views unless otherwise specified. The figures are not necessarily drawn to scale. It is appreciated that these drawings depict only some embodiments according to the disclosure and are not therefore to be considered limiting of its scope.

Fig. 1 is a schematic flow chart of steps of a lecture manuscript management method based on a cloud according to an embodiment of the application;

fig. 2 is a block diagram illustrating a cloud-based lecture management apparatus according to an embodiment of the present application;

Fig. 3 is a block diagram of an electronic device according to an embodiment of the application.

Detailed Description

Hereinafter, only certain exemplary embodiments are briefly described. As will be recognized by those of skill in the pertinent art, the described embodiments may be modified in various different ways without departing from the spirit or scope of the present application. Accordingly, the drawings and description are to be regarded as illustrative in nature and not as restrictive.

Referring to fig. 1, a flowchart of a cloud-based lecture management method according to an embodiment of the present application is shown, where the cloud-based lecture management method at least includes steps S100 to S500:

s100, acquiring an original lecture, carrying out at least one talent dimension talent analysis on the original lecture, and determining a talent dimension score of the talent dimension.

And S200, determining the overall quality score of the original lecture according to the talent dimension score, the dimension weight of the talent dimension and the scene factor parameter.

And S300, generating an optimization suggestion according to the overall quality score and the talent dimension score, and displaying the optimization suggestion on a display page of the cloud server.

S400, responding to an editing instruction of at least one object on the display page, editing the original lecture manuscript and generating a revised lecture manuscript.

S500, determining comprehensive scores of the revised lectures, and generating correction suggestions according to the comprehensive scores.

The lecture management method based on the cloud terminal in the embodiment of the application can be executed by an electronic control unit, a controller, a processor and the like of a terminal such as a computer, a mobile phone, a tablet, a vehicle-mounted terminal and the like, and also can be executed by a cloud server, for example, by a system in the cloud server.

According to the technical scheme, the original lecture is obtained, at least one talent dimension talent analysis is carried out on the original lecture, the talent dimension score of the talent dimension is determined, the overall quality score of the original lecture is determined according to the talent dimension score, the dimension weight of the talent dimension and the scene factor parameter, the optimization suggestion is generated according to the overall quality score and the talent dimension score, and the optimization suggestion is displayed on a display page of a cloud server, and is output based on the talent dimension score, so that the high-quality optimization suggestion can be conveniently and efficiently provided for an object; and responding to an editing instruction of at least one object on the display page, editing the original lecture, generating a revised lecture, determining the comprehensive score of the revised lecture, generating a correction suggestion according to the comprehensive score, and automatically scoring when one or more objects are edited, so as to provide the correction suggestion for the revision, thereby being beneficial to the co-collaboration of the lecture.

In one embodiment, the user may upload the original lecture to the cloud server for the cloud server to acquire the original lecture and then process the cloud server. Optionally, step S100 performs at least one spoken analysis of the original lecture for a spoken dimension, and determining a spoken dimension score for the spoken dimension includes at least one of steps S110-S180:

s110, calculating emotion intensity scores of all sentences and emotion continuity scores of all sentences in the original lecture, and determining a first talent dimension score of emotion conveying force dimension according to the number of sentences, the emotion intensity scores and the emotion continuity scores.

Optionally, a first talent dimension score of the emotion conveying force dimensionThe calculation formula of (2) is as follows:

Wherein, For original lectures,/>For the number of sentences,/>For the emotion intensity score of sentence j, calculated by emotion analysis algorithm,/>And scoring the emotion continuity of the sentence j, and taking the rationality and continuity of emotion fluctuation in the whole lecture draft into consideration and calculating by a deep learning algorithm.

S120, calculating effectiveness scores of the correction techniques applied to each sentence in the original lecture, and determining second talent dimension scores of talent dimensions of the correction effects according to the effectiveness scores, the total word number of the original lecture and the variety number of the correction techniques.

Optionally, a second spoken dimension score of the spoken dimension of the grooming effectThe calculation formula of (2) is as follows:

Wherein, For the variety and quantity of the techniques of the repair/instructionScore for effectiveness of kth method of tutorial applied in sentence k,/>The total word number of the original lecture is used for normalizing the score of the correction effect.

S130, calculating the structure score of each paragraph in the original lecture, and determining a third talent dimension score of the structure rationality dimension according to the structure score and the number of the paragraphs.

Optionally, a third spoken dimension score of the structural rationality dimensionThe calculation formula of (2) is as follows:

Wherein, For the number of paragraphs,/>For the structure score (structure rationality score) of paragraph j, it is calculated by analyzing the relationships and transitions between sentences inside the paragraph, for example, by considering the logical relationships between sentences, the quality and suitability of transition words and phrases, the logical continuity of sentences, the fluency of sentences inside the paragraph, or user-customized rules.

And S140, calculating the consistency score of each statement in the original lecture, and determining a fourth talent dimension score of the consistency dimension of the statement according to the consistency score and the number of the statement.

Optionally, a fourth spoken dimension score that argues for a coherence dimensionThe calculation formula of (2) is as follows:

Wherein, For the number of arguments,/>The consistency score of the arguments k is calculated by a deep learning algorithm, for example, by considering the logical relevance, the link, the grammar and the language style among the arguments or the custom rules.

And S150, calculating the definition score of each argument in the original lecture, and determining a fifth talent dimension score of the definition dimension of the argument according to the definition score and the number of arguments.

Optionally, a fifth spoken dimension score of the argument specificity dimensionThe calculation formula of (2) is as follows:

Wherein, For the number of arguments,/>And evaluating the definition and the definition of the discussion point for the definition score of the discussion point l, and calculating by a deep learning algorithm. For example, consider grammar and logic, vocabulary and expressions, context, or user-customized rule calculations.

S160, calculating the score of each avatar expression in the original lecture, and determining a sixth talent dimension score of the avatar dimension according to the score of the avatar expression and the number of the avatar expressions.

Optionally, a sixth spoken dimension score of the visual dimensionThe calculation formula of (2) is as follows:

Wherein, For the number of visual expressions,/>For visual expression/>The score of (2) is calculated by a deep learning algorithm, and the vividness and the infectivity of the expression are considered. For example, consider vocabulary and expression choices, metaphor and metaphor analysis, emotion and emotion expressions, or custom rules.

S170, calculating the scores of the metaphors in the original lecture, and determining a seventh talent dimension score of the metaphors using dimension according to the scores of the metaphors and the number of the metaphors.

Optionally, metaphor uses a seventh spoken dimension score of the dimensionThe calculation formula of (2) is as follows:

Wherein, For the number of metaphors,/>The score of metaphor n is calculated by a deep learning algorithm, and the ingenious degree and the expression effect of the metaphor are evaluated. For example, consider emotion, keywords and phrases, emotion intensity assessment, emotion sequence, resonance assessment, etc. factor calculations.

S180, calculating the score of each emotion guiding point in the original lecture, and determining an eighth talent dimension score of the emotion guiding dimension according to the score of the emotion guiding point and the number of the emotion guiding points.

Optionally, an eighth spoken dimension score of the emotion guiding dimensionThe calculation formula of (2) is as follows:

Wherein, For the number of emotion guide points,/>Guide points/>The score of the emotion expression is calculated through a deep learning algorithm, and the emotion expression effect and emotion resonance degree are considered.

It should be noted that, in the embodiment of the present application including the steps S110 to S180, in some implementations, one or more of the steps may be included, and the spoken dimension may also include other dimensions, which are not limited specifically.

In one embodiment, step S200 includes steps S210-S220:

S210, carrying out weighted calculation according to at least one of the first talent dimension score, the second talent dimension score, the third talent dimension score, the fourth talent dimension score, the fifth talent dimension score, the sixth talent dimension score, the seventh talent dimension score and the eighth talent dimension score, and carrying out weighted calculation according to the dimension weight and the dimension weight coefficient of the talent dimension to obtain a talent dimension quality score.

S220, determining the overall quality score of the original lecture according to the sum value of the talent dimension quality score and the scene factor parameter.

In the embodiment of the application, the score corresponding to each talent dimension i is provided with a corresponding dimension weightReflecting the inherent importance of this dimension in the overall score, each talent dimension i has a corresponding dimension weight coefficient/>The system can be dynamically adjusted according to different user demands and occasions, and allows the user to customize the optimization direction in a personalized way. Optionally, scene factor parameter/>Score characterizing influence factors other than the spoken dimension,/>For additional correction parameters, for taking into account other influencing factors than the spoken dimension,/>Scoring for special factors, such as relevance and adaptability to a particular occasion or topic, to ensure that the lecture content has good adaptability and pertinency to a particular environment.

Optionally, an overall quality score of the original lecture is calculatedThe formula of (2) is:

Wherein, For the number of spoken dimensions,/>A mouth dimension score for the i-th mouth dimension, e.g. i=1,The first mouth dimension score.

Alternatively, wi may be updated according to feedback information using gradient descent or other optimization algorithms, so that the optimized lecture may obtain a higher quality Q (T) in the target scenario. Or using a reinforcement learning algorithm: designing an agent to try to modify the lecture content in a simulation environment, giving rewards or punishment signals (such as adjusting Q (T) values according to audience feedback or expert opinion) by the system, and finally generating a higher-quality lecture by continuously iterating and learning an optimal editing strategy; or using genetic algorithms: taking a lecture as an individual, taking editing operation as genetic variation, iteratively generating new lecture versions in a population through operations such as crossing, mutation and the like, selecting the version with higher Q (T) score under the current environment as the next generation, and continuously optimizing until a preset termination condition is reached. Or deep neural network assisted optimization: and (5) using a pre-trained deep neural network to understand and reconstruct the content of the lecture. And the network parameters are adjusted through back propagation, so that the lecture fragments generated by the network are excellent in each talent dimension, and are further integrated into the original lecture, thereby realizing content optimization.

In one embodiment, in step S300, an optimization suggestion is generated according to the overall quality score and the spoken dimension score, and the optimization suggestion is displayed in a list form on a display page of the cloud server, so that a user can watch the optimization suggestion, refer to the optimization suggestion, and edit an original lecture. For example, the formula for generating the optimization suggestion O (T) is: o (T) =dnn [ ]，[/>，...../>The DNN is a pre-trained deep neural network model for outputting generated targeted optimization suggestions, which may include, but are not limited to: the vocabulary is replaced to enhance the expressive force, adjust the paragraph sequence to improve the logic, insert the interactive links to improve the audience participation, and the like.

In the embodiment of the application, the intelligent lecture content optimization engine fuses the Natural Language Processing (NLP) technology and the professional knowledge in the field of talent training, performs refined processing on the original lecture input by the user by utilizing a deep learning algorithm and a semantic analysis technology, determines the score, can detect and correct grammar errors, and can also provide optimization suggestions from the aspect of talent expression, such as improving sentence structure to enhance logic and consistency, recommending words and sentence patterns with more infectivity to improve emotion resonance degree, and improving the effectiveness of information transmission by using simple and clear language. The optimization suggestions can be fed back in real time by the engine, and guidance is embedded in the user writing process, so that the user is supported to gradually optimize the content quality of the lecture, and the user is enabled to better meet the excellent oral expression requirements.

In one embodiment, in response to an edit instruction of the display page by at least one object, the original lecture is edited to generate a revised lecture in step S400. In the embodiment of the application, the cloud server provides a function of editing one or more objects in parallel, for example, the object A performs editing operation on a display page of the cloud server to generate an editing instruction, and the cloud server responds to the editing instruction to edit an original lecture and generate a revised lecture. If editing is performed by a plurality of objects, when there is no conflict in editing contents of different objects, a corresponding plurality of revised lectures may be generated at this time.

In one embodiment, the step S500 of determining the composite score of the revised lecture includes steps S510-S550:

s510, determining a logical dimension score of the logical dimension of the revised lecture.

Optionally, a logical dimension scoreThe calculation formula of (2) is as follows:

Wherein, To revise lectures,/>To revise the number of sentences in the lecture,Is a weight parameter in the logical dimension,/>Is sentence/>Is used to determine the continuity score of the (c),Is sentence/>And the consistency score with the context is calculated by a deep learning calculation algorithm.

S520, determining a language mobility dimension score of the language mobility dimension of the revised lecture.

Alternatively, a formula for calculating a language mobility dimension scoreThe method comprises the following steps:

Wherein, To revise the total word count of lectures,/>To revise the metaphors and the number of images of the lecture,、/>Weight parameter for language mobility dimension,/>For/>Vivid score of individual image,/>For/>The effectiveness score of the metaphor is calculated by a deep learning calculation algorithm.

S530, determining emotion conveying force dimension scores of emotion conveying force dimensions of the revised lectures.

Optionally, emotion conveying force dimension scoringThe calculation formula of (2) is as follows:

Wherein, 、/>As a weight parameter for emotion conveying force dimension,Is sentence/>Emotion polarity score,/>Is sentence/>The emotion expression degree score of (2) is calculated by a deep learning calculation algorithm.

S540, determining the grade of the grade effect of the revision lecture.

Optionally, the calculation formula of the score of the pedigree effect dimension is:

Wherein, 、/>For the weight parameter of the pedigree effect dimension,/>Is sentence/>The application score of the middle and repair method,/>Is sentence/>The effect score of the middle and the middle congratulation is calculated by a deep learning calculation algorithm.

S550, carrying out weighted calculation according to the logical dimension score, the language liveness dimension score, the emotion conveying force dimension score, the congratulation effect dimension score and the preset weight to obtain the comprehensive score of the revised lecture.

Optionally, revise the composite score of the lectureThe calculation formula of (2) is as follows:

Wherein, 4 Dimensions of 4 corresponding logical dimensions, language vividness dimensions, emotion conveying force dimensions and pedigree effect dimensions,/>Preset weight for dimension i,/>For the Z-th revised lecture, if there is only one revised lecture, Z is 1,And (3) scoring the ith dimension of the Z revised lecture, wherein when i is 1,2, 3 and 4, the scores respectively correspond to the logical dimension score, the language liveness dimension score, the emotion conveying force dimension score and the pedigree effect dimension score.

In one embodiment, when the composite score of the revised lecture is less than the first score threshold in step S500, a revised proposal is generated according to the composite score, including steps S560 or S570:

S560, determining a first target score of the lowest score among the logical dimension score, the language liveness dimension score, the emotion conveying force dimension score and the pedigree effect dimension score, determining a first target dimension corresponding to the first target score, and generating a first modifier suggestion according to the first target dimension.

Optionally, correction suggestions corresponding to each dimension may be configured in advance in the cloud server, or correction models of each dimension may be trained in advance, and when it is determined that improvement is required in a certain dimension, the corresponding correction suggestions are matched, or the revised lecture is processed based on the corresponding correction models, so as to obtain the corresponding correction suggestions.

For example, a revised lecture determines a corresponding logical dimension score, language liveness dimension score, emotion conveying force dimension score, and a correction effect dimension score, and the correction effect dimension score is the lowest, where the first target score is the correction effect dimension score, and the corresponding first target dimension is the correction effect dimension, and then matches the correction suggestion (i.e., the first correction sub-suggestion) according to the correction effect dimension or generates the first correction sub-suggestion based on the correction model.

S570, determining a second target score lower than a second score threshold in the logical dimension score, the language liveness dimension score, the emotion conveying force dimension score and the pedigree effect dimension score, determining a second target dimension corresponding to the second target score, and generating a second modifier suggestion according to the second target dimension.

Optionally, a second score threshold may be set in advance, and a second target score lower than the second score threshold is determined, for example, the logical dimension score and the language vividness dimension score are all lower than the second score threshold, and then the logical dimension score and the language vividness dimension score are used as the second target score, and the corresponding second target dimension is determined to be the logical dimension and the language vividness dimension. Similarly, a second syndrome recommendation may be obtained by matching or modifying the model.

In one implementation manner, the cloud-based lecture management method of the embodiment of the present application may further include steps S610 to S650:

And S610, when the number of the objects is a plurality of, determining corresponding revisions of a plurality of different versions.

Optionally, when the number of objects is plural, a plurality of different versions of revised lectures corresponding to different objects are determined. Wherein, in order to determine the final revised lecture, the revised lecture of which version is in conflict at the time of editing can be determined. For example, there are three objects, A, B, C, A, B, for which there is a conflict, thus determining whether to merge the revised lectures with the A, C object or the revised lecture with the B, C object.

S620, determining a first score of a structural rationality dimension, a second score of a continuity dimension, a third score of a visual dimension, a fourth score of a metaphor usage dimension, and a fifth score of an emotion guiding dimension of different versions of the revised lecture.

Alternatively, a first score of a structural rationality dimension of the revised lecture may be calculated based on the principle of step S130, a second score of a continuity dimension of the argument of the revised lecture may be calculated based on the principle of step S140, a third score of a visual dimension of the revised lecture may be calculated based on the principle of step S160, a fourth score of a metaphor usage dimension of the revised lecture may be calculated based on the principle of step S170, and a fifth score of an emotion guiding dimension of the revised lecture may be calculated based on the principle of step S180, which will not be described again.

S630, determining a logical dimension merging score according to the first preset weight parameter, the first score and the second score, determining a language mobility dimension merging score according to the second preset weight parameter, the third score and the fourth score, and determining an emotion conveying force dimension merging score according to the third preset weight parameter and the fifth score.

Optionally, the first preset weight parameter includes a weight sub-parameter corresponding to the first score and the second score, and the second preset weight parameter includes a weight sub-parameter corresponding to the third score and the fourth score, so that a logical dimension merging score can be obtained by performing weighted calculation based on the first preset weight parameter, the first score and the second score, a language liveness dimension merging score can be determined by performing weighted calculation based on the second preset weight parameter, the third score and the fourth score, and an emotion transmission force dimension merging score can be determined by performing weighted calculation based on the third preset weight parameter and the fifth score.

And S640, determining the total merging score of the merged lectures obtained after the revised lectures of different versions are merged according to the logical dimension merging score, the language mobility dimension merging score and the emotion conveying force dimension merging score.

Optionally, a total aggregate score function Q 'is defined, such as the example described above, and the total aggregate score of the merged lectures resulting from this merging approach is determined by calculating A, C the sum of the logical dimension merge score, the language-activity dimension merge score, and the emotion-conveying-force dimension merge score for the version of the revised lecture using the score function Q'. Similarly, the sum of the logical dimension merge scores, language mobility dimension merge scores, and emotion conveying dimension merge scores for the revised lectures of the version of the B, C object is calculated using a scoring function Q', and the total merge score for the merged lecture resulting from this merge is determined.

And S650, synchronizing the combined lectures with the highest total combined score on the cloud server.

And then, synchronizing the combined lectures with the highest total combined score as the final lectures on a cloud server for the user (i.e. the object) to watch.

In one implementation manner, the cloud-based lecture management method of the embodiment of the present application may further include steps S710 to S730:

S710, when the number of the objects is a plurality of and the editing contents of different objects have conflicts, determining editing scores or score differences of different editing contents in at least one talent dimension.

For example, assuming that when the object a and the object B edit the original lecture, a conflict exists between the edited content of the object a and the edited content of the object B, the detected conflict may be detected by a conflict detection function ConflictDetected (Aedit, bedit), where Aedit is the edited content of the user a, bedit is the edited content of the user B, the output of this function may be a boolean value, based on which it may be determined whether the conflict is detected, and if the conflict exists, the score difference of the different edited contents in at least one talent dimension is determined, in the embodiment of the present application, the score difference in a plurality of talent dimensions is determined, and other embodiments may determine the score difference in one or more talent dimensions.

For example, edit scores may be calculated that determine the spoken dimensions of edit content Aedit, logical dimensions of edit content Bedit, language-activity dimensions, and the like, respectively, based on the principles of S510-S540. Or further determine a score difference Δ Scorei across different mouth dimensions i based on the edit score:

ΔScorei=DAi-DBi

where DAi is the edit score of user a editing content Aedit in the spoken dimension i and DBi is the edit score of user B editing content Bedit in the spoken dimension i.

S720, determining a target conflict resolution strategy according to the score difference or the editing score so as to determine target editing content.

Alternatively, a first total edit score of the edit content Aedit and a second total edit score DB of the edit content Bedit may be determined based on the edit scores DA of the respective dictation dimensions, and then a target conflict resolution policy may be determined from among preset conflict resolution policies based on the conflict resolution policy functions ConflictResolution (Aedit, bedit, DA, DB) to determine target edit content from the edit contents Aedit and Bedit by the target conflict resolution policy. Alternatively, the conflict resolution policy function may analyze both text differences based on the edit content Aedit and Bedit, such as if at least one of the newly added and modified portions exists, merging the newly added portion directly as the target edit content if the newly added portion exists, and further determining if the modified portion (including the deletion or modification) exists. In the embodiment of the application, the important priorities of different users can be set in advance, if the priority of the user A is higher than the priority of the user B, the target editing content is determined according to the modification part of the user A, and if the priorities of the user A and the user B are the same, the modification content updated in the modification time can be used as the target editing content according to the information of the additional data such as the time stamp, the editing purpose and the like. Or in some embodiments, the target edit content may also be determined based on user preferences set by the user personalization, e.g., the user may set compliance with the system or preference to which user modification, etc., when a conflict exists.

Or the importance of the versions corresponding to the different editing contents can be determined based on the score differences in the dimensions of the different portraits, when the sum of the score differences is greater than 0, the editing content Aedit is used as the target editing content, and otherwise, the editing content Bedit is used as the target editing content.

And S730, combining the target editing content with the original lecture manuscript to obtain the target combined lecture manuscript.

Optionally, after determining the target editing content, the target editing content and the original lecture draft are combined, so as to obtain a target combined lecture draft, and the user may review the target combined lecture draft on the cloud server or edit the target combined lecture next time.

Optionally, after the target combined lecture is determined, adaptive adjustment can be performed 1, and according to preset target audience characteristics (such as a cultural background, education level, interest preference and the like) and expected lecture scenes (formal or informal, on-line or off-line), the model can perform adaptive adjustment on the combined lecture, so that the content of the combined lecture is ensured to have universality and the heart chord of a target audience can be accurately touched; 2. iterative feedback and optimization: after generating the preliminary target combined lecture, the model may further optimize the quality of the lecture by evaluating through simulation or accepting user feedback. For example, using natural language processing technology to analyze emotion colors, intonation and influence of text, then fine-tuning each spoken dimension score weight according to the evaluation result, and iterating until reaching the optimal state; 3. output diversity and personalization options: in addition to providing an optimal target combined lecture, the model may also generate multiple alternative versions based on different combining strategies and optimization schemes for the user to select the most appropriate lecture version according to actual needs and personal style preferences.

Meanwhile, the dimension comparison of the talents among different versions can be intuitively displayed in the display page, so that a user can conveniently and quickly understand and adopt the suggestion of the system, and meanwhile, the manual intervention decision making process of the user is supported.

The method of the embodiment of the application can also utilize advanced AI voice synthesis technology to convert the words edited by the user into realistic voice output immediately, simulate the sound characteristics of different lecturers, including tone, rhythm, pause, emotion colors and the like, and enable each object to hear the preview effect similar to the real lecture, thereby adjusting the text content according to the actual hearing sense and ensuring that the lecture manuscript achieves ideal states in terms of language, rhythm and emotion rendering. In addition, the preview function supports personalized settings, such as selecting different sound models to adapt to the preference of specific occasions or audience groups, so that each object of the team can predict the actual performance effect of the lecture in the editing stage.

In the system provided by the embodiment of the application, in a multi-person cooperation mode, an on-line guidance and peer review mechanism of a talent expert is introduced, the expert can directly provide the expertise through a platform, and team members are guided to pay attention to talent points possibly ignored in a lecture; meanwhile, mutual evaluation activities can be developed in the team, and the lecture contents are improved and perfected jointly by means of respective professional viewing angles and experiences. Meanwhile, the system has the functions of real-time discussion and annotation, team members can leave messages and communicate on specific paragraphs of the lecture, deep discussion and efficient collaboration among teams are promoted, and the overall expression effect of the lecture is further optimized.

The system of the embodiment of the application integrates a rich scene library and audience models, and a user can select a corresponding simulation environment according to actual lecture occasions (such as business meetings, academic discussions, public lectures and the like) and expected audience characteristics (such as ages, industry backgrounds, interests and the like). In addition, when editing lectures, a user can switch different situation modes at any time to view the presentation effect of the lecture content under the corresponding scene, which is helpful for the user to better grasp the pertinence and the effectiveness of the lecture, and adjust the lectures in time to meet the requirements of different occasions and audiences, thereby showing stronger adaptability and expressive force in the actual lecture. Specifically, a talent database is established, various successful speech cases are collected and analyzed to form a talent dimension evaluation model, and the talent dimension evaluation model is applied to real-time auxiliary suggestions in the editing process; training a voice generation model by adopting a deep learning algorithm, so that the voice generation model can simulate a real speech scene more accurately, and an immersive preview experience is provided for a user; the cloud service and the big data analysis capability are integrated, real-time collaborative editing of large-scale users is realized, and the best-fit collaboration partner is matched according to the historical behaviors and the talent characteristics of the users.

Optionally, the system of the application has a development customized collaborative editor, an intelligent analysis and optimization function is embedded, online synchronous editing of multiple persons is supported, real-time scoring and suggestion based on talent dimension are displayed, an AI voice synthesis platform interface is integrated, a real-time voice preview function is realized, voice model parameters are continuously optimized according to user feedback, talent experts and user communities are constructed, sharing communication and mutual evaluation are encouraged, and co-creation sharing of high-quality lectures is promoted. Specifically, the application has 1. Real-time collaborative editing technique: real-time data synchronization is realized through WebSocket or other two-way communication technologies, and when a user edits a lecture draft, the lecture draft is immediately uploaded to the cloud and is pushed to interfaces of other collaborators in real time; 2. version control algorithm (e.g., git) integration: the method comprises the steps that a Git merging strategy is applied to a non-code text environment, each modification of each user is tracked, conflicts can be effectively resolved and data consistency is kept even when multiple users edit the same part of content at the same time, generated lectures of each version have corresponding IDs, all editing operations are based on the IDs for version tracking and synchronization, and when the condition that multiple users edit the same part of content occurs, a system automatically processes conflicts by using a version control algorithm or prompts the users to manually resolve the conflicts; 3. rich text editor and collaborative tagging system: providing visual user experience by using a rich text editor supporting multi-person collaboration, and displaying the editing state and operation record of each collaborator; designing a set of collaborative marking systems, such as color coding or highlighting, to make the user clear which parts are being edited by others or have been edited by others; 4. rights management mechanism: the method realizes the fine authority management, ensures that different roles (such as a master, a contributor and a reviewer) can edit, comment, view and the like according to preset authorities, and 5. Online discussion and feedback tools: integrating an instant messaging function, allowing team members to develop discussions about a specific paragraph or the whole manuscript, and providing annotation and suggestion functions directly on the lecture; 6. intelligent prompting and auxiliary writing functions: and by combining with an artificial intelligence technology, language optimization suggestions, such as guidance on sentence fluency, congratulation techniques, emotion expression and the like, are provided according to the characteristics and requirements of the talent expression industry.

The system of the embodiment of the application is a decentralised version control system, fully utilizes the characteristics of a blockchain technology, records the modification operation of each user as a transaction on the blockchain to form a transparent and non-tamperable version chain, and each block not only contains modification content, but also comprises metadata such as operation time stamp, operator information and the like, so that all editing histories of the whole lecture can be completely traced and audited. In practical application, each node participating in editing (such as a team member's device or server) stores a complete copy of the lecture version information, thereby realizing distributed storage. For example, each time a user editing operation is converted into a transaction, and the transaction is recorded on the blockchain to form a block containing a timestamp, an editor identifier and a content digest; the Merkle tree structure is adopted to store and verify the history modification of each part of the lecture manuscript, so that the data integrity and consistency are ensured; and maintaining a local lecture version library on each node participating in editing, and synchronizing all version information in real time through a P2P network.

Meanwhile, the system adopts an efficient synchronization mechanism, and the version states on all relevant nodes are updated in real time while the user edits. When multiple users edit the same paragraph at the same time, the blockchain-based consensus algorithm can automatically detect concurrent modifications and identify potential conflict areas. And simultaneously has a concurrency control mechanism: to ensure data consistency and integrity, the module employs advanced concurrency control strategies, such as optimistic or pessimistic locking schemes based on lock mechanisms, and higher-level lock-free concurrency control algorithms. The mechanisms can accurately lock the text paragraphs being edited, and prevent other users from modifying the same content in the same time period, so that the problems of dirty reading (reading of uncommitted updates) and update loss are effectively avoided. Once a conflict is detected, the system does not wait for all editing to finish before starting the conflict resolution flow, but immediately triggers the conflict resolution mechanism, so that the real-time collaboration experience of the system is greatly improved, team members can know and resolve the conflict at the first time, and the progress of a project is prevented from being delayed. For example, the system implements fine-grained locking at the text paragraph level at the editor level based on lock-free concurrency algorithms such as distributed lock or CAS (computer-and-Swap). When a user begins to edit an area, the system will acquire the lock for that area, and other users attempting to edit the same area at the same time will not be able to acquire the lock, thereby avoiding dirty reading and update loss problems. Meanwhile, the system utilizes an optimistic lock strategy to allow parallel editing in a short time, but can detect whether conflict occurs at the time of submission and immediately start a conflict resolution flow.

Meanwhile, the intelligent conflict analysis engine analyzes the changes of semantic structures, emotion colors, congratulation methods and the like of the conflict text by utilizing an NLP technology and a deep learning model, and quantifies the influence of the intelligent conflict analysis engine on the dimension of a lecture statement; establishing a set of conflict resolution strategies which are specially optimized for lectures, for example, evaluating the quality of conflict versions according to the mouth indexes such as logical continuity, language vividness, infectivity and the like, and preferentially selecting revision schemes with higher scores; a visual interface is provided to help the user understand and resolve the conflict, and the user can also manually intervene in the conflict resolution result to customize the merged content.

The system of the embodiment of the application further comprises:

1. And the talent sensitive load monitoring and predicting subsystem: the method comprises the steps of monitoring the use condition of each resource of a server cluster in real time, and carrying out refined analysis on related services related to a talent element, such as tasks with higher demands on CPU and memory, such as voice synthesis, NLP processing and the like; and (3) deeply mining user behavior data by using a machine learning model, and accurately predicting future system resource requirements by combining key time nodes (such as large-scale editing activities when the lecture date is close) and specific talent dimension indexes (such as emotion rendering and logic optimization requirement increase) in the lecture creation process.

2. The spoken priority dynamic load balancing subsystem: according to the real-time load and the characteristic of each talent dimension work load, server resources are dynamically allocated, so that high-priority talent related calculation tasks (such as complex emotion analysis and paraphrase suggestion generation) are ensured to be supported by sufficient calculation power, and high-quality lecture draft is ensured to be produced; in the process of collaborative editing by multiple persons, the problem of resolving the conflict of the mouth possibly generated by concurrent editing on the same paragraph is solved by timely scheduling additional computing resources. For example, the use condition of each resource of the server cluster is monitored in real time by adopting technologies such as deep learning, time sequence analysis and the like, and service loads related to the dimension of the talents, such as a speech synthesis engine, NLP processing tasks and the like, are particularly concerned.

3. Talent optimization edge computation framework: the talent optimization function, which depends in part on immediate feedback, is deployed to edge nodes, such as based on user input facts. The voice simulation effect of the time preview obviously reduces network delay through edge calculation, and improves the real-time adjustment experience of the user on the presentation of the lecture script; the edge node customizes and caches common materials and model data according to the user group characteristics and the target lecture scene, reduces the pressure of the central server, and improves the response speed of personalized services aiming at different talent styles and audience characteristics. For example, the edge computing technology is utilized to sink part of the mouth optimizing functions which are computationally intensive and have high real-time requirements to edge nodes to execute, such as real-time voice preview, local mouth score and the like, so that network delay is reduced, and user experience is improved. And a refined caching strategy is implemented on the edge node, and is used for caching common materials, model data and lecture fragments accessed in recent high frequency, so that the pressure of a central server is reduced, and the response speed of the related talent optimization service is increased.

4. Resource elastic telescoping mechanism driven by mouth: the system automatically increases and decreases cloud server instances by adopting a self-adaptive algorithm according to the demand fluctuation of real-time talent optimizing service, and particularly ensures that the service quality of a core talent optimizing function is not influenced in a dense editing stage of large-scale speech projects or in a user access peak caused by sudden hot events.

5. Talent dimension service quality guarantee strategy: setting SLA (service level agreement) based on talent dimension, providing higher priority resource guarantee for core functions (such as version control, conflict resolution, content optimization engine and the like) directly influencing speech effect, and ensuring that the requirements of users on high-quality lecture creation can be met under any condition.

The system of the embodiment of the application comprises the following steps: continuously collecting and analyzing various resource usage data of the system, predicting possible resource bottleneck points and demand peaks in the future through an intelligent algorithm, and pre-allocating resources according to prediction results, such as adding a temporary server instance or reserving computing capacity for optimizing service for a specific talent; 2. resource scheduling and load balancing: adding new talent optimizing tasks into a queue according to a priority order, and distributing the tasks to proper nodes for execution through a load balancing algorithm according to the current load condition of each server node; 3. edge computing and caching application stage: when a user initiates the requirements of real-time voice preview and the like, pushing a task to an edge node for execution, and performing cache management on frequently accessed data by utilizing the local cache acceleration response speed, so that the data reading pressure of a central server is reduced; 4. elastic expansion and automatic operation and maintenance stage: according to the current workload and the predicted future trend, cloud server examples are automatically increased and decreased, dynamic expansion or contraction of resources is achieved, service performance is regularly estimated and optimized, operations such as version upgrading and fault recovery are conducted through operation and maintenance automation tools, and stable operation of the system is maintained.

The method provided by the embodiment of the application has at least the following effects:

1) Allowing a plurality of objects (users and lectures) to cooperatively edit lectures in real time, realizing seamless experience of Distributed service cooperation through Distributed Lock service, solving the problem of difficult traditional decentralized editing and cooperation, and realizing efficient cooperation editing; and in collaboration, the system can monitor and evaluate the quality of lectures in real time, and provide immediate feedback and improvement suggestions for users, including but not limited to, from the core index expressed by the talents: evaluating language fluency, such as sentence length, complexity and transitional word use condition; checking information definition such as punctuation definition, supporting strength of arguments and conclusion induction capability; analyzing emotion resonance degree, such as positive and negative emotion distribution, story telling skills, emotion guiding strategies and the like; in addition, the system can provide personalized guidance, such as recommending expression modes and pedigree skills suitable for the styles of different types of lecturers (such as inward type and outward type), so that the system helps the lecturers to maintain personal characteristics and simultaneously effectively improve the overall quality and final field expression effect of the lecture. By the method, the capacity of the cloud collaborative lecture making and managing system is optimized in all directions from the dimension of the talents;

2) By means of the intelligent lecture content optimizing engine, the optimizing advice of the talents can be automatically provided according to the dimension index of the talents, so that the final manuscript is ensured to reach the optimal level in the dimension of the talents such as logicality, language mobility, emotion transmission force and the like, the problem of uneven expression quality of the talents is solved, and the dimension optimization of the talents is realized;

3) The editing conflict can be automatically detected and solved, the merged manuscript is ensured to be kept high in quality on each talent dimension based on an intelligent merging strategy of the talent dimension, the problem of low merging efficiency in the traditional method is solved, and efficient version management and merging are realized;

4) In conflict resolution and conflict determination strategies, natural Language Processing (NLP) technology and a spoken language expression theory are fused, the text content of the lecture is deeply interpreted through a deep learning model, the system can recognize literal text differences, and further semantic changes and correction and manipulation adjustment can be captured according to context information. For example, in the face of a user adding or modifying metaphors, ranks, etc. in a lecture, the system can analyze whether these changes enhance the emotional expressive power of the lecture or improve the logic of information delivery. Meanwhile, aiming at the important talent dimension of emotion intensity, the system can evaluate the change of emotion color in each revised version by combining emotion analysis technology, and consider the influence of emotion consistency and audience resonance of the whole lecture, when version conflict is solved, the system can keep the modification which is helpful for keeping the consistency of the whole emotion venation of the lecture as much as possible, a machine learning algorithm and a customized talent evaluation model are adopted to quantitatively evaluate the value of each revised version, and the revisions which can enhance the logic tightness of the lecture, improve the infectivity, adapt to the requirements of specific occasions and the characteristics of target audience are preferentially adopted, so that the original emotion atmosphere and expression effect are prevented from being damaged due to simple text replacement, for example, if one user's modification leads the lecture to be more suitable for the professional strict style of business lecture, and the other user's modification strengthens entertainment and interactivity, then the system can automatically select an optimal integration scheme according to the actual lecture scene and purpose, and generate the merging result which is most beneficial to the lecture; according to the comprehensive scores of lectures of different versions, the conflict of which parts can be determined, and further conflict analysis and correction are needed;

5) The micro-service architecture and the container technology are used, so that each service related to the talent optimization can be independently expanded and dynamically scheduled, and computing, storage and network resources are flexibly allocated according to actual traffic and talent dimension requirements. For example, the containerization technology such as Docker is adopted, and each service related to the talent optimization is modularized into an independent micro-service, so that flexible scheduling and rapid capacity expansion and contraction of resources are realized; the user can optimize the service workload to automatically stretch and retract resources according to actual ports by using a container arrangement tool such as Kubernetes, so that the maximum utilization rate of the resources is ensured, and the service quality is maintained at the same time;

6) And the special API interface is developed and integrated in cooperation with the cloud service provider, so that the function of dynamically adjusting the resource quota according to the requirement of the talent optimization is realized. For example, the API interface provided by the cloud computing platform is integrated, so that automation of operations such as adjusting the number of cloud server instances, configuring resources and the like in real time is realized, and the change of peak resource requirements is effectively treated. The CI/CD flow is matched, the service performance is continuously monitored and optimized, and the resource scheduling strategy is ensured to be continuously and iteratively updated along with the service development and the technical progress;

7) Designing a resource request queue based on the talent dimension priority, preferentially meeting service requests which directly affect key talent elements such as improving the logic property, the infectivity and the like of speech contents, combining a load balancing algorithm (such as polling, minimum connection number, weight distribution and the like) in a distributed system, dynamically distributing tasks to proper server nodes according to actual resource requirements and node performances, and ensuring that a talent optimization function with high priority is supported by sufficient computing resources.

Referring to fig. 2, a block diagram illustrating a cloud-based lecture management apparatus according to an embodiment of the present application may include:

the display module is used for generating an optimization suggestion according to the overall quality score and the talent dimension score and displaying the optimization suggestion on a display page of the cloud server;

And the generation module is used for determining the comprehensive score of the revised lecture and generating the revised suggestion according to the comprehensive score.

In one embodiment, the generating module is further configured to:

when the number of the objects is a plurality of, determining a plurality of corresponding revisions of the lecture script with different versions;

Determining a logical dimension merging score according to the first preset weight parameter, the first score and the second score, determining a language mobility dimension merging score according to the second preset weight parameter, the third score and the fourth score, and determining an emotion conveying force dimension merging score according to the third preset weight parameter and the fifth score;

In one embodiment, the generating module is further configured to:

When the number of the objects is a plurality of and the editing contents of different objects have conflicts, determining the score difference of the different editing contents in at least one talent dimension;

and combining the target editing content with the original lecture manuscript to obtain the target combined lecture manuscript.

The functions of each module in each device of the embodiments of the present application may be referred to the corresponding descriptions in the above methods, and are not described herein again.

Referring to fig. 3, a block diagram of an electronic device according to an embodiment of the present application is shown, the electronic device including: the cloud-based lecture management method in the above embodiment is implemented by the processor 320 and the memory 310, and instructions executable on the processor 320 are stored in the memory 310. Wherein the number of memory 310 and processors 320 may be one or more.

In one embodiment, the electronic device further includes a communication interface 330 for communicating with an external device for data interactive transmission. If the memory 310, the processor 320 and the communication interface 330 are implemented independently, the memory 310, the processor 320 and the communication interface 330 may be connected to each other and communicate with each other through buses. The bus may be an industry standard architecture (Industry Standard Architecture, ISA) bus, an external device interconnect (PERIPHERAL COMPONENT INTERCONNECT, PCI) bus, or an extended industry standard architecture (Extended Industry Standard Architecture, EISA) bus, among others. The bus may be classified as an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown in fig. 3, but not only one bus or one type of bus.

Alternatively, in a specific implementation, if the memory 310, the processor 320, and the communication interface 330 are integrated on a chip, the memory 310, the processor 320, and the communication interface 330 may communicate with each other through internal interfaces.

An embodiment of the present application provides a computer-readable storage medium storing a computer program which, when executed by a processor, implements the cloud-based lecture management method provided in the above embodiment.

The embodiment of the application also provides a chip, which comprises a processor and is used for calling the instructions stored in the memory from the memory and running the instructions stored in the memory, so that the communication equipment provided with the chip executes the method provided by the embodiment of the application.

The embodiment of the application also provides a chip, which comprises: the input interface, the output interface, the processor and the memory are connected through an internal connection path, the processor is used for executing codes in the memory, and when the codes are executed, the processor is used for executing the method provided by the application embodiment.

It should be appreciated that the processor may be a central processing unit (Central Processing Unit, CPU), other general purpose processor, digital signal processor (DIGITAL SIGNAL processing, DSP), application Specific Integrated Circuit (ASIC), field programmable gate array (fieldprogrammablegate array, FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, etc. A general purpose processor may be a microprocessor or any conventional processor or the like. It is noted that the processor may be a processor supporting an advanced reduced instruction set machine (ADVANCED RISC MACHINES, ARM) architecture.

Further, optionally, the memory may include a read-only memory and a random access memory, and may further include a nonvolatile random access memory. The memory may be volatile memory or nonvolatile memory, or may include both volatile and nonvolatile memory. The nonvolatile memory may include a read-only memory (ROM), a Programmable ROM (PROM), an erasable programmable ROM (erasable PROM), an electrically erasable programmable EPROM (EEPROM), or a flash memory, among others. Volatile memory can include random access memory (random access memory, RAM), which acts as external cache memory. By way of example, and not limitation, many forms of RAM are available. For example, static random access memory (STATIC RAM, SRAM), dynamic random access memory (dynamic random access memory, DRAM), synchronous Dynamic Random Access Memory (SDRAM), double data rate synchronous dynamic random access memory (double DATA DATE SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (ENHANCED SDRAM, ESDRAM), synchronous link dynamic random access memory (SYNCHLINK DRAM, SLDRAM), and direct memory bus random access memory (direct rambus RAM, DR RAM).

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions in accordance with the present application are fully or partially produced. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable devices. Computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another.

In the description of the present specification, a description referring to the terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

Furthermore, the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present application, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.

Any process or method description in a flowchart or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process. And the scope of the preferred embodiments of the present application includes additional implementations in which functions may be performed in a substantially simultaneous manner or in an opposite order from that shown or discussed, including in accordance with the functions that are involved.

Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions.

It is to be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the methods of the embodiments described above may be performed by a program that, when executed, comprises one or a combination of the steps of the method embodiments, instructs the associated hardware to perform the method.

In addition, each functional unit in the embodiments of the present application may be integrated in one processing module, or each unit may exist alone physically, or two or more units may be integrated in one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules described above, if implemented in the form of software functional modules and sold or used as a stand-alone product, may also be stored in a computer-readable storage medium. The storage medium may be a read-only memory, a magnetic or optical disk, or the like.

The foregoing is merely illustrative of the present application, and the present application is not limited thereto, and any person skilled in the art will readily recognize that various changes and substitutions are possible within the scope of the present application. Therefore, the protection scope of the application is subject to the protection scope of the claims.

Claims

1. The lecture manuscript management method based on the cloud is characterized by comprising the following steps of:

determining the comprehensive score of the revised lecture, and generating a revised suggestion according to the comprehensive score;

the step of performing at least one spoken dimension of the original lecture, and the step of determining a spoken dimension score of the spoken dimension includes:

Calculating the score of each emotion guiding point in the original lecture, and determining an eighth talent dimension score of emotion guiding dimension according to the score of the emotion guiding point and the number of the emotion guiding points;

the determining the overall quality score of the original lecture according to the talent dimension score, the dimension weight of the talent dimension and the scene factor parameter comprises:

According to the first talent dimension score, the second talent dimension score, the third talent dimension score, the fourth talent dimension score, the fifth talent dimension score, the sixth talent dimension score, the seventh talent dimension score and the eighth talent dimension score, and according to the dimension weight and the dimension weight coefficient of the talent dimension, carrying out weighted calculation to obtain a talent dimension quality score;

2. The cloud-based lecture management method of claim 1, wherein: the determining the composite score of the revised lecture includes:

3. The cloud-based lecture management method of claim 2, characterized in that: the generating correction advice according to the composite score comprises:

Or alternatively

4. The cloud-based lecture management method according to any one of claims 1 to 3, characterized in that: the method further comprises the steps of:

5. The cloud-based lecture management method of claim 1, wherein: the method further comprises the steps of:

6. A lecture management apparatus based on a cloud, comprising:

The generation module is used for determining the comprehensive score of the revised lecture and generating a revised suggestion according to the comprehensive score;

7. An electronic device, comprising: a processor and a memory in which instructions are stored, the instructions being loaded and executed by the processor to implement the method of any one of claims 1 to 5.

8. A computer readable storage medium having stored therein a computer program which when executed implements the method of any of claims 1-5.