Multi-person document annotation management method and device, equipment and storage medium

Abstract

The invention provides an annotation management method and device for a multi-person document, equipment and a storage medium. The annotation management method comprises the steps of obtaining a to-be-processed document; performing text classification on the annotation information to obtain annotation classification labels; generating a to-be-handled annotation task according to the annotation classification label; and then, according to the annotation classification label, distributing the annotation task to be handled. According to the method, the annotation information is subjected to intelligent classification, priority evaluation and semantic understanding, the annotation task to be handled is automatically generated, and accurate allocation and processing state synchronization are realized in combination with task content and personnel responsibilities, so that the systematization, high efficiency and traceability of annotation processing in a multi-person collaboration document are effectively improved, and the annotation processing efficiency is improved. The problems of dispersed annotations, fuzzy task allocation, response lag and the like in the prior art are solved, so that the collaborative efficiency and management quality of document review are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of document processing, and in particular to a method, device, equipment and storage medium for managing annotations of documents of multiple persons. Background Art

[0002] In existing multi-person collaborative document systems, annotation functionality is typically limited to adding simple text comments to documents for others to review and respond to manually. This approach lacks structured information representation and intelligent task management. In practical applications, especially in document review scenarios involving multiple roles, the large number and diversity of annotations often lead to fragmented content, difficulty in aggregating them, inability to identify annotation types and priorities, and unclear responsibility for annotations.

[0003] In summary, the problems existing in the prior art need to be solved urgently. Summary of the Invention

[0004] The present invention provides a method, device, equipment and storage medium for multi-person document annotation management, which are used to solve the defects in the existing technology, generate to-do tasks with priority and processing status from classified annotations, and improve processing efficiency and execution controllability.

[0005] The present invention provides a method for managing annotations of documents by multiple people, comprising: Acquire a document to be processed, wherein the document to be processed includes at least one annotation information created by a user; Performing text classification on the annotation information to obtain annotation classification labels; Generate to-do annotation tasks according to the annotation classification labels; The to-do annotation tasks are allocated according to the annotation classification labels.

[0006] According to a method for managing annotations of documents for multiple persons provided by the present invention, the step of performing text classification on the annotation information to obtain annotation classification labels specifically includes: Inputting the annotation information into a text classification model to obtain the predicted probability of each text category; When there is a text type whose prediction probability is greater than the classification confidence threshold, the annotation information is classified into this category.

[0007] According to the method for managing annotations of multi-person documents provided by the present invention, after the step of inputting the annotation information into the text classification model to obtain the predicted probability of each text category, When there is no text type with a prediction probability greater than the classification confidence threshold, matching the annotation information with the keyword list of each text category; When the number of matched keywords is greater than a preset matching threshold, the annotation information is classified into the category.

[0008] According to a method for managing annotations of documents for multiple persons provided by the present invention, after the step of matching the annotation information with a keyword list of each text category when there is no text type with a prediction probability greater than a classification confidence threshold, the method further comprises: When the number of matched keywords is less than or equal to a preset matching threshold, a manual annotation window is generated. The manual annotation window is used to obtain annotation classification information input by the user, and the annotation classification information is used to classify the annotation information.

[0009] According to a method for managing annotations of documents for multiple people provided by the present invention, the step of assigning the pending annotation tasks according to the annotation classification labels specifically includes: Determine the corresponding processor type according to the annotation classification label; Query the list of handlers corresponding to the handler type; Determine the target handler based on the current task load.

[0010] According to a method for managing annotations for multiple documents provided by the present invention, the pending annotation tasks include priority information, and the step of generating the pending annotation tasks based on the annotation classification labels specifically includes: Performing feature extraction on the annotation information to obtain annotation semantic features; Inputting the annotation classification label and the annotation semantic feature into a priority evaluation model to obtain priority information; The to-do annotation task is generated according to the priority information, the annotation classification label and the annotation information.

[0011] According to a method for managing annotations of documents for multiple persons provided by the present invention, after the step of assigning the pending annotation tasks according to the annotation classification labels, the method further comprises: Acquire annotation task processing information, wherein the annotation task processing information includes annotation task status, processing time and processor; The annotation database is updated according to the annotation task processing information and pushed to relevant users.

[0012] The present invention also provides a multi-person document annotation management device, comprising: A document acquisition module, configured to acquire a document to be processed, wherein the document to be processed includes at least one annotation information created by a user; An annotation and marking module, used to classify the annotation information into text and obtain annotation classification labels; A task generation module, used to generate to-do annotation tasks according to the annotation classification labels; The task allocation module is used to allocate the to-be-done annotation tasks according to the annotation classification labels.

[0013] The present invention also provides an electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for managing annotations of multi-person documents as described above is implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned methods for managing annotations of multi-person documents.

[0015] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned methods for managing annotations of multi-person documents.

[0016] The annotation management method, device, equipment and storage medium for multi-person documents provided by the present invention obtains the document to be processed; and performs text classification on the annotation information to obtain annotation classification labels; then generates pending annotation tasks based on the annotation classification labels; and then assigns the pending annotation tasks based on the annotation classification labels. The present invention automatically generates pending annotation tasks by intelligently classifying, prioritizing and semantically understanding the annotation information, and achieves precise assignment and synchronization of processing status based on the task content and personnel responsibilities, effectively improving the systematization, efficiency and traceability of annotation processing in multi-person collaborative documents, and solving the problems of scattered annotations, ambiguous task assignment, and delayed response in the prior art, thereby significantly improving the collaborative efficiency and management quality of document review. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 It is a flowchart of the annotation management method for multiple documents provided by the present invention; Figure 2 It is a structural diagram of the device for managing annotations of documents for multiple persons provided by the present invention; Figure 3 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0019] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0020] In order to solve the problems in the prior art, the present invention proposes a method for managing annotations of multi-person documents, so as to realize the generation of to-do tasks with priority and processing status from the classified annotations, thereby improving processing efficiency and execution controllability. The following describes the method for managing annotations of multi-person documents. Figure 1 As shown, including but not limited to the following steps: Step 110: Obtain a document to be processed, where the document to be processed includes at least one annotation information created by a user.

[0021] In this step, the target document for annotation management can be obtained by connecting to the document collaboration platform or office system interface. This document contains annotations generated by one or more users during the editing or review process. Annotations can include the comment text, the time of the annotation, the annotator's ID, the location of the associated paragraph, and other content.

[0022] Step 120: Perform text classification on the annotation information to obtain annotation classification labels.

[0023] In this step, the text content of each annotation is semantically analyzed and automatically categorized using a pre-trained text classification model. The classification model used can be a deep learning-based neural network model, such as BERT or its variants. This model can identify the category to which the annotation content belongs, such as "formatting issue," "logic issue," "content supplement suggestion," or "unclear semantics." The corresponding annotation classification label is generated based on the model's output.

[0024] Step 130: Generate to-do annotation tasks based on the annotation classification labels.

[0025] In this step, annotations are converted into structured to-do tasks based on their classification tags. Each task includes at least: task content (i.e., annotation text), classification tag, associated documents and location, task status (initially "pending"), and priority information. Priority information is automatically assessed by combining factors such as the annotation's classification type, the annotation's vocabulary (e.g., "must" or "urgent"), and context.

[0026] Step 140: Allocate the pending annotation tasks according to the annotation classification labels.

[0027] In this step, the corresponding handler type is determined by mapping the classification labels to organizational roles. For example, "formatting issues" can be assigned to a typesetter, while "content suggestions" can be assigned to a copywriter. The system then searches the list of available handlers based on the handler type and, taking into account factors such as the current workload and urgency of the task, determines the final target handler and completes the task assignment. Once the task is assigned, a task notification is sent to the handler and the task status is updated to "Assigned."

[0028] By executing the above steps, the automatic classification, task-based processing and intelligent allocation of document annotations are achieved, thereby improving the annotation processing efficiency and document management level in the multi-person collaborative review process.

[0029] As a further optional embodiment, the step of performing text classification on the annotation information to obtain annotation classification labels specifically includes: Inputting the annotation information into a text classification model to obtain the predicted probability of each text category; When there is a text type whose prediction probability is greater than the classification confidence threshold, the annotation information is classified into this category.

[0030] First, the text content of the annotation information is input into a pre-trained text classification model. This text classification model can be constructed using supervised learning, with the training samples being historical annotation data labeled with classification labels. The model architecture can utilize natural language processing models such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), and bidirectional encoder representations (such as BERT). This model can output the predicted probabilities of the annotation information across multiple predefined text categories.

[0031] Next, we determine whether a prediction probability exceeds a preset classification confidence threshold (e.g., 0.7). If so, we classify the annotation as belonging to the text type with the highest probability value exceeding the threshold and assign the corresponding classification label. This classification label will serve as the basis for subsequent processing, such as task generation and assignment.

[0032] For example, if the probability that an annotation is predicted by the model to be a "content supplement suggestion" is 0.81, and the probability of this category exceeds the set threshold of 0.7, the system will classify the annotation as a "content supplement suggestion".

[0033] This method achieves high-accuracy automatic classification of annotation information, reduces the burden of manual labeling, and improves the intelligence level of annotation management.

[0034] As a further optional embodiment, after the step of inputting the annotation information into the text classification model to obtain the predicted probability of each text category, When there is no text type with a prediction probability greater than the classification confidence threshold, matching the annotation information with the keyword list of each text category; When the number of matched keywords is greater than a preset matching threshold, the annotation information is classified into the category.

[0035] In this embodiment, after the step of inputting the annotation information into the text classification model to obtain the predicted probability of each text category, if there is no category whose predicted probability is greater than the preset classification confidence threshold, a backup classification mechanism based on keyword matching is enabled to enhance the fault tolerance and applicability of the model classification results.

[0036] Specifically, a keyword list is pre-maintained for each text category. This keyword list can be compiled based on typical words in historical annotation data and can be supplemented and revised using manual rules. Then, the text content in the annotation information is matched against the keyword list corresponding to each text category, and the number of matching keywords is counted.

[0037] If the number of keywords matching a certain text category exceeds the preset matching threshold (e.g., 3), the annotation information will be classified into that text category and assigned a corresponding classification label. This label will be used in the subsequent task generation and assignment process.

[0038] For example, if an annotation is not hit by the model's confidence prediction, but contains multiple keywords such as "perfect", "details", "explanation", and these keywords belong to the keyword library of "supplementary explanation category", then the system can classify the annotation as "supplementary explanation category".

[0039] This supplementary mechanism improves the robustness of the text classification module, ensuring that classification labels can still be reasonably determined under model uncertainty and guaranteeing the consistency of the system's overall automated processing flow.

[0040] As a further optional embodiment, after the step of matching the annotation information with the keyword lists of each text category when there is no text type with a prediction probability greater than the classification confidence threshold, the method further includes: When the number of matched keywords is less than or equal to a preset matching threshold, a manual annotation window is generated. The manual annotation window is used to obtain annotation classification information input by the user, and the annotation classification information is used to classify the annotation information.

[0041] In this embodiment, if the number of matched keywords is less than or equal to a preset matching threshold, the system determines that the annotation information cannot be accurately classified in an automatic manner.

[0042] To avoid misclassification and omissions, a manual annotation window will be generated on the interface to guide users in manually classifying the current annotation. Specifically, users can select the category that best matches the semantics of the annotation from several preset classification options provided by the system, or enter a custom label if necessary.

[0043] After receiving the user's confirmed annotation classification information, the classification label of the annotation is updated and the classification result is used in the subsequent task generation and assignment process. At the same time, this manually annotated information can be used as one of the data samples for subsequent training or fine-tuning of the text classification model to continuously optimize the model's classification performance.

[0044] This implementation ensures that even in extreme cases where the automated mechanism cannot reliably determine the classification, the system can still achieve complete closed-loop processing through human-computer collaboration, enhancing the system's generalization and adaptability.

[0045] As a further optional embodiment, the step of allocating the to-do annotation tasks according to the annotation classification labels specifically includes: Determine the corresponding processor type according to the annotation classification label; Query the list of handlers corresponding to the handler type; Determine the target handler based on the current task load.

[0046] In this embodiment, the corresponding handler type is first determined based on the annotation classification label. For example, if a comment is classified as "Legal Compliance," the corresponding handler type may be "Legal Affairs Personnel"; if the comment is classified as "Technical Implementation," the handler type may be "R&D Engineer."

[0047] Next, a list of processors corresponding to the processor type is queried. The list can be dynamically generated from a user role database, project collaborator information, or an organizational structure.

[0048] The system then compares each handler's current task load (e.g., number of uncompleted tasks, average processing time, current status, and other indicators) and, while maintaining a balanced task load, determines the target handler. Preferably, a weighted scheduling algorithm or a minimum load priority strategy can be employed to improve the rationality of task allocation and response efficiency.

[0049] Through the above steps, the system implements an intelligent task assignment mechanism based on classification labels, ensuring that annotation tasks are assigned to the most appropriate personnel, thereby improving the efficiency and collaborative quality of annotation processing.

[0050] As a further optional embodiment, the to-do annotation task includes priority information, and the step of generating the to-do annotation task according to the annotation classification label specifically includes: Performing feature extraction on the annotation information to obtain annotation semantic features; Inputting the annotation classification label and the annotation semantic feature into a priority evaluation model to obtain priority information; The to-do annotation task is generated according to the priority information, the annotation classification label and the annotation information.

[0051] In this embodiment, first, feature extraction is performed on the annotation information. The extracted features may include but are not limited to: keywords, syntactic structures, emotional tendencies, context positions, involved chapters, historical processing records, etc. in the annotation, so as to obtain semantic features of the annotation.

[0052] The extracted annotation semantic features and the annotation classification labels are then fed into a priority assessment model. This model can be trained using supervised learning methods, such as XGBoost, neural networks, or Transformer architectures, and outputs annotation priority information, such as "high," "medium," "low," or a specific numerical score.

[0053] Finally, based on the generated priority information, annotation classification labels and original annotation content, to-do annotation tasks are comprehensively generated. The task objects include processing suggestions, responsible person suggestions, expected response time limits, etc., which facilitate subsequent task scheduling and processing.

[0054] Through the above processing method, the system can realize intelligent priority management of annotation tasks, improve the response efficiency of important annotations, and reasonably guide the team's review resources.

[0055] As a further optional embodiment, after the step of allocating the to-do annotation tasks according to the annotation classification labels, the method further includes: Acquire annotation task processing information, wherein the annotation task processing information includes annotation task status, processing time and processor; The annotation database is updated according to the annotation task processing information and pushed to relevant users.

[0056] In this embodiment, after the step of assigning the to-do annotation tasks according to the annotation classification labels, it is also necessary to obtain the processing information of the annotation tasks. The annotation task processing information includes: the annotation task status, such as "not started", "processing", "completed", etc.; the processing time, that is, the start processing time, end processing time or estimated processing time of the task; the processor, that is, the identification of the person responsible for executing the task.

[0057] After receiving the above-mentioned annotation task processing information, the corresponding record in the annotation database is updated in real time to reflect the latest processing progress, status and responsible person information.

[0058] At the same time, the system will push updates on the processing status to relevant users, including the annotation initiator, task handler, and other collaborators, through message push, email notification, and changes in document markup, thereby ensuring the team's transparency, synchronization, and traceability of the annotation processing progress.

[0059] By introducing the above operations, the method can achieve dynamic synchronization of annotation task status and improve team collaboration efficiency.

[0060] As a further optional embodiment, whenever a user adds a new annotation or modifies an existing annotation, the system will automatically record the version history information of the annotation and generate a unique version number for the annotation.

[0061] The version history information includes but is not limited to: the identity of the editor who performed the operation; the modification timestamp; the original content and the modified content of the annotation (ie, content difference information).

[0062] Users can view the historical modification history of each annotation in the interface to understand its evolution. The system also provides a responsibility tracking mechanism to ensure that each annotation can be accurately traced back to the creator and the person who modified it at each stage, forming a complete chain of responsibility.

[0063] This implementation method helps to ensure the transparency, traceability and clear responsibility of document annotations during multi-person collaboration, and avoid information confusion caused by unclear responsibilities or erroneous operations.

[0064] The following describes the annotation management device for multiple documents provided by the present invention. Figure 2 As shown, the apparatus for managing annotations of multi-person documents described below and the method for managing annotations of multi-person documents described above can refer to each other.

[0065] A device for managing annotations of documents of multiple persons, comprising: The document acquisition module 210 is used to acquire a document to be processed, wherein the document to be processed includes at least one annotation information created by a user; Annotation marking module 220, used to classify the annotation information into text and obtain annotation classification labels; A task generation module 230 is used to generate to-do annotation tasks according to the annotation classification labels; The task assignment module 240 is configured to assign the pending annotation tasks according to the annotation classification labels.

[0066] Figure 3 An example of a physical structure diagram of an electronic device is shown below. Figure 3As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330, and a communication bus 340. The processor 310, the communications interface 320, and the memory 330 communicate with each other via the communication bus 340. The processor 310 may call the logic instructions in the memory 330 to execute the annotation management method for multi-person documents, which includes: Acquire a document to be processed, wherein the document to be processed includes at least one annotation information created by a user; Performing text classification on the annotation information to obtain annotation classification labels; Generate to-do annotation tasks according to the annotation classification labels; The to-do annotation tasks are allocated according to the annotation classification labels.

[0067] Furthermore, the logic instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0068] On the other hand, the present invention further provides a computer program product, comprising a computer program, which may be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can perform the method for managing annotations for documents by multiple persons provided by the above methods, the method comprising: Acquire a document to be processed, wherein the document to be processed includes at least one annotation information created by a user; Performing text classification on the annotation information to obtain annotation classification labels; Generate to-do annotation tasks according to the annotation classification labels; The to-do annotation tasks are allocated according to the annotation classification labels.

[0069] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for managing annotations of documents for multiple persons provided by the above methods, the method comprising: Acquire a document to be processed, wherein the document to be processed includes at least one annotation information created by a user; Performing text classification on the annotation information to obtain annotation classification labels; Generate to-do annotation tasks according to the annotation classification labels; The to-do annotation tasks are allocated according to the annotation classification labels.

[0070] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0071] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for managing annotations of documents by multiple people, characterized in that: include: Acquire a document to be processed, wherein the document to be processed includes at least one annotation information created by a user; Performing text classification on the annotation information to obtain annotation classification labels; Generate to-do annotation tasks according to the annotation classification labels; The to-do annotation tasks are allocated according to the annotation classification labels.

2. The method for managing annotations of multi-person documents according to claim 1, characterized in that: The step of performing text classification on the annotation information to obtain annotation classification labels specifically includes: Inputting the annotation information into a text classification model to obtain the predicted probability of each text category; When there is a text type whose prediction probability is greater than the classification confidence threshold, the annotation information is classified into this category.

3. The method for managing annotations of multi-person documents according to claim 2, characterized in that: After the step of inputting the annotation information into the text classification model to obtain the predicted probability of each text category, When there is no text type with a prediction probability greater than the classification confidence threshold, matching the annotation information with the keyword list of each text category; When the number of matched keywords is greater than a preset matching threshold, the annotation information is classified into the category.

4. The method for managing annotations of multi-person documents according to claim 3, characterized in that: After the step of matching the annotation information with the keyword lists of each text category when there is no text type with a prediction probability greater than the classification confidence threshold, the method further includes: When the number of matched keywords is less than or equal to a preset matching threshold, a manual annotation window is generated. The manual annotation window is used to obtain annotation classification information input by the user, and the annotation classification information is used to classify the annotation information.

5. The method for managing annotations of documents by multiple persons according to claim 1, wherein: The step of allocating the to-be-done annotation tasks according to the annotation classification labels specifically includes: Determine the corresponding processor type according to the annotation classification label; Query the list of handlers corresponding to the handler type; Determine the target handler based on the current task load.

6. The method for managing annotations of multi-person documents according to claim 1, characterized in that: The pending annotation tasks include priority information. The step of generating the pending annotation tasks according to the annotation classification labels specifically includes: Performing feature extraction on the annotation information to obtain annotation semantic features; Inputting the annotation classification label and the annotation semantic feature into a priority evaluation model to obtain priority information; The to-do annotation task is generated according to the priority information, the annotation classification label and the annotation information.

7. The method for managing annotations of documents for multiple persons according to claim 1, characterized in that: After the step of allocating the to-do annotation tasks according to the annotation classification labels, the method further includes: Acquire annotation task processing information, wherein the annotation task processing information includes annotation task status, processing time and processor; The annotation database is updated according to the annotation task processing information and pushed to relevant users.

8. A device for managing annotations of documents of multiple persons, characterized in that: include: A document acquisition module, configured to acquire a document to be processed, wherein the document to be processed includes at least one annotation information created by a user; An annotation and marking module, used to classify the annotation information into text and obtain annotation classification labels; A task generation module, used to generate to-do annotation tasks according to the annotation classification labels; The task allocation module is used to allocate the to-be-done annotation tasks according to the annotation classification labels.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method for managing annotations of multi-person documents as claimed in any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for managing annotations of multi-person documents as claimed in any one of claims 1 to 7 is implemented.

Images