Document parsing method and device, storage medium and processor

Abstract

The present invention discloses a document parsing method and device, storage medium and processor. The method comprises: obtaining a document to be parsed; parsing the document to obtain a document representation; processing the document representation by natural language to generate a hierarchical structure tree corresponding to the document; converting the hierarchical structure tree into a parsing result in a target representation form, and outputting the parsing result. The present invention solves the technical problem of low document parsing efficiency in the related art.

Description

Technical Field

[0001] The present invention relates to the field of information processing technology, and in particular to a document parsing method and device, a storage medium and a processor. Background Art

[0002] There are many types of corporate documents and various storage media. A large amount of document information often needs to be analyzed and used in various intelligent applications to create higher value. However, when electronic or paper documents need to be digitized, most of them are processed manually, which has low efficiency in document parsing.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present invention provide a document parsing method and device, a storage medium and a processor, so as to at least solve the technical problem of low efficiency in parsing documents in the related art.

[0005] According to one aspect of an embodiment of the present invention, a document parsing method is provided, comprising: obtaining a document to be parsed; parsing the document to obtain a document representation; processing the document representation through natural language to generate a hierarchical structure tree corresponding to the document; converting the hierarchical structure tree into a parsing result in a target representation form, and outputting the parsing result.

[0006] Furthermore, the type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo.

[0007] Furthermore, after outputting the parsing result, the method further includes: generating parsed content of a target version based on the parsing result; and pushing the parsed content of the target version to a target object.

[0008] Furthermore, the document representation is processed by natural language processing including at least one of the following: text classification, layout analysis, and title extraction.

[0009] Furthermore, rich text information in the document is identified while parsing the document, wherein the rich text information includes at least: font information, alignment information, tables, and pictures.

[0010] Furthermore, the document is any one of the following: patent documents, judicial documents, contract documents, and chat records.

[0011] Furthermore, after outputting the analysis result, the method also includes: receiving a query instruction, wherein the query instruction carries query information and the query instruction is a text instruction or a voice instruction; responding to the query instruction, obtaining target information matching the query information from the analysis result; and displaying the target information.

[0012] According to one aspect of an embodiment of the present invention, a method for processing a contract document is provided, comprising: obtaining a contract document to be parsed; processing the contract document to generate a hierarchical structure tree corresponding to the contract document; converting the hierarchical structure tree into a parsing result in a target representation form, and outputting the parsing result.

[0013] Furthermore, after outputting the parsing result, the method also includes: analyzing the parsing result to determine the extracted content from the parsing result; analyzing the extracted content and identifying the target content in the extracted content, wherein the target content includes at least: content classification and clause number; and outputting the identified target content.

[0014] Furthermore, before outputting the marked target content, the method further includes: detecting whether an editing instruction is received; if an editing instruction is received, adjusting the clauses in the extracted content in response to the editing instruction.

[0015] Furthermore, the method also includes: determining the approval object corresponding to the content of each structure in the contract document; transmitting the content of each structure to the corresponding approval object; and obtaining the approval result of the approval object.

[0016] Furthermore, transmitting the contents of each structure to the corresponding approval object includes: marking the content related to the corresponding approval object in the content of each structure using a preset marking method, wherein the preset marking method is one of the following: highlighting, bolding, font enlargement, font change; transmitting the marked content to the corresponding approval object.

[0017] Furthermore, after obtaining the approval result of the approval object, the method also includes: determining the performance matters involved in the contents of each structure in the contract document; generating a performance plan based on the performance matters; determining the target object related to the performance plan; and transmitting the performance plan to the target object related to the performance plan.

[0018] Furthermore, after obtaining the approval result of the approval object, the method also includes: determining the performance risk involved in the content of each structure in the contract document; generating a risk management plan based on the performance risk; determining the target object involved in the risk management plan; and transmitting the risk management plan to the target object involved in the risk management plan.

[0019] Furthermore, after determining the performance risk involved in the contents of each structure in the contract document, the method further includes: adding a risk label to the contents of each structure in the contract document based on the performance risk, wherein the risk label is used for risk reminder.

[0020] According to one aspect of an embodiment of the present invention, a document parsing method is provided, comprising: obtaining a document to be parsed; processing the document to generate a structure tree corresponding to the document; splitting the document based on the structure tree to obtain multiple information fragments, wherein each information fragment includes at least a document fragment and key information of the document fragment; extracting keywords from each information fragment, and storing the extracted keywords, the information fragments corresponding to the keywords, and the document into a search engine.

[0021] Furthermore, the method also includes: if a search term input by a user is detected, matching the search term with a keyword in the search engine to determine a target keyword that matches the search term; and returning an information segment corresponding to the target keyword.

[0022] Furthermore, after splitting the document based on the structure tree to obtain multiple information fragments, the method also includes: converting each information fragment into a corresponding image; storing the converted image in a non-relational database; returning the information fragment corresponding to the target keyword also includes: extracting the image corresponding to the information fragment from the non-relational database, and returning the image corresponding to the information fragment while returning the information fragment corresponding to the target keyword.

[0023] According to one aspect of an embodiment of the present invention, a document parsing method is provided, comprising: receiving a document to be parsed uploaded by a client and a service call request sent by the client; parsing the document in a server in response to the service call request to obtain a document representation; processing the document representation through natural language to generate a hierarchical structure tree corresponding to the document; converting the hierarchical structure tree into a parsing result in a target representation form; and returning the parsing result to the client.

[0024] According to one aspect of an embodiment of the present invention, a document retrieval method is provided, comprising: receiving a document retrieval instruction, wherein the document retrieval instruction is used to instruct analysis of a target document set, wherein the documents in the target document set have the same category, or the document category and the document subcategory are the same, and the document category includes patent documents, judicial documents, contract documents, and chat records; obtaining a document analysis result based on the parsing result of each document in the target document set and the document retrieval instruction, wherein the parsing result is obtained by parsing the documents in the target document set by any of the methods described above; and displaying the analysis result in a target display form.

[0025] According to one aspect of an embodiment of the present invention, a document parsing device is provided, comprising: a first acquisition unit, used to acquire a document to be parsed; a first parsing unit, used to parse the document to obtain a document representation; a first generation unit, used to process the document representation through natural language processing to generate a hierarchical structure tree corresponding to the document; and a first processing unit, used to convert the hierarchical structure tree into a parsing result in a target representation form, and output the parsing result.

[0026] According to one aspect of an embodiment of the present invention, a contract document parsing device is provided, comprising: a third acquisition unit, used to acquire the contract document to be parsed; a third generation unit, used to process the contract document and generate a hierarchical structure tree corresponding to the contract document; and a first output unit, used to convert the hierarchical structure tree into a parsing result in a target representation form, and output the parsing result.

[0027] According to one aspect of an embodiment of the present invention, a document parsing device is provided, comprising: a fifth acquisition unit, used to acquire a document to be parsed; a seventh generation unit, used to process the document and generate a structure tree corresponding to the document; a splitting unit, used to split the document based on the structure tree to obtain multiple information fragments, wherein each information fragment includes at least a document fragment and key information of the document fragment; an extraction unit, used to extract keywords from each information fragment, and store the extracted keywords, the information fragments corresponding to the keywords and the document into a search engine.

[0028] According to one aspect of an embodiment of the present invention, a document parsing device is provided, comprising: a second receiving unit, used to receive a document to be parsed uploaded by a client and a service call request sent by the client; a second parsing unit, used to respond to the service call request, parse the document in a server, and obtain a document representation; a generating unit, used to process the document representation through natural language to generate a hierarchical structure tree corresponding to the document; a second conversion unit, used to convert the hierarchical structure tree into a parsing result in a target representation form; and a fourth processing unit, used to return the parsing result to the client.

[0029] According to one aspect of an embodiment of the present invention, a document retrieval device is provided, comprising: a third receiving unit, configured to receive a document retrieval instruction, wherein the document retrieval instruction is used to instruct to analyze a target document set, wherein the documents in the target document set have the same category, or the document category and the document subcategory are the same, and the document categories include patent documents, judicial documents, contract documents, and chat records; a sixth acquisition unit, configured to obtain a document analysis result based on the parsing result of each document in the target document set and the document retrieval instruction, wherein the parsing result is obtained by parsing the documents in the target document set by one of the methods described above; and a display unit, configured to display the analysis result in a target display form.

[0030] According to one aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the storage medium includes a stored program, wherein when the program is run, the device where the storage medium is located is controlled to execute any one of the document parsing methods described above, or any one of the contract document processing methods described above, or any one of the document review methods described above.

[0031] According to one aspect of an embodiment of the present invention, a processor is provided, wherein the processor is used to run a program, wherein when the program is run, the program executes any one of the document parsing methods described above, or any one of the contract document processing methods described above, or any one of the document review methods described above.

[0032] In an embodiment of the present invention, a document to be parsed is obtained; the document is parsed to obtain a document representation; the document representation is processed by natural language to generate a hierarchical structure tree corresponding to the document; the hierarchical structure tree is converted into a parsing result in a target representation form, and the parsing result is output. That is, the hierarchical structure tree corresponding to the document is obtained quickly and accurately using natural language processing technology, and finally the parsing result is output in the target representation form, thereby avoiding the use of manual parsing of the document, thereby solving the technical problem of low document parsing efficiency in related technologies and achieving the purpose of improving the document parsing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0034] Figure 1 is a hardware structure block diagram of a computer terminal according to an embodiment of the present invention;

[0035] Figure 2 is a flowchart of a document parsing method provided according to Embodiment 1 of the present invention;

[0036] Figure 3 The following is a schematic diagram of an optional document parsing method provided according to the first embodiment of the present invention. Figure 1 ;

[0037] Figure 4 is a schematic diagram of a hierarchical structure tree in a document parsing method provided in Embodiment 1 of the present invention;

[0038] Figure 5 is a flowchart of a method for processing a contract document provided in Embodiment 2 of the present invention;

[0039] Figure 6 The following is a schematic diagram of an optional contract document processing method provided according to the second embodiment of the present invention. Figure 1 ;

[0040] Figure 7 The following is a schematic diagram of an optional contract document processing method provided according to the second embodiment of the present invention. Figure 2 ;

[0041] Figure 8 is a flowchart of a document parsing method provided according to Embodiment 3 of the present invention;

[0042] Fig. 9 Schematic diagram of an optional document parsing method provided according to Embodiment 3 of the present invention Figure 1 ;

[0043] Fig.10 Schematic diagram of an optional document parsing method provided according to Embodiment 3 of the present invention Figure 2 ;

[0044] Fig.11 is a flowchart of a document parsing method provided according to a fourth embodiment of the present invention;

[0045] Fig.12 is a flowchart of a document access method provided in Embodiment 5 of the present invention;

[0046] Fig.13 is a schematic diagram of a document parsing device provided according to a sixth embodiment of the present invention;

[0047] Fig.14 is a schematic diagram of a contract document processing device provided according to Embodiment 7 of the present invention;

[0048] Fig.15 is a schematic diagram of a document parsing device provided according to Embodiment 8 of the present invention;

[0049] Fig.16 is a schematic diagram of a document parsing device provided according to Embodiment 9 of the present invention;

[0050] Fig.17 is a schematic diagram of a document access device provided according to a tenth embodiment of the present invention;

[0051] Fig.18 is a structural block diagram of an optional computer terminal according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0053] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0054] For the convenience of description, some nouns or terms involved in the embodiments of the present application are explained below:

[0055] Contract Templates: Standardized contract templates for quick online contract drafting.

[0056] Intelligent extraction: Text extraction based on Word, PDF and other documents using optical character recognition technology.

[0057] Intelligent splitting: Based on intelligent algorithms, the contract title, contracting parties, main text, signing area, attachments, etc. are identified; the algorithm analyzes the clause numbers and performs deep tree structured stratification.

[0058] Natural language processing, or NLP for short, is an important research direction in the fields of computer science and artificial intelligence. It studies the use of computers to process, understand, and apply human languages ​​(such as Chinese, English, etc.) to achieve effective communication between people and computers. In general, users may not be familiar with machine language, so natural language processing technology can help such users communicate with machines using natural language. From a modeling perspective, in order to facilitate computer processing, natural language can be defined as a set of rules or symbols. We combine the symbols in the set to convey various information. The understanding of natural language is a comprehensive system engineering, which includes many sub-disciplines, including phonology for sound, morphology for word formation, syntax for sentence structure, and semantic syntax and pragmatics for understanding.

[0059] Example 1

[0060] According to an embodiment of the present invention, an embodiment of a document parsing method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0061] The method embodiment provided in the first embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 FIG. 1 shows a hardware structure block diagram of a computer terminal (or mobile device) for implementing a document parsing method. Figure 1 As shown, the computer terminal 10 (or mobile device 10) may include one or more (102a, 102b, ..., 102n are used to illustrate) processors 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission device for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 1The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0062] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10 (or mobile device). As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0063] The memory 104 can be used to store software programs and modules of application software, such as program instructions / data storage devices corresponding to the document parsing method in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, the document parsing method of the above-mentioned application program is realized. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0064] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0065] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0066] Under the above operating environment, this application provides Figure 2The parsing method of the document shown. Figure 2 4 is a flowchart of a document parsing method according to the first embodiment of the present invention.

[0067] Step S101, obtaining the document to be parsed.

[0068] The document to be parsed may be of various types, for example, the document may be any of the following: Word, Excel, PDF, scanned copy, etc. The document may be a patent document or a judicial document. The document to be parsed is obtained on the client, which may be uploaded by the user or stored locally on the client.

[0069] Step S102, parse the document to obtain a document representation.

[0070] The above-mentioned document representation can be understood as the intermediate information of document parsing, for example, the rich text information of the document. That is, when parsing the document, the information in the document is identified, wherein the information includes rich text information, wherein the rich text information includes at least: font information, alignment information, table, picture. Among them, the font information can also include: font style, font size, whether bold, etc.

[0071] For different document types, different open source tools can be used to parse the document and obtain the corresponding document representation. For example, for Word type documents, the underlying protocol can be parsed through Apache POI, where Apache POI is an open source library of the Apache Software Foundation. POI provides an API for Java programs to read and write Office format documents. For PDF type documents, the underlying protocol can be parsed through Apache PDFBox, where Apache PDFbox is an open source, Java-based tool library that supports PDF document generation. It can be used to create new PDF documents, modify existing PDF documents, and extract required content from PDF documents. For scanned documents, the underlying protocol can be parsed through optical character recognition technology.

[0072] Step S103: Process the document representation through natural language processing to generate a hierarchical structure tree corresponding to the document.

[0073] In the above steps, the processing method of the document representation through natural language processing includes at least one of the following: text classification, layout analysis, and title extraction. By using NLP algorithm technology to represent the document, a hierarchical structure tree (which can be understood as an expression of a tree structure) can be obtained. It should be noted that the hierarchical depth of the hierarchical structure tree is determined according to the requirements and the depth of the document classification. For example, for a purchase order agreement, the corresponding hierarchical structure tree is generated, and the first level includes: contract subject, text, signature and attachments, and the text in the second level includes: definition, payment of fees, delivery, other agreements, etc., among which the payment part will generate a clause tree, and the clause tree includes the relevant clauses in the document.

[0074] Step S104: convert the hierarchical structure tree into a parsing result in a target representation form, and output the parsing result.

[0075] The specific form of the above target representation is not limited, and the representation of the parsing result can be pre-configured according to the needs of downstream tasks. The purpose of the above output target representation is to output the parsing result in a unified representation so that different types of text can obtain the same interface output, which is convenient for downstream users to use.

[0076] In summary, in the document parsing method according to the first embodiment of the present invention, a document to be parsed is obtained; the document is parsed to obtain a document representation; the document representation is processed by natural language to generate a hierarchical structure tree corresponding to the document; the hierarchical structure tree is converted into a parsing result in the target representation form, and the parsing result is output. That is, the natural language processing technology is used to quickly and accurately obtain the hierarchical structure tree corresponding to the document, and finally the parsing result is output in the target representation form, thereby avoiding the use of manual parsing of the document, thereby solving the technical problem of low document parsing efficiency in related technologies and achieving the purpose of improving the document parsing efficiency.

[0077] like Figure 3 As shown, for different document types (e.g., Word, Excel, PDF, and scanned documents), the underlying protocol is parsed through different open source tools including Apache POI and PDFBox. Based on the existing open source tools, the intermediate information of document parsing (corresponding to the above-mentioned document representation) is first obtained, such as text, font, alignment, etc., and further analyzed on this basis. Combining rules and NLP algorithms, the hierarchical structure tree of the document is constructed and the document is uniformly expressed, that is, all the contents in the document are parsed through a unified standard parsing protocol to obtain text information and structure tree information, wherein the text information may include rich text information, and a hierarchical structure tree is constructed based on the rich text information and the structure tree information, so that the hierarchical structure tree is output in a unified representation form.

[0078] At the same time, the technical solution of the present application can not only parse the normal paragraph text content in the document, but also identify and locate the table and image information, and realize the storage of the table HTML representation and image content, so as to facilitate the use of downstream tasks and ensure the consistency of the parsing result output with the original document as much as possible. Therefore, the technical solution of the present application will parse the rich text meta information of the document itself while parsing the plain text and output the hierarchical structured parsing result of the document.

[0079] After the underlying parsing is implemented, the parsed results and NLP algorithms are used to comprehensively analyze the rich text information and obtain hierarchical structure tree results. This includes text label module recognition and fine hierarchical structure recognition, which mainly utilize text classification, text boundary detection, and heuristic fine hierarchical methods. The following introduces the implementation principles and effects of the above algorithms.

[0080] A "Purchase Order Agreement" includes the contract title, the main part of the contract, the body, the signature part, and the attachments. The method of this solution uses Fast Text (a word vector calculation and text classification tool) and a pre-trained model of word vectors to perform text classification and identify the boundary text paragraphs of each module. For the body part, it is necessary to obtain fine hierarchical structure information, where the above-mentioned fine hierarchical structure information refers to the information in each hierarchical structure in the body part. First, identify the title number, and use a heuristic algorithm according to different sequence number categories to generate a structure tree. For the contract, a clause tree is generated (corresponding to the above-mentioned hierarchical structure tree, such as Figure 4 As shown). When there is no explicit sequence information in the text, the boundary paragraphs can also be identified with the help of sequence labeling methods. Finally, the hierarchical structure tree result is output in a unified representation form, so that different types of texts can obtain the output of the same interface, so that downstream tasks can focus on their respective tasks. It should be noted that the content, display and storage in the hierarchical structure tree can be pre-defined when defining the structure of the structure tree and determined by pre-defined rules.

[0081] Optionally, in the document parsing method provided in Example 1 of the present application, after outputting the parsing results, the method also includes: receiving a query instruction, wherein the query instruction carries query information, and the query instruction is a text instruction or a voice instruction; responding to the query instruction, obtaining target information matching the query information from the parsing results; and displaying the target information.

[0082] For example, if a user-triggered voice question is detected: When is the validity period of contract document A? Information matching the question content can be extracted from the parsed results. For example, if the validity period of contract document A is from August 30, 2019 to August 30, 2020, the extracted content will be displayed or the extracted content will be announced in voice broadcast. This allows users to quickly obtain the information they need, improving user experience.

[0083] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0084] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0085] Example 2

[0086] In the operating environment of the above embodiment 1, the present application provides the following Figure 5 The method of handling the contract documents shown. Figure 5 It is a flowchart of a method for processing contract documents according to the second embodiment of the present invention.

[0087] Step S501, obtaining the contract document to be parsed.

[0088] Step S502: Process the contract document to generate a hierarchical structure tree corresponding to the contract document.

[0089] The contract document can be processed by first parsing the intermediate information of the contract document, such as the rich text information of the contract document, and then generating a hierarchical structure tree corresponding to the contract document based on the intermediate information of the contract document.

[0090] Step S503: convert the hierarchical structure tree into a parsing result in a target representation format, and output the parsing result.

[0091] The specific form of the above target representation is not limited in this application, and the representation of the parsing result can be pre-configured according to the needs of downstream tasks. Through the above steps, the contract document can be quickly parsed and output in a unified representation (target representation).

[0092] Optionally, in the contract document processing method of embodiment 2 of the present invention, after outputting the parsing results, the method also includes: analyzing the parsing results to determine the extracted content from the parsing results; analyzing the extracted content and identifying the target content in the extracted content, wherein the target content includes at least: content classification and clause number; and outputting the identified target content.

[0093] For example, after a user uploads a contract document, the contract document is parsed to obtain the parsing result, and then the extracted content is determined from the parsing result, and the target content in the extracted content is marked to highlight the target content so that the user can see the marked target content more quickly and accurately.

[0094] The process of parsing and splitting the contract documents can be as follows: Figure 6 As shown, after uploading the file, you can parse the file to determine the text, style, serial number and other information in the file (corresponding to the above-mentioned extraction content). Then perform algorithmic labeling, which may involve clause classification, clause depth analysis, clause labeling, etc. Structuring the clauses in the contract, merging certain similar clauses, and adding corresponding labels to the number, level, and each part in the contract. If there are other contents in the contract that need to be configured, configure them.

[0095] Optionally, in the contract document processing method of the second embodiment of the present invention, before outputting the marked target content, the method also includes: detecting whether an editing instruction is received; if an editing instruction is received, adjusting the terms in the extracted content in response to the editing instruction.

[0096] The process of drafting and publishing the template can be as follows: Figure 7 As shown, for example, when a template operator creates a new template, he first uploads the original, intelligently extracts and splits the text online, edits the generated template, and splits the parsed conditions. At this time, it will detect whether an editing instruction has been received; if an editing instruction has been received, the clauses in the extracted content will be adjusted in response to the editing instruction. In other words, some clauses that need to be edited are edited. After the modification of the clause content is completed, the historical log is recorded, and the modified template is submitted for approval. If the approval is not passed, the template is withdrawn and reprocessed. If the approval is passed, the contract template takes effect.

[0097] Optionally, in the contract document processing method of embodiment 2 of the present invention, the method also includes: determining the approval object corresponding to the content of each structure in the contract document; transmitting the content of each structure to the corresponding approval object; and obtaining the approval result of the approval object.

[0098] For example, after the contract template is generated, the content of different structures is pushed to different approval objects. For example, the approval objects include: department leaders, legal affairs and management. The content that department leaders need to pay attention to, such as the content of the rights and obligations structure, will be transmitted to the department leaders for review. The structural part involving legal terms will be transmitted to the legal department for review, and the structural part involving performance risks will be transmitted to the management for review. In addition, the content of each structure related to the corresponding approval object can also be marked using a preset marking method, where the preset marking method is one of the following: highlight, bold, font enlargement, font change; the marked content is transmitted to the corresponding approval object, so that the review object can quickly obtain the content in the contract that needs to be focused on, quickly approve, and improve approval efficiency.

[0099] like Figure 7 As shown, after the contract template is created, the template contract is approved, the contract is signed and archived, and finally the contract is fulfilled. Specifically, the process of template creation has been described above and will not be repeated here. After the template is approved, the template is stored in the template library (corresponding to the template supermarket mentioned above) so that subsequent users can download or browse the template. When it is necessary to draft a contract, the contract is created, the clause content in the contract is edited, the associated clauses are associated, and the component values ​​are entered. Then the contract is submitted, the structured fields are extracted based on the components of the contract template, and the drafted contract is submitted for approval. In the contract approval stage, the approval process is started, and the contract personalized approval process can be carried out. For example, the content that needs to be approved for each different object in the contract is marked, and each object can intuitively obtain the content that needs to be approved. In addition, the approval node can also be dynamically adjusted based on the clause labels in the contract. When approving the contract, the template can also be edited online, the differences of each contract template can be compared, and other contract templates in the template library can be downloaded or browsed. After the approval is completed, the final draft of the contract is generated.

[0100] Add an electronic seal to the final contract online, stamp it with a physical seal offline, and then proceed to the contract performance stage. Add corresponding labels to the risks in the contract and predict the risks in the contract.

[0101] Optionally, in the contract document processing method of embodiment 2 of the present invention, after obtaining the approval result of the approval object, the method also includes: determining the performance matters involved in the contents of each structure in the contract document; generating a performance plan based on the performance matters; determining the target object related to the performance plan; and transmitting the performance plan to the target object related to the performance plan.

[0102] Through the above solution, the performance plan generated based on the performance items in the contract can be transmitted to the target objects related to the performance plan, so as to better implement the performance after the contract is signed. For example, the performance plan includes the task of generating 1,000 heaters from September 1, 2020 to December 1, 2020. The task is transmitted to the person in charge of the production line so that timely follow-up, timely scheduling of production tasks and monitoring of production conditions can be carried out.

[0103] Optionally, in the contract document processing method of embodiment 2 of the present invention, after obtaining the approval result of the approval object, the method also includes: determining the performance risk involved in the content of each structure in the contract document; generating a risk management plan based on the performance risk; determining the target object involved in the risk management plan; and transmitting the risk management plan to the target object involved in the risk management plan.

[0104] For example, performance risks include: contract price, settlement method, contract duration, project payment and other risks. The performance risks are generated into a risk management plan. For example, the risk management plan involves the financial party, the project implementation party, etc. The risk management plan is transmitted to the financial personnel and the project implementation responsible personnel to better monitor the risks in the contract performance process. In addition, after determining the performance risks involved in the content of each structure in the contract document, the method also includes: adding risk tags to the content of each structure in the contract document based on the performance risk, wherein the risk tags are used for risk reminders. For example, there is a possibility that the contract duration may be exceeded, and a risk tag is added to the contract duration in the contract to provide risk reminders.

[0105] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0106] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0107] Example 3

[0108] In the operating environment of the above embodiment 1, the present application provides the following Figure 8 The parsing method of the document shown. Figure 8 4 is a flowchart of a document parsing method according to Embodiment 3 of the present invention.

[0109] Step S801, obtaining the document to be parsed.

[0110] Step S802: Process the document to generate a structure tree corresponding to the document.

[0111] The document can be processed by first parsing the intermediate information of the document, such as the rich text information of the document, and then generating a structure tree corresponding to the document based on the intermediate information of the document.

[0112] Step S803, splitting the document based on the structure tree to obtain multiple information fragments, wherein each information fragment includes at least a document fragment and key information of the document fragment.

[0113] For example, the document fragment is the first paragraph of the document, and the key information of the document fragment is the key information extracted from the first paragraph of the document, such as information such as tags, time, title, paragraph ID, etc.

[0114] Step S804: extract keywords from each information segment, and store the extracted keywords, the information segments corresponding to the keywords, and the documents into a search engine.

[0115] Keywords can be extracted from key information in the document fragment of the information fragment, for example, the keywords: antitrust, challenge. The extracted keywords, the information fragments corresponding to the keywords and the documents are stored in the search engine, which solves the technical problem of low document parsing efficiency, achieves the purpose of improving the document parsing efficiency, and also enriches the content stored in the search engine.

[0116] like Fig. 9As shown, if the document type is word or PDF, the document is split through OneDoc processing (the document can be processed by first parsing the intermediate information of the document, and then generating a structure tree corresponding to the document based on the intermediate information of the document) to obtain multiple information fragments, and then keyword extraction is performed, and the extracted keywords, information fragments, documents, etc. are stored in the search engine.

[0117] Optionally, in the document parsing method of embodiment three of the present invention, the method also includes: if a search term input by a user is detected, matching the search term with keywords in a search engine to determine a target keyword that matches the search term; and returning an information fragment corresponding to the target keyword.

[0118] Since the search engine stores the information fragments corresponding to the keywords, the documents to which the information fragments corresponding to the keywords belong, etc., when the user enters the search term in the search engine to search, more types of relevant information can be returned to the user. For example, the information fragments corresponding to the keywords, or the documents to which the information fragments corresponding to the keywords belong, etc. are returned, so as to help the user quickly and accurately search for the information they want, greatly improving the user's work efficiency.

[0119] Optionally, in the document parsing method of the third embodiment of the present invention, after splitting the document based on the structure tree to obtain multiple information fragments, the method further includes: converting each information fragment into a corresponding image; and storing the converted image in a non-relational database. Returning the information fragment corresponding to the target keyword also includes: extracting the image corresponding to the information fragment from the non-relational database, and returning the image corresponding to the information fragment while returning the information fragment corresponding to the target keyword.

[0120] The above steps are as follows Fig.10As shown, if the document type is PPT (including ppt and pptx), the document is processed by OneDoc technology (for example, the document can be processed by first parsing the intermediate information of the document, and then generating the structure tree corresponding to the document based on the intermediate information of the document), the document is split to obtain multiple knowledge cards (it can also be understood as splitting each page of PPT into a knowledge card), each page of PPT is converted into an image and saved in a non-relational database, and then keywords are extracted from each knowledge card, and the extracted keywords and knowledge cards (which can also be understood as information fragments), documents, etc. are stored in the search engine. When the user subsequently searches for the search terms entered in the search engine, in addition to returning the information fragments corresponding to the keywords, or the documents to which the information fragments corresponding to the keywords belong, etc., the images corresponding to the information fragments can also be extracted from the non-relational database, and the images corresponding to the information fragments are returned at the same time as the information fragments corresponding to the target keywords, so as to help users quickly and accurately search for the information they want, greatly improving the user's work efficiency.

[0121] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0122] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0123] Example 4

[0124] In the operating environment of the above embodiment 1, the present application provides the following Fig.11 The parsing method of the document shown. Fig.11 4 is a flowchart of a document parsing method according to a fourth embodiment of the present invention.

[0125] Step 1101: Receive a document to be parsed uploaded by a client and a service call request sent by the client.

[0126] Step 1102, in response to the service call request, the document is parsed in the server to obtain a document representation; the document representation is processed by natural language to generate a hierarchical structure tree corresponding to the document; and the hierarchical structure tree is converted into a parsed result in a target representation form.

[0127] Step 1103, returning the parsing result to the client.

[0128] Through the above steps, calling the service in the server to parse the document can quickly and accurately obtain the hierarchical structure tree corresponding to the document by using natural language processing technology on the server side, and finally output the parsing result in the target representation form, avoiding the manual parsing of the document, thereby solving the technical problem of low document parsing efficiency in related technologies and achieving the purpose of improving the parsing efficiency of the document.

[0129] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0130] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0131] Example 5

[0132] In the operating environment of the above embodiment 1, the present application provides the following Fig.12 The document reading method shown. Fig.12 4 is a flowchart of a document access method according to Embodiment 5 of the present invention.

[0133] Step 1201, receiving a document retrieval instruction, wherein the document retrieval instruction is used to instruct an analysis of a target document set, wherein the documents in the target document set have the same category, or the document category and the document subcategory are the same, and the document category includes patent documents, judicial documents, contract documents, and chat records.

[0134] It should be noted that the subcategories of the above-mentioned documents refer to different classifications within the same type of documents. For example, contract documents include multiple subcategories such as sales contracts, purchase contracts, loan contracts, lease contracts, etc., and civil judgment documents in judicial documents include: civil judgments, civil rulings, civil mediation documents, etc.

[0135] Step 1202, obtaining a document analysis result according to the parsing result of each document in the target document set and the document retrieval instruction, wherein the parsing result is obtained by parsing the documents in the target document set using the method in the first embodiment.

[0136] Step 1203, display the analysis results in a target display format.

[0137] Through the above steps S1201 to S1203, documents of the same category can be quickly analyzed to obtain analysis results, which are then displayed in a target display format so that users can intuitively view the analysis results, thereby improving user experience.

[0138] For example, the document category in the target document collection is a contract document, and the subcategory of the document is a sales contract. When receiving a document retrieval instruction indicating that the sales contract in the contract document is to be analyzed, the document retrieval instruction is responded to, and the analysis is performed according to the parsing results of each sales contract. Since the parsing results of each sales contract include the information in the contract, such as the sales price, contract requirements, legal terms involved, etc., the analysis results of the sales contract can be quickly analyzed to obtain. For example, the sales prices involved in the sales contract are compared and analyzed, and the sales prices and sales volumes of similar products in different periods can be obtained in the analysis results, and the data can also be intuitively displayed in the form of charts for intuitive viewing by users.

[0139] For another example, the document category in the target document collection is civil judgment documents, and the subcategory of the document is civil ruling documents. When receiving a document retrieval instruction instructing to analyze the civil ruling documents in the civil judgment documents, the document retrieval instruction is responded to, and the analysis is performed according to the parsing results of the civil judgment documents. Since the parsing results of each civil judgment document include the information in the civil judgment document, such as work-related injury compensation, legal clauses involved, etc., the analysis results of the civil judgment documents can be quickly analyzed to obtain the analysis results. For example, the work-related injury compensation involved in the civil judgment documents is compared and analyzed, and the work-related injury compensation corresponding to different work-related injury situations in different years can be obtained in the analysis results, and the data can also be intuitively displayed in the form of a bar chart so that users can intuitively view the analysis results and improve the user experience.

[0140] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0141] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0142] Example 6

[0143] According to an embodiment of the present invention, a device for implementing the document parsing method in the first embodiment is also provided. Fig.13 As shown, the device includes: a first acquisition unit 1301, a first parsing unit 1302, a first generation unit 1303 and a first processing unit 1304.

[0144] Specifically, the first acquisition unit 1301 is used to acquire the document to be parsed;

[0145] A first parsing unit 1302 is used to parse the document to obtain a document representation;

[0146] The first generating unit 1303 is used to generate a hierarchical structure tree corresponding to the document by natural language processing on the document representation;

[0147] The first processing unit 1304 is used to convert the hierarchical structure tree into a parsing result in a target representation form, and output the parsing result.

[0148] In summary, in the document parsing device provided in Example 5 of the present application, the first acquisition unit 1301 is used to acquire the document to be parsed; the first parsing unit 1302 is used to parse the document to obtain a document representation; the first generation unit 1303 is used to generate a hierarchical structure tree corresponding to the document through natural language processing of the document representation; the first processing unit 1304 is used to convert the hierarchical structure tree into a parsing result in the target representation form, and output the parsing result, thereby avoiding the use of manual parsing of the document, thereby solving the technical problem of low document parsing efficiency in related technologies and achieving the purpose of improving the document parsing efficiency.

[0149] Optionally, in the document parsing device provided in Example 5 of the present application, the type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo.

[0150] Optionally, in the document parsing device provided in Example 5 of the present application, the device also includes: a second generation unit, used to generate the parsed content of the target version based on the parsing result after outputting the parsing result; and a first push unit, used to push the parsed content of the target version to the target object.

[0151] Optionally, in the document parsing device provided in Example 5 of the present application, the processing method of the document representation through natural language processing includes at least one of the following: text classification, layout analysis, and title extraction.

[0152] Optionally, in the document parsing device provided in Example 5 of the present application, rich text information in the document is identified while parsing the document, wherein the rich text information includes at least: font information, alignment information, tables, and pictures.

[0153] Optionally, in the document parsing device provided in Example 5 of the present application, the document is any one of the following: a patent document, a judicial document, a contract document, and a chat record.

[0154] Optionally, in the document parsing device provided in Example 5 of the present application, the device also includes: a first receiving unit, used to receive a voice instruction after outputting a parsing result, wherein the voice instruction carries question information; a second acquisition unit, used to respond to the voice instruction and acquire target information matching the question information from the parsing result; and a display unit, used to display the target information.

[0155] It should be noted that the first acquisition unit 1301, the first parsing unit 1302, the first generation unit 1303 and the first processing unit 1304 correspond to steps S101 to S104 in Example 1, and the four units and the corresponding steps implement the same examples and application scenarios, but are not limited to the contents disclosed in the above-mentioned Example 1. It should be noted that the above-mentioned units and modules as part of the device can be run in the computer terminal 10 provided in Example 1.

[0156] Example 6

[0157] According to an embodiment of the present invention, a device for implementing the method for processing contract documents in the second embodiment is also provided. Fig.14 As shown, the device includes: a third acquisition unit 1401, a third generation unit 1402 and a first output unit 1403.

[0158] The third acquisition unit 1401 is used to acquire the contract document to be parsed;

[0159] The third generating unit 1402 is used to process the contract document and generate a hierarchical structure tree corresponding to the contract document;

[0160] The first output unit 1403 is used to convert the hierarchical structure tree into a parsing result in a target representation form, and output the parsing result.

[0161] In summary, in the document parsing device provided in the fifth embodiment of the present application, the third acquisition unit 1401 acquires the contract document to be parsed; the third generation unit 1402 processes the contract document to generate a hierarchical structure tree corresponding to the contract document; the first output unit 1403 converts the hierarchical structure tree into a parsing result in the target representation form, and outputs the parsing result. This avoids the manual parsing of the contract document, thereby solving the technical problem of low parsing efficiency of the document in the related art, and achieving the purpose of improving the parsing efficiency of the document.

[0162] Optionally, in the document parsing device provided in Example 5 of the present application, the device also includes: a first analysis unit, used to analyze the parsing results after outputting the parsing results to determine the extracted content from the parsing results; a second analysis unit, used to analyze the extracted content and identify the target content in the extracted content, wherein the target content includes at least: content classification and clause number; and a fourth generation unit, used to output the identified target content.

[0163] Optionally, in the document parsing device provided in Example 5 of the present application, the device also includes: a detection unit, used to detect whether an editing instruction is received before outputting the marked target content; and a second response unit, used to adjust the terms in the extracted content in response to the editing instruction if the editing instruction is received.

[0164] Optionally, in the document parsing device provided in Example 5 of the present application, the device also includes: a first determination unit, used to determine the approval object corresponding to the content of each structure in the contract document; a first transmission unit, used to transmit the content of each structure to the corresponding approval object; and a fourth acquisition unit, used to obtain the approval result of the approval object.

[0165] Optionally, in the document parsing device provided in Example 5 of the present application, the first transmission unit 1403 includes: a marking sub-unit, used to mark the content related to the corresponding approval object in the content of each structure using a preset marking method, wherein the preset marking method is one of the following: highlighting, bolding, font enlargement, font change; a first transmission sub-unit, used to transmit the marked content to the corresponding approval object.

[0166] Optionally, in the document parsing device provided in Example 5 of the present application, the device also includes: a second determination unit, used to determine the performance matters involved in the contents of each structure in the contract document after obtaining the approval result of the approval object; a fifth generation unit, used to generate a performance plan based on the performance matters; a third determination unit, used to determine the target object related to the performance plan; and a second transmission unit, used to transmit the performance plan to the target object related to the performance plan.

[0167] Optionally, in the document parsing device provided in Example 5 of the present application, the device also includes: a fourth determination unit, used to determine the performance risk involved in the content of each structure in the contract document after obtaining the approval result of the approval object; a sixth generation unit, used to generate a risk management plan based on the performance risk; a fifth determination unit, used to determine the target object involved in the risk management plan; and a third transmission unit, used to transmit the risk management plan to the target object involved in the risk management plan.

[0168] Optionally, in the document parsing device provided in Example 5 of the present application, the device also includes: a second processing unit, which is used to add risk labels to the contents of each structure in the contract document based on the performance risk after determining the performance risk involved in the contents of each structure in the contract document, wherein the risk label is used for risk reminder.

[0169] It should be noted that the above-mentioned units and modules as part of the device can be run in the computer terminal 10 provided in the first embodiment.

[0170] Example 7

[0171] According to an embodiment of the present invention, a device for implementing the document parsing method in the third embodiment is also provided. Fig.15 As shown, the device includes: a fifth acquisition unit 1501, a seventh generation unit 1502, a splitting unit 1503 and an extraction unit 1504.

[0172] Specifically, the fifth acquisition unit 1501 is used to acquire the document to be parsed;

[0173] The seventh generating unit 1502 is used to process the document and generate a structure tree corresponding to the document;

[0174] A splitting unit 1503 is used to split the document based on the structure tree to obtain multiple information fragments, wherein each information fragment includes at least a document fragment and key information of the document fragment;

[0175] The extraction unit 1504 is used to extract keywords from each information segment, and store the extracted keywords, the information segments corresponding to the keywords and the documents into the search engine.

[0176] In summary, in the document parsing device provided in the sixth embodiment of the present application, the fifth acquisition unit 1501 acquires the document to be parsed; the seventh generation unit 1502 processes the document to generate a structure tree corresponding to the document; the splitting unit 1503 splits the document based on the structure tree to obtain multiple information fragments, wherein each information fragment includes at least a document fragment and key information of the document fragment; the extraction unit 1504 extracts keywords from each information fragment, and stores the extracted keywords, the information fragment corresponding to the keyword, and the document into the search engine. This avoids the manual parsing of documents, thereby solving the technical problem of low parsing efficiency of documents in related technologies, and achieving the purpose of improving the parsing efficiency of documents.

[0177] Optionally, in the document parsing device provided in Example 6 of the present application, the device also includes: a sixth determination unit, which is used to determine a target keyword that matches the search word based on matching the search word with keywords in the search engine if a search word input by the user is detected; and a third processing unit, which is used to return an information fragment corresponding to the target keyword.

[0178] Optionally, in the document parsing device provided in Example 6 of the present application, the device also includes: a first conversion unit, used to split the document based on a structure tree to obtain multiple information fragments, convert each information fragment into a corresponding image, and store the converted image into a non-relational database; the third processing unit is also used to extract the image corresponding to the information fragment from the non-relational database, and return the image corresponding to the information fragment while returning the information fragment corresponding to the target keyword.

[0179] It should be noted that the above-mentioned units and modules as part of the device can be run in the computer terminal 10 provided in the first embodiment.

[0180] Example 8

[0181] According to an embodiment of the present invention, a device for implementing the document parsing method in the fourth embodiment is also provided. Fig.16 As shown, the device includes: a second receiving unit 1601, a second parsing unit 1602, a generating unit 1603, a second converting unit 1604 and a fourth processing unit 1605.

[0182] Specifically, the second receiving unit 1601 is used to receive a document to be parsed uploaded by a client and a service call request sent by the client;

[0183] The second parsing unit 1602 is used to respond to the service call request and parse the document in the server to obtain a document representation;

[0184] A generating unit 1603 is used to generate a hierarchical structure tree corresponding to the document by natural language processing on the document representation;

[0185] The second conversion unit 1604 is used to convert the hierarchical structure tree into a parsing result in a target representation form;

[0186] The fourth processing unit 1605 is used to return the parsing result to the client.

[0187] Through the above scheme, the service in the server is called to parse the document. The natural language processing technology can be used on the server to quickly and accurately obtain the hierarchical structure tree corresponding to the document, and finally the parsing result is output in the target representation form, avoiding the manual parsing of the document, thereby solving the technical problem of low parsing efficiency of documents in related technologies and achieving the purpose of improving the parsing efficiency of documents.

[0188] It should be noted that the above-mentioned unit is a part of the device and can be run in the computer terminal 10 provided in the first embodiment.

[0189] Example 9

[0190] According to an embodiment of the present invention, a device for implementing the document access method in the fifth embodiment is also provided. Fig.17 As shown, the device includes: a third receiving unit 1701, a sixth acquiring unit 1702 and a display unit 1703.

[0191] Specifically, the third receiving unit 1701 is used to receive a document retrieval instruction, wherein the document retrieval instruction is used to instruct an analysis of a target document set, wherein the documents in the target document set have the same category, or the document category and the document subcategory are the same, and the document category includes patent documents, judicial documents, contract documents, and chat records.

[0192] It should be noted that the subcategories of the above-mentioned documents refer to different classifications within the same type of documents. For example, contract documents include multiple subcategories such as sales contracts, purchase contracts, loan contracts, lease contracts, etc., and civil judgment documents in judicial documents include: civil judgments, civil rulings, civil mediation documents, etc.

[0193] A sixth acquisition unit 1702, configured to obtain a document analysis result according to the parsing result of each document in the target document set and the document retrieval instruction, wherein the parsing result is obtained by parsing the documents in the target document set by one of the methods described above;

[0194] The display unit 1703 is used to display the analysis result in a target display form.

[0195] Through the document retrieval device provided in Example 9 of the present application, documents of the same category can be quickly analyzed to obtain analysis results, and displayed in a target display format so that users can intuitively view the analysis results, thereby improving user experience.

[0196] For example, the document category in the target document collection is a contract document, and the subcategory of the document is a sales contract. When receiving a document retrieval instruction indicating that the sales contract in the contract document is to be analyzed, the document retrieval instruction is responded to, and the analysis is performed according to the parsing results of each sales contract. Since the parsing results of each sales contract include the information in the contract, such as the sales price, contract requirements, legal terms involved, etc., the analysis results of the sales contract can be quickly analyzed to obtain. For example, the sales prices involved in the sales contract are compared and analyzed, and the sales prices and sales volumes of similar products in different periods can be obtained in the analysis results, and the data can also be intuitively displayed in the form of charts for intuitive viewing by users.

[0197] For another example, the document category in the target document collection is civil judgment documents, and the subcategory of the document is civil ruling documents. When receiving a document retrieval instruction instructing to analyze the civil ruling documents in the civil judgment documents, the document retrieval instruction is responded to, and the analysis is performed according to the parsing results of the civil judgment documents. Since the parsing results of each civil judgment document include the information in the civil judgment document, such as work-related injury compensation, legal clauses involved, etc., the analysis results of the civil judgment documents can be quickly analyzed to obtain the analysis results. For example, the work-related injury compensation involved in the civil judgment documents is compared and analyzed, and the work-related injury compensation corresponding to different work-related injury situations in different years can be obtained in the analysis results, and the data can also be intuitively displayed in the form of a bar chart so that users can intuitively view the analysis results and improve the user experience.

[0198] It should be noted that the above-mentioned units and sub-units as part of the device can be run in the computer terminal 10 provided in the first embodiment.

[0199] Example 10

[0200] The embodiment of the present invention can provide a computer terminal, which can be any computer terminal device in a computer terminal group. Optionally, in this embodiment, the computer terminal can also be replaced by a terminal device such as a mobile terminal.

[0201] Optionally, in this embodiment, the computer terminal may be located in at least one network device among a plurality of network devices of the computer network.

[0202] The computer terminal provided in the embodiment of the present application includes a computer program product, including a computer program / instruction, which performs the following method when executed by a processor: obtaining a document to be parsed; parsing the document to obtain a document representation; generating a hierarchical structure tree corresponding to the document through natural language processing of the document representation; converting the hierarchical structure tree into a parsing result in a target representation form, and outputting the parsing result.

[0203] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: the type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo.

[0204] Optionally, when the computer program / instruction is executed by a processor, the following method is performed: after outputting the parsing result, the method further includes: generating parsed content of a target version based on the parsing result; and pushing the parsed content of the target version to a target object.

[0205] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: the document representation is processed by natural language processing in a manner including at least one of the following: text classification, layout analysis, and title extraction.

[0206] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: while parsing the document, rich text information in the document is identified, wherein the rich text information includes at least: font information, alignment information, tables, and pictures.

[0207] Optionally, when the computer program / instructions are executed by the processor, the following method is performed: the document is any one of the following: a patent document, a judicial document, a contract document, and a chat record.

[0208] Optionally, the computer program / instruction performs the following method when executed by the processor: after outputting the analysis result, the method also includes: receiving a query instruction, wherein the query instruction carries query information and the query instruction is a text instruction or a voice instruction; responding to the query instruction, obtaining target information matching the query information from the analysis result; and displaying the target information.

[0209] Optionally, when the computer program / instructions are executed by the processor, the following method is performed: obtaining a contract document to be parsed; processing the contract document to generate a hierarchical structure tree corresponding to the contract document; converting the hierarchical structure tree into a parsing result in a target representation form, and outputting the parsing result.

[0210] Optionally, the computer program / instruction performs the following method when executed by the processor: after outputting the parsing result, the method also includes: analyzing the parsing result to determine the extracted content from the parsing result; analyzing the extracted content and identifying the target content in the extracted content, wherein the target content includes at least: content classification and clause number; and outputting the identified target content.

[0211] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: before outputting the marked target content, the method also includes: detecting whether an editing instruction is received; if an editing instruction is received, adjusting the terms in the extracted content in response to the editing instruction.

[0212] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: the method also includes: determining the approval object corresponding to the content of each structure in the contract document; transmitting the content of each structure to the corresponding approval object; and obtaining the approval result of the approval object.

[0213] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: transmitting the content of each structure to the corresponding approval object includes: marking the content related to the corresponding approval object in the content of each structure using a preset marking method, wherein the preset marking method is one of the following: highlighting, bolding, font enlargement, font change; transmitting the marked content to the corresponding approval object.

[0214] Optionally, the computer program / instructions perform the following method when executed by the processor: after obtaining the approval result of the approval object, the method also includes: determining the performance matters involved in the contents of each structure in the contract document; generating a performance plan based on the performance matters; determining the target object related to the performance plan; and transmitting the performance plan to the target object related to the performance plan.

[0215] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: after obtaining the approval result of the approval object, the method also includes: determining the performance risk involved in the content of each structure in the contract document; generating a risk management plan based on the performance risk; determining the target object involved in the risk management plan; and transmitting the risk management plan to the target object involved in the risk management plan.

[0216] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: after determining the performance risk involved in the content of each structure in the contract document, the method also includes: adding a risk label to the content of each structure in the contract document based on the performance risk, wherein the risk label is used for risk reminder.

[0217] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: obtaining a document to be parsed; processing the document to generate a structure tree corresponding to the document; splitting the document based on the structure tree to obtain multiple information fragments, wherein each information fragment includes at least a document fragment and key information of the document fragment; extracting keywords from each information fragment, and storing the extracted keywords, the information fragments corresponding to the keywords, and the document into a search engine.

[0218] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: if a search term input by a user is detected, a target keyword matching the search term is determined based on matching the search term with a keyword in the search engine; and an information segment corresponding to the target keyword is returned.

[0219] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: after splitting the document based on the structure tree to obtain multiple information fragments, the method also includes: converting each information fragment into a corresponding image; storing the converted image in a non-relational database; returning the information fragment corresponding to the target keyword also includes: extracting the image corresponding to the information fragment from the non-relational database, and returning the image corresponding to the information fragment while returning the information fragment corresponding to the target keyword.

[0220] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: receiving a document to be parsed uploaded by a client and a service call request sent by the client; in response to the service call request, parsing the document in the server to obtain a document representation; processing the document representation through natural language to generate a hierarchical structure tree corresponding to the document; converting the hierarchical structure tree into a parsing result in a target representation form; and returning the parsing result to the client.

[0221] Optionally, when the computer program / instruction is executed by the processor, the following method is performed: receiving a document retrieval instruction, wherein the document retrieval instruction is used to instruct analysis of a target document set, wherein the document categories in the target document set are the same, or the document categories and document subcategories are the same, and the document categories include patent documents, judicial documents, contract documents, and chat records; obtaining a document analysis result based on the parsing result of each document in the target document set and the document retrieval instruction, wherein the parsing result is obtained by parsing the documents in the target document set by any of the methods described above; and displaying the analysis result in a target display form.

[0222] Optionally, Fig.18 is a structural block diagram of a computer terminal according to an embodiment of the present invention. Fig.18 As shown, the computer terminal may include: one or more ( Fig.18 Only one processor and memory are shown.

[0223] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the document parsing method and device in the embodiment of the present invention. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizing the above-mentioned document parsing method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories can be connected to the terminal A via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0224] The processor can call the information and application stored in the memory through the transmission device to execute the following steps: obtain the document to be parsed; parse the document to obtain a document representation; generate a hierarchical structure tree corresponding to the document through natural language processing of the document representation; convert the hierarchical structure tree into a parsing result in the target representation form, and output the parsing result.

[0225] Optionally, the processor may also execute program code of the following steps: the type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo.

[0226] Optionally, the processor may also execute program code of the following steps: after outputting the parsing result, the method further includes: generating parsed content of a target version based on the parsing result; and pushing the parsed content of the target version to a target object.

[0227] Optionally, the processor may also execute program code of the following steps: the document representation is processed by natural language processing including at least one of the following: text classification, layout analysis, and title extraction.

[0228] Optionally, the processor may also execute the program code of the following steps: parsing the document and identifying rich text information in the document, wherein the rich text information includes at least: font information, alignment information, tables, and pictures.

[0229] Optionally, the processor may also execute the program code of the following steps: the document is any one of the following: a patent document, a judicial document, a contract document, or a chat record.

[0230] Optionally, the processor may also execute program code of the following steps: after outputting the analysis result, the method further includes: receiving a query instruction, wherein the query instruction carries query information, and the query instruction is a text instruction or a voice instruction; responding to the query instruction, obtaining target information matching the query information from the analysis result; and displaying the target information.

[0231] Optionally, the processor may also execute the program code of the following steps: obtaining the contract document to be parsed; processing the contract document to generate a hierarchical structure tree corresponding to the contract document; converting the hierarchical structure tree into a parsing result in a target representation form, and outputting the parsing result.

[0232] Optionally, the processor may also execute program code of the following steps: after outputting the parsing result, the method further includes: analyzing the parsing result to determine extracted content from the parsing result; analyzing the extracted content and identifying target content in the extracted content, wherein the target content includes at least: content classification and clause number; and outputting the identified target content.

[0233] Optionally, the processor may also execute program code of the following steps: before outputting the identified target content, the method further includes: detecting whether an editing instruction is received; if an editing instruction is received, adjusting the terms in the extracted content in response to the editing instruction.

[0234] Optionally, the processor may also execute program code of the following steps: the method also includes: determining the approval object corresponding to the content of each structure in the contract document; transmitting the content of each structure to the corresponding approval object; and obtaining the approval result of the approval object.

[0235] Optionally, the processor may also execute the program code of the following steps: transmitting the contents of each structure to the corresponding approval object includes: marking the contents of each structure related to the corresponding approval object using a preset marking method, wherein the preset marking method is one of the following: highlighting, bolding, enlarging the font, changing the font; transmitting the marked content to the corresponding approval object.

[0236] Optionally, the processor may also execute program code for the following steps: after obtaining the approval result of the approval object, the method further includes: determining the performance matters involved in the contents of each structure in the contract document; generating a performance plan based on the performance matters; determining the target object related to the performance plan; and transmitting the performance plan to the target object related to the performance plan.

[0237] Optionally, the processor may also execute program code of the following steps: after obtaining the approval result of the approval object, the method further includes: determining the performance risk involved in the content of each structure in the contract document; generating a risk management plan based on the performance risk; determining the target object involved in the risk management plan; and transmitting the risk management plan to the target object involved in the risk management plan.

[0238] Optionally, the processor may also execute program code for the following steps: after determining the performance risk involved in the contents of each structure in the contract document, the method further includes: adding risk tags to the contents of each structure in the contract document based on the performance risk, wherein the risk tags are used for risk reminders.

[0239] Optionally, the processor may also execute the following program code steps: obtaining a document to be parsed; processing the document to generate a structure tree corresponding to the document; splitting the document based on the structure tree to obtain multiple information fragments, wherein each information fragment includes at least a document fragment and key information of the document fragment; extracting keywords from each information fragment, and storing the extracted keywords, the information fragments corresponding to the keywords, and the document into a search engine.

[0240] Optionally, the processor may also execute program code of the following steps: the method further includes: if a search term input by a user is detected, determining a target keyword matching the search term based on matching the search term with a keyword in the search engine; and returning an information fragment corresponding to the target keyword.

[0241] Optionally, the processor may also execute program code for the following steps: after splitting the document based on the structure tree to obtain a plurality of information fragments, the method further includes: converting each information fragment into a corresponding image; storing the converted image in a non-relational database; returning the information fragment corresponding to the target keyword further includes: extracting the image corresponding to the information fragment from the non-relational database, and returning the image corresponding to the information fragment while returning the information fragment corresponding to the target keyword.

[0242] Optionally, the processor may also execute program codes for the following steps: receiving a document to be parsed uploaded by a client and a service call request sent by the client; parsing the document in a server in response to the service call request to obtain a document representation; processing the document representation through natural language to generate a hierarchical structure tree corresponding to the document; converting the hierarchical structure tree into a parsing result in a target representation form; and returning the parsing result to the client.

[0243] Optionally, the processor may also execute program code for the following steps: receiving a document retrieval instruction, wherein the document retrieval instruction is used to instruct analysis of a target document set, wherein the document categories in the target document set are the same, or the document categories and document subcategories are the same, and the document categories include patent documents, judicial documents, contract documents, and chat records; obtaining a document analysis result based on the parsing result of each document in the target document set and the document retrieval instruction, wherein the parsing result is obtained by parsing the documents in the target document set by any of the methods described above; and displaying the analysis result in a target display form.

[0244] By adopting the embodiment of the present invention, a solution of a document parsing method is provided, which comprises the following steps: obtaining a document to be parsed; parsing the document to obtain a document representation; processing the document representation through natural language to generate a hierarchical structure tree corresponding to the document; converting the hierarchical structure tree into a parsing result in a target representation form, and outputting the parsing result, that is, utilizing natural language processing technology to quickly and accurately obtain the hierarchical structure tree corresponding to the document, and finally outputting the parsing result in the target representation form, thereby avoiding the manual parsing of the document, thereby solving the technical problem of low parsing efficiency of the document in the related technology, and achieving the purpose of improving the parsing efficiency of the document.

[0245] It can be understood by those skilled in the art that Fig.18 The structure shown is for illustration only, and the computer terminal may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a handheld computer, a mobile Internet device (Mobile Internet Devices, MID), a PAD, or other terminal devices. Fig.18 The structure of the electronic device is not limited. For example, the computer terminal 10 may also include Fig.18 More or fewer components (such as network interfaces, display devices, etc.) shown in, or having Fig.18 Different configurations are shown.

[0246] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0247] Embodiment 11

[0248] The embodiment of the present invention further provides a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the method provided in the above embodiment.

[0249] Optionally, in this embodiment, the storage medium may be located at any location in a computer terminal group in a computer network.

[0250] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0251] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0252] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0253] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0254] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0255] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0256] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A document parsing method, It is characterized in that include: Get the document to be parsed; Parsing the document to obtain a document representation; Processing the document representation through natural language to generate a hierarchical structure tree corresponding to the document; Convert the hierarchical structure tree into a parsing result in a target representation form, and output the parsing result; The type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo; The document representation is processed by natural language processing including at least one of the following: text classification, layout analysis, and title extraction; The document representation is processed by natural language to generate a hierarchical structure tree corresponding to the document, including: If there is explicit serial number information in the text, the word vector calculation and text classification tools and the pre-trained model of the word vector are used to perform text classification to identify the boundary text paragraphs of each module, wherein the module includes at least one of the following: title, body, text, signature, and attachment; If there is no explicit sequence number information in the text, a sequence labeling method is used to identify the boundary text paragraphs of each module; For the main text, a heuristic algorithm is used according to different sequence number categories to generate a hierarchical structure tree corresponding to the document.

2. The analysis method according to claim 1, It is characterized in that After outputting the parsing result, the method further includes: Based on the parsing result, generating parsed content of the target version; Push the parsed content of the target version to the target object.

3. The analysis method according to claim 1, It is characterized in that The rich text information in the document is identified while parsing the document, wherein the rich text information at least includes: font information, alignment information, tables, and pictures.

4. The analysis method according to claim 1, It is characterized in that The document is any one of the following: patent documents, judicial documents, contract documents, and chat records.

5. The analysis method according to claim 1, It is characterized in that After outputting the parsing result, the method further includes: Receiving a query instruction, wherein the query instruction carries query information and the query instruction is a text instruction or a voice instruction; In response to the query instruction, obtaining target information matching the query information from the parsing result; The target information is displayed.

6. A method for handling contract documents. It is characterized in that include: Get the contract file to be parsed; Processing the contract document to generate a hierarchical structure tree corresponding to the contract document; Convert the hierarchical structure tree into a parsing result in a target representation form, and output the parsing result; The type of the contract document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo; Processing the contract document to generate a hierarchical structure tree corresponding to the contract document includes: parsing the contract document to obtain a contract document representation; processing the contract document representation through natural language to generate a hierarchical structure tree corresponding to the contract document; The contract document is represented by natural language processing, including at least one of the following: text classification, layout analysis, and title extraction; The contract document is represented by natural language processing to generate a hierarchical structure tree corresponding to the contract document, including: If there is explicit serial number information in the text, the word vector calculation and text classification tools and the pre-trained model of the word vector are used to perform text classification to identify the boundary text paragraphs of each module, wherein the module includes at least one of the following: title, body, text, signature, and attachment; If there is no explicit sequence number information in the text, a sequence labeling method is used to identify the boundary text paragraphs of each module; For the main text, a heuristic algorithm is used according to different serial number categories to generate a hierarchical structure tree corresponding to the contract document.

7. The processing method according to claim 6, It is characterized in that After outputting the parsing result, the method further includes: Analyzing the parsing results to determine the extracted content from the parsing results; Analyze the extracted content and identify target content in the extracted content, wherein the target content at least includes: content classification and clause number; Output the marked target content.

8. The processing method according to claim 7, It is characterized in that Before outputting the marked target content, the method further includes: Detect whether an editing instruction is received; If an editing instruction is received, the clauses in the extracted content are adjusted in response to the editing instruction.

9. The processing method according to claim 8, It is characterized in that The method further comprises: Determine the approval object corresponding to the content of each structure in the contract document; Transfer the contents of each structure to the corresponding approval object; Get the approval result of the approval object.

10. The processing method according to claim 9, It is characterized in that The contents of each structure are transferred to the corresponding approval objects including: Marking the content related to the corresponding approval object in the content of each structure using a preset marking method, wherein the preset marking method is one of the following: highlighting, bolding, enlarging the font, and changing the font; Transfer the marked content to the corresponding approval object.

11. The processing method according to claim 9, It is characterized in that After obtaining the approval result of the approval object, the method further includes: Determine the performance matters involved in the contents of each structure in the contract documents; Generate a performance plan based on the performance matters; Identify the target audiences relevant to the implementation plan; The fulfillment plan is transmitted to a target object associated with the fulfillment plan.

12. The processing method according to claim 9, It is characterized in that After obtaining the approval result of the approval object, the method further includes: Determine the performance risks involved in the content of each structure in the contract documents; Generate a risk management plan based on the performance risks; Identify the target objects involved in the risk management plan; The risk management plan is transmitted to the target objects involved in the risk management plan.

13. The processing method according to claim 12, It is characterized in that After determining the performance risk involved in the contents of each structure in the contract document, the method further includes: Based on the performance risk, a risk label is added to the content of each structure in the contract document, wherein the risk label is used for risk reminder.

14. A document parsing method, It is characterized in that include: Get the document to be parsed; Processing the document to generate a structure tree corresponding to the document; Splitting the document based on the structure tree to obtain multiple information fragments, wherein each information fragment includes at least a document fragment and key information of the document fragment; Extracting keywords from each information fragment, and storing the extracted keywords, the information fragment corresponding to the keywords and the document in a search engine; The type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo; Processing the document to generate a structure tree corresponding to the document includes: parsing the document to obtain a document representation; processing the document representation through natural language to generate a structure tree corresponding to the document; The document representation is processed by natural language processing including at least one of the following: text classification, layout analysis, and title extraction; The document representation is processed by natural language to generate a structure tree corresponding to the document, including: If there is explicit serial number information in the text, the word vector calculation and text classification tools and the pre-trained model of the word vector are used to perform text classification to identify the boundary text paragraphs of each module, wherein the module includes at least one of the following: title, body, text, signature, and attachment; If there is no explicit sequence number information in the text, a sequence labeling method is used to identify the boundary text paragraphs of each module; For the main text, a heuristic algorithm is used according to different sequence number categories to generate a structure tree corresponding to the document.

15. The analysis method according to claim 14, It is characterized in that The method further comprises: If a search term input by a user is detected, a target keyword matching the search term is determined based on matching the search term with a keyword in the search engine; Return the information fragment corresponding to the target keyword.

16. The analysis method according to claim 15, It is characterized in that After the document is split based on the structure tree to obtain multiple information fragments, the method further includes: converting each information fragment into a corresponding image; storing the converted image in a non-relational database; Returning the information segment corresponding to the target keyword further includes: extracting a picture corresponding to the information segment from the non-relational database, and returning the picture corresponding to the information segment while returning the information segment corresponding to the target keyword.

17. A document parsing method, It is characterized in that include: Receiving a document to be parsed uploaded by a client and a service call request sent by the client; In response to the service call request, the document is parsed in the server to obtain a document representation; the document representation is processed by natural language to generate a hierarchical structure tree corresponding to the document; and the hierarchical structure tree is converted into a parsing result in a target representation form; Return the parsing result to the client; The type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo; The document representation is processed by natural language processing including at least one of the following: text classification, layout analysis, and title extraction; The document representation is processed by natural language to generate a hierarchical structure tree corresponding to the document, including: If there is explicit serial number information in the text, the word vector calculation and text classification tools and the pre-trained model of the word vector are used to perform text classification to identify the boundary text paragraphs of each module, wherein the module includes at least one of the following: title, body, text, signature, and attachment; If there is no explicit sequence number information in the text, a sequence labeling method is used to identify the boundary text paragraphs of each module; For the main text, a heuristic algorithm is used according to different sequence number categories to generate a hierarchical structure tree corresponding to the document.

18. A document access method, It is characterized in that include: Receiving a document retrieval instruction, wherein the document retrieval instruction is used to instruct to analyze a target document set, wherein the documents in the target document set have the same category, or the document category and the document subcategory are the same, and the document category includes patent documents, judicial documents, contract documents, and chat records; Obtaining a document analysis result according to the parsing result of each document in the target document set and the document retrieval instruction, wherein the parsing result is obtained by parsing the documents in the target document set according to the method of any one of claims 1 to 5; The analysis results are presented in a target presentation format.

19. A document parsing device, It is characterized in that include: A first acquisition unit, used to acquire a document to be parsed; A first parsing unit, used to parse the document to obtain a document representation; A first generating unit, configured to generate a hierarchical structure tree corresponding to the document by natural language processing on the document representation; A first processing unit, configured to convert the hierarchical structure tree into a parsing result in a target representation form, and output the parsing result; The type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo; The document representation is processed by natural language processing including at least one of the following: text classification, layout analysis, and title extraction; The first generation unit is also used for: if there is explicit serial number information in the text, using word vector calculation and text classification tools as well as a pre-trained model of word vectors to perform text classification to identify the boundary text paragraphs of each module, wherein the module includes at least one of the following: title, main body, body, signature, and attachment; if there is no explicit serial number information in the text, using a sequence labeling method to identify the boundary text paragraphs of each module; for the body, using a heuristic algorithm according to different serial number categories to generate a hierarchical structure tree corresponding to the document.

20. A contract document parsing device, It is characterized in that include: A third acquisition unit is used to acquire the contract document to be parsed; A third generating unit is used to process the contract document and generate a hierarchical structure tree corresponding to the contract document; A first output unit, configured to convert the hierarchical structure tree into a parsing result in a target representation form, and output the parsing result; The type of the contract document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo; The third generating unit is further used to: parse the contract document to obtain a contract document representation; and generate a hierarchical structure tree corresponding to the contract document by natural language processing of the contract document representation; The contract document is represented by natural language processing, including at least one of the following: text classification, layout analysis, and title extraction; The third generation unit is also used for: if there is explicit serial number information in the text, using word vector calculation and text classification tools as well as a pre-trained model of word vectors to perform text classification to identify the boundary text paragraphs of each module, wherein the module includes at least one of the following: title, main body, body, signature, and attachment; if there is no explicit serial number information in the text, using a sequence labeling method to identify the boundary text paragraphs of each module; for the body, using a heuristic algorithm according to different serial number categories to generate a hierarchical structure tree corresponding to the contract document.

21. A document parsing device, It is characterized in that include: A fifth acquisition unit, used to acquire a document to be parsed; A seventh generating unit, used for processing the document to generate a structure tree corresponding to the document; A splitting unit, used for splitting the document based on the structure tree to obtain a plurality of information fragments, wherein each information fragment includes at least a document fragment and key information of the document fragment; An extraction unit, used to extract keywords from each information fragment, and store the extracted keywords, the information fragment corresponding to the keywords and the document into a search engine; The type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo; The seventh generating unit is further used to: parse the document to obtain a document representation; and generate a structure tree corresponding to the document by natural language processing the document representation; The document representation is processed by natural language processing including at least one of the following: text classification, layout analysis, and title extraction; The seventh generation unit is also used for: if there is explicit serial number information in the text, then using word vector calculation and text classification tools as well as a pre-trained model of word vectors to perform text classification to identify the boundary text paragraphs of each module, wherein the module includes at least one of the following: title, main body, body, signature, attachment; if there is no explicit serial number information in the text, then using a sequence labeling method to identify the boundary text paragraphs of each module; for the body, using a heuristic algorithm according to different serial number categories to generate a structure tree corresponding to the document.

22. A document parsing device, It is characterized in that include: A second receiving unit, configured to receive a document to be parsed uploaded by a client and a service call request sent by the client; A second parsing unit, configured to respond to the service call request, parse the document in the server, and obtain a document representation; A generating unit, configured to generate a hierarchical structure tree corresponding to the document by natural language processing on the document representation; A second conversion unit, used for converting the hierarchical structure tree into a parsing result in a target representation form; A fourth processing unit, used for returning the parsing result to the client; The type of the document is any one of the following: Word, Excel, PDF, scanned copy, picture, photo; The document representation is processed by natural language processing including at least one of the following: text classification, layout analysis, and title extraction; The generation unit is also used for: if there is explicit serial number information in the text, using word vector calculation and text classification tools as well as a pre-trained model of word vectors to perform text classification to identify the boundary text paragraphs of each module, wherein the module includes at least one of the following: title, main body, body, signature, and attachment; if there is no explicit serial number information in the text, using a sequence labeling method to identify the boundary text paragraphs of each module; for the body, using a heuristic algorithm according to different serial number categories to generate a hierarchical structure tree corresponding to the document.

23. A document access device, It is characterized in that include: A third receiving unit is used to receive a document retrieval instruction, wherein the document retrieval instruction is used to instruct to analyze a target document set, the documents in the target document set have the same category, or the document category and the document subcategory are the same, and the document category includes patent documents, judicial documents, contract documents, and chat records; a sixth acquisition unit, configured to obtain a document analysis result according to the parsing result of each document in the target document set and the document retrieval instruction, wherein the parsing result is obtained by parsing the documents in the target document set according to the method of any one of claims 1 to 5; The display unit is used to display the analysis result in a target display form.

24. A computer-readable storage medium, It is characterized in that The storage medium includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the document parsing method described in any one of claims 1 to 5, or the contract document processing method described in any one of claims 6 to 13, or the document parsing method described in any one of claims 14 to 16, or the document parsing method described in claim 17, or the document retrieval method described in claim 18.

25. A processor, It is characterized in that The processor is used to run a program, wherein the program, when running, executes the document parsing method described in any one of claims 1 to 5, or the contract document processing method described in any one of claims 6 to 13, or the document parsing method described in any one of claims 14 to 16, or the document parsing method described in claim 17, or the document retrieval method described in claim 18.

26. A computer program product comprising a computer program / instructions, It is characterized in that When the computer program / instructions are executed by a processor, the computer program / instructions perform the document parsing method described in any one of claims 1 to 5, or the contract document processing method described in any one of claims 6 to 13, or the document parsing method described in any one of claims 14 to 16, or the document parsing method described in claim 17, or the document retrieval method described in claim 18.

Images