Document data processing method and device and related equipment

Abstract

The embodiment of the invention discloses a document data processing method and device and related equipment, and can be applied to the technical field of data processing. The method comprises the steps of displaying a summary interaction page associated with a first business object; in response to a document import operation for the document resource, displaying a first session message which is sent by the first business object and carries the document resource on the summary interaction page; in a first association display area associated with the first session message, traversing and displaying node progress prompt information of each flow processing node in the plurality of flow processing nodes; and when the displayed node progress prompt information is the node progress prompt information of the abstract summarizing node and the node progress prompt information of the abstract summarizing node indicates that the first summarizing abstract is obtained through summarizing, displaying a second session message carrying the first summarizing abstract on a summarizing interaction page. By adopting the embodiment of the invention, the information display mode can be enriched in the process of extracting the summary information of the document resource.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a document data processing method, apparatus, and related equipment. Background Art

[0002] Currently, when obtaining relevant summary information of a document resource, a control for triggering extraction of summary information of a document is usually clicked on a resource display page, thereby displaying the extracted summary information for the document resource displayed on the resource display page.

[0003] However, the inventors discovered in practice that, in the current process of extracting the summary information of document resources and displaying it on the page, information is extracted from the document resources directly displayed on the current resource display page, and then the summary information of the entire document resource currently extracted is directly displayed on the control page of the resource display interface. This means that for the existing summary information display method, it is necessary to display it only when the summary information of the entire document resource is obtained, so that in the process of extracting the summary information of the document resource, there is a phenomenon that the information display method is relatively single. Therefore, how to enrich the information display method in the process of extracting the summary information of the document resource is an urgent problem to be solved. Summary of the Invention

[0004] The embodiments of the present application provide a document data processing method, apparatus, and related equipment, which can display the process processing progress of document resources during the process of extracting summary information of document resources, thereby enabling business objects to perceive the processing progress of document resources in real time, and helping to enrich the information display method during the process of extracting summary information of document resources.

[0005] On the one hand, an embodiment of the present application provides a document data processing method, the method comprising:

[0006] displaying a summary interaction page associated with the first business object;

[0007] In response to a document import operation on a document resource, displaying a first conversation message carrying the document resource sent by the first business object on a summary interaction page;

[0008] In the first associated display area associated with the first conversation message, node progress prompt information of each process processing node in the plurality of process processing nodes associated with the document resource is traversed and displayed; the plurality of process processing nodes include a summary node for summarizing resource fragment contents in the document resource;

[0009] When the node progress prompt information displayed is the node progress prompt information of the summary summary node, and the node progress prompt information of the summary summary node indicates that the first summary summary associated with the resource slice content is obtained through the summary, a second session message carrying the first summary summary is displayed on the summary interaction page.

[0010] The method further includes:

[0011] In response to the message triggering operation, displaying a third conversation message indicated by the message triggering operation on the summary interaction page; the third conversation message includes an interaction instruction initiated for the document resource;

[0012] The fourth conversation message is displayed on the summary interaction page; the fourth conversation message includes an interaction reply to the interaction instruction.

[0013] Wherein, any process processing node among the multiple process processing nodes is a target process processing node;

[0014] The node progress prompt information of each process processing node in multiple process processing nodes associated with the document resource is traversed and displayed, including:

[0015] When the node progress prompt information displayed is the node progress prompt information of the target process processing node, the node progress prompt information of the target process processing node is dynamically displayed; the node progress prompt information of the target process processing node changes as the node processing progress when the target process processing node processes the document resources changes.

[0016] The node progress prompt information of the target process processing node is dynamically displayed, including one or more of the following:

[0017] When the node progress prompt information of the target process processing node includes the time progress prompt information of the target process processing node, dynamically display the time progress prompt information of the target process processing node; or

[0018] When the node progress prompt information of the target process processing node includes the numerical progress prompt information of the target process processing node, the numerical progress prompt information of the target process processing node is dynamically displayed.

[0019] The method further includes:

[0020] In response to a node viewing operation for multiple process processing nodes, process node prompt information corresponding to the multiple process processing nodes is displayed in the first associated display area; one process processing node corresponds to one process node prompt information; the process node prompt information of any process processing node includes node status information for prompting the node processing status of any process processing node.

[0021] The plurality of process processing nodes further include a document upload node for uploading document resources; the summary interaction page includes a document processing status control, which is used to adjust the document processing status of the document resource, the document processing status including the document upload status, which is the document processing status determined when the document resource is processed through the document upload node;

[0022] The method also includes:

[0023] When the node progress prompt information displayed in the first associated display area is the node progress prompt information of the document upload node, the first status prompt information is displayed in the document processing status control; the first status prompt information is used to indicate that the document processing status of the document resource is the document upload status;

[0024] In response to the first state switching operation on the document processing state control, the document processing state of the document resource is adjusted from the document uploading state to the upload canceling state, and a first state prompting conversation message is output to prompt that the document processing state of the document resource is the upload canceling state.

[0025] Among them, the summary interaction page includes a document processing status control, which is used to adjust the document processing status of the document resource; the multiple process processing nodes also include a document content processing node, which includes a summary summary node, a content parsing node for determining the document content of the document resource, a content fragmentation node for determining the resource fragment content corresponding to the document resource, and a fragment summary extraction node for determining the content summary of the resource fragment content; the document processing status includes a document content processing status, which is the document processing status determined when the document resource is processed by the document content processing node;

[0026] The method also includes:

[0027] When the node progress prompt information displayed in the first associated display area is the node progress prompt information of the document content processing node, second status prompt information is displayed in the document processing status control; the second status prompt information is used to indicate that the document processing status of the document resource is the document content processing status;

[0028] In response to the second state switching operation on the document processing state control, the document processing state of the document resource is adjusted from the document content processing state to the content processing cancellation state, and a second state prompting session message is output to prompt that the document processing state is the content processing cancellation state.

[0029] The method further includes:

[0030] When displaying the second session message on the summary interaction page, displaying a summary re-extraction control for instructing to regenerate the summary summary of the document resource;

[0031] In response to a triggering operation on the summary re-extraction control, a second summary regenerated for the document resource is displayed in the second conversation message.

[0032] The method further includes:

[0033] Displaying a summary switching control for a summary of the document resource in a second associated display area associated with the second conversation message;

[0034] In response to a triggering operation on the summary switching control, the first summary and the second summary are switched and displayed in the second conversation message.

[0035] The method further includes:

[0036] In response to a session sharing operation in a summary interaction page, the session message in the summary interaction page is shared to a second business object, so that the second business object triggers the display of a sharing interaction page associated with the second business object through a message sharing link associated with the session sharing operation, and the sharing interaction page includes the session message displayed in the summary interaction page.

[0037] The document resources include a first document resource and a second document resource;

[0038] The method also includes:

[0039] When displaying a conversation message carrying a summary summary of a first document resource and a conversation message carrying a summary summary of a second document resource in a summary interaction page, a conversation message carrying document merge and comparison information between the first document resource and the second document resource is displayed; the document merge and comparison information is determined based on the content summary of the resource fragment content of the first document resource and the content summary of the resource fragment content of the second document resource.

[0040] The document import operation is a combined import operation for the first document resource and the second document resource; or

[0041] The document import operation includes a first import operation for a first document resource and a second import operation for a second document resource.

[0042] The first conversation message is a conversation message that carries both the first document resource and the second document resource; or

[0043] The first session message includes a first sub-session message carrying a first document resource and a second sub-session message carrying a second document resource.

[0044] The method further includes:

[0045] Perform document parsing on the document resource to obtain the document text information corresponding to the document content of the document resource;

[0046] Perform document fragmentation processing on the document text information to obtain the resource fragment content corresponding to the document resource;

[0047] Performing summary extraction on the resource slice content to obtain a content summary of the resource slice content;

[0048] The content summaries of the resource fragments are summarized to obtain a first summary of the document resource.

[0049] The document resource is parsed to obtain the document text information corresponding to the document content of the document resource, including:

[0050] When the document resource includes multimedia data information, performing text extraction processing on the multimedia data information to obtain translated text information associated with the multimedia data information; the multimedia data information includes one or more of the following: image data, video data, and voice data;

[0051] When the initial text information in the document resource is obtained, the document text information corresponding to the document content of the document resource is determined based on the spatial position relationship between the initial text information and the transferred text information in the document resource.

[0052] The document text information is segmented to obtain the resource segment content corresponding to the document resource, including:

[0053] When the number of text segmentations in the document text information is greater than the slice length threshold, the first text segmentation in the document text information is determined as the starting segmentation position, and text segmentations corresponding to the slice length threshold are searched backward from the starting segmentation position, and the content of the first resource slice is determined based on the text segmentation at the starting segmentation position and the found text segmentations;

[0054] When the number of text segmentations other than the first resource fragment content in the document text information is greater than the slice length threshold, the starting segmentation position is updated based on the next text segmentation in the first resource fragment content, and text segmentations corresponding to the number of segmentations corresponding to the slice length threshold are searched backward from the updated starting segmentation position, and the second resource fragment content is determined based on the text segmentation at the updated starting segmentation position and the searched text segmentation;

[0055] The resource segment content corresponding to the document resource is determined based on the first resource segment content and the second resource segment content.

[0056] The document resources include a first document resource and a second document resource;

[0057] The method also includes:

[0058] Performing topic extraction processing on the content summary of the resource fragment content of the first document resource to obtain the content topic of the first document resource;

[0059] Performing topic extraction processing on the content summary of the resource fragment content of the second document resource to obtain the content topic of the second document resource;

[0060] Based on the content theme of the first document resource and the content theme of the second document resource, document merging and comparison information between the first document resource and the second document resource is determined.

[0061] The method further includes:

[0062] Performing document sentence processing on the document text information to obtain at least one sentence content corresponding to the document resource;

[0063] Performing feature extraction processing on each sentence content in at least one sentence content to obtain sentence content features corresponding to each sentence content;

[0064] When an interaction instruction initiated by the first business object for the document resource is obtained, feature extraction processing is performed on the interaction instruction to obtain instruction features corresponding to the interaction instruction;

[0065] When determining the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content, an interactive response to the interactive instruction is determined based on the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content.

[0066] The step of determining an interactive response to an interactive instruction based on feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content includes:

[0067] When a sentence content having a feature similarity greater than a first similarity threshold is found in at least one sentence content, determining an interactive reply to the interactive instruction based on the found sentence content having a feature similarity greater than the first similarity threshold;

[0068] When no sentence content with a feature similarity greater than the first similarity threshold is found in at least one sentence content, if a sentence content with a feature similarity less than the first similarity threshold and greater than the second similarity threshold is found, the search component is called to search for content associated with the interaction instruction, and an interactive response to the interaction instruction is determined based on the searched associated content.

[0069] On the one hand, an embodiment of the present application provides a document data processing method, the method comprising:

[0070] Display the summary interaction page;

[0071] When a first conversation message carrying a document resource is displayed on the summary interaction page, document analysis is performed on the document resource to obtain a document directory of the document resource; the paragraph analysis identifier in the document directory is associated with the document paragraph in the document resource;

[0072] When the second conversation message carrying the document directory is displayed on the summary interaction page, in response to a triggering operation on a paragraph analysis identifier in the document directory, the document paragraph corresponding to the paragraph analysis identifier is jumped to display.

[0073] Wherein, the paragraph analysis identifier includes the document page number;

[0074] In response to a trigger operation on a paragraph analysis identifier in a document directory, jumping to display the document paragraph corresponding to the paragraph analysis identifier includes:

[0075] In response to a trigger operation on a document page number in a document directory, a document resource display page associated with the document resource is displayed, and a document paragraph corresponding to the document page number is displayed on the document resource display page.

[0076] Wherein, the paragraph analysis identifier includes the document page number;

[0077] In response to a trigger operation on a paragraph analysis identifier in a document directory, jumping to display the document paragraph corresponding to the paragraph analysis identifier includes:

[0078] In response to a trigger operation on a document page number in the document directory, a document paragraph corresponding to the document page number is displayed on the summary interaction page.

[0079] On the one hand, an embodiment of the present application provides a document data processing device, the device comprising:

[0080] A first interaction page display module, configured to display a summary interaction page associated with the first business object;

[0081] A document import module, configured to, in response to a document import operation on a document resource, display a first conversation message carrying the document resource sent by the first business object on a summary interaction page;

[0082] a progress display module configured to traverse and display, in a first associated display area associated with the first conversation message, node progress prompt information of each process processing node among a plurality of process processing nodes associated with the document resource; the plurality of process processing nodes including a summary node configured to summarize resource fragment contents in the document resource;

[0083] A summary output module is used to display a second session message carrying the first summary summary on the summary interaction page when the node progress prompt information displayed is the node progress prompt information of the summary summary node, and the node progress prompt information of the summary summary node indicates that the first summary summary associated with the resource fragment content is obtained through the summary.

[0084] On the one hand, an embodiment of the present application provides a document data processing device, the device comprising:

[0085] The second interactive page display module is used to display the summary interactive page;

[0086] A directory acquisition module is configured to display a conversation message carrying a document directory of the document resource when a conversation message carrying a document resource is displayed on the summary interaction page; the document directory is determined by performing document analysis on at least one document paragraph in the document resource;

[0087] The paragraph jump module is used to jump and display the target document paragraph in response to a trigger operation on the target document paragraph in the document directory.

[0088] On the one hand, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. The computer program is suitable for being loaded and executed by a processor so that a computer device having the processor executes the method provided by the embodiment of the present application.

[0089] On one hand, an embodiment of the present application provides a computer program product, including a computer program / instruction, which, when executed by a processor, performs the method provided in the embodiment of the present application.

[0090] In an embodiment of the present application, information display can be performed in the form of conversational interaction during the process of extracting summary information of document resources, and document resources can be imported through a document import operation, without being limited to extracting information from document resources directly displayed on the page. In addition, during the process of extracting summary information of document resources, the document resource process processing progress can be displayed in the associated display area of ​​the session message carrying the document resource, and the node progress prompt information of multiple process processing nodes can be traversed and displayed, that is, the processing progress can be displayed step by step. Therefore, by displaying the process processing progress of the document resource during the process of extracting summary information of the document resource, the business object can perceive the processing progress of the document resource in real time, reducing the waiting feeling of the business object during the summary extraction process, and helping to enrich the information display method during the process of extracting summary information of the document resource. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0092] Figure 1 This is a structural diagram of a data processing system provided in an embodiment of the present application;

[0093] Figure 2 This is a scenario diagram of a document data processing method provided by an embodiment of the present application;

[0094] Figure 3 This is a flowchart of a document data processing method provided by an embodiment of the present application;

[0095] Figure 4 This is a schematic diagram of a document resource service portal provided in an embodiment of the present application;

[0096] Figure 5 This is a schematic diagram of the effect of a prompt message provided in an embodiment of the present application;

[0097] Figure 6 This is a flow chart of a document import operation provided by an embodiment of the present application;

[0098] Figure 7 This is a flowchart of a document data processing method provided by an embodiment of the present application;

[0099] Figure 8 This is a schematic diagram of the effect of a process review process provided by an embodiment of the present application;

[0100] Figure 9 This is a schematic diagram of the effect of a state adjustment operation provided by an embodiment of the present application;

[0101] Figure 10 This is a schematic diagram of the effect of a state adjustment operation provided by an embodiment of the present application;

[0102] Figure 11 This is a schematic diagram of the effect of a process processing node provided by an embodiment of the present application;

[0103] Figure 12 This is a schematic diagram of the effect of abstract re-extraction provided by an embodiment of the present application;

[0104] Figure 13 This is a schematic diagram of the effect of a session sharing operation provided in an embodiment of the present application;

[0105] Figure 14This is a schematic diagram of the effect of a page exit prompt page provided in an embodiment of the present application;

[0106] Figure 15 This is a schematic diagram of the effect of a historical conversation page provided by an embodiment of the present application;

[0107] Figure 16 This is a flowchart of a document data processing method provided by an embodiment of the present application;

[0108] Figure 17 This is a schematic diagram of the effect of a document content processing process provided by an embodiment of the present application;

[0109] Figure 18 This is a flowchart of a multi-document comparison process provided by an embodiment of the present application;

[0110] Figure 19 This is a flowchart of a dialogue response process provided by an embodiment of the present application;

[0111] Figure 20 This is a schematic diagram of the effect of a processing flow of a document resource service provided by an embodiment of the present application;

[0112] Figure 21 This is a flowchart of a document data processing method provided by an embodiment of the present application;

[0113] Figure 22 This is a schematic diagram of the display effect of the document directory provided in an embodiment of the present application;

[0114] Figure 23 This is a schematic diagram of a document paragraph jump process provided by an embodiment of the present application;

[0115] Figure 24 This is a schematic diagram of a document paragraph jump process provided by an embodiment of the present application;

[0116] Figure 25 This is a schematic diagram of a document paragraph jump process provided by an embodiment of the present application;

[0117] Figure 26 This is a schematic diagram of the structure of a document data processing device provided in an embodiment of the present application;

[0118] Figure 27 This is a schematic diagram of the structure of another document data processing device provided in an embodiment of the present application;

[0119] Figure 28 It is a structural diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0120] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0121] See Figure 1 , Figure 1 This is a structural diagram of a data processing system provided by an embodiment of the present application. Figure 1 As shown, the data processing system may include terminal devices (such as device 11a, device 12a, device 13a) and server 100a. It is understandable that Figure 1 The number of terminal devices and servers in the example is merely illustrative; any number of terminal devices and servers may be used depending on implementation requirements. Terminal devices (e.g., device 11a, device 12a, and device 13a) may communicate with servers via a network (i.e., a medium providing a communication link via a wired or wireless communication link or fiber optic cable, etc.) to transmit data.

[0122] It is understandable that a client may be running on a terminal device (such as device 12a), and the client may be a program that provides local services to a user (also known as a business object or an operation object). The server 100a may be the server corresponding to the client, and the server 100a may be running a program for providing resources, service data and other services to the client. It is understandable that the client running on the terminal device may also be called an application client, a business client, and so on. For example, the client running on the terminal device may be a client for providing social communication, and user A may communicate with user B through the client on the corresponding terminal device 11a. For example, user A sends a conversation message F1 to user B in the conversation page with user B, and the terminal device 11a corresponding to user A sends a conversation message F11 to the server 100a. The server 100a may then forward the conversation message F1 to the terminal device 12a corresponding to user B, thereby enabling user A to have a conversation with user B.

[0123] It is understood that terminal devices (such as device 12a) may include, but are not limited to, mobile phones, computers, intelligent voice interaction devices, smart home appliances, vehicle-mounted terminals, aircraft, smart speakers, smart home appliances, etc., and are not limited here. The server 100a can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms, and are not limited here.

[0124] It is understandable that the embodiments of the present application can be applied to a computer device, which can be the above-mentioned Figure 1 The terminal device in the Figure 1 Servers in the Figure 1 A data processing device that processes business data within a data processing system in a data processing system. The data processing device can be a server or a terminal device, wherein the terminal device may include but is not limited to mobile phones, computers, intelligent voice interaction devices, smart home appliances, vehicle-mounted terminals, aircraft, smart speakers, smart home appliances, etc., without limitation here. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms, without limitation here.

[0125] See Figure 2 , Figure 2 This is a scenario diagram of a document data processing method provided by an embodiment of the present application. Figure 2 As shown, the business object can trigger the display of page 200a, which can be a summary interaction page. In the summary interaction page that is triggered to display, a prompt conversation message (such as Figure 2 The prompt in the conversation message 201a can be determined according to actual needs, such as "Import document resources to extract document summary!", which is not limited here. The summary interaction page can also include a control for indicating the import of document resources (such as Figure 2 It is understood that the summary interaction page can be understood as a page for the business object to interact with the business processing object, and the prompt conversation message can be a conversation message associated with the business processing object.

[0126] Furthermore, the business object can perform a document import operation, and then display a conversation message carrying the imported document resource on the summary interaction page (eg Figure 2 The conversation message 203a has an associated display area (as shown in 203a). Figure 2 204a in the figure), the node progress prompt information of the current process processing node can be displayed in the associated display area 204a. For example, the node progress prompt information can be "Uploading document 75%", which means that in the process of resource processing of document resources, the process processing node of uploading document is being processed, and the process processing node of uploading document has reached 75% progress.

[0127] Furthermore, after the document resource is uploaded, the resource content of the uploaded document resource can be processed to extract a summary of the document resource. Throughout this process, node progress prompt information for multiple process processing nodes associated with the document resource, including the process processing node for the uploaded document resource, can be displayed in the associated display area of ​​the conversation message 203a. In other words, the displayed node progress prompt information can be changed in real time based on the progress of the document resource processing, allowing business objects to understand the current status of the document resource processing.

[0128] Furthermore, as the document resource is processed, the displayed node progress prompt information may be changed to the node progress prompt information of the summary node (a process processing node for summarizing the resource fragment content in the document resource), and the node progress prompt information of the summary node indicates that the summary summary of the document resource is obtained, such as Figure 2 As shown in 205a, the node progress prompt information is "summarize document 100%", which means that the current node is in the summary node and the processing of the summary node is completed.

[0129] Furthermore, when the displayed node progress prompt information may be the node progress prompt information of the summary node, and the node progress prompt information of the summary node indicates that the summary summary of the document resource is obtained by summarizing, a session message carrying the summary summary obtained by summarizing may be displayed (e.g., Figure 2 Optionally, when the display time of the node progress prompt information indicating the summary of the document resource reaches a certain time threshold (such as 1 second), the node prompt information displayed in the associated display area of ​​the conversation message 203a can be hidden.

[0130] In an embodiment of the present application, information display can be performed in the form of conversational interaction during the process of extracting summary information of document resources, and document resources can be imported through a document import operation, without being limited to extracting information from document resources directly displayed on the page. In addition, during the process of extracting summary information of document resources, the document resource process processing progress can be displayed in the associated display area of ​​the session message carrying the document resource, and the node progress prompt information of multiple process processing nodes can be traversed and displayed, that is, the processing progress can be displayed step by step. Therefore, by displaying the process processing progress of the document resource during the process of extracting summary information of the document resource, the business object can perceive the processing progress of the document resource in real time, reducing the waiting feeling of the business object during the summary extraction process, and helping to enrich the information display method during the process of extracting summary information of the document resource.

[0131] In the content extraction process of the present invention, the content is first segmented by an algorithm into multiple segments of content with a limited number of words (i.e., multiple resource segmented contents), and then the model service is called in parallel. The document resources are summarized through the business model provided by the model service to obtain the summarized summaries (i.e., content summaries) corresponding to the multiple segments of content (i.e., multiple resource segmented contents); then the summarized summaries of the multiple segments of content are aggregated, and the model service is called again to obtain the final content summary, which is fed back to the user. In this way, there is no need to directly extract the summary from the complete resource content, but the summaries of the multiple segmented contents are extracted separately and then aggregated, which reduces the difficulty of summary extraction and controls the time of the content summary generation process within a short time, thereby improving the efficiency of extracting summary information of document resources.

[0132] It is understood that the embodiments of the present application can be applied to the field of artificial intelligence technology, such as extracting content summaries of multiple resource slices of a document resource through a pre-trained business model, and aggregating the content summaries of the multiple resource slices to obtain a summary. Artificial intelligence (AI) refers to the theories, methods, techniques, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use knowledge to achieve optimal results. In other words, AI is a comprehensive technology within computer science that attempts to understand the essence of intelligence and produce new intelligent machines that can respond in a manner similar to human intelligence. AI is the study of the design principles and implementation methods of various intelligent machines, enabling them to have the capabilities of perception, reasoning, and decision-making. AI technology is an interdisciplinary discipline covering a wide range of fields, encompassing both hardware and software technologies. Basic AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technologies, pre-trained model technologies, operating / interaction systems, and mechatronics. Pre-trained models, also known as large models or basic models, can be widely applied to downstream tasks in various AI fields after fine-tuning. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0133] It is understandable that the business model in the embodiment of the present application can be a pre-training model (Pre-training model), also known as a cornerstone model or a large model, which refers to a deep neural network (DNN) with a large number of parameters. It is trained on a large amount of unlabeled data, and the function approximation ability of the large-parameter DNN is used to enable the PTM to extract common features from the data. After fine-tuning, efficient parameter fine-tuning (PEFT), prompt-tuning and other technologies, it is suitable for downstream tasks. Therefore, the pre-training model can achieve ideal results in a few-shot or zero-shot scenario. According to the data modality processed, PTM can be divided into language models (ELMO, BERT, GPT), visual models (swin-transformer, ViT, V-MOE), speech models (VALL-E), multimodal models (ViBERT, CLIP, Flamingo, Gato), etc., where a multimodal model refers to a model that establishes feature representations of two or more data modalities. The pre-trained model is an important tool for outputting artificial intelligence generated content (AIGC), and can also be used as a general interface to connect multiple specific task models. For example, the business model mentioned in this application can be a mixed model. It is understandable that since each slice of the document resource is generally long, it may be necessary to use a large language model of more than 6B to extract the summary of the slice content. For example, a large LLaMA 13B model can be specially trained for summary extraction.

[0134] It is understood that the above scenarios are merely examples and do not limit the application scenarios of the technical solutions provided in the embodiments of this application. The technical solutions of this application can also be applied to other scenarios. For example, those skilled in the art will appreciate that with the evolution of system architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of this application will also be applicable to similar technical problems.

[0135] Further, see Figure 3 , Figure 3 This is a flowchart of a document data processing method provided by an embodiment of the present application. The method can be executed by a computer device, for example, the computer device can be Figure 1 The method may include at least the following steps S101-S102.

[0136] S101: Display a summary interaction page associated with a first business object.

[0137] Among them, the summary interaction page can be an interaction page used by the first business object to perform resource processing of document resources. It can be understood that the summary interaction page can be used to provide document resource services (also known as resource processing services, document resource processing services) in the form of a session, and the document resource services can be used to let business objects understand the content of document resources. Among them, the first business object can be any business object that enables the document resource service, such as user A, user B, etc. It should be understood that the summary interaction page can be understood as a page for the first business object to interact with the business processing object. In other words, the summary interaction page can be used to display the session messages sent by the first business object and the business processing object, and the business processing object can be the object that provides the document resource service.

[0138] Optionally, the summary interaction page associated with the first business object may be displayed in response to a business trigger operation of the first business object on a document resource business portal in a business selection page. The business selection page may be a page provided in an application for business selection (i.e., the document resource business portal is one of multiple business portals involved in an application), or may be an application selection page in a computer device (i.e., the document resource business portal is a portal to an application), without limitation herein.

[0139] The business trigger operation can be a touch operation on the document resource business entrance (such as clicking on the document resource business entrance), a voice command operation (such as entering a voice command to indicate the activation of business resources), or a gesture operation (such as drawing an "S" gesture on the business selection page), etc., which is not limited here.

[0140] For example, see Figure 4 , Figure 4 This is a schematic diagram of a document resource service entry provided by an embodiment of the present application. Figure 4 As shown, page 401a may be a page for providing service selection in an application. In the page 401a, multiple service entrances may be included, such as service entrance 1, service entrance 2, ..., service entrance 6, etc. Among them, service entrance 1 (such as Figure 4 411a in FIG) is a business entry for providing document resource services (ie, document resource business entry), and when the business object clicks on the business entry 1, the summary interaction page corresponding to the document resource business is displayed (eg, Figure 4 As shown in page 402a in the summary interaction page, the summary interaction page may include a session message carrying prompt information for prompting the summary interaction page to learn the resource content of the document resource, also known as a prompt message for prompting the summary interaction page to learn the resource content of the document resource, such as Figure 4 As shown in 412a.

[0141] Optionally, in response to a business trigger operation of the first business object on the document resource business entry in the business selection page, the displayed summary interaction page may include a document import control, which may be a control for indicating the import of document resources to be processed, and the business object may import document resources through the document import control.

[0142] Optionally, in response to a business triggering operation of the first business object on the document resource business entry in the business selection page, the displayed summary interaction page may include a prompt message for prompting that the resource content of the document resource can be learned on the summary interaction page, and the prompt message may be displayed in the form of a conversation. Optionally, the prompt message may include at least one resource processing flow example. When the first business object triggers the target resource processing flow example in the at least one resource processing flow example, a conversation message carrying the example document resource associated with the resource processing flow example may be displayed, and a conversation message carrying the summary summary of the example document resource may be displayed. The resource processing flow example is an example of a processing flow for displaying the summary summary of the document resource obtained by the resource processing business, and the example document resource may be a document summary for exemplarily displaying the processing flow for displaying the summary summary of the document resource obtained by the resource processing business. It is understandable that when displaying the resource processing process example, the resource processing process of the resource processing business can be displayed in the form of a video, or the resource processing process of the resource processing business can be displayed in the form of a graphic or text, or the session message associated with the resource processing process example can be directly displayed in the summary interaction page, and the session message carrying the summary summary of the example document resource can be displayed. There is no limitation here.

[0143] For example, see Figure 5 , Figure 5 This is a schematic diagram of the effect of a prompt message provided by an embodiment of the present application. Figure 5 As shown, page 501a is a summary interaction page, which may display a prompt message for prompting that the resource content of the document resource can be learned on the summary interaction page, such as Figure 5 In the prompt message 502a, multiple resource processing flow examples may be included, such as resource processing flow example 1 (eg Figure 5 521a in FIG), resource processing flow example 2 (as shown in FIG. Figure 5 522a in FIG), resource processing flow example 3 (as shown in FIG. Figure 5Furthermore, when the business triggers (e.g., clicks) resource processing flow example 1, the resource processing flow example 1 can be used to demonstrate the process of resource processing for the document resource. Optionally, the resource processing flow example cannot be triggered while resource processing is in progress for the document resource.

[0144] S102: In response to a document import operation on a document resource, display a first conversation message carrying the document resource and sent by the first business object on a summary interaction page.

[0145] Wherein, the document import operation may be an operation for indicating the import of a document resource to be processed. It is understandable that the document import operation may be a confirmation import operation for the selected document resource in the resource management page when the resource management page associated with the document import control is displayed in response to a touch operation on the document import control (such as clicking the document import control). Wherein, the resource management page may be a page for managing document resources in a computer device. It is understandable that the resource management page may be a page for managing all document resources in a computer device, or a page for managing document resources in an application associated with the resource processing business (also referred to as a document resource business) in the computer device. The application associated with the resource processing business may be some social communication applications, and the document resources in the application may be document resources sent or received when the first business object interacts in the social communication application.

[0146] The document resource may be in various formats, including but not limited to text, image, video, code, table, URL (Uniform Resource Locator), etc., which are not limited here.

[0147] It is understood that the first conversation message may be a conversation message carrying a document resource. It is also understood that the first conversation message may be a conversation message associated with a first business object, i.e., a conversation message sent by the first business object. It is also understood that when the first conversation message is displayed, the document resource carried by the first conversation message may be displayed in the form of a document card (such as a file card, link card, etc.) or a thumbnail, meaning that the entire resource content of the document resource may not be displayed directly in the conversation message.

[0148] It is understandable that when the document resource is a file in a format such as text, table, video, voice, PPT (slide), code, etc., a file card of the document resource can be displayed in the first conversation message. The file card can include the file title, or a file identification image of the file format corresponding to the document resource can be displayed (for example, document resources in PPT and text formats are identified by different file identification images). When the document resource is a URL (Uniform Resource Locator), the URL can be directly displayed in the first conversation message, or a link card corresponding to the URL can be displayed. The link card can include the resource title and thumbnail of the web page corresponding to the URL, which is not limited here. When the document resource is an image, the thumbnail corresponding to the image can be displayed, or the image access address corresponding to the image can be displayed. It is understandable that when the first business object triggers the document resource in the first conversation message, the document content of the document resource can be viewed. For example, when the first business object clicks a file card in the first conversation message, the file corresponding to the file card can be opened to view the file content; when the first business object clicks a link card in the first conversation message, the web page corresponding to the resource link (i.e., URL) indicated by the link card can be opened to view the web page content of the resource link; when the first business object clicks an image in the first conversation message, the specific content of the image can be viewed.

[0149] For example, see Figure 6 , Figure 6 This is a flow chart of a document import operation provided by an embodiment of the present application. Figure 6 As shown, the business page 601a is a summary interaction page displayed when the summary interaction page is triggered to be displayed. The page 601a may include a document import control (such as a document import control) for importing document resources. Figure 6 ), when the business object clicks the document import control 611a, a page 602a may be displayed. The page 602a may be a resource management page. The page 602a may include multiple document resources, such as document resource 1, document resource 2, document resource 3, and other document resources. When the business object confirms the selection of document resource 1 (as shown in FIG. Figure 6 621a in FIG), a page 603a may be displayed, which may be a summary interaction page for displaying a conversation message carrying the imported document resource 1, as shown in FIG. Figure 6 As shown in the page 603a in FIG, the session message 631a is a message carrying document resource 1 (such as Figure 6 632a in the session message).

[0150] S103. In the first associated display area associated with the first conversation message, traverse and display the node progress prompt information of each process processing node in the multiple process processing nodes associated with the document resource; the multiple process processing nodes include a summary summary node for summarizing the resource fragment content in the document resource.

[0151] The first associated display area may be a display area associated with the first conversation message. It is understood that the first associated display area may be above or below the first conversation message, or to the left or right of the first associated display area, without limitation. It is understood that the size of the first associated display area may be determined based on the actual size of the information to be displayed, or may be a display area of ​​a preset size (i.e., the size of the first associated display area is the preset size), without limitation.

[0152] The resource fragment content can be a fragment of a document resource, or can be fragment content, document slice, document fragment, etc. A document resource can include one or more resource fragment contents. The method for determining the resource fragment content of a document resource can refer to the subsequent related description and will not be repeated here.

[0153] It is understandable that the multiple process processing nodes also include a document upload node for uploading document resources, and a document content processing node for processing the document content of the document resources. The document content processing node may include a content parsing node for determining the document content of the document resource, a content slicing node for determining the resource slicing content corresponding to the document resource, a slicing summary extraction node for determining the content summary of the resource slicing content, and a summary aggregation node for summarizing the content summary of the resource slicing content associated with the document resource. In other words, the document upload node may be a process processing node for uploading document resources, the content parsing node may be a process processing node for parsing to obtain the document content of the document resource, the content slicing node may be a process processing node for slicing the document resource to obtain the resource slicing content corresponding to the document resource, the slicing summary extraction node may be a process processing node for extracting the content summary of the resource slicing content, and the summary aggregation node may be a process processing node for summarizing the content summary of the resource slicing content associated with the document resource to obtain a summary summary.

[0154] It is understood that a document content processing node may involve multiple process processing nodes, namely, the aforementioned content parsing node, content fragmentation node, fragment summary extraction node, and summary aggregation node. It is also understood that the process processing nodes under the document content processing node may be further divided into more fine-grained process processing nodes. For example, the content parsing node may include an image processing node, a video processing node, an audio processing node, and so on, but this is not limited here.

[0155] It is understandable that the node progress prompt information of each process processing node in the multiple process processing nodes associated with the document resource can be displayed in sequence according to the process node sequence of the resource processing process of the resource document, and the node progress prompt information of the displayed process processing node is determined according to the progress of resource processing of the resource document (that is, the node progress prompt information displayed is determined according to the process processing node to which the resource processing of the resource document is executed). The process node sequence can be the execution order of each process processing node in the resource processing process of the resource document, for example, the process node sequence can be: document upload node - content parsing node - content fragmentation node - fragment summary extraction node - summary aggregation node.

[0156] It is understandable that the node progress prompt information can be used to prompt the progress of the process processing node that processes the document resource. One process processing node corresponds to one node progress prompt information. The node progress prompt information corresponding to a process processing node includes process node text information associated with the process processing node, and can also include prompt information for indicating the node processing progress of the process processing node corresponding to the displayed node progress prompt information. The node processing progress can be used to indicate the processing progress of a process processing node, such as the remaining processing time of a process processing node (such as a countdown of 10 seconds), the proportion of the current processing progress to the total progress (such as 75% progress), etc. It should be understood that the node progress prompt information corresponding to a process processing node can be dynamically changed according to the execution progress of the process processing node, such as continuously updating the prompt information for prompting the node processing progress in the node progress prompt information according to the execution progress.

[0157] Specifically, any one of the multiple process processing nodes is a target process processing node; then, traversing and displaying the node progress prompt information of each process processing node in the multiple process processing nodes associated with the document resource may include the following steps: when the displayed node progress prompt information is the node progress prompt information of the target process processing node, dynamically displaying the node progress prompt information of the target process processing node; the node progress prompt information of the target process processing node changes as the node processing progress of the target process processing node changes when the target process processing node processes the document resource.

[0158] Among them, the target process processing node can be any process processing node among the above-mentioned multiple process processing nodes. It is understandable that the dynamic display of the node progress prompt information of the target process processing node can change as the node processing progress changes when the target process processing node processes the document resources. For example, when the node progress prompt information includes the progress execution ratio of the current processing progress to the total progress (that is, the ratio of the node processing progress that characterizes the target process processing node), the progress execution ratio is continuously updated. For example, as the processing progress of the target process processing node progresses, the ratio can be continuously updated to 1%, 10%, 30%, ..., 100%, etc. When it is updated to 100%, it means that the target process processing node has been executed.

[0159] Specifically, the node progress prompt information of the target process processing node is dynamically displayed, including one or more of the following: when the node progress prompt information of the target process processing node includes the time progress prompt information of the target process processing node, the time progress prompt information of the target process processing node is dynamically displayed; or, when the node progress prompt information of the target process processing node includes the numerical progress prompt information of the target process processing node, the numerical progress prompt information of the target process processing node is dynamically displayed.

[0160] Among them, the time progress prompt information can be information that prompts the processing progress of the target process processing node through time. The time progress prompt information can include an estimated remaining processing time (also called a countdown), and the estimated remaining processing time can be determined based on the difference between the estimated total processing time and the actual execution time. The estimated total processing time can be determined based on the data volume of the resource document in the target process processing node. For example, the estimated total processing time of the document upload node can be determined based on the data volume of the document resource, and the content segmentation node can be determined based on the number of text segmentations of the document resource. This is not limited here. The estimated remaining processing time can also be determined based on the data volume of the remaining data of the document resource (that is, the part of the total data of the document resource excluding the processed data).

[0161] It should be understood that the time progress prompt information can be continuously updated as the target process processing node is executed, such as continuously reducing the estimated remaining processing time as the target process processing node is executed.

[0162] The numerical progress prompt information may be information indicating the processing progress of the target process processing node through a numerical value. The numerical progress prompt information may be a progress execution ratio, which may be determined based on the ratio of the current processing progress (e.g., the amount of data currently being processed) to the total progress (e.g., the total amount of data to be processed).

[0163] It is understood that the node progress prompt information may include only time progress prompt information, only numerical progress prompt information, or both time progress prompt information and numerical progress prompt information, which is not limited here. Optionally, the node progress prompt information may also include a progress prompt animation, which may dynamically change according to the numerical progress prompt information and may also be called a progress bar.

[0164] S104. When the displayed node progress prompt information is the node progress prompt information of the summary summary node, and the node progress prompt information of the summary summary node indicates that a first summary associated with the resource shard content is obtained through the summary, a second session message carrying the first summary is displayed on the summary interaction page.

[0165] It is understood that the first summary can be a summary extracted from the document resource, which can be used to describe the main document content of the document resource. It should be understood that the summary node can be the last process processing node among the multiple process processing nodes mentioned above.

[0166] It is understood that when the node progress prompt information of the summary node indicates that the first summary of the document resource has been obtained, it means that the resource processing flow for the document resource has been completed. It is understood that the node progress prompt information of the summary node indicating that the first summary of the document resource has been obtained can be an estimated remaining processing time of 0 as indicated by the time progress prompt information in the node progress prompt information, or a progress execution ratio of 100% as indicated by the numerical progress prompt information, or can be prompt information used to indicate that the summary of the document resource has been completed.

[0167] It is understood that the second session message can be a session message that carries a summary of the document resource (such as the first summary). It is also understood that the second session message can be a session message associated with the above-mentioned business processing object, that is, the first session message can be displayed as a session message sent by the business processing object.

[0168] It is understandable that when the second conversation message is displayed on the summary interaction page, the first summary can be directly displayed in the second conversation message, and is displayed here in the form of text, image, table, etc., which is not limited here.

[0169] Optionally, when displaying the second conversation message on the summary interaction page, the first summary can be displayed in batches in the second conversation message. The number of characters displayed in each batch can be one character (or multiple characters) or one sentence (or multiple sentences), which is not limited here. It is understood that the output time interval between each batch can be a preset time interval, such as 0.2 seconds. For example, one additional character or sentence can be displayed in the second conversation message at a time until the first summary is output, thereby creating an effect similar to outputting characters word by word or sentence by sentence.

[0170] Optionally, when determining the summary, the document resource's document directory may be extracted and then displayed in the second conversation message. Specifically, the summary and document directory may be displayed in the second conversation message, thereby enabling the business object to quickly understand the content distribution within the document resource and improving the business object's understanding of the document resource's content. The document directory typically includes paragraph summaries (i.e., paragraph titles), corresponding page numbers, and serial numbers for each paragraph in the document.

[0171] Further, see Figure 7 , Figure 7 This is a flowchart of a document data processing method provided by an embodiment of the present application. The method can be executed by a computer device, for example, the computer device can be Figure 1 The method may include at least the following steps S201 to S206.

[0172] S201: Display a summary interaction page associated with a first business object.

[0173] S202: In response to a document import operation on a document resource, display a first conversation message carrying the document resource sent by the first business object on a summary interaction page.

[0174] Optionally, in the document import operation, the document content that needs to be summarized can be specified, such as the document page number that needs to be summarized (such as specifying page 1 as the document page number that needs to be summarized), the document paragraph (such as specifying the first document paragraph as the document paragraph that needs to be summarized), etc. Specifically, in the resource management page indicated by the document import operation, in response to the confirmation operation for the selected document resource, a content designation page for specifying the document content that needs to be summarized is displayed, and in response to the confirmation operation for the selected document page number or document content indicated by the content designation page, a session message carrying the document resource is displayed. Among them, the content designation page can be a page for selecting the document content that needs to be summarized, and the content designation page can be a pop-up page or a page different from the resource management page, which is not limited here.

[0175] Furthermore, the summary carried in the output second conversation message may be the selected document content for which summary extraction is required. It is understood that if the business object selects a document page or paragraph for which summary extraction is required, summary extraction can be performed on the selected document page or paragraph. If the business object does not specify a document page or paragraph for which summary extraction is required, summary extraction can be performed on the entire document resource. This is not limited here.

[0176] S203. In the first associated display area associated with the first conversation message, traverse and display the node progress prompt information of each process processing node in the multiple process processing nodes associated with the document resource; the multiple process processing nodes include a summary summary node for summarizing the resource fragment content in the document resource.

[0177] Among them, it can be understood that in an embodiment of the present application, the business object can also view the processing status of multiple process processing nodes associated with the document resource in the first associated display area, so that the business object can understand the processing flow for the document resource in advance, and let the business object understand the current process node, reducing the business object's sense of unknown about the resource processing flow, and thereby reducing the business object's sense of waiting.

[0178] Specifically, in response to the node viewing operation for multiple process processing nodes, the process node prompt information corresponding to the multiple process processing nodes is displayed in the first associated display area; one process processing node corresponds to one process node prompt information; the process node prompt information of any process processing node includes node status information for prompting the node processing status of any process processing node.

[0179] The node viewing operation can be an operation for viewing multiple process processing nodes associated with a document resource. It is understood that the node viewing operation can be a triggering operation within the first association display area, such as clicking anywhere within the first association display area, or clicking a control within the first association display area for triggering the display of multiple process processing nodes, without limitation herein.

[0180] It should be understood that the process node prompt information can be information used to indicate the process processing node. For example, the process node prompt information can include textual description information of the process processing node. It is understood that when displaying multiple process processing nodes, the multiple process processing nodes can be arranged in sequence according to their process node order to indicate the execution order between the multiple process processing nodes. It is understood that each process processing node corresponds to a process node prompt information. In other words, each process processing node has corresponding process node prompt information.

[0181] It should be understood that the process node prompt information of any process processing node includes node status information for indicating the node processing status of any process processing node. The node status information can be node processing status information for indicating the process processing node. The node processing status of a process processing node can include a processing status (also called a first node status), a processed status (also called a second node status), and an unprocessed status (also called a third node status).

[0182] It can be understood that in the first associated display area, process processing nodes displaying different node processing states can be distinguished by different display effects, such as by different symbols, texts, font colors (or font thickness, font size), etc., which are not limited here.

[0183] Optionally, the estimated total processing time for an unprocessed process node can be displayed in the process node prompt information. This can indicate the time required for the unprocessed process node, thus reducing the wait time for the business object. Optionally, the corresponding estimated total processing time can be displayed for each unprocessed process node, or only for process nodes with an estimated total processing time greater than a certain threshold, thereby avoiding the display of excessive information and improving the sophistication and simplicity of the page.

[0184] For example, see Figure 8 , Figure 8 This is a schematic diagram of the effect of a process viewing process provided by the embodiment of this application. Figure 8As shown, page 801a is a summary interaction page for processing the imported document resource 1. In page 801a, an associated display area (such as Figure 8 ), in which the node progress prompt information indicating the process processing node that is processing the image information is displayed, such as "75% of the image information is being processed". After the business object performs a node viewing operation (such as clicking anywhere in area 802a), the display effect of the interactive page that can be summarized is shown in page 801b. In page 801b, the associated display area 802a is expanded to be displayed as the area shown in area 802b. In the area shown in area 802b, node progress prompt information of multiple process processing nodes can be included. Among them, as shown in FIG. Figure 8 The node progress prompt information in the area shown in 803a is the process node prompt information of the processed process processing node, such as Figure 8 The area shown in 804a is the process node prompt information of the process processing node being processed, such as Figure 8 The process node prompt information in the area 805a shown in FIG is the process node prompt information of the unprocessed process processing node. Optionally, the process node prompt information of the unprocessed process processing node can display the estimated total processing time of the unprocessed process processing node, for example, Figure 8 The process node prompt information shown in 806a is "organization and merging summary 11s", which means that the estimated total processing time corresponding to the process processing node (ie, summary node) is 11s.

[0185] S204. When the displayed node progress prompt information is the node progress prompt information of the summary summary node, and the node progress prompt information of the summary summary node indicates that a first summary associated with the resource slice content is obtained through the summary, a second session message carrying the first summary is displayed on the summary interaction page.

[0186] It is understood that the summary interaction page may also include a document processing status control, which is used to adjust the document processing status of the document resource. For example, during the process of uploading a document, the document processing status control can be triggered to cancel or pause the upload of the document resource; for another example, during the processing of the document content of the document resource, the document processing status control can be triggered to cancel or temporarily pause the processing of the document content of the document resource, without limitation here.

[0187] Specifically, multiple process processing nodes also include a document upload node for uploading document resources; the summary interaction page includes a document processing status control, which is used to adjust the document processing status of the document resource, and the document processing status includes a document upload status, which is the document processing status determined when the document resource is processed through the document upload node; then, the embodiment of the present application may also include the following steps: when the node progress prompt information displayed in the first associated display area is the node progress prompt information of the document upload node, first status prompt information is displayed in the document processing status control; the first status prompt information is used to indicate that the document processing status of the document resource is the document upload status; in response to the first status switching operation on the document processing status control, the document processing status of the document resource is adjusted from the document upload status to the cancel upload status, and a first status prompt session message is output to prompt that the document processing status of the document resource is the cancel upload status.

[0188] Among them, the document processing status includes the document upload status, and the document upload status is the document processing status determined when the document resource is processed through the document upload node. In other words, the document upload status can be the document processing status when the document resource is uploaded.

[0189] It should be understood that the first status prompt information may be used to indicate that the document processing status of the document resource is the document uploading status. Optionally, the first status prompt information may also include prompt information for instructing to execute a trigger operation for the document processing status control to cancel the document uploading.

[0190] It should be understood that the first state switching operation can be an operation for switching the document upload state to a cancel upload state. The cancel upload state can be a state for indicating the cancellation of document upload. It should be understood that the first state switching operation can be a touch operation on the document processing state control (such as clicking the document processing state control), or a voice command operation (such as inputting a voice command for instructing to cancel document upload), etc., which is not limited here.

[0191] It can be understood that the first status prompt conversation message can be used to prompt that the document processing status of the document resource is the cancel upload status. In other words, the first status prompt conversation message can be used to prompt that the upload of the document resource has been canceled, that is, the first status prompt conversation message can include the text of the conversation message used to prompt that the upload of the document has been canceled.

[0192] It is understood that the associated display area of ​​the first status prompt conversation message may include a control for instructing to re-upload the document resource. If the business object performs a touch operation on the control for re-uploading the document resource, the process processing node for uploading the document resource is re-entered, and then the process for processing the document resource is re-entered.

[0193] For example, see Figure 9 , Figure 9 This is a schematic diagram of the effect of a state adjustment operation provided by an embodiment of the present application. Figure 9 As shown, page 901a is a schematic diagram of the effect of the summary interaction page when uploading a document. In page 901a, a document processing status control (such as Figure 9 In the control 902a, a status prompt message "Uploading, click to cancel" may be included to indicate that the document resource 1 is in the document uploading state. The business object can perform an operation to cancel the uploading of the document. For example, if the control 902a is clicked, the summary interaction page may be as follows: Figure 9 In the page 901b, a status prompting session message for prompting that the document processing status of the document resource is the cancel upload status may be included, such as Figure 9 Optionally, the page 901a may also include a control for instructing to re-upload the document resource (e.g., Figure 9 If the business object clicks the control 904a, the document resource 1 can be re-uploaded. Optionally, the page 901a may also include a control for instructing to import a new document resource (such as Figure 9 As shown in the control 905a in the figure), the business object can click the control 905a to trigger the import of a new document resource. The new document resource can be the same as the document resource imported for the first time or different, which is not limited here.

[0194] Optionally, if the document resource fails to be uploaded due to network issues or other problems during the document upload process, a prompt indicating the failure can be displayed in the display area associated with the first conversation message. Optionally, a control for instructing the user to re-upload the document resource carried in the first conversation message can also be displayed, so that the business object can re-import the document resource by clicking the control for instructing the user to re-upload the document resource carried in the first conversation message.

[0195] Specifically, the summary interaction page includes a document processing status control, which is used to adjust the document processing status of the document resource; the multiple process processing nodes also include a document content processing node, which includes a summary summary node, a content parsing node for determining the document content of the document resource, a content fragmentation node for determining the resource fragment content corresponding to the document resource, and a fragment summary extraction node for determining the content summary of the resource fragment content; the document processing status includes the document content processing status, which is the document processing status determined when the document resource is processed by the document content processing node; then, the embodiment of the present application may also include the following steps: when the node progress prompt information displayed in the first associated display area is the node progress prompt information of the document content processing node, a second status prompt information is displayed in the document processing status control; the second status prompt information is used to indicate that the document processing status of the document resource is the document content processing status; in response to the second state switching operation on the document processing status control, the document processing status of the document resource is adjusted from the document content processing status to the content processing cancellation status, and a second state prompt session message is output to prompt that the document processing status is the content processing cancellation status.

[0196] The introduction to the document content processing node can be found in the above description and will not be repeated here. The document processing state includes the document content processing state, which is the document processing state determined when a document resource is processed by the document content processing node. In other words, the document content processing state can be the document processing state of the document resource's document content.

[0197] It should be understood that the second status prompt information may be used to indicate that the document processing status of the document resource is the document content processing status. Optionally, the second status prompt information may also include prompt information for instructing to execute a trigger operation for the document processing status control to cancel the processing of the document content of the document resource.

[0198] It should be understood that the second state switching operation can be an operation for switching the document content processing state to a content processing cancellation state. The content processing cancellation state can be a state for indicating the cancellation of the document content processing of the document resource. It should be understood that the second state switching operation can be a touch operation on the document processing state control (such as clicking the document processing state control), or a voice command operation (such as inputting a voice command for instructing to cancel the upload of the document), etc., which is not limited here.

[0199] It can be understood that the second status prompt conversation message can be used to prompt that the document processing status of the document resource is a content processing cancellation status. In other words, the second status prompt conversation message can be a conversation message used to prompt that the processing of the resource content of the document resource has been canceled, that is, the second status prompt conversation message can include text used to prompt that the processing of the resource content of the document resource has been canceled.

[0200] For example, see Figure 10 , Figure 10 This is a schematic diagram of the effect of a state adjustment operation provided by an embodiment of the present application. Figure 10 As shown, page 1001a is a schematic diagram of the effect of the summary interaction page when the resource content of the document resource is being processed. In page 1001a, a document processing status control (such as Figure 10 In the control 1002a, a status prompt message "Content processing, click to cancel" can be included to indicate that the document resource 1 is in the document content processing state. The business object can perform an operation to cancel the resource content processing of the document resource. For example, if the control 1002a is clicked, the summary interaction page can be as follows: Figure 10 In the page 1001b, a status prompting session message for prompting that the document processing status of the document resource is the content processing canceled state may be included, such as Figure 10 Optionally, the page 1001a may also include a control for instructing to reprocess the document resource (e.g., Figure 10 If the business object clicks the control 1004a, the resource content of the document resource 1 can be reprocessed. Optionally, the page 1001a may also include a control for instructing to import a new document resource (such as Figure 10 As shown in the control 1005a in the figure), the business object can click the control 1005a to trigger the import of a new document resource. The new document resource can be the same as the document resource imported for the first time or different, which is not limited here.

[0201] Optionally, if processing of the document resource's content fails due to network or other issues, a prompt indicating the failure of the corresponding process processing node may be displayed in the display area associated with the first conversation message. Optionally, a control instructing the reprocessing of the document resource's content carried in the first conversation message may also be displayed, so that the business object can click the control to reprocess the document resource's content and obtain a summary of the document resource.

[0202] For example, see Figure 11 , Figure 11 This is a schematic diagram of the effect of a process processing node provided by the embodiment of this application. Figure 11 As shown, Figure 11 As shown, page 1011a is a summary interaction page for the failed processing of document resource 1 by the image parsing node. In page 1011a, an associated display area (such as Figure 11 In the area 1012a, a node progress prompt message indicating that the image information processing failed is displayed, such as "Processing image information failed". In the page 1011a, a control for instructing to re-process the resource content of the document resource carried by the first conversation message (such as Figure 11 1013a in the document resource), and then after the business object clicks the control, the resource content of the document resource can be reprocessed to obtain a summary of the document resource. It can be understood that after the business object performs a node viewing operation (such as clicking anywhere in area 1012a), the display effect of the interactive page that can be summarized is shown in page 1011b. In page 1011b, the associated display area 1012a is expanded to be displayed as the area shown in area 1012b. In the area shown in area 1012a, process node prompt information of multiple process processing nodes can be included. Among them, as shown in FIG. Figure 11 The process node prompt information indicated by 1014a can be used to indicate that the corresponding process processing node has failed, such as "processing image information failed."

[0203] It is understandable that the associated display area of ​​the second state prompt conversation message may include a control for indicating that the document content of the document resource is to be reprocessed. If the business object performs a touch operation on the control for indicating that the document content of the document resource is to be reprocessed, the process of processing the resource content of the document resource is re-entered. Optionally, when re-entering the processing process, the processing can be re-performed from the starting position of the document content processing node, or the processing can be re-performed from an unfinished process processing node in the document content processing node. For example, when executing the second state switching operation, the processing is being performed through the content segmentation node. Then, when resuming the processing of the resource content of the document resource, the processing can be started from the content segmentation node. This is not limited here.

[0204] Optionally, after displaying the second conversation message carrying the first summary, the summary of the document resource can also be regenerated in response to a summary re-extraction operation. The summary re-extraction operation can be a triggering operation for a summary re-extraction control on the summary interaction page for instructing the regeneration of the summary summary of the document resource, or it can be a conversation message sent in the summary interaction page carrying an instruction for instructing the regeneration of the summary summary of the document resource, which is not limited here. The summary re-extraction control can be a control for instructing the regeneration of the summary summary of the document resource, which can be displayed at any position in the summary interaction page, or it can be displayed in an associated display area (referred to as a second associated display area) associated with the second conversation message, which is not limited here.

[0205] Specifically, when the second conversation message is displayed on the summary interaction page, a summary re-extraction control is displayed for indicating the regeneration of the summary summary of the document resource; in response to the triggering operation of the summary re-extraction control, the second summary summary regenerated for the document resource is displayed in the second conversation message.

[0206] The triggering operation for the summary re-extraction control may be used to trigger the operation of the summary re-extraction control, such as clicking to trigger the operation of the summary re-extraction control.

[0207] It should be understood that the second summary can be a summary generated from the document resource. It should be understood that the method for generating the second summary can refer to the above description of determining the first summary, which will not be repeated here.

[0208] Optionally, a summary switching control for the summary of the document resource is displayed in the second associated display area associated with the second conversation message; in response to a triggering operation on the summary switching control, the first summary and the second summary are switched and displayed in the second conversation message.

[0209] The summary switching control may be a control for instructing to switch the summary displayed in the second conversation message. The triggering operation on the summary switching control may be an operation for triggering the summary switching control, such as clicking the summary switching control.

[0210] For example, the summary switching control may include a control for displaying each summary generated, such as using identifiers such as 1, 2, and 3 to represent different summary summaries, and the business object selects the corresponding identifier to display the corresponding summary summary; or, the summary switching control may include controls for switching the previous summary and the next summary, and when the control for switching the previous summary is clicked, the previous summary is displayed, and when the control for switching the next summary is clicked, the next summary is displayed; or, the summary switching control may include a control for switching the next summary, and when the control for switching the next summary is clicked, the next summary is displayed.

[0211] For example, see Figure 12 , Figure 12 This is a schematic diagram of the effect of abstract re-extraction provided by the embodiment of this application. Figure 12 As shown, page 1101a may be a summary interaction page when outputting the summary extracted for the first time, and the page 1101a may include a conversation message carrying the summary extracted for the first time, such as Figure 12 The page 1101a may also display a summary re-extraction control (e.g., a summary re-extraction control) for instructing to regenerate the summary of the document resource in the associated display area associated with the conversation message 1102a. Figure 12 When the business object clicks the control 1103a, the summary of the document resource can be re-extracted. When the summary of the document resource is re-extracted, the summary interaction page can refer to the page 1101b. In page 1101b, the conversation message for outputting the summary of the document resource (i.e., the second conversation message) can display the re-extracted summary, as shown in FIG. Figure 12 The conversation message shown in 1102b in the figure includes summary summary 2. Optionally, in the associated display area associated with the conversation message 1102b in the page 1101b, a summary re-extraction control (such as a summary re-extraction control) indicating the regeneration of the summary summary of the document resource may be displayed. Figure 12 As shown in 1104a in FIG) Optionally, in the associated display area associated with the conversation message 1102b, a control for switching the summary displayed in the second conversation message may also be displayed, such as Figure 121105a and 1106a, wherein control 1105a can be used to switch the display of the previous summary (if the previous summary does not exist, the control 1106a cannot be triggered), and control 1106a can be used to switch the display of the next summary (if the next summary does not exist, the control 1106a cannot be triggered). Optionally, the associated display area associated with the conversation message 1102b can also display a control for indicating the satisfaction of the business object with the displayed summary, such as Figure 12 1107a and 1108a, wherein if the business object is satisfied with the displayed summary, the business object can click control 1107a to provide feedback; if the business object is not satisfied with the displayed summary, the business object can click control 1108a to provide feedback.

[0212] It is understood that the newly generated summary and the previously generated summary can be displayed in the same second conversation message, i.e., displayed as described above. Optionally, when displaying the newly generated summary, the newly generated summary (e.g., the second summary) can be displayed in a second conversation message different from the second conversation message containing the previously generated summary. In other words, each generated summary can be displayed separately across multiple conversation messages.

[0213] It is understandable that in the embodiment of the present application, the conversation message displayed in the summary interaction page associated with the first business object can be shared with another business object (such as a second business object different from the first business object).

[0214] Specifically, in response to the session sharing operation in the summary interaction page, the session message in the summary interaction page is shared to the second business object, so that the second business object triggers the display of the sharing interaction page associated with the second business object through the message sharing link associated with the session sharing operation, and the sharing interaction page includes the session message displayed in the summary interaction page.

[0215] The session sharing operation can be an operation for sharing session messages on the summary interaction page. It is understood that the session sharing operation can be a triggering operation for a control on the summary interaction page for session sharing (which can default to sharing all session messages on the summary interaction page), or a confirmation operation for a selected session message when triggering the selection of session messages to be forwarded. It is understood that when performing a session operation, the first business object can select the session messages to be shared.

[0216] It is understood that when a conversation message in a summary interaction page is shared with a second business object, the shared conversation message can be displayed in the conversation page between the first business object and the second business object, or a message sharing link associated with the shared conversation message can be displayed in the conversation page between the first business object and the second business object. The message sharing link can be used to indicate that when the second business object triggers the message sharing link, a sharing interaction page associated with the second business object is displayed, and the sharing interaction page includes the conversation message displayed in the summary interaction page. The sharing interaction page can be used to display the conversation message in the summary interaction page shared by the first business object, and the display effect of the conversation message in the sharing interaction page is consistent with the display effect of the conversation message in the summary interaction page.

[0217] For example, see Figure 13 , Figure 13 This is a schematic diagram of the effect of a session sharing operation provided by an embodiment of the present application. Figure 13 As shown, in the device of business object A (corresponding to the first business object), page 1201a can be displayed. Page 1201a can be a summary interaction page for determining the summary summary of the document resource, including a conversation message 1202a carrying summary summary 1. When the business object triggers the conversation message 1202a (such as long pressing any area in the conversation message 1202a), function options for the conversation message in the summary interaction page can be displayed (for example, Figure 13 Function options in the area shown in 1203a), such as copy (such as Figure 13 1206a in), excerpt copy (as shown in Figure 13 1207a in), multiple choice (such as Figure 13 1205a in), forwarding (as shown in Figure 13 Furthermore, business object A can select the conversation message to be shared by clicking the function option shown in 1204a. When business object A confirms to share the selected conversation message (i.e., the conversation sharing operation), the conversation message in the summary interaction page can be shared with business object B (i.e., the second business object). Furthermore, the page of the conversation between business object A and business object B can be displayed on the device of business object B (i.e., Figure 13 Page 1208a may include a session message carrying a message shared by business object A to business object B (e.g., a message sharing connection 1). Figure 13 Further, when the business object triggers the message sharing connection 1, a page 1210a may be displayed. The page 1210a may be the above-mentioned sharing interaction page. The page 1210a may include the conversation message carrying the document resource 1 sent by the business object A in the page 1201a (as shown in FIG. Figure 13 1211a in FIG), and the conversation message (as shown in FIG) replied by the business processing object carrying the summary of the document resource 1 Figure 13 Optionally, in the page 1210a, a control for indicating the start of a new session may also be displayed (such as Figure 13 As shown in the control 1213a in FIG), when business object B clicks the control 1213a, a new summary interaction page can be triggered to display, so that business object B can import any document resource in the new summary interaction page to obtain a summary summary of any document resource.

[0218] It is understood that in the embodiment of the present application, it can be used to compare multiple document resources. Then, the document merge comparison information between the multiple document resources can be displayed on the summary interaction page. The merge comparison information can include information indicating the common points between the multiple document resources, or information indicating the differences between the multiple document resources.

[0219] Specifically, the document resources include a first document resource and a second document resource; then, the embodiment of the present application may further include the following steps: when displaying a conversation message carrying a summary summary of the first document resource and a conversation message carrying a summary summary of the second document resource in a summary interaction page, displaying a conversation message carrying document merge and comparison information between the first document resource and the second document resource; the document merge and comparison information is determined based on the content summary of the resource fragment content of the first document resource and the content summary of the resource fragment content of the second document resource.

[0220] Among them, the first document resource and the second document resource can be different document resources. It is understandable that the session message carrying the summary of the first document resource and the session message carrying the summary of the second document resource can be the same session message or different session messages, which is not limited here. In other words, the same session message can include one document resource or multiple document resources (such as the first document resource and the second document resource), which is not limited here. It is understandable that the document merge comparison information is determined based on the content summary of the resource fragment content of the first document resource and the content summary of the resource fragment content of the second document resource, which is not elaborated here.

[0221] It can be understood that the document import operation is a combined import operation for the first document resource and the second document resource; or, the document import operation includes a first import operation for the first document resource and a second import operation for the second document resource.

[0222] The merge import operation may be an operation for importing multiple document resources at once. For example, multiple document resources (such as a first document resource and a second document resource) may be selected at once on the resource management page, and then, in response to a confirmation operation on the selected multiple document resources, the import of the first document resource and the second document resource may be confirmed.

[0223] The first import operation may be an import operation for a first document resource, and the second import operation may be an import operation for a second document resource, which will not be described in detail here.

[0224] Optionally, the first session message is a session message that carries both the first document resource and the second document resource; or, the first session message includes a first sub-session message that carries the first document resource and a second sub-session message that carries the second document resource.

[0225] The first session message may include the merged and imported multiple document resources. Alternatively, the first session message may include a document resource corresponding to each imported document resource, i.e., one document resource corresponds to one session message. The first sub-session message is a session message that carries the first document resource, and the second sub-session message is a session message that carries the second document resource.

[0226] Optionally, the resource content of the document resource can be processed when the summary interaction page is displayed (i.e., processing stops when exiting the page), or it can be processed in the background after exiting the page, which is not limited here. It is understandable that in the process of resource processing of document resources, if the business object triggers the exit of the interaction summary page, a prompt page for indicating whether to exit the page (i.e., the page exit prompt page) can be displayed, and the prompt page will prompt that the task will be terminated after exiting the page (i.e., processing of the document resource will be stopped), or the task will continue in the background, so that the business object needs to confirm twice before exiting the summary interaction page, to avoid the business object accidentally touching the exit page, which will cause the processing of the resource document to fail.

[0227] For example, see Figure 14 , Figure 14 This is a schematic diagram of the effect of a page exit prompt page provided by the embodiment of the present application. Figure 14 As shown, page 1221a can be a prompt page for indicating whether to confirm exiting the page (i.e., page exit prompt page). In this page 1221a, prompt information can be included, such as "Are you sure to leave? The task is in progress. Leaving the current page will terminate the current task." This page 1221a can also include a control for indicating confirmation to exit the summary interaction page (such as Figure 14When the business object clicks the control 1222a, it can leave the summary interaction page; in addition, the page 1221a can also include a control for indicating the continuation of the task (such as Figure 14 As shown in 1223a in the figure), when the business object clicks the control 1223a, the summary interaction page can continue to be displayed and the document resource can continue to be processed.

[0228] S205 . In response to the message triggering operation, display a third conversation message indicated by the message triggering operation on the summary interaction page; the third conversation message includes an interaction instruction initiated for the document resource.

[0229] The message triggering operation may be an operation for triggering an interactive message. The third conversation message may be a conversation message including an interactive instruction initiated with respect to a document resource. The interactive instruction may be an instruction for asking a question about the content of the document resource, or an instruction for specifying a business processing object to output a corresponding reply, which is not limited here.

[0230] S206: Display the fourth conversation message on the summary interaction page; the fourth conversation message includes an interaction reply to the interaction instruction.

[0231] The fourth conversation message may be a conversation message including an interactive reply to the interactive instruction. It is understandable that the fourth conversation message may be understood as a reply from the business processing object to the interactive instruction initiated by the first business object.

[0232] It is understood that after the summary of the document resource is output, the business object can ask further questions based on the summary content (i.e., input interactive instructions), and the model will further generate answers (i.e., interactive replies). It is also understood that the business object can repeatedly modify the summary by sending interactive instructions and ask questions based on the content in the summary.

[0233] It can be understood that the conversation messages displayed in the summary interaction page can be saved, and the conversation messages generated in the same summary interaction page can be saved as a summary interaction dialogue, and then the historical summary interaction dialogue can be displayed in the historical dialogue page. In response to the trigger operation for the target historical summary interaction dialogue, the summary interaction page corresponding to the target historical summary interaction dialogue can be displayed, and the summary interaction page can display the corresponding conversation messages.

[0234] For example, see Figure 15 , Figure 15 This is a schematic diagram of the effect of a historical conversation page provided by an embodiment of the present application. Figure 15As shown, page 1231a may be a history dialogue page, which may include three history summary interaction dialogues, for example, history summary interaction dialogue 1, history summary interaction dialogue 2, and history summary interaction dialogue 3. When the business object triggers the history summary interaction dialogue 1 (such as Figure 15 When 1232a) is reached in the above process, a summary interaction page corresponding to the historical summary interaction dialogue 1 may be displayed. This summary interaction page may display corresponding conversation messages, such as conversation messages carrying document resources involved in the summary interaction page, conversation messages carrying summary summaries of corresponding document resources, conversation messages carrying interaction instructions sent by business objects, and conversation messages responding to interaction instructions, without limitation. It is understood that the historical conversation page may display the document name and corresponding summary summary of each document resource associated with each historical summary interaction dialogue, with the display being truncated if the number of characters exceeds a certain threshold.

[0235] Further, see Figure 16 , Figure 16 This is a flowchart of a document data processing method provided by an embodiment of the present application. The method can be executed by a computer device, for example, the computer device can be Figure 1 The method may include at least the following steps S301 to S308.

[0236] S301: Display a summary interaction page associated with a first business object.

[0237] S302: In response to a document import operation on a document resource, display a first conversation message carrying the document resource sent by the first business object on the summary interaction page.

[0238] S303. In the first associated display area associated with the first conversation message, node progress prompt information of each process processing node in the multiple process processing nodes associated with the document resource is traversed and displayed; the multiple process processing nodes include a summary summary node for summarizing the resource fragment content in the document resource.

[0239] S304: Perform document parsing on the document resource to obtain document text information corresponding to the document content of the document resource.

[0240] Among them, parsing the document can be a process of obtaining the document text information corresponding to the document content of the document resource. The document text information can be the information of all texts associated with the document resource. It can be understood that in an embodiment of the present application, the document text information obtained by parsing can be saved, so that when multiple rounds of dialogue are subsequently conducted through interactive instructions, the saved document text information can be directly used for corresponding processing without the need to re-parse the document resource, which helps to improve the efficiency of the question-and-answer process of the document resource (i.e., question-and-answer through interactive instructions).

[0241] Specifically, performing document parsing processing on a document resource to obtain document text information corresponding to the document content of the document resource may include the following steps: when the document resource includes multimedia data information, performing text extraction processing on the multimedia data information to obtain transferred text information associated with the multimedia data information; the multimedia data information includes one or more of the following: image data, video data, voice data; when the initial text information in the document resource is obtained, determining the document text information corresponding to the document content of the document resource based on the spatial position relationship between the initial text information in the document resource and the transferred text information in the document resource.

[0242] The multimedia data information may include one or more of the following: image data, video data, and voice data.

[0243] The text extraction process can be the process of converting text information from multimedia data. For example, text can be extracted from image data (i.e., OCR); subtitles and text can be extracted from video data; or speech can be converted to text from video data or speech data, etc., without limitation.

[0244] The converted text information may be text information extracted from multimedia data information, and the initial text information may be text information originally existing in a document resource.

[0245] The spatial position relationship can be used to indicate the relationship between the positions of the original text information and the converted text information in the document resource. For example, the document text information of a document resource can be represented as T = {text1, text1, ..., OCR1, OCR2, ..., ASR1}, where text1 and text2 represent the original text information of two document paragraphs, OCR1 and OCR2 represent the converted text information extracted from two images, and ASR1 represents the converted text information extracted from the speech data.

[0246] S305: Perform document segmentation processing on the document text information to obtain resource segmentation content corresponding to the document resource.

[0247] Among them, document segmentation processing can be a processing process for determining the resource segmentation content in the document resource. Among them, as mentioned above, the resource segmentation content in the document resource can be one or more, which is not limited here. It is understandable that document segmentation processing can include but is not limited to slicing by the number of text segmentations, slicing by semantics, slicing by document paragraphs, etc., which is not limited here. It is understandable that after document segmentation processing is performed on the document text information and the resource segmentation content corresponding to the document resource is obtained, the resource segmentation content obtained by the segmentation can be saved, so that if it is necessary to regenerate the summary of the document resource later, the resource segmentation content obtained by the segmentation can be directly used to avoid repeated processing, improve the processing efficiency when regenerating the summary of the document resource, and avoid waste of computing resources.

[0248] Specifically, document fragmentation processing is performed on document text information to obtain resource fragment content corresponding to the document resource, which may include the following steps: when the number of text fragments in the document text information is greater than a slice length threshold, the first text fragment in the document text information is determined as the starting fragment position, and the number of text fragments corresponding to the slice length threshold is searched backward from the starting fragment position, and the first resource fragment content is determined based on the text fragment at the starting fragment position and the found text fragment; when the number of text fragments other than the first resource fragment content in the document text information is greater than the slice length threshold, the starting fragment position is updated based on the next text fragment in the first resource fragment content, and the number of text fragments corresponding to the slice length threshold is searched backward from the updated starting fragment position, and the second resource fragment content is determined based on the text fragment at the updated starting fragment position and the found text fragment; based on the first resource fragment content and the second resource fragment content, the resource fragment content corresponding to the document resource is determined.

[0249] The slice length threshold can be used to indicate the maximum number of text words that can be included in the content of a resource slice, for example, the slice length threshold is 200. It is understandable that the slice length threshold can be a preset fixed value, or can be determined based on the total number of text words corresponding to the document text information, which is not limited here.

[0250] It is understood that the starting segmentation position can be the position of the first text segmentation word used to determine the content of a resource segment. It is understood that during the resource segmentation process of the document resource, the segmentation position can be continuously updated, so as to determine the content of a resource segment based on the starting segmentation position and the number of text segmentations corresponding to the backward search segment length threshold.

[0251] It is understandable that if the text segmentation number corresponding to the slice length threshold found is in a complete sentence, the nearest target symbol (such as a period, question mark or exclamation mark) is searched forward, and the text segmentation between the nearest target symbol and the starting segmentation position is determined as a resource segmentation content. If the text segmentation number corresponding to the slice length threshold found is the last text segmentation of a sentence, the text segmentation number corresponding to the slice length threshold found and the text segmentation between the starting segmentation position are determined as a resource segmentation content.

[0252] It is understandable that the first resource fragment content can be the first resource fragment content in the resource document, and the second resource fragment content can be the next resource fragment content of the first resource fragment content. It is understandable that if the number of text fragments (also called text fragments to be segmented) other than the first resource fragment content and the second resource fragment content in the document text information is greater than the slice length threshold, the remaining text fragments to be segmented can continue to be segmented. The specific process can refer to the relevant description of determining the second resource fragment content, so that a larger number of text fragments can be obtained until the number of remaining text fragments to be segmented is less than or equal to the slice length threshold. The remaining text fragments to be segmented in the document text information are determined as a resource fragment content, and then the segmentation can be stopped. Furthermore, after all the document text information of the document resource is segmented, the resource fragment content obtained by segmentation (such as the first resource fragment content and the second resource fragment content) can be determined as the resource fragment content corresponding to the document resource.

[0253] It is understood that when determining the resource slice content of a document resource, it can also be determined based on document paragraphs. In other words, a document paragraph can be determined as a resource slice content, thereby obtaining at least one resource slice content in the document resource. Optionally, when determining the resource slice content of a document resource, it can also be determined based on document semantics. In other words, the semantic information of each sentence in the document resource can be extracted through the business model, and then based on the semantic information of each sentence, sentences with similar semantic information can be divided into a resource slice content, thereby obtaining at least one resource slice content in the document resource.

[0254] It is understandable that if the number of text words in the document resource is less than or equal to the slice length threshold, all text words in the document resource can be directly determined as one resource slice content.

[0255] It is understood that when the number of text segmentations other than the first resource segmentation content in the document text information is less than or equal to the slice length threshold, the text segmentations other than the first resource segmentation content in the document text information are determined as the third resource segmentation content, and then the resource segmentation content corresponding to the document resource is determined based on the first resource segmentation content and the third resource segmentation content. The third resource segmentation content is the resource segmentation content determined as the text segmentations other than the first resource segmentation content in the document text information under certain conditions.

[0256] S306: Perform summary extraction processing on the resource slice content to obtain a content summary of the resource slice content.

[0257] It is understood that the summary extraction process can be a process for extracting a summary of the resource slice content. The summary extraction process for the resource slice content can call a business model to perform summary extraction, thereby obtaining a content summary of the resource slice content. However, it should be understood that if there are multiple resource slice contents, a content summary for each resource slice content can be determined separately.

[0258] S307: Summarize the content summaries of the resource fragments to obtain a first summary of the document resource.

[0259] Among them, summary aggregation can be a process of aggregating the content summaries of the resource slice contents into a summary summary. It is understandable that if there is only one resource slice content, then summarizing is to determine the content summary of the one resource slice content as the first summary summary. If there are multiple resource slice contents, then summarizing is to aggregate the content summaries of the multiple resource slice contents into a summary summary. It is understandable that when determining the first summary summary, the content summaries of the multiple resource slice contents can be simply combined into one summary to obtain the first summary summary; or, when determining the first summary summary, the content summaries of the multiple resource slice contents can be subjected to semantic feature extraction through the business model to fuse the semantics of the content summaries of the content of each resource slice to generate the first summary summary, which is not limited here.

[0260] Here, we illustrate the document resource processing process with examples. After uploading a document resource, you can perform document parsing on the document resource. This involves parsing the document content, converting the text in the document, including text, text extracted from images (OCR), subtitles and text from videos, and spoken audio from videos, into text, organizing it in spatial order, and sending it to the document slicing step below. Furthermore, the document resource can be sliced ​​(i.e., document fragmentation). Specifically, the text obtained in the document parsing step (i.e., document text information) can be segmented into text segments with a length of no more than L (L = 3000 in the implementation) tokens (text segmentation). The specific segmentation method is: starting from the starting token (the first text segmentation), extract L tokens at a time. If a complete sentence is split, look forward to find the last period, question mark, or exclamation mark. Furthermore, you can extract a summary of the fragments, and use the business model to extract the summary of the text segments (i.e., resource fragment content) in segments. The extraction method is to add instructions to the information input into the business model, requiring the business model to extract content summaries, and specify the word count and summary format of the summary, such as "Summarize the article as concisely and clearly as possible, no more than 100 words, and cover the important information of the article". The instruction information can be automatically generated by the application background, and the specific content can be pre-set by the developer or by the business object, which is not limited here. The business model will summarize each slice according to such instructions and return the main content of each slice (that is, the content summary of the slice content). Furthermore, the summary can be summarized, that is, all summaries (that is, the content summary of each slice content) are summarized together, and then summarized using the business model to obtain the overall summary of the document resource (that is, the summary summary). The specific approach is to add instructions to the information input to the business model, requiring the business model to summarize the content, and specify the number of words and summary format of the summary, such as "Summarize the above content, no more than 300 words, need to include the central idea of ​​the article, and list 3 to 5 important information of the article." This instruction information can be automatically generated by the application background, and its specific content can be pre-set by the developer or by the business object, and is not limited here. Optionally, the business model will summarize the fragmented content (i.e., the summary) again according to such instructions, and return the main content of each document paragraph (i.e., the paragraph summary of the document paragraph mentioned above).

[0261] Optionally, the embodiment of the present application can also be used to merge and compare multiple texts. Specifically, the document resources include a first document resource and a second document resource; then, the embodiment of the present application can also include: performing a topic extraction process on the content summary of the resource fragment content of the first document resource to obtain the content theme of the first document resource; performing a topic extraction process on the content summary of the resource fragment content of the second document resource to obtain the content theme of the second document resource; and determining the document merging and comparison information between the first document resource and the second document resource based on the content theme of the first document resource and the content theme of the second document resource.

[0262] The topic extraction process may be a process of extracting the content topics of a document resource. The content topics may be information extracted from the document resource to represent the key content of the document, for example, the content topics may be health, beauty, fitness, etc.

[0263] It is understandable that the document merge comparison information can be determined by comparing the content themes of each document resource. For example, the document merge comparison information can include common point information between multiple document resources (such as the first document resource and the second document resource), or include difference point information, or can include both common point information and difference point information. For example, the content themes corresponding to document 1 include: health preservation, beauty, and fitness, and the content themes corresponding to document 2 are: health preservation, beauty, and massage. Then the common point information between document 1 and document 2 can include health preservation and beauty, and the difference point information can be fitness and massage.

[0264] For example, see Figure 17 , Figure 17 This is a schematic diagram of the effect of a document content processing process provided by an embodiment of the present application. Figure 17 As shown, you can get the parsed document resources (such as Figure 17 As shown in 1401a in the figure, the parsed document resource can be the document text information corresponding to the document resource. Then, the parsed document resource can be processed into document fragments to obtain multiple resource fragment contents (also called resource fragment contents). For example, resource fragment content 1 (such as Figure 17 1402a in FIG), resource fragment content 2 (such as Figure 17 1403a in), resource fragment content n (as shown in Figure 17 Further, the business model (such as Figure 17 1405a in the figure) extracts the content summary of each resource fragment content, for example, the content summary 1 corresponding to the resource fragment content 1 can be extracted (as shown in FIG. Figure 17 1406a in FIG), content summary 2 corresponding to resource fragment content 2 (as shown in FIG. Figure 171407a in FIG), content summary n corresponding to resource fragment content n (as shown in FIG. Figure 17 Further, the content summaries can be summarized (such as Figure 17 1409a in FIG), thereby inputting a business model based on the summary of each content (such as Figure 17 1405b in the figure), and the integration result is obtained (as shown in Figure 17 Specifically, it may include summarizing the content summaries of the document resources through the business model, thereby obtaining a summary summary of the document resources. At the same time, the content theme of the document resources (such as Figure 17 1411a in the figure), for example, it may include content topics such as Topic 1, Topic 2, Topic 3, etc.

[0265] For further information, see Figure 18 , Figure 18 This is a flowchart of a multi-document comparison process provided by the embodiment of the present application. Figure 18 As shown, multiple document resources can be obtained, such as document resource 1 (such as Figure 18 1501a in ), document resource 2 (as shown in Figure 18 1502a in ), document resource 3 (as shown in Figure 18 Further, each document resource may be processed for content (e.g., Figure 18 1504a in FIG), thereby obtaining the integration result corresponding to each document resource, such as integration result 1 corresponding to document resource 1 (as shown in FIG. Figure 18 1505a in FIG), the integration result 2 corresponding to the document resource 2 (as shown in FIG. Figure 18 1506a in FIG), the integration result 3 corresponding to the document resource 3 (as shown in FIG. Figure 18 Each integration result may include a summary and content theme corresponding to the document resource. For example, integration result 1 may include content theme 1 (e.g., Figure 18 1508a in the figure), which may include summarized topics 1, 2, 3, etc. Further, each integration result may be summarized (as shown in FIG. Figure 18 Specifically, it may include determining document merging comparison information (such as Figure 18 1510a in FIG), wherein the document merge comparison information may include merged information (such as Figure 18 1511a in FIG), the merge information can be used to indicate the common points between the various documents (such as Figure 181513a in FIG), the document merge comparison information may also include comparison information (such as Figure 18 1512a in FIG), the comparison information can be used to indicate the differences between the documents (eg Figure 18 1514a in the figure).

[0266] It is understandable that the embodiments of the present application can be applied to different citation scenarios. After fine-tuning the model and encapsulating the upper-layer product applications for different file scenarios in different industries, more product forms can be derived. Specific application scenarios include but are not limited to the following: financial report assistant: extracting financial report summaries, paper assistant: extracting academic papers, textbook assistant: extracting textbook summaries, novel assistant: extracting novel summaries, extracting novel character relationships, etc. It is understandable that in the embodiments of the present application, the summary of each fragment content (i.e., resource fragment content) can be extracted by first fragmenting, and then the summary corresponding to each fragment content can be summarized. The key information of multiple fragments can be integrated, and the features of the document structure are retained, avoiding the loss of information at different positions of the document caused by directly extracting the summary of the entire document. Therefore, when the embodiments of the present application are used to extract summaries in scenarios of different classifications, their correctness, completeness, simplicity, intuitiveness, and coherence are all excellent. For example, the performance of each application scenario under the evaluation criteria of different dimensions can be reflected with reference to the following table (Table 1).

[0267] Table 1

[0268]

[0269] S308. When the displayed node progress prompt information is the node progress prompt information of the summary summary node, and the node progress prompt information of the summary summary node indicates that a first summary associated with the resource slice content is obtained through the summary, a second session message carrying the first summary is displayed on the summary interaction page.

[0270] It is understandable that in an embodiment of the present application, the first business object can conduct multiple rounds of dialogue in the summary interaction page so that the business object can better understand the content of the document resource. The present application implements support for multiple rounds of question and answer by storing text features of the results of file parsing (such as document text information obtained by parsing document resources, resource fragment content obtained after fragmentation, etc.). When the user enters multiple rounds of questions and answers, the system will perform similarity matching on the interactive instructions entered by the user and the text features of the file parsing results (such as the features corresponding to each sentence), find the content in the file that is most relevant to the user's question, and then summarize it through the model and output the answer. This approach can achieve multiple rounds of questions and answers, and the performance is also very good.

[0271] It is understandable that the embodiments of the present application can support answering subsequent questions based on the content of the document starting from the second round of the conversation, that is, generating interactive replies for interactive instructions initiated by the business object. Among them, in order to achieve this capability, on the one hand, the questions are rewritten in context, and on the other hand, the document needs to be split into paragraphs at a more fine-grained level, or even sentence-level. Specifically, feature extraction can be performed on the context questions, and multiple rounds of rewriting can be performed on the context questions and previous answers to maintain the original meaning of the question and answer content, and finally the rewritten questions are obtained. For example, the summary can be further rewritten by inputting instructions, specifying the number of words in the summary, so as to generate a summary with the number of words specified by the business object. Further, feature extraction can be performed on the rewritten sentences. In addition, document features can be extracted, specifically, feature extraction can be performed on the document content at the sentence granularity, and the correlation between sentences is also modeled, that is, the features corresponding to the sentences include information indicating the position of the sentences in the document resources. During feature extraction, the structural information of the document can also be retained, including page numbers, chapters and other information. In addition, relevant content can be retrieved from document resources. Based on the question (i.e., the interaction instruction), document content can be searched to find documents related to the question, that is, sentences with feature similarity greater than a certain threshold. Furthermore, business models can be used to answer the question based on the relevant questions and the retrieved content, that is, to determine the interaction response.

[0272] Specifically, the embodiments of the present application also include: performing document sentence processing on the document text information to obtain at least one sentence content corresponding to the document resource; performing feature extraction processing on each sentence content in the at least one sentence content to obtain sentence content features corresponding to each sentence content; when obtaining an interaction instruction initiated by the first business object for the document resource, performing feature extraction processing on the interaction instruction to obtain instruction features corresponding to the interaction instruction; when determining the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content, determining an interaction response to the interaction instruction based on the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content.

[0273] Among them, document sentence processing can be a process of segmenting each sentence of a document resource. It is understandable that document sentence processing can be divided based on the identification symbols of the document resource. For example, it can be divided based on identification symbols such as commas, periods, exclamation marks, question marks, etc., which are not limited here. It should be understood that a document resource can include at least one sentence content. In other words, the number of sentence contents in a document resource can be one or more. It is understandable that each sentence content obtained by segmentation can be the same as or different from the method of resource segmentation content in the document resource, which is not limited here.

[0274] The sentence content feature may be a feature used to characterize the content of a sentence, and the sentence content feature may be represented as a feature vector or a feature matrix, which is not limited here. It should be understood that the feature extraction process may be performed by calling a business model.

[0275] It can be understood that the instruction feature can be a feature used to characterize an interactive instruction. The instruction feature can be represented as a feature vector or a feature matrix, which is not limited here.

[0276] It is understood that feature similarity is used to characterize the similarity between an instruction feature and a sentence content feature. It should be understood that a feature similarity can be determined based on the distance (such as Euclidean distance) between the instruction feature and the sentence content feature. The smaller the distance, the greater the feature similarity, indicating that the similarity between the instruction feature and the sentence content feature is greater; conversely, the larger the distance, the smaller the feature similarity, indicating that the similarity between the instruction feature and the sentence content feature is less.

[0277] It is understandable that when determining the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content, different methods can be used to determine the interactive response to the interactive instruction based on the size of the feature similarity.

[0278] Specifically, determining the interactive response to the interactive instruction based on the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content can include the following steps: when sentence content with a feature similarity greater than a first similarity threshold is found in at least one sentence content, determining the interactive response to the interactive instruction based on the sentence content with the feature similarity greater than the first similarity threshold found; when sentence content with a feature similarity greater than the first similarity threshold is not found in at least one sentence content, if sentence content with a feature similarity less than the first similarity threshold and greater than the second similarity threshold is found, calling the search component to search for content associated with the interactive instruction, and determining the interactive response to the interactive instruction based on the searched associated content.

[0279] Among them, the first similarity threshold can be the minimum feature similarity required for determining an interactive reply based on the content in the document resource. In other words, when a sentence content with a feature similarity greater than the first similarity threshold is found in at least one sentence content, the interactive reply can be determined based on the content in the document resource.

[0280] The second similarity threshold may be the minimum feature similarity required when determining an interactive reply based on the search component. In other words, if a sentence content is found whose feature similarity is less than the first similarity threshold and greater than the second similarity threshold, the interactive reply may be determined based on the search component. The search component may be a component for searching for information, for example, the search component may be some search engine. The associated content may be content related to the interactive instruction searched by calling the search component, and the interactive reply may be determined based on the associated content.

[0281] Optionally, if no sentence content with a feature similarity greater than a second similarity threshold is found in at least one sentence content, a prompt message may be displayed to indicate that a reply cannot be made.

[0282] It can be understood that the embodiments of the present application can greatly improve the utilization of model resources through a scheduling method, that is, the concurrency of the model can be increased by allowing prompts (i.e., information input into the business model) with similar time to access the same model service; and ensure that the newly arrived prompt always accesses the available model resources when there are available model resources, avoiding the user's poor experience due to the blockage when the prompt accesses the resource being inferred.

[0283] Optionally, this application will conduct real-time security audits on documents and extracted summary summaries to prevent business objects from maliciously guiding the model to generate sensitive and harmful information. For example, it can detect whether there are sensitive words, harmful information, etc. in document resources and summary summaries. It is understandable that if there are sensitive words or harmful information in the document resource or summary summary, it means that the document resource or summary summary has failed the audit, and a corresponding prompt message is output to indicate that there are sensitive words or harmful information in the document resource or summary summary. For example, after the document resource uploaded by the business object is parsed, harmful information or sensitive words are detected based on the document text information of the document resource. If harmful information or sensitive words are detected in the document text information, a prompt message is output to indicate that there is harmful information or sensitive words in the document resource, and the resource content of the document resource is not subsequently processed. For another example, after the business model generates a summary summary, harmful information or sensitive words are detected based on the summary summary. If harmful information or sensitive words are detected in the summary summary, a prompt message is output to indicate that there is harmful information or sensitive words in the summary summary, and the summary containing harmful information or sensitive words is no longer output.

[0284] See Figure 19 , Figure 19 This is a flowchart of a dialogue reply process provided by an embodiment of the present application. Figure 19 As shown, after obtaining the parsed document resource 1 (such as Figure 191601a in FIG), that is, the document text information of the document resource 1, and then multiple sentences in the parsed document resource 1 can be determined, such as sentence 1 (such as Figure 19 1602a in), clause 2 (as shown in Figure 19 1603a in), clause m (as shown in Figure 19 Further, each sentence can be encoded by the encoder 1605a (ie, feature extraction processing) to obtain the sentence content features corresponding to each sentence. For example, feature 1 corresponding to sentence 1 can be obtained (eg, Figure 19 1606a in ), feature 2 corresponding to clause 2 (such as Figure 19 1607a in ), feature m corresponding to clause m (such as Figure 19 1608a in the figure).

[0285] After obtaining the interaction instruction initiated by the first business object (such as Figure 19 1609a in FIG), the interactive instruction can be encoded by the encoder 1610a (ie, feature extraction processing) to obtain the instruction feature corresponding to the interactive instruction, namely, feature q (as shown in FIG). Figure 19 Then, the feature similarities corresponding to each sentence content feature (i.e., feature 1, feature 2, feature m, etc.) and the instruction feature (i.e., feature q) can be used to filter out sentence content features whose feature similarities meet certain conditions. For example, feature 2 can be filtered out. Then, based on feature 2, the search component (e.g., Figure 19 1612a in FIG), and feature q is input into the business model (as shown in FIG. Figure 19 1613a in FIG), to determine the instruction response to the interactive instruction based on the business model (such as Figure 19 1614a in the text).

[0286] Here, with the help of diagrams, we explain the processing flow of the document resource business. Figure 20 , Figure 20 This is a schematic diagram of the effect of the processing flow of a document resource service provided by the embodiment of this application. Figure 20 As shown in the figure, in the process of document resource business processing, service nodes such as front-end page, engineering service, parsing service, pre-processing and model service may be involved. Among them, in the summary interaction page, the first round of dialogue can be carried out. In the first round of dialogue, the business object can import document resources on the front-end page (such as Figure 20 As shown in 1701a in ), the engineering service can then perform classification analysis through intent recognition (such as Figure 20Specifically, the corresponding parsing method can be selected according to the format of the document resource, which will not be described in detail here. Further, the parsing service can determine the text or convert the text from the document resource (such as Figure 20 ), such as directly extracting the original text, or extracting the converted text from the image, video, voice and other data, or, if the document resource is a web page, obtaining the web page content (such as Figure 20 ), determine text or convert text from the web page content. Further, the document resource can be segmented. Specifically, it can be determined whether the number of remaining text segmentations (i.e., the number of text segmentations to be segmented) is greater than a threshold (e.g., Figure 20 As shown in 1705a in FIG), the remaining text segmentation may be a text segmentation in which the fragmentation content is not determined in the document text information of the document resource, and the remaining text segmentation may be subjected to content slicing processing (as shown in FIG. Figure 20 1706a in FIG), until the fragment content is determined for all the document text information of the document resource, and then the fragment content obtained by slicing can be saved, that is, the cached parsed content (such as Figure 20 Furthermore, the model service can be used to determine whether the number of summaries is greater than 1 (e.g. Figure 20 If it is greater than 1, the content summary of the fragment content such as summary 1, summary 2, ..., summary n can be determined (as shown in 1707a in FIG). Figure 20 1708a in FIG), and then the content summaries of the fragmented contents such as summary 1, summary 2, ..., summary n can be summarized (as shown in FIG). Figure 20 1708b in FIG), further, the summary summary obtained by the summary can be cached, that is, the summary summary can be cached (as shown in FIG. Figure 20 1710a in FIG), further, the summary can be displayed (as shown in FIG. Figure 20 (as shown in 1711a in the figure). When caching the summary and parsing the content, a unique resource identifier for the document resource can be generated, and the parsed document resource, the summary, and the conversation identifier can be associated and stored. The conversation identifier can be a unique identifier for the summary interaction conversation corresponding to a summary interaction page.

[0287] Furthermore, business objects can conduct multiple rounds of dialogue on the summary interaction page, for example, they can upload new document resources (such as Figure 20 The processing process for the newly uploaded document resources can refer to the processing process for the document resources uploaded in the first round of dialogue. Further, context questions (such as Figure 201713a in ), i.e., interactive instructions. Further, the model service can perform semantic parsing and keyword extraction (e.g. Figure 20 As shown in 1714a in FIG, the instruction features of the interactive instruction and the sentence content features of the sentence content in the document resource can be determined. Then, the retrieval object and quantity (such as Figure 20 1715a in the figure), that is, the feature similarity is calculated by the instruction feature of the interactive instruction and the sentence feature of each sentence, so that the sentence with feature similarity meeting a certain threshold is determined as the fixed search object, and its number is counted. Further, the engineering service can retrieve the document history source content (such as Figure 20 1716a in ), that is, it is retrieved from the parsed content of the cached document resource, and then the model service replies through the business model (such as Figure 20 1717a in FIG), that is, generating an interactive response corresponding to the interactive instruction, and then displaying the content (as shown in FIG. Figure 20 1718a in the figure), that is, outputting the conversation message carrying the interaction reply in the summary interaction page.

[0288] For further information, see Figure 21 , Figure 21 This is a flowchart of a document data processing method provided by an embodiment of the present application. The method can be executed by a computer device, for example, the computer device can be Figure 1 The method may include at least the following steps S401 to S403.

[0289] S401: Display a summary interaction page.

[0290] The method for displaying the summary interaction page may refer to the relevant description in the above step S101 and will not be elaborated here.

[0291] S402: When a first conversation message carrying a document resource is displayed on the summary interaction page, a document analysis is performed on the document resource to obtain a document directory of the document resource; and a paragraph analysis identifier in the document directory is associated with a document paragraph in the document resource.

[0292] It can be understood that the first conversation message carrying the document resource is displayed on the summary interaction page. This may be in response to the document import operation of the first business object on the document resource, and the first conversation message carrying the document resource is displayed on the summary interaction page. For a specific introduction to the first conversation message, please refer to the relevant description in the above step S102, which will not be repeated here.

[0293] It should be understood that the document directory may refer to a directory of document paragraphs in a document resource. The document directory is determined by performing document analysis on the document resource, that is, by summarizing and aggregating the paragraph summaries (also called paragraph abstracts or paragraph titles) corresponding to the document paragraphs in the document resource.

[0294] It is understandable that the number of document paragraphs in a document resource can be one or more, which is not limited here. The document paragraph can be determined based on the symbols in the document content in the document resource. For example, the document paragraph in the document resource can be determined by a paragraph marking symbol, or the document paragraph in the document resource can be determined based on a period in the document resource. The paragraph marking symbol can be a symbol used to indicate the end of a paragraph, then all the paragraph marking symbols in the document resource can be obtained, and the document paragraphs can be divided at the position of each paragraph marking symbol, thereby obtaining at least one document paragraph of the document resource. It should be understood that the method of dividing the document paragraphs in the document resource can be the same as or different from the method of dividing the resource fragment content, which is not limited here. It is understandable that the document paragraph in the document resource can be a type of resource fragment content.

[0295] It is understandable that the document directory may generally include a paragraph summary of a document paragraph (also known as a paragraph title, paragraph summary), the page number corresponding to the document paragraph (i.e., the document page number), a number (such as 1, 2, 3, etc., to indicate which document paragraph the document paragraph is), and the like. The paragraph summary may be a summary description determined for a document paragraph, and may also be called a paragraph title. The paragraph summary of a document paragraph may be determined based on the paragraph content of the document paragraph. It is understandable that the paragraph summary of each document paragraph may be determined by calling a business model, and the description of the business model may refer to the above-mentioned related description, which will not be elaborated here. The document resource may include at least one document page, and each document page may be associated with a corresponding document page number, such as page 1, page 2, and the like. Each document page is used to record the corresponding document content, and a document paragraph may be recorded on a document page. The document page number corresponding to the document page where the document paragraph is located is the document page number corresponding to the document paragraph. A document paragraph may also span multiple document pages, and the document page where the starting position or the ending position of the document paragraph is located may be preset as the document page corresponding to the document paragraph.

[0296] The document directory may also include a paragraph analysis identifier, which may be an identifier for indicating a jump to a document paragraph. The paragraph analysis identifier may be a document page number or paragraph summary in the document directory, or may be prompt information for indicating a jump to a document paragraph, which is not limited here. It is understood that the paragraph analysis identifier is associated with the document paragraph. For example, the paragraph analysis identifier may be bound to a jump instruction for triggering a jump to a document paragraph, and then when a business object triggers the paragraph analysis identifier, the document paragraph may be jumped to.

[0297] It should be understood that the process of determining the document directory of a document resource can refer to the description of determining the summary of a document resource described above. Specifically, the document resource can be divided into paragraphs to obtain document paragraphs in the document resource; the business model can be invoked to perform paragraph summarization on each document paragraph to obtain a paragraph summary of each document paragraph; and summarization based on the paragraph summaries of each document paragraph can be performed to obtain the document directory of the document resource.

[0298] Among them, the paragraph summary processing can be a related description for determining the paragraph summary (also known as paragraph title, paragraph abstract) of the document paragraph. It can be understood that the processing process of the paragraph summary processing can be a process of determining the content summary of the fragmented content. Among them, summarizing and aggregating based on the paragraph summaries of each document paragraph can be used to indicate the processing process of integrating the paragraph summaries of each document paragraph into a document directory. When determining the document directory based on the paragraph summaries of each document paragraph, the directory items corresponding to each document paragraph can be determined based on the paragraph summaries of each document paragraph and the document page numbers corresponding to each document paragraph, and then the document directory of the document resource can be determined based on the directory items corresponding to each document paragraph.

[0299] It should be understood that in an embodiment of the present application, after importing a document resource, the node progress prompt information of each process processing node in the multiple process processing nodes associated with the document resource can be traversed and displayed in the first associated display area associated with the first session message; the multiple process processing nodes include a directory summary node for determining the document directory; when the displayed node progress prompt information is the node progress prompt information of the directory summary node, and the node progress prompt information of the directory summary node indicates that the document directory associated with the document resource is summarized, a second session message carrying the document directory is displayed on the summary interaction page.

[0300] Among them, the multiple process processing nodes here are process processing nodes for determining the document directory of the document resource. The directory summary node can be a process processing node for determining the document directory of the document resource based on the paragraph summary of each document paragraph. It can be understood that the multiple process processing nodes for determining the document directory can also include the above-mentioned document upload node for uploading the document resource (specifically, refer to the relevant description of the above-mentioned document upload node), the content parsing node for parsing the document resource (specifically, refer to the relevant description of the above-mentioned content parsing node), the paragraph division node for determining at least one document paragraph in the document resource (the method for dividing at least one document paragraph can refer to the relevant description above), and the paragraph summary node for generating a paragraph summary of each document paragraph (the method for generating a paragraph summary of each document paragraph can refer to the relevant description above). Among them, the multiple process processing nodes have a certain process node sequence, and the node progress prompt information can be displayed in the process node sequence, such as the process node sequence can be document upload node-content parsing node-paragraph division node-paragraph summary node-directory summary node. It is understandable that the display method for each process processing node can refer to the above description of traversing and displaying node progress prompt information in the process of determining the summary of document resources, which is not repeated here.

[0301] S403: When the second conversation message carrying the document directory is displayed on the summary interaction page, in response to a triggering operation on a paragraph analysis identifier in the document directory, jump to display the document paragraph corresponding to the paragraph analysis identifier.

[0302] Among them, the triggering operation for the paragraph analysis identifier in the document directory can be an operation for triggering a jump to display the document paragraph corresponding to the paragraph analysis identifier. For example, the triggering operation can be a touch operation for the paragraph analysis identifier, such as clicking on the paragraph analysis identifier. It should be understood that when a document resource includes multiple document paragraphs, the document directory includes paragraph analysis identifiers corresponding to the multiple document paragraphs, and then in response to a target paragraph analysis identifier among the paragraph analysis identifiers corresponding to the multiple document paragraphs, the document paragraph corresponding to the target paragraph analysis identifier can be jumped to display, and the target paragraph analysis identifier can be the paragraph analysis identifier required to be triggered.

[0303] As mentioned above, if the paragraph analysis identifier can be a document page number in the document directory, the trigger operation for the paragraph analysis identifier can be a trigger operation for the document page number in the document directory. Similarly, if the paragraph analysis identifier can be a paragraph summary in the document directory, the trigger operation for the paragraph analysis identifier can be a trigger operation for the paragraph summary in the document directory. Similarly, if the paragraph analysis identifier can be a prompt message in the document directory for indicating a jump to display a document paragraph, the trigger operation for the paragraph analysis identifier can be a trigger operation for the prompt message in the document directory for indicating a jump to display a document paragraph.

[0304] Among them, the paragraph analysis mark displayed in the document directory can be distinguished by special display effects (such as special font color, font size, font weight, underline, etc.) to prompt the business object that the paragraph analysis mark is triggerable.

[0305] See Figure 22 , Figure 22 This is a schematic diagram of the display effect of the document directory provided by the embodiment of this application. Figure 22 As shown, in the conversation message, when displaying the document directory, it can be displayed in at least the following ways. For example, page 2210a can be a summary interactive page that displays a document directory. In this page 2210a, the conversation message carrying the document directory can be shown in 2212a. In this document directory, directory items corresponding to paragraphs 1, 2, and 3 can be included. For example, the directory item for paragraph 1 can be "①, Paragraph 1 Title... [Page 1]", where "①" identifies the serial number of paragraph 1, "Paragraph 1 Title" identifies the paragraph summary of paragraph 1 (also called paragraph title), and "Page 1" indicates the document page number where paragraph 1 is located. The introduction of the directory items of other paragraphs refers to the relevant description of the directory item of paragraph 1, which will not be repeated here. It can be seen that, as Figure 22 As shown in 2211a, the document page number is distinguished from other text by underscores and "[ ]", so as to prompt the business object that the document page number can trigger a jump to display the corresponding document paragraph, that is, the paragraph analysis is identified as the document page number, and then the business object clicks the document page number indicated by 2211a, which can trigger a jump to display document paragraph 1.

[0306] For another example, page 2220a may be another summary interactive page that displays a document directory. In this page 2220a, the conversation message of the document directory can be seen as shown in 2222a. In this document directory, the directory items corresponding to paragraph 1, paragraph 2, and paragraph 3 may be included. The introduction of the directory items of each paragraph refers to the relevant description of the directory items in the conversation message shown in 2212a above, which will not be repeated here. It can be seen that, as Figure 22As shown in 2221a, the paragraph title is distinguished from other text by underscores and "[ ]" to prompt the business object that the paragraph title can trigger a jump to display the corresponding document paragraph, that is, the paragraph analysis is marked as a paragraph summary, and then the business object clicks the paragraph 1 title indicated by 2221a to trigger a jump to display document paragraph 1.

[0307] For another example, page 2230a may be another summary interactive page that displays a document directory. In this page 2230a, the conversation message of the document directory can be shown in 2232a. In this document directory, directory items corresponding to paragraphs 1, 2, and 3 may be included. For example, the directory item for paragraph 1 may be "①, Paragraph 1 Title... Page 1 [Jump to Paragraph 1]", where [Jump to Paragraph 1] may be a text prompt information for indicating the jump to the displayed document paragraph. It can be seen that, as Figure 22 As shown in 2231a, the text prompt information for indicating the jump to display the document paragraph is distinguished from other text by underline and "[ ]", so as to prompt the business object that the text prompt information can trigger the jump to display the corresponding document paragraph, that is, the paragraph analysis is identified as the text prompt information for indicating the jump to display the document paragraph, and then the business object clicks the text prompt information for indicating the jump to display the document paragraph indicated by 2231a, which can trigger the jump to display document paragraph 1.

[0308] It is understood that when jumping to display the target document section, a page for displaying the document resource can be opened, and the target document section can be displayed on the opened page. Alternatively, a pop-up window can be displayed on the summary interaction page to display the target document section, or a conversation message containing the target document section can be summarized, etc., which are not limited here.

[0309] The following describes a method for jumping to and displaying the document paragraph corresponding to the paragraph analysis identifier, taking the paragraph analysis identifier as an example of a document page number.

[0310] Optionally, the paragraph analysis identifier includes a document page number; then, in response to a trigger operation on the paragraph analysis identifier in the document directory, jumping to display the document paragraph corresponding to the paragraph analysis identifier may include the following steps: in response to a trigger operation on the document page number in the document directory, displaying a document resource display page associated with the document resource, and displaying the document paragraph corresponding to the document page number in the document resource display page.

[0311] The triggering operation for the document page number in the document directory may be a touch operation on the document page number in the document directory, such as single-clicking, double-clicking, etc. It should be understood that, as described above, the document page number may be associated with an instruction for jumping to display a corresponding document paragraph, and thus, when the document page number is triggered, the corresponding document paragraph may be jumped to display.

[0312] The document resource display page may be a page for displaying document resources. In other words, a business object can view document resources on the document resource display page. It is understood that a document resource may include at least one document page, with each document paragraph displayed on a corresponding document page. When a jump to displaying a document paragraph is triggered, the corresponding document paragraph may be displayed on the triggered document resource display page. In other words, the content displayed on the document resource display page is the corresponding document paragraph.

[0313] For example, see Figure 23 , Figure 23 This is a schematic diagram of a document paragraph jump process provided by an embodiment of the present application. Figure 23 As shown, page 2310a can be a summary interaction page showing a conversation message with a document directory. Among them, the document page number corresponding to document paragraph 1 is page 1 (such as Figure 23 ), the document page number is distinguished from other characters by "[ ]" and underscores to indicate that the document page number is associated with an instruction for jumping to the corresponding document paragraph. When the business object clicks the document page number 2311a, page 2320a may be displayed. Page 2320a may be a document resource display page. The document page where paragraph 1 is located, i.e., page 1 (as shown in FIG. Figure 23 2322a in ), that is, displaying paragraph 1 in the document resource display page (as shown in Figure 23 2321a in the figure).

[0314] Optionally, the paragraph analysis identifier includes a document page number; then, in response to a trigger operation on the paragraph analysis identifier in the document directory, jumping to display the document paragraph corresponding to the paragraph analysis identifier may include the following steps: in response to a trigger operation on the document page number in the document directory, displaying the document paragraph corresponding to the document page number in the summary interaction page.

[0315] It can be understood that, as mentioned above, when displaying a document paragraph in the summary interaction page, a pop-up window can be displayed in the summary interaction page to display the document paragraph to be jumped to, or a conversation message carrying the document paragraph corresponding to the paragraph analysis identifier can be displayed in the summary interaction page.

[0316] For example, see Figure 24 , Figure 24 This is a schematic diagram of a document paragraph jump process provided by an embodiment of the present application. Figure 24 As shown, page 2410a can be a summary interaction page showing a conversation message with a document directory. Among them, the document page number corresponding to document paragraph 1 is page 1 (such as Figure 24When the business object clicks the document page number 2411a, a page 2420a may be displayed. The page 2420a may be a pop-up window (such as Figure 24 2421a in the summary interaction page), wherein the pop-up window 2421a is used to display the content of document paragraph 1.

[0317] For example, see Figure 25 , Figure 25 This is a schematic diagram of a document paragraph jump process provided by an embodiment of the present application. Figure 25 As shown, page 2510a can be a summary interaction page showing a conversation message with a document directory. Among them, the document page number corresponding to document paragraph 1 is page 1 (such as Figure 25 When the business object clicks the document page number 2511a, a page 2520a may be displayed. The page 2520a may be a conversation message containing paragraph 1 (as shown in FIG. Figure 25 2521a in the summary interaction page), where the conversation message 2521a is used to display the content of document paragraph 1.

[0318] Optionally, in an embodiment of the present application, after importing a document resource, only a summary of the document resource may be determined, and a second conversation message carrying the summary of the document resource may be displayed; or, after importing a document resource, only a document directory of the document resource may be determined, and a second conversation message carrying the document directory of the document resource may be displayed. Alternatively, after importing a document resource, both a summary of the document resource and a document directory may be determined, and a second conversation message carrying the summary of the document resource and the document directory may be displayed. This second conversation message may be a single conversation message carrying both the summary and the document directory, or may be two conversation messages carrying the summary and the document directory, respectively. This is not limited here.

[0319] It is understood that when displaying a conversation message carrying a document directory, a control for instructing to regenerate the document directory may also be displayed in the associated display area of ​​the conversation message. The document directory may then be regenerated in response to a touch operation on the control for instructing to regenerate the document directory. It is understood that the method for displaying the regenerated document directory can refer to the above description of displaying the regenerated summary, and will not be elaborated on here.

[0320] See Figure 26 , Figure 26 This is a schematic diagram of the structure of a document data processing device provided by an embodiment of the present application. Figure 26 As shown, the document data processing apparatus 1 may be a computer device (eg, Figure 1A computer program (including program code) of the terminal device 11a in the embodiment of the present application, for example, the document data processing device 1 is an application software; it can be understood that the document data processing device 1 can be used to execute the corresponding steps of the document data processing method provided in the embodiment of the present application. Figure 26 As shown, the document data processing device 1 may include: a first interactive page display module 11, a document import module 12, a progress display module 13, and a summary output module 14;

[0321] A first interaction page display module 11 is configured to display a summary interaction page associated with the first business object;

[0322] A document import module 12 is configured to display, in response to a document import operation on a document resource, a first conversation message carrying the document resource sent by the first business object on a summary interaction page;

[0323] A progress display module 13 is configured to traverse and display, in a first associated display area associated with the first conversation message, node progress prompt information of each process processing node among a plurality of process processing nodes associated with the document resource; the plurality of process processing nodes includes a summary node for summarizing resource fragment contents in the document resource;

[0324] The summary output module 14 is used to display a second session message carrying the first summary summary on the summary interaction page when the node progress prompt information displayed is the node progress prompt information of the summary summary node, and the node progress prompt information of the summary summary node indicates that the first summary summary associated with the resource slice content is obtained by summary.

[0325] The document data processing device 1 includes: an instruction interaction module 15;

[0326] The instruction interaction module 15 is used to:

[0327] In response to the message triggering operation, displaying a third conversation message indicated by the message triggering operation on the summary interaction page; the third conversation message includes an interaction instruction initiated for the document resource;

[0328] The fourth conversation message is displayed on the summary interaction page; the fourth conversation message includes an interaction reply to the interaction instruction.

[0329] Wherein, any process processing node among the multiple process processing nodes is a target process processing node;

[0330] The progress display module 13 is used to:

[0331] When the node progress prompt information displayed is the node progress prompt information of the target process processing node, the node progress prompt information of the target process processing node is dynamically displayed; the node progress prompt information of the target process processing node changes as the node processing progress when the target process processing node processes the document resources changes.

[0332] The node progress prompt information of the target process processing node is dynamically displayed, including one or more of the following:

[0333] When the node progress prompt information of the target process processing node includes the time progress prompt information of the target process processing node, dynamically display the time progress prompt information of the target process processing node; or

[0334] When the node progress prompt information of the target process processing node includes the numerical progress prompt information of the target process processing node, the time progress prompt information of the target process processing node is dynamically displayed.

[0335] The progress display module 13 further includes: a node viewing module 16;

[0336] Among them, the node viewing module 16 is used to respond to the node viewing operation for multiple process processing nodes, and display the process node prompt information corresponding to multiple process processing nodes in the first associated display area; one process processing node corresponds to one process node prompt information; the process node prompt information of any process processing node includes node status information for prompting the node processing status of any process processing node.

[0337] Wherein, the document resource includes at least one document paragraph;

[0338] The summary output module 14 is specifically configured to:

[0339] In the second conversation message, a document directory of the document resource is displayed; the document directory includes a paragraph summary of at least one document paragraph, and the paragraph summary of each document paragraph is determined based on the paragraph content of each document paragraph.

[0340] The document data processing device 1 further includes: a first state switching module 17;

[0341] The first state switching module 17 is specifically configured to:

[0342] When the node progress prompt information displayed in the first associated display area is the node progress prompt information of the document upload node, the first status prompt information is displayed in the document processing status control; the first status prompt information is used to indicate that the document processing status of the document resource is the document upload status;

[0343] In response to the first state switching operation on the document processing state control, the document processing state of the document resource is adjusted from the document uploading state to the upload canceling state, and a first state prompting conversation message is output to prompt that the document processing state of the document resource is the upload canceling state.

[0344] Among them, the summary interaction page includes a document processing status control, which is used to adjust the document processing status of the document resource; the multiple process processing nodes also include a document content processing node, which includes a summary summary node, a content parsing node for determining the document content of the document resource, a content fragmentation node for determining the resource fragment content corresponding to the document resource, and a fragment summary extraction node for determining the content summary of the resource fragment content; the document processing status includes a document content processing status, which is the document processing status determined when the document resource is processed by the document content processing node;

[0345] The document data processing device 1 further includes: a second state switching module 18;

[0346] The second state switching module 18 is specifically configured to:

[0347] When the node progress prompt information displayed in the first associated display area is the node progress prompt information of the document content processing node, second status prompt information is displayed in the document processing status control; the second status prompt information is used to indicate that the document processing status of the document resource is the document content processing status;

[0348] In response to the second state switching operation on the document processing state control, the document processing state of the document resource is adjusted from the document content processing state to the content processing cancellation state, and a second state prompting session message is output to prompt that the document processing state is the content processing cancellation state.

[0349] The document data processing device 1 further includes: a summary re-extraction module 19;

[0350] The summary re-extraction module 19 is specifically used for:

[0351] When displaying the second session message on the summary interaction page, displaying a summary re-extraction control for instructing to regenerate the summary summary of the document resource;

[0352] In response to a triggering operation on the summary re-extraction control, a second summary regenerated for the document resource is displayed in the second conversation message.

[0353] The document data processing device 1 further includes: a summary switching module 20;

[0354] The summary switching module 20 is specifically configured to:

[0355] Displaying a summary switching control for a summary of the document resource in a second associated display area associated with the second conversation message;

[0356] In response to a triggering operation on the summary switching control, the first summary and the second summary are switched and displayed in the second conversation message.

[0357] The document data processing device 1 further includes: a conversation sharing module 21;

[0358] The session sharing module 21 is used to share the session messages in the summary interaction page to the second business object in response to the session sharing operation in the summary interaction page, so that the second business object triggers the display of the sharing interaction page associated with the second business object through the message sharing link associated with the session sharing operation, and the sharing interaction page includes the session messages displayed in the summary interaction page.

[0359] The document resources include a first document resource and a second document resource;

[0360] Wherein, the document data processing device 1 further includes: a document comparison module 22;

[0361] The document comparison module 22 is used to display a conversation message carrying a summary summary of the first document resource and a conversation message carrying a summary summary of the second document resource in a summary interaction page, and to display a conversation message carrying document merge comparison information between the first document resource and the second document resource; the document merge comparison information is determined based on the content summary of the resource fragment content of the first document resource and the content summary of the resource fragment content of the second document resource.

[0362] The document import operation is a combined import operation for the first document resource and the second document resource; or

[0363] The document import operation includes a first import operation for a first document resource and a second import operation for a second document resource.

[0364] The first conversation message is a conversation message that carries both the first document resource and the second document resource; or

[0365] The first session message includes a first sub-session message carrying a first document resource and a second sub-session message carrying a second document resource.

[0366] The document data processing device 1 further includes: a document content processing module 23;

[0367] The document content processing module 23 is specifically used to:

[0368] Perform document parsing on the document resource to obtain the document text information corresponding to the document content of the document resource;

[0369] Perform document fragmentation processing on the document text information to obtain the resource fragment content corresponding to the document resource;

[0370] Performing summary extraction on the resource slice content to obtain a content summary of the resource slice content;

[0371] The content summaries of the resource fragments are summarized to obtain a first summary of the document resource.

[0372] The document content processing module 23 includes: a text parsing unit 231;

[0373] The text parsing unit 231 is specifically configured to:

[0374] When the document resource includes multimedia data information, performing text extraction processing on the multimedia data information to obtain translated text information associated with the multimedia data information; the multimedia data information includes one or more of the following: image data, video data, and voice data;

[0375] When the initial text information in the document resource is obtained, the document text information corresponding to the document content of the document resource is determined based on the spatial position relationship between the initial text information and the transferred text information in the document resource.

[0376] The document content processing module 23 includes: a text segmentation unit 232;

[0377] The text segmentation unit 232 is specifically configured to:

[0378] When the number of text segmentations in the document text information is greater than the slice length threshold, the first text segmentation in the document text information is determined as the starting segmentation position, and text segmentations corresponding to the slice length threshold are searched backward from the starting segmentation position, and the content of the first resource slice is determined based on the text segmentation at the starting segmentation position and the found text segmentations;

[0379] When the number of text segmentations other than the first resource fragment content in the document text information is greater than the slice length threshold, the starting segmentation position is updated based on the next text segmentation in the first resource fragment content, and text segmentations corresponding to the number of segmentations corresponding to the slice length threshold are searched backward from the updated starting segmentation position, and the second resource fragment content is determined based on the text segmentation at the updated starting segmentation position and the searched text segmentation;

[0380] The resource segment content corresponding to the document resource is determined based on the first resource segment content and the second resource segment content.

[0381] The document resources include a first document resource and a second document resource;

[0382] The document content processing module 23 further includes: a document comparison unit 233;

[0383] The document comparison unit 233 is specifically used to:

[0384] Performing topic extraction processing on the content summary of the resource fragment content of the first document resource to obtain the content topic of the first document resource;

[0385] Performing topic extraction processing on the content summary of the resource fragment content of the second document resource to obtain the content topic of the second document resource;

[0386] Based on the content theme of the first document resource and the content theme of the second document resource, document merging and comparison information between the first document resource and the second document resource is determined.

[0387] The document content processing module 23 further includes: an instruction interaction unit 234;

[0388] The instruction interaction unit 234 is used to:

[0389] Performing document sentence processing on the document text information to obtain at least one sentence content corresponding to the document resource;

[0390] Performing feature extraction processing on each sentence content in at least one sentence content to obtain sentence content features corresponding to each sentence content;

[0391] When an interaction instruction initiated by the first business object for the document resource is obtained, feature extraction processing is performed on the interaction instruction to obtain instruction features corresponding to the interaction instruction;

[0392] When determining the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content, an interactive response to the interactive instruction is determined based on the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content.

[0393] The instruction interaction unit 234 is further configured to:

[0394] When a sentence content having a feature similarity greater than a first similarity threshold is found in at least one sentence content, determining an interactive reply to the interactive instruction based on the found sentence content having a feature similarity greater than the first similarity threshold;

[0395] When no sentence content with a feature similarity greater than the first similarity threshold is found in at least one sentence content, if a sentence content with a feature similarity less than the first similarity threshold and greater than the second similarity threshold is found, the search component is called to search for content associated with the interaction instruction, and an interactive response to the interaction instruction is determined based on the searched associated content.

[0396] See Figure 27 , Figure 27 This is a schematic diagram of the structure of another document data processing device provided by the embodiment of the present application. Figure 27 As shown, the document data processing device 2 may be a computer device (eg, Figure 1 A computer program (including program code) of the terminal device 11a in the embodiment of the present invention, for example, the document data processing device 2 is an application software; it can be understood that the document data processing device 2 can be used to execute the corresponding steps of the document data processing method provided in the embodiment of the present application. Figure 27 As shown, the document data processing device 2 may include: a second interactive page display module 31, a directory acquisition module 32, and a paragraph jump module 33;

[0397] The second interaction page display module 31 is used to display the summary interaction page;

[0398] The directory acquisition module 32 is configured to, when a first conversation message carrying a document resource is displayed on the summary interaction page, perform document analysis on the document resource to obtain a document directory of the document resource; the paragraph analysis identifier in the document directory is associated with the document paragraph in the document resource;

[0399] The paragraph jump module 33 is configured to jump to display the document paragraph corresponding to the paragraph analysis identifier in response to a triggering operation on the paragraph analysis identifier in the document directory when displaying the second conversation message carrying the document directory on the summary interaction page.

[0400] Wherein, the paragraph analysis identifier includes the document page number;

[0401] The paragraph jump module 33 is specifically used to:

[0402] In response to a trigger operation on a document page number in a document directory, a document resource display page associated with the document resource is displayed, and a document paragraph corresponding to the document page number is displayed on the document resource display page.

[0403] Wherein, the paragraph analysis identifier includes the document page number;

[0404] The paragraph jump module 33 is specifically used to:

[0405] In response to a trigger operation on a document page number in the document directory, a document paragraph corresponding to the document page number is displayed on the summary interaction page.

[0406] See Figure 28 , Figure 28 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. Figure 28 As shown, the computer device 1000 may include: a processor 1001, a network interface 1004 and a memory 1005. In addition, the above-mentioned computer device 1000 may also include: a user interface 1003, and at least one communication bus 1002. The communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), a keyboard (Keyboard), and the user interface 1003 may optionally include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk memory. The memory 1005 may optionally also be at least one storage device located away from the aforementioned processor 1001. As Figure 28 As shown, the memory 1005 as a computer-readable storage medium may include an operating system, a network communication module, a user interface module, and a device control application.

[0407] In such Figure 28 In the illustrated computer device 1000, the network interface 1004 provides network communication functionality; the user interface 1003 primarily provides an input interface for the user; and the processor 1001 is configured to invoke a device control application stored in the memory 1005 to execute the document data processing method described in any of the preceding embodiments, which will not be repeated here. Furthermore, the beneficial effects of employing the same method will not be repeated here.

[0408] In addition, it should be pointed out here that: the embodiment of the present application also provides a computer-readable storage medium, and the computer-readable storage medium stores the computer program executed by the document data processing device 1 and the data processing device 2 mentioned above, and the computer program includes program instructions. When the processor executes the program instructions, it can execute the description of the document data processing method in the above embodiment. Therefore, it will not be repeated here. In addition, the description of the beneficial effects of using the same method will not be repeated. For technical details not disclosed in the computer-readable storage medium embodiment involved in this application, please refer to the description of the method embodiment of this application.

[0409] The computer-readable storage medium may be the data processing device provided in any of the aforementioned embodiments or the internal storage unit of the computer device, such as the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device. Furthermore, the computer-readable storage medium may also include both the internal storage unit of the computer device and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by the computer device. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0410] In addition, it should be noted that the embodiments of the present application also provide a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the method provided by any of the corresponding embodiments above. In addition, the description of the beneficial effects of using the same method will not be repeated. For technical details not disclosed in the computer program product or computer program embodiments involved in this application, please refer to the description of the method embodiments of this application.

[0411] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program that has a predetermined function and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0412] The terms "first", "second", etc. in the description, claims, and drawings of the embodiments of the present application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "comprising" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or device comprising a series of steps or units is not limited to the listed steps or modules, but may optionally include steps or modules not listed, or may optionally include other step units inherent to these processes, methods, apparatuses, products, or devices.

[0413] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0414] The above disclosure is only a preferred embodiment of the present application, and certainly cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. A document data processing method, characterized in that: The method comprises: displaying a summary interaction page associated with the first business object; In response to a document import operation on a document resource, displaying a first conversation message carrying the document resource sent by the first business object on the summary interaction page; In a first associated display area associated with the first conversation message, node progress prompt information of each process processing node in a plurality of process processing nodes associated with the document resource is traversed and displayed; the plurality of process processing nodes includes a summary node for summarizing resource fragment contents in the document resource; When the node progress prompt information displayed is the node progress prompt information of the summary summary node, and the node progress prompt information of the summary summary node indicates that a first summary associated with the resource slice content is obtained by summary, a second session message carrying the first summary is displayed on the summary interaction page.

2. The method according to claim 1, characterized in that The method further comprises: In response to a message triggering operation, displaying a third conversation message indicated by the message triggering operation on the summary interaction page; the third conversation message includes an interaction instruction initiated for the document resource; A fourth conversation message is displayed on the summary interaction page; the fourth conversation message includes an interaction reply to the interaction instruction.

3. The method according to claim 1, characterized in that Any process processing node among the multiple process processing nodes is a target process processing node; The traversal and display of node progress prompt information of each process processing node among the plurality of process processing nodes associated with the document resource includes: When the displayed node progress prompt information is the node progress prompt information of the target process processing node, the node progress prompt information of the target process processing node is dynamically displayed; the node progress prompt information of the target process processing node changes as the node processing progress when the target process processing node performs resource processing on the document resource changes.

4. The method according to claim 3, characterized in that The dynamically displaying the node progress prompt information of the target process processing node includes one or more of the following: When the node progress prompt information of the target process processing node includes the time progress prompt information of the target process processing node, dynamically displaying the time progress prompt information of the target process processing node; or, When the node progress prompt information of the target process processing node includes the numerical progress prompt information of the target process processing node, the numerical progress prompt information of the target process processing node is dynamically displayed.

5. The method according to claim 1, wherein The method further comprises: In response to the node viewing operation for the multiple process processing nodes, the process node prompt information corresponding to the multiple process processing nodes is displayed in the first associated display area; one process processing node corresponds to one process node prompt information; the process node prompt information of any process processing node includes node status information for prompting the node processing status of any process processing node.

6. The method according to claim 1, characterized in that The multiple process processing nodes further include a document upload node for uploading the document resource; the summary interaction page includes a document processing status control, the document processing status control is used to adjust the document processing status of the document resource, the document processing status includes a document upload status, and the document upload status is the document processing status determined when the document resource is processed by the document upload node; The method further comprises: When the node progress prompt information displayed in the first associated display area is the node progress prompt information of the document upload node, first status prompt information is displayed in the document processing status control; the first status prompt information is used to indicate that the document processing status of the document resource is the document upload status; In response to a first state switching operation on the document processing state control, the document processing state of the document resource is adjusted from the document uploading state to the cancel upload state, and a first state prompting conversation message is output to prompt that the document processing state of the document resource is the cancel upload state.

7. The method according to claim 1, characterized in that The summary interaction page includes a document processing status control, which is used to adjust the document processing status of the document resource; the multiple process processing nodes also include a document content processing node, which includes the summary summary node, a content parsing node for determining the document content of the document resource, a content fragmentation node for determining the resource fragment content corresponding to the document resource, and a fragment summary extraction node for determining the content summary of the resource fragment content; the document processing status includes a document content processing status, which is the document processing status determined when the document resource is processed by the document content processing node; The method further comprises: When the node progress prompt information displayed in the first associated display area is the node progress prompt information of the document content processing node, second status prompt information is displayed in the document processing status control; the second status prompt information is used to indicate that the document processing status of the document resource is the document content processing status; In response to a second state switching operation on the document processing state control, the document processing state of the document resource is adjusted from the document content processing state to the content processing cancellation state, and a second state prompting session message is output to prompt that the document processing state is the content processing cancellation state.

8. The method according to claim 1, characterized in that The method further comprises: When the second conversation message is displayed on the summary interaction page, a summary re-extraction control is displayed for instructing to regenerate the summary summary of the document resource; In response to a triggering operation on the summary re-extraction control, a second summary regenerated for the document resource is displayed in the second conversation message.

9. The method according to claim 8, characterized in that The method further comprises: Displaying a summary switching control for a summary of the document resource in a second associated display area associated with the second conversation message; In response to a triggering operation on the summary switching control, the first summary and the second summary are switched and displayed in the second conversation message.

10. The method according to claim 1, characterized in that The method further comprises: In response to a session sharing operation in the summary interaction page, the session message in the summary interaction page is shared to a second business object, so that the second business object triggers the display of a sharing interaction page associated with the second business object through a message sharing link associated with the session sharing operation, and the sharing interaction page includes the session message displayed in the summary interaction page.

11. The method according to claim 1, wherein The document resources include a first document resource and a second document resource; The method further comprises: When displaying a conversation message carrying a summary summary of the first document resource and a conversation message carrying a summary summary of the second document resource in the summary interaction page, a conversation message carrying document merge and comparison information between the first document resource and the second document resource is displayed; the document merge and comparison information is determined based on the content summary of the resource fragment content of the first document resource and the content summary of the resource fragment content of the second document resource.

12. The method according to claim 11, characterized in that The document import operation is a combined import operation for the first document resource and the second document resource; or, The document import operation includes a first import operation for the first document resource and a second import operation for the second document resource.

13. The method according to claim 1, wherein The method further comprises: Performing document parsing processing on the document resource to obtain document text information corresponding to the document content of the document resource; Performing document fragmentation processing on the document text information to obtain resource fragment content corresponding to the document resource; Performing summary extraction processing on the resource slice content to obtain a content summary of the resource slice content; The content summaries of the resource slice contents are summarized and aggregated to obtain the first summary of the document resource.

14. The method according to claim 13, characterized in that The performing document parsing processing on the document resource to obtain document text information corresponding to the document content of the document resource includes: When the document resource includes multimedia data information, performing text extraction processing on the multimedia data information to obtain translated text information associated with the multimedia data information; the multimedia data information includes one or more of the following: image data, video data, and voice data; When the original text information in the document resource is obtained, the document text information corresponding to the document content of the document resource is determined based on the spatial position relationship between the original text information in the document resource and the transferred text information in the document resource.

15. The method according to claim 13, characterized in that The performing document fragmentation processing on the document text information to obtain resource fragment content corresponding to the document resource includes: When the number of text segmentations in the document text information is greater than the slice length threshold, the first text segmentation in the document text information is determined as the starting segmentation position, and text segmentations corresponding to the number of segmentations corresponding to the slice length threshold are searched backward from the starting segmentation position, and the content of the first resource slice is determined based on the text segmentation at the starting segmentation position and the searched text segmentations; When the number of text segmentations other than the first resource fragment content in the document text information is greater than a slice length threshold, the starting segmentation position is updated based on the next text segmentation in the first resource fragment content, and text segmentations having a number of segmentations corresponding to the slice length threshold are searched backward from the updated starting segmentation position, and the second resource fragment content is determined based on the text segmentation at the updated starting segmentation position and the searched text segmentation; The resource slice content corresponding to the document resource is determined based on the first resource slice content and the second resource slice content.

16. The method according to claim 13, characterized in that The document resources include a first document resource and a second document resource; The method further comprises: Performing a topic extraction process on the content summary of the resource fragment content of the first document resource to obtain the content topic of the first document resource; Performing a topic extraction process on the content summary of the resource fragment content of the second document resource to obtain the content topic of the second document resource; Document merging and comparison information between the first document resource and the second document resource is determined based on a content theme of the first document resource and a content theme of the second document resource.

17. The method according to claim 13, wherein The method further comprises: Performing document sentence processing on the document text information to obtain at least one sentence content corresponding to the document resource; Performing feature extraction processing on each sentence content in the at least one sentence content to obtain sentence content features corresponding to each sentence content; When acquiring the interaction instruction initiated by the first business object for the document resource, performing feature extraction processing on the interaction instruction to obtain an instruction feature corresponding to the interaction instruction; When determining the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content, an interactive response to the interactive instruction is determined based on the feature similarity between the instruction feature and the sentence content feature corresponding to each sentence content.

18. A document data processing method, characterized in that: The method comprises: Display the summary interaction page; When a first conversation message carrying a document resource is displayed on the summary interaction page, document analysis is performed on the document resource to obtain a document directory of the document resource; the paragraph analysis identifier in the document directory is associated with the document paragraph in the document resource; When the second conversation message carrying the document directory is displayed on the summary interaction page, in response to a triggering operation on the paragraph analysis identifier in the document directory, the document paragraph corresponding to the paragraph analysis identifier is displayed in a jump manner.

19. The method according to claim 18, characterized in that The paragraph analysis identifier includes a document page number; The step of responding to a triggering operation on the paragraph analysis identifier in the document directory and jumping to display a document paragraph corresponding to the paragraph analysis identifier includes: In response to a triggering operation on a document page number in the document directory, a document resource display page associated with the document resource is displayed, and a document paragraph corresponding to the document page number is displayed on the document resource display page.

20. The method according to claim 18, wherein The paragraph analysis identifier includes the document page number; The step of responding to a triggering operation on the paragraph analysis identifier in the document directory and jumping to display a document paragraph corresponding to the paragraph analysis identifier includes: In response to a triggering operation on a document page number in the document directory, a document paragraph corresponding to the document page number is displayed in the summary interaction page.

21. A document data processing device, characterized in that: The device comprises: A first interaction page display module, configured to display a summary interaction page associated with the first business object; a document import module, configured to, in response to a document import operation on a document resource, display a first conversation message carrying the document resource, sent by the first business object, on the summary interaction page; a progress display module configured to traverse and display, in a first associated display area associated with the first conversation message, node progress prompt information for each of a plurality of process processing nodes associated with the document resource; the plurality of process processing nodes including a summary node configured to summarize resource fragment contents in the document resource; A summary output module is used to display a second session message carrying the first summary summary on the summary interaction page when the node progress prompt information displayed is the node progress prompt information of the summary summary node, and the node progress prompt information of the summary summary node indicates that a first summary summary associated with the resource fragment content is obtained by summary.

22. A document data processing device, characterized in that: The device comprises: The second interactive page display module is used to display the summary interactive page; a directory acquisition module configured to, when a first conversation message carrying a document resource is displayed on the summary interaction page, perform document analysis on the document resource to obtain a document directory of the document resource; wherein the paragraph analysis identifier in the document directory is associated with the document paragraph in the document resource; The paragraph jump module is used to jump to display the document paragraph corresponding to the paragraph analysis identifier in response to a triggering operation on the paragraph analysis identifier in the document directory when displaying the second conversation message carrying the document directory on the summary interaction page.

23. A computer device, characterized in that: including memory and processor; The memory is connected to the processor, the memory is used to store a computer program, and the processor is used to call the computer program so that the computer device executes the method according to any one of claims 1 to 20.

24. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which is suitable for being loaded and executed by a processor, so that a computer device having the processor executes the method according to any one of claims 1 to 20.

25. A computer program product, characterized in that The method comprises a computer program / instruction, which implements the method according to any one of claims 1 to 20 when executed by a processor.

Images