Automatic labeling, modifying method, device

Abstract

The application discloses a patent-related text automatic labeling and modifying method and device. The automatic labeling method comprises the following steps: determining a claim or a specification range in the patent-related text; defining structural names in the claim or the specification range; performing hierarchical division on at least two defined structural names; and labeling the at least two structural names according to the hierarchical division. The application can automatically label the patent-related text, is convenient and fast, and reduces the risk of manual errors in the process.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to an automatic labeling and modification method and device. Background Technology

[0002] Product-related patent documents generally include a specification with accompanying drawings, and the accompanying drawings contain... Figure One Generally, structures need to be numbered, and the corresponding structural names in the instruction manual also need to be numbered. Matching the structural names in the instruction manual to their corresponding numbers is usually a time-consuming and meticulous process, especially when the product structure is complex and there are many types of structural names. The numbers also need to be distinguished according to the relationships between the structural names. Currently, the numbers are usually found manually one by one, which is a tedious process and prone to errors. Summary of the Invention

[0003] This application provides an automatic labeling and modification method and device for patent-related texts, which can improve the processing efficiency and accuracy of patent-related texts.

[0004] To address the aforementioned technical problems, the first aspect of this application provides an automatic labeling method for patent-related texts. This method includes: determining the scope of claims or specification in the patent-related texts; defining structural names within the scope of claims or specification; hierarchically classifying at least two defined structural names; and labeling the at least two structural names according to the hierarchical classification.

[0005] To address the aforementioned technical problems, a second aspect of this application provides an automatic labeling method for patent-related texts. The method includes: determining the scope of claims or specification in the patent-related texts; defining the structural names in the scope of claims or specification; and labeling at least two of the defined structural names using random labeling or predefined labeling.

[0006] To address the aforementioned technical problems, a third aspect of this application provides an automatic modification method for patent-related text. The method includes: receiving a manual revision termination instruction for the patent-related text; in response to the manual revision termination instruction, or in response to meeting preset conditions, determining the pre-revision content in the patent-related text and defining the scope of the claims or the structural names in the specification; in response to matching the pre-revision content with the structural names, matching the remaining or selected portions of the patent-related text with the structural names or the pre-revision content; and revising all or part of the matched target content based on the revised content.

[0007] To solve the above-mentioned technical problems, the fourth aspect of this application provides a computer device, which includes: a processor; a memory connected to the processor for storing program instructions, which can be executed by the processor to implement the automatic labeling method for patent-related texts provided in the first or second aspect, or the automatic modification method for patent-related texts provided in the third aspect.

[0008] The beneficial effects of this application are as follows: Unlike the prior art, the embodiments of this application define the structural names in the claims or specification; and classify the at least two defined structural names hierarchically, thereby replacing manual identification and relationship classification of structural names, and automatically labeling them according to the relationship classification, which can improve the processing efficiency and accuracy of patent-related texts. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein:

[0010] Figure 1 This is a flowchart illustrating an embodiment of the automatic labeling method for texts related to this patent application;

[0011] Figure 2 This is a schematic diagram of the first interface of an embodiment of the automatic modification method for the relevant text of this patent application;

[0012] Figure 3 This is a schematic diagram of the second interface of an embodiment of the automatic modification method for the relevant text of this patent application;

[0013] Figure 4 This is a schematic diagram of the structure of an embodiment of the computer device provided in this application;

[0014] Figure 5 This is a schematic diagram of the structure of an embodiment of the computer-readable storage medium provided in this application;

[0015] Figure 6 This is a schematic diagram of the client and server structure in one embodiment of the automatic modification method for the relevant text of this patent application. Detailed Implementation

[0016] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0017] In this document, the reference to "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0018] Currently, part of the work of patent professionals involves drafting patent documents, creating patent-related texts, and then submitting them as patent applications through official channels. Patent applications generally include invention applications, design applications, or patent-related documents. Invention applications typically include an abstract, claims, and a specification. Due to patent law requirements, parts of the abstract, claims, and specification may be substantially identical. Furthermore, parts of different claims or different sections of the same claim may also be substantially identical, as may parts of the specification in different locations. Therefore, if a sentence or a few words in the patent-related text are modified, the user may need to find and modify the rest of the text to ensure consistency. Current text editing tools offer "find" and "replace" functions. Users need to operate the text editing software interface, open the "find" or "replace" window, enter the content to be found and replaced in the input field, and then click "find" or "replace all."

[0019] The aforementioned "Find" and "Replace" functions offer a faster method than users searching one by one. However, as patent work moves towards higher quality requirements and the work of applicants and patent agencies becomes more standardized, patent-related texts may need to undergo multiple revisions and improvements before being submitted as patent applications through official channels. Each revision and improvement may require several or even more find and replace operations. Users need to frequently operate the text editing software interface, open the "Find" and "Replace" windows, enter the content to be found and replaced in the input fields, and then click "Find" or "Replace All," resulting in relatively low work efficiency.

[0020] Furthermore, the "find" and "replace" functions of current word processing software can only perform 100% matching. For example, when replacing "abc", it can only find other content containing the string "abc", but cannot find content with labels like "a11bc" or content like "a said bc", and vice versa. This creates many blind spots in the matching process, requiring users to find them one by one and make consistent modifications, which cannot further improve work efficiency.

[0021] On the other hand, matching the corresponding structural names with their corresponding labels in the instruction manual is a time-consuming and meticulous process, especially when the product structure is complex and there are many types of structural names. The labels also need to be distinguished according to the relationships between the structural names. Currently, the labels are usually found manually one by one, which is a tedious process and prone to errors.

[0022] The following describes in detail, with reference to embodiments, the automatic labeling method, device and storage medium for patent-related texts provided by the present invention, which can partially or completely solve the above-mentioned technical problems.

[0023] Please see Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the automatic labeling method for text related to this patent application. Figure 1 The example only shows the operation flow for the claims; the automatic numbering method for patent-related text includes:

[0024] S100: Determine the scope of the claims or specification in the patent-related text;

[0025] The purpose of defining the scope of the claims or specification is to determine the structural names of each component of the product, facilitating subsequent labeling of these structural names. From a computer's perspective, it cannot see or know which words in a sentence within a document are correct structural names, or whether a sentence contains any structural names at all. A document in a computer's view is only a set of binary encoded data; it cannot distinguish which are structural names and which are not, nor can it discern the left and right boundaries of structural names. Therefore, it is essential to first determine the scope of the claims or specification in the patent-related text.

[0026] An invention application and the patent-related texts involved in it generally include an abstract, claims, a description, abstract drawings, and description drawings. If the claims or description are separate documents, then the scope of the claims or description is the entire document.

[0027] If the claims and at least one of the abstract, description, abstract drawings, and description drawings constitute a single document, or if the description and at least one of the abstract, claims, abstract drawings, and description drawings constitute a single document, then means are needed to determine the scope of the claims or description. The following example illustrates how to determine the scope of the claims:

[0028] S110: Locate the "claims" in the patent-related texts;

[0029] Patent-related texts have a fixed format, with each section beginning with a module name. "Claims" is the module name; finding this module name means that the claims will follow this module name.

[0030] S120: In response to finding the "claims", take the found result as the first boundary and search for "technical field", "abstract", "abstract drawing", "drawings", "description" or document end mark after the "claims";

[0031] If there is a next module after the "Claims" module in the patent-related text, then the next module is one of "Technical Field", "Abstract", "Abstract Drawings", "Drawings", or "Specification". If there is no next module, then the end of the document is a file end marker.

[0032] S130: In response to finding "technical field", "specification abstract", "abstract drawing", "specification drawing", "specification" or document end mark, the found result shall be used as the second boundary, and the area between the first boundary and the second boundary shall be used as the scope of the claims.

[0033] S200: Define the structural names in the claims or scope of the specification;

[0034] As mentioned earlier, computers cannot distinguish between structural names and non-structural names, nor can they differentiate between the left and right boundaries of structural names. Therefore, it is necessary to determine which structural names are correct from the claims to facilitate subsequent labeling.

[0035] There are several methods for defining the structural names in the scope of the claims, such as:

[0036] By using the first word of the structural name introduced in the claims and the last word after the structural name, the first content between the first word and the last word, or the second content between adjacent first words, is searched within the scope of the claims. The first content and the second content are used as the structural name, or the structural name is further determined from the first content and the second content.

[0037] The first word includes any of the following: containing, having, having, including, and the last word includes punctuation marks.

[0038] As mentioned above, the format of the claims is fixed, and the description introducing the structure name is also fixed. Therefore, the structure name can be located by identifying the formatted content introducing the structure name, i.e., the first word mentioned above. The following is an exemplary example of a claim:

[0039] 1 .A sort of advertising module The group, characterized in that it comprises:

[0040] B1 ;

[0041] B3 ,include c3 The c3 have d3 ;

[0042] B6 ,include c7 The c7 have d7 The b6 is connected at one end The b1, the other end passing through The d3 ;

[0043] B8 , interposed between The d7, and located The c3 faces away from The b1 side .

[0044] In the aforementioned claims, the advertising module includes B1, B3, B6, and B8. Here, "including" serves as the initial word, introducing B1, B3, B6, and B8. Therefore, the initial word is defined as a word that can introduce the structural name. After B1, B3, B6, and B8 are introduced, they are generally followed by punctuation marks, which serve as suffixes. Suffixes are defined as symbols that can stop the structural name from continuing. With the initial word and suffix, the first content and the second content containing the structural name can generally be defined.

[0045] Furthermore, for B3 and other structures such as c3, c3 is usually followed by a punctuation mark, so we can further determine the structure names such as c3.

[0046] However, the first and second parts of the content may contain more than just structural names; they may also include modifiers and other words. Therefore, it is necessary to filter out these irrelevant words. Methods for filtering out these irrelevant words, that is, methods for further determining the structural names from the first and second parts of the content, include:

[0047] Find the content after the last "de" and before the end symbol from the first content and the second content as the structure name.

[0048] When there are multiple attributives for the structure name, find the content after the last "de" and before the end symbol as the structure name. When there is only one attributive, the first "de" is also the last "de".

[0049] The method for defining the structure name in the scope of the claim book is also as follows:

[0050] Identify at least two identical words in the scope of the claim book, and at least one of the identical words is preceded by the definite article "the", and the other is not preceded by the definite article "the". Use the identical words as the structure nouns.

[0051] As mentioned above, there is a fixed format for writing claims. For the structure name that appears the second time, "the" needs to be added in front. Therefore, using this fixed format, as long as two identical words are found, and the previous word is not preceded by "the" or "this", and the latter word is preceded by "the" or "this", then this identical word is used as a noun.

[0052] Optionally, identifying at least two identical words in the scope of the claim book includes:

[0053] S210: Identify a two-character word in the scope of the claim book, and traverse all two-character identical words that are the same as the word in the scope of the claim book;

[0054] A one-character word is not operable. Therefore, first set the number of characters of the word to two, read two-character words at the beginning of the claim book, and traverse within the scope of the claim book; of course, two-character words can also be randomly read in the claim book, and some formatted words such as "according to", "claim", etc. are filtered out;

[0055] S220: In response to finding more than one identical word, add one character to the left or right of the word to form another three-character word, and continue to refer to the step of traversing all two-character identical words that are the same as the word in the scope of the claim book, traverse and perform subsequent steps for the other word, and continue to loop;

[0056] In the claims, the structural name generally appears at least twice. If more than one identical word can be found, it means that the number of characters in the identical word may be more. So, the scope of this word is expanded. For example, other words are searched to the left or right based on the identical word. If found, the word is added to the identical word, so that the number of characters in the new identical word becomes three. Then, the steps of traversing all identical words with two characters and the same as the word in the scope of the claims are continued. The traversal and subsequent steps are carried out. As long as more than one identical word can be found, the loop continues.

[0057] S230: In response to the inability to find more than one identical word, determine another word with two characters in the scope of the claims, and continue to refer to the step of traversing all identical words with two characters in the scope of the claims and the same as the word, traversing the other word and subsequent steps, and continuing the loop until the scope of the claims has been traversed.

[0058] If, during the iterative search for two or more identical words, it is found that no other identical word can be found after increasing the word count, it indicates that the new word with increased word count is not a correct structural name. In this case, the iterative search terminates, and the last identical word found is used as the structural name. Returning to the first step, the system reads the remaining two words from the claims and searches for identical words again, continuing until the entire claims have been read and all structural names have been identified.

[0059] S300: Hierarchical division of at least two defined structure names;

[0060] To ensure hierarchical labeling, during or after defining the structure names, different structure names can be hierarchically divided according to the relationships between them, for example:

[0061] Divide the two structural names that are immediately adjacent to each other into upper and lower levels;

[0062] For example, in the aforementioned claim example, the technical feature... B3 ,include c3 If the first word inside is "include", then the two structural names immediately before and after "include" are B3 and c3. Therefore, B3 is the superior level of c3, and c3 is the inferior level of B3.

[0063] S400: Number at least two structure names according to the hierarchy.

[0064] That is, assign corresponding labels to the two structure names at the upper and lower levels according to their hierarchical designations. For example, if B3 is labeled 10, then c3 is labeled 11. Number all structure names in the same way.

[0065] Optionally, the above steps S300 and S400 can also be replaced in the following manner:

[0066] Label at least two defined structure names in a random labeling manner or in a predefined labeling manner.

[0067] That is to say, it is not necessary to hierarchically divide the structure names. Instead, all structure names can be directly labeled in a random labeling manner, or all structure names can be labeled in a predefined labeling manner. For example, in the order of the appearance of the structure names, they can be labeled one by one in the order of Roman numerals 1, 2, 3...

[0068] Optionally, if the user finds that some structure names cannot be recognized by the program, manual supplementary labeling can be performed. Therefore, after labeling at least two structure names, it includes:

[0069] S410: Receive a first instruction to insert a label in the patent-related text;

[0070] The user selects to insert a label after a certain unlabeled structure name in the patent-related text, thus triggering the first instruction.

[0071] S420: In response to the first instruction, insert a label in the patent-related text;

[0072] The user's action of supplementary labeling is completed, and the label has been inserted after the unlabeled structure name.

[0073] S430: Determine a structure name immediately adjacent to and in front of the above-mentioned one label as the third boundary;

[0074] Although the user can clearly know which is the unlabeled structure name, the computer does not know. However, the computer can be made to use the above-mentioned action of inserting a label for recognition so as to perform consistent labeling of the same word in other places. Since the label is on the right side of the structure name, the above-mentioned inserted one label can be used as the third boundary, that is, the right boundary, to enable the computer to obtain a structure name immediately adjacent to and in front of the third boundary. Generally, the left boundary of the structure name can be "of" or "the", so the unlabeled structure name can be defined by using the left boundary and the right boundary.

[0075] S440: Search for the above-mentioned defined unlabeled structure name in the patent-related text, and insert the above-mentioned one label after the found content;

[0076] For example, search for the defined unlabeled structure name in the specification or claims, and insert the above-mentioned one label on the right side of the structure name one by one. Or

[0077] Optionally, if the user is dissatisfied with the program's automatic labeling, or if an incorrect label has been applied to a location that should not be labeled, labeling at least two structure names will include:

[0078] S440: Receive a second instruction to delete or replace a label;

[0079] When a user deletes or replaces a label, a second instruction is triggered.

[0080] S450: In response to the second instruction, delete or replace the label;

[0081] The program responds to the second instruction to delete or replace the label.

[0082] S460: Use the deletion or replacement label as the fourth boundary to determine the name of a structure immediately preceding the fourth boundary;

[0083] After deleting or replacing a label, any remaining identical labels also need to be deleted or replaced. You can directly search the entire text and then delete or replace them. Alternatively, for greater accuracy, you can use the deleted or replaced label as the fourth boundary to determine the structure name immediately preceding that fourth boundary, and then locate the corresponding structure name.

[0084] S470: Search for a structure name in patent-related text, and delete or replace the label of the found content.

[0085] Search the specified section of the patent-related text to locate all identical words. Once the structure name is located, you can delete or replace the numbers to its right.

[0086] Optionally, if it is necessary to rename the structure name, the definition of the structure name within the scope of the claims includes:

[0087] S480: Receive a third instruction to amend the patent-related text;

[0088] A user modified a structure name, triggering the generation of a third instruction.

[0089] S490: In response to a third directive, determine whether the revision applies to a structure name or an adjacent structure name;

[0090] When users make modifications to patent-related texts, the computer cannot directly determine whether the modification pertains to a specific structure name or whether a structure name has been renamed. Therefore, it is necessary to determine whether the revision targets a specific structure name or is adjacent to one. One method to determine whether a revision targets a specific structure name is to directly search among all structure names using the content of the revision. If a structure name is found, then the content at the revision location is searched using that structure name. If a match is also found, then the modification is determined to have targeted the structure name.

[0091] S491: In response to an amendment targeting a structure name, or an adjacent structure name, replace at least one identical word in the patent-related text with the amended structure name. After determining that the amendment is targeting a structure name, perform a full-text search and replace.

[0092] Optionally, the scope of labeling at least two structural names can be limited to the specific embodiments described in the specification, or it can include both the claims and the specific embodiments described in the specification, or it can be limited to only the claims, or even a user-selected area or a selection of the preceding or following text. Structural name labels in the claims must be enclosed in parentheses.

[0093] Optionally, the claims, specification, and full text numerals can be removed with a single click.

[0094] Optionally, the timing for performing partial or full-text search and replace, and how to automatically modify for consistency once the timing has elapsed, can be determined using the following automatic modification method:

[0095] S11: Receive the instruction to end manual revisions of patent-related text.

[0096] A manual revision completion command is a command that indicates the completion of at least one revision action by the user. This command can be issued proactively by the user or passively during the user's operation.

[0097] In one embodiment, receiving a manual revision end instruction for patent-related text includes: triggering a manual revision end instruction in response to at least one of the following: a user clicking an operation item in the user interface, pressing a preset button, or entering a preset instruction.

[0098] Specifically, a user can confirm the completion of the current revision action and request a manual revision end command by actively clicking an action item, pressing a preset key, or entering a preset command. Action items can be buttons in the user interface, such as the "Run" button on a software interface. Users can interact with action items through common methods like clicking and swiping. Pressing a preset key can be a physical key on the keyboard or other input method. For example, the "F1" key on the keyboard can be set as a preset key. Preset keys can also be key combinations; for example, pressing Ctrl+F1 simultaneously can trigger a manual revision end command. Entering a preset command can be a specific string entered by the user to trigger the manual revision end command. Preset commands can be entered within a specific preset area; only input within this area will trigger the manual revision end command. Preset commands can also be entered in real-time by detecting user input at any location, triggering the manual revision end command whenever the condition is met. For example, it could be set to trigger a manual revision end command whenever a specific formatted command, such as "Run, Modify, Follow," is detected.

[0099] With the above settings, users can repeatedly adjust the content that needs revision, and only actively trigger a manual revision end command once, for example, after all manual revisions are completed or after a certain stage is finished. Content before revision can be processed in batches in a single operation, resulting in higher processing efficiency. Furthermore, since the manual revision end command is always issued by the user, the user can clearly control the content and number of revisions, reducing the likelihood of unexpected operations beyond the user's expectations and ensuring stable operation.

[0100] In another embodiment, receiving the $ for patent-related text includes: triggering a manual revision end command in response to a preset change in the user's cursor position.

[0101] Specifically, when a user changes the position of the text input cursor using the keyboard or mouse, such as moving the cursor from the current line to another line, it represents an operation performed by the user. Each user operation is detected and triggers a manual revision end command, meaning the manual revision end command is passively triggered in real time during user operations.

[0102] With the above settings, the manual revision end command is triggered in real time after the user completes the revision. The user does not need to trigger the manual revision end command again after each revision. The user operation steps are reduced and it is more convenient to use.

[0103] The preset change can be moving to the next or previous character position, or jumping to another character position. In one embodiment, the preset change is detecting a newline character break at the user's cursor. Therefore, while maintaining real-time triggering, this reduces unnecessary triggering attempts and balances processing speed.

[0104] S12: In response to a manual revision end instruction or in response to the fulfillment of preset conditions, determine the pre-revision content in the patent-related text.

[0105] In response to the fulfillment of preset conditions, including the arrival of a preset delay time after receiving the manual revision end instruction, such as 1 minute after receiving the manual revision end instruction, the preset conditions are met. This allows processing of the pre-revision content in the patent-related text to be performed when the user is not editing the file or not operating the computer, without affecting the user experience.

[0106] Specifically, the content before revision refers to the content before it was changed by the user. The content before revision should include at least one character or formatting character. The content before revision can include changes resulting from operations such as adding, deleting, or modifying. For example, if a user revises "abcd" in patent-related text to "abzz", the content before revision should at least include "cd". Or, if a user revises "abcd" in patent-related text to "ab", the content before revision should at least include "cd". Another example is if a user revises "ab" in patent-related text to "abcd", the content before revision should at least include "ab". Alternatively, a user might rename a structure or renumber the label after a structure name. When a manual revision end command is triggered and no content before revision is detected, an end command can be triggered, and subsequent steps will not continue.

[0107] The method for determining the pre-revision content as the structure name has been described earlier, so it will not be repeated here.

[0108] In one embodiment, determining the pre-amendment content in the patent-related text specifically includes:

[0109] In addition, identify revised third-party content in the patent-related text, such as by recording the most recent or several revised third-party content in the background, or

[0110] Identify the third content with revised format in the patent-related text.

[0111] Revision formatting can be a format generated by revision operations in other software. For example, revision formatting can be a format generated by the "Revision" function in word processing software. Determining third content can be done by taking several consecutive characters with revision formatting as a third content. For example, if patent-related text contains "abcdefg", where "ab" and "efg" have revision formatting, then "ab" is taken as one third content and "efg" is taken as another third content.

[0112] The fourth content is determined based on the third content, and if the length of the fourth content is greater than or equal to the length of the third content, the fourth content is used as the content before revision.

[0113] Determining the fourth content can be done by directly using each part containing the third content as the fourth content. For example, if the third content is "ab", the fourth content is also directly determined to be "ab". Alternatively, the fourth content can be determined by finding a complete word or phrase corresponding to the position of the third content. When determining the fourth content, the location of the third content can be detected, and the text can be expanded in both directions based on the location of the third content until a complete word or phrase is obtained as the fourth content. For example, if the patent-related text contains "abcdefg", where "de" is determined to be the third content, the text can be expanded forward and backward based on "de" until "abcdefg" is obtained as the fourth content. When several third contents are adjacent, causing the complete words or phrases obtained by expanding from these third contents to overlap, the results of merging the words or phrases obtained from expanding these third contents can be combined to form only one fourth content. For example, if the patent-related text contains "abcdefg", where "bc" and "de" are identified as third content, then "abcdefg" can be obtained by expanding "bc" with the surrounding text, and "abcdefg" can also be obtained by expanding "de" with the surrounding text. If the results of expanding "bc" and "de" with the surrounding text are found to be the same, then two separate fourth content entries for "abcdefg" are not formed; only one fourth content entry for "abcdefg" is created.

[0114] Understandably, one purpose of determining the fourth content is to make the matching more accurate. Therefore, the specific strategy for obtaining the fourth content based on the third content can be set according to the actual situation. In addition, when the revision type is "change," the determination of the fourth content should be based on the text before the revision, and the revised content can be ignored.

[0115] With the above settings, users can directly indicate the location of revisions using existing software's revision formats and background records. Furthermore, determining the fourth content based on the third content and using the fourth content as the original content enhances the uniqueness of the original content within the patent-related text, reduces noise during matching, and ensures accurate positioning.

[0116] In one specific embodiment, determining the fourth content of the minimum matching unit based on the third content includes:

[0117] Starting from the third content, find the first first separator that appears in the preceding and following text. Using the first first separator that appears in the preceding and following text as the ending point, take the content between the ending point of the preceding and following text as the fourth content.

[0118] The first separator can be a comma, period, semicolon, space, paragraph break, or other special symbol. For example, if the patent-related text contains "abc,defgh,ijk.", where "fg" is the third content, then the text is expanded forward and backward until the first separator comma is encountered as the endpoint, thus determining the fourth content as "defgh".

[0119] Therefore, by extracting the text containing the third content based on the first delimiter appearing in the preceding and following text, it is easy to obtain a sentence segment containing certain unique features, making it less likely to generate incorrect matching noise when using the fourth content as the smallest matching unit.

[0120] In another specific embodiment, the fourth aspect of determining the minimum matching unit includes:

[0121] Centered on the third content and with the preset length of the preceding and following text as the boundary, the fourth content is determined as the smallest matching unit between the preset length of the preceding and following text and the third content.

[0122] The preset length can be a pre-defined value, such as 3, 4, or 5. The preset length can be set by the user or be a fixed value. The fourth content is obtained by concatenating the characters of the preceding preset length, the third content, and the characters of the following preset length. Additionally, when the length of the preceding and following text in the paragraph containing the third content is less than the preset length, the maximum length of the preceding and following text that can be extracted from the paragraph containing the third content can be used as the cutoff point to determine the smallest matching unit for the fourth content. For example, when the third content is at the end of a paragraph, the preceding text is truncated according to the preset length, and the following text is truncated according to the actual remaining character length of the paragraph.

[0123] With the above settings, the fourth content of the smallest matching unit can be obtained easily and quickly. The function is simple to implement and has a fast processing speed.

[0124] In another embodiment, determining the pre-amendment content in the patent-related text specifically includes:

[0125] The content selected by the user in the patent-related text will be used as the content before revision.

[0126] When the selected range contains content that has been modified by the user, the entire selected range represents the content before revision and serves as the smallest matching unit. When the selected range does not contain content that has been modified by the user, meaning the selected range contains only the original text, an end command can be triggered, and subsequent steps will not be performed.

[0127] By using the above settings, the pre-revision content in patent-related texts can be determined without relying on the revision format of other software, thus broadening its applicability.

[0128] In another embodiment, determining the unrevised content in the patent-related text specifically includes:

[0129] The third content, determined by comparing the current version of the patent-related text with historical versions, is the content before revision.

[0130] Historical versions of patent-related texts can be another older document, or older versions of patent-related texts located in other parts of the same document. By comparing historical versions of patent-related texts with the current patent-related texts, the parts of the current patent-related texts that have changed relative to historical versions can be identified. These changed parts can then be considered the pre-revision content.

[0131] In addition, optionally, the scope of the claims or the structural name in the specification can be defined. The method for defining the structural name can refer to the foregoing embodiments, and will not be repeated here.

[0132] Using the above method, one-click fully automatic numbering can be achieved without manual labeling, manual identification of structure names, or manual organization of structure names, greatly improving editing efficiency and reducing the error rate of manual numbering.

[0133] In addition, it can quickly and consistently add missing labels and make quick and consistent modifications to labels by automatically recognizing the structure name, which can easily cope with various labeling scenarios.

[0134] S13: Match the remaining or selected portions of the patent-related text by the structure name or the content before the revision.

[0135] Specifically, by matching the remaining or selected portions of the patent-related text against the original content, the positions of the corresponding content that needs to be modified synchronously can be determined. For example, if the original content is "abc", then all instances of "abc" will be matched in the remaining or selected portions of the patent-related text. Matching can be performed on the entire remaining portion of the patent-related text, i.e., full-text matching. Alternatively, matching can be performed only on the user-selected portion, leaving other portions unprocessed. The specific implementation of the matching can be any character matching technique capable of locating positions, which will not be elaborated upon here.

[0136] In one embodiment, matching the remaining or selected portions of the patent-related text with the pre-amendment content includes:

[0137] In response to the lack of a match in the original content, the second delimiter in the original content is searched.

[0138] Specifically, during the matching process of the remaining or selected portions of the patent-related text based on the pre-revision content, situations may arise where no matching result is found. In this embodiment, the existence of corresponding content in the remaining portions of the patent-related text based on the pre-revision content is considered a high-probability scenario. The absence of a matching result is initially identified as an anomaly. The failure to match may be due to factors such as character changes, sentence splitting, or grammatical changes. Therefore, this embodiment first searches for the second separator in the pre-revision content as a basis for further processing. The second separator can be a comma, semicolon, period, space, etc.

[0139] In response to the absence of a second separator, replace the conjunction in the original content with the subject of the preceding sentence, or insert the subject of the preceding sentence at the beginning of the original content.

[0140] When no second separator is found, it means the probability of no match is low due to sentence splitting. It's more likely that the lack of a match is due to grammatical changes. For example, a sentence in the original text might have had its subject replaced with a conjunction due to writing habits, but the subject in the remaining part of the patent-related text might not have been replaced with a conjunction. In this case, although the meaning is the same, the text is inconsistent. Replacing the conjunction in the original text with the subject of the preceding sentence can include this situation in the matching scope, making it less likely to miss a match. For example, the preceding sentence in the original text might be "a includes b.", and the original text might be "also includes c.", while the remaining part of the patent-related text might contain "a includes b" and "a includes c." Directly using "also includes c." might not yield a match. Replacing "also" with the subject "a" of the preceding sentence will result in a correct match.

[0141] In response to the presence of a second delimiter, the original content is split into a first part and several second parts based on the second delimiter. The keywords in each second part are replaced with the subject of the first part, or the subject of the first part is inserted at the beginning of each second part.

[0142] When the second separator is found, it indicates a high probability that the lack of a match is due to the splitting of a long sentence. For example, the original content consisted of two short sentences connected by commas or "and," but in the remaining part of the patent-related text, these were split into two short sentences, with the commas and "and" in the latter sentence replaced by the subject. While the meaning remains the same, the text is inconsistent. Replacing each keyword in the latter part with the subject in the former part can bring these situations into the matching scope, reducing the likelihood of missing matches. For example, the original content was "a includes b, and includes cd.", while the remaining part of the patent-related text reads "a includes b. a includes cd." In this case, the splitting of the long sentence might result in a missed match. Replacing the "and" in the latter part with the subject "a" in the former part will ensure a correct match. Keywords can be set according to the actual situation and can be common words or symbols used to separate long sentences; there are no restrictions here.

[0143] The positions of the unamended content and the remaining portion of the patent-related text can be set according to the actual situation. In one specific embodiment, the unamended content is located in the claims section of the patent-related text. The remaining portion or selected portion of the patent-related text is at least partially located in the specification section of the patent-related text. In another specific embodiment, the unamended content is located in the specification section of the patent-related text, and the remaining portion or selected portion of the patent-related text is at least partially located in the claims section of the patent-related text.

[0144] In addition, optionally, the triggering condition for matching the remaining or selected parts of the patent-related text with the structure name or the content before revision may include: the content before revision matches the structure name, that is, the remaining or selected parts of the patent-related text are matched with the structure name or the content before revision only when the content before revision matches the structure name.

[0145] S14: Revise all or part of the target content obtained from the matching based on the revised content.

[0146] Optionally, all or part of the target content will be revised to match the revised content. All matched target content can be revised uniformly, or only a portion of the matched target content can be revised. When the content before revision is determined by the revision format, the target content can also be revised using the revision format, allowing users to easily observe the location of the revised target content. In one embodiment, a positioning mark is formed at the revised target content. Users can quickly locate the revised target content using the positioning mark. The positioning mark can be an annotation in text editing software, etc.

[0147] In one embodiment, before or simultaneously matching the remaining or selected portions of the patent-related text with the pre-amendment content, the following is included:

[0148] The original content, remaining content, and selected portion of the matched text are denoised. Denoising includes removing punctuation marks, quotation marks, definite articles, and demonstrative pronouns, or at least all of them.

[0149] The original content may interfere with the matching process due to differences in wording compared to the target content. For example, if the original content is located in the claims and reads "the a includes b", the word "the" needs to be removed. If the remaining content in the patent-related text contains reference numerals, such as "a11 includes b12", these reference numerals also need to be removed from the remaining content in the patent-related text before matching the original content with the remaining content in the patent-related text.

[0150] With the above settings, when matching the remaining or selected parts of the patent-related text with the original content, it is less likely to be interfered with by noise, and the matching is accurate.

[0151] Optionally, in one embodiment, all or part of the matched target content is revised based on the revised content, wherein the revised content is the content obtained based on the content before revision, including:

[0152] The revised content is determined based on the content before the revision.

[0153] The revised content refers to the portion of the patent-related text that has changed based on the original content. When the original content was modified by the user, the revised content can be the new content after the change. For example, if a user changes "abcd" to "abzz" in the patent-related text, the revised content will be "zz". When the original content was deleted by the user, the revised content can be empty. For example, if a user changes "abcd" to "ab" in the patent-related text, the revised content will be empty. When the original content was added by the user, the revised content can be the newly added content. For example, if a user changes "ab" to "abcd" in the patent-related text, the revised content will be "cd".

[0154] The first content, after being denoised, replaces the second content in the target content, without modifying the remaining content in the target content other than the second content.

[0155] In this embodiment, the second content refers to the content within the target content that needs to be modified. When revising the target content, only the modified portions corresponding to the original content are changed, not the entire target content. For example, if a patent-related text is changed from "the a includes b" to "the c includes b", the target content becomes "a11 includes b12". In this case, the revised content is denoised to obtain "c", the second content is confirmed to be "a11", and the revised content replaces the second content, resulting in "c includes b12". This ensures that the remaining content within the target content is less likely to undergo unexpected changes that could lead to errors, resulting in accurate revisions.

[0156] In one embodiment, replacing the second content in the target content with the first content that has undergone noise reduction processing of the revised content, without modifying the remaining content in the target content other than the second content, further includes:

[0157] Add an icon number to the revised content.

[0158] For example, after obtaining "c includes b12" in the above embodiment, the corresponding reference numeral "13" will be added to "c", resulting in "c13 includes b12". This allows the reference numerals to be automatically added after revision, making it less likely to be omitted and cause formal errors.

[0159] To more clearly illustrate the workflow of the automatic modification method for the relevant text of this patent application, the following example demonstrates the process using revisions as an example. Please refer to [link / reference]. Figure 2 , Figure 2 This is a schematic diagram of the first interface of an embodiment of the method for automatically modifying patent-related text according to this application. In one embodiment, the method for modifying patent-related text may have a first interface 100. Users can edit patent-related text within the first interface 100. The patent-related text includes at least a first part 10 and a second part 20. For example... Figure 2 As shown, the user can edit the original text "ABCD" in Part 10, and then click the "Run" button in Operation Item 30. At this time, "CD" in Part 10 has revision formatting, and "CD" is confirmed as the third content 11, and "ZZ" is the revised content 13. Next, based on the third content 11, the text is expanded forward and backward. The expansion process is based on the text before revision, and the revised content 13 "ZZ" will be skipped until it is expanded to the first comma position as the end, resulting in the fourth content 12 "ABCD" as the text before revision.

[0160] Then, based on the fourth content 12, the target content 22 "ABCD" in the second part 20 is obtained, and the second content 21 "CD" in the second part 20 is located. The revised content 13 replaces the second content 21, thus completing the content synchronization revision. After the content synchronization revision, the revised content 13 that is replaced in the second part 10 can also be labeled.

[0161] The above matching and replacement process will perform noise reduction processing, which includes removing punctuation marks, removing labels, removing definite articles, and removing demonstrative pronouns.

[0162] As can be seen from the above embodiments, this application eliminates the need for users to input keywords for searching in the "Find" or "Replace" interface each time. The computer can determine the original content (i.e., the keywords used for searching) in the patent-related text and perform subsequent matching and replacement, which greatly improves the efficiency of revising patent-related text and reduces the chance of human error.

[0163] Furthermore, since the embodiments of this application can perform noise reduction processing on the content before revision, the remaining part, and the selected part for matching, the "search" or "replace" function has a generalization function, which can find the matching target more comprehensively and accurately, greatly reducing the trouble of manually searching one by one due to the inability to find a match and the problem of missing the modification when the match cannot be found.

[0164] Secondly, since the embodiments of this application replace the second content in the target content with the first content after the revised content has undergone noise reduction processing, and do not modify the remaining content in the target content other than the second content, only the parts of the target content that need to be modified will be affected by the changes, and the remaining content will not be processed. This makes it less likely for unexpected changes to occur and form errors to occur, and the revision is accurate.

[0165] Furthermore, in this embodiment, during matching, the changed third content can be expanded into a fourth content by considering the preceding and following text. Compared to the third content, which may consist of only scattered words, the fourth content can contain more unique features, making the matching more accurate and less prone to noise interference. When determining the fourth content, the text can be truncated using the first separator between the preceding and following text as the cutoff point, or a preset length can be directly set for truncating the text. The method for constructing the fourth content can be flexibly chosen, and the fourth content formed by truncating the text can all possess good uniqueness.

[0166] In addition, the content before revision in this application embodiment can be determined by the revision format, by the user selecting a range in the patent-related text, or by comparing the current version of the patent-related text with historical versions of the patent-related text. There are various ways to determine the content before revision, and you can freely choose to adapt to different occasions.

[0167] When the content before revision is determined through the revision format, the positioning is not only precise, but the revision format is also easy to observe, allowing users to intuitively grasp the content to be revised. When the content before revision is determined by selecting a range, users can easily decide which part of the text to process according to their needs, giving them greater control over the revision process. When the content before revision is determined by comparing the current version of the patent-related text with historical versions, users only need to store a copy of the historical version of the patent-related text in advance, or even have it automatically stored by the software in this application embodiment. This eliminates the need for editing under the revision format and repeated selection of the range, saving time and effort, and simplifying the operation. It is especially suitable for situations where users do not want to enable or forget to enable the revision format during the revision of patent-related text.

[0168] When matching a target, this embodiment also identifies whether the content before revision contains a second separator when a match fails. If the second separator is not found, it means the probability of not matching a result due to long sentence splitting is low, and the failure to match a result should be considered primarily as a grammatical change. In this case, this embodiment replaces the conjunctions in the content before revision with the main clause of the preceding sentence before matching, including the above situation in the matching range. If the second separator is found, it means the probability of not matching a result due to long sentence splitting is high, and this embodiment replaces each keyword in the latter part with the subject of the former part, including the above situation in the matching range. Therefore, this embodiment can also adapt to situations where there are character changes, long sentence splitting, grammatical changes, etc., during matching, and is less likely to miss matching results. In this embodiment, the matching and revision process can be actively triggered by the user by responding to at least one of the following: clicking an operation item in the user interface, pressing a preset key, or inputting a preset command; or it can be passively triggered based on a preset change in the user's cursor position. The setting method is flexible and can effectively respond to the user's revision operations, making it convenient to use.

[0169] Optionally, all or part of the matched target content can be revised based on the revised content. The revised content is the content obtained based on the unrevised content, which includes:

[0170] S21: Generate a history record, which includes at least some information related to revising all or part of the target content obtained from the match based on the revised content.

[0171] History records contain information related to the corresponding revision actions. For example, history records may include the time of the revision action, a summary of the revision content, and the scope of the revision.

[0172] S22: In response to user click history, restore the target content to the state before revision corresponding to the history.

[0173] When a user clicks on the history, the changes made during the corresponding revision will be restored to the state prior to that revision. Restoring the target content to its pre-revision state in the history can be done in two ways: either by storing the old version of the text, which overwrites the current text when the user clicks on it; or by storing the old file, which overwrites the current file when the user clicks on it.

[0174] Please see Figure 3 , Figure 3 This is a schematic diagram of the second interface of an embodiment of the automatic modification method for patent-related text in this application. In one embodiment, the automatic numbering method for patent-related text includes a second interface 200, on which a history record is displayed. The history record may record the revision number, revision location, revision content, and operation time, etc. The revision location may be the paragraph number in the patent-related text where the revision occurred. The revision content may record a summary of the content of the revision operation. The operation time may be the time when the revision operation was triggered. The second interface 200 is clickable, such as... Figure 3 As shown, when a user clicks on the history record numbered "2" in the second interface 200, the patent-related text will revert to the state before the "delete aa" operation.

[0175] Please see Figure 4 , Figure 4 This is a schematic diagram of the structure of an embodiment of the computer device 50 provided in this application. In one embodiment, the computer device 50 includes a memory 51 and a processor 52 coupled to each other. The processor 52 is used to execute program instructions stored in the memory 51 to implement the steps of any of the above-described embodiments of the automatic labeling method for patent-related texts. In a specific implementation scenario, the computer device 50 may include, but is not limited to, a microcomputer or a server. In addition, the computer device 50 may also include mobile devices such as laptops and tablets, which are not limited here.

[0176] Specifically, processor 52 controls itself and memory 51 to implement the steps of the automatic labeling method embodiment for any of the aforementioned patent-related texts. Processor 52 may also be referred to as a CPU (Central Processing Unit). Processor 52 may be an integrated circuit chip with signal processing capabilities. Processor 52 may also be a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component. A general-purpose processor may be a microprocessor or any conventional processor. Furthermore, processor 52 may be implemented using integrated circuit chips.

[0177] Please see Figure 5 , Figure 5 This is a schematic diagram of an embodiment of the computer-readable storage medium provided in this application. The computer-readable storage medium 60 of this application embodiment stores program instructions 61. When executed, these program instructions 61 implement the method provided by any embodiment of the automatic labeling method for related texts of this application, and any non-conflicting combination thereof. The program instructions 61 can form a program file and be stored in the aforementioned computer-readable storage medium 60 in the form of a software product, so that a computer device (which may be a personal computer, server, or network device, etc.) executes all or part of the steps of the methods of various embodiments of this application. The aforementioned computer-readable storage medium 60 includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, or terminal devices such as computers, servers, mobile phones, and tablets.

[0178] Please see Figure 6 , Figure 6This is a schematic diagram of the client and server structures in one embodiment of the automatic modification method for patent-related text. This application provides an automatic modification method for patent-related text, comprising: a client 80 receiving a revision trigger end command for the patent-related text; the client 80 sending the revision trigger end command to a server 70, causing the server 70 to determine the pre-revision content in the patent-related text, define the scope of the claims or structural names in the specification, and then match the remaining or selected portions of the patent-related text with the pre-revision content; revising all or part of the matched target content based on the revised content or structural names, wherein the revised content is based on the pre-revision content; and the client 80 receiving the revision results and displaying them on an interface.

[0179] Specifically, the automatic modification method for patent-related text can also be executed by server 70. In this embodiment, client 80 is mainly used for interaction with the user. The patent-related text can reside on server 70 and be displayed by client 80. That is, client 80 may not store the patent-related text, but only serve a display function; all processing of the patent-related text is performed by server 70. For example, the patent-related text can be an online cloud operation client 80 that interacts with server 70. After receiving a revision trigger end command from client 80, server 70 executes the relevant steps of the automatic modification method for patent-related text in this embodiment, completes the revision of the patent-related text, and then sends the result to client 80 for display.

[0180] All embodiments of this application are applicable not only to patent-related texts, but also to articles, reports, papers, novels, or scientific and technological literature.

[0181] In one embodiment, the automatic labeling or modification method for the patent-related text of this application is executed by program instructions. The program instructions are installed in the form of a plug-in into existing text editing software and are applied to the patent-related text opened by the text editing software to realize the above-mentioned automatic labeling method for the patent-related text.

[0182] In one embodiment, the method for automatically labeling or modifying the patent-related text of this application is executed through program instructions. These program instructions run independently as computer software and include at least text editing functionality. The program instructions act on the patent-related text within the computer software to implement the aforementioned method for automatically labeling the patent-related text.

[0183] In one embodiment, the method for automatically labeling or modifying patent-related text is executed through program instructions, which include at least two parts, one part of which is located on the user's computer device and the other part is located on a server. At least one step in the above embodiment of the automatic labeling or modification method for patent-related text is executed by the server and can return a result. The user performs other steps by communicating with the server and obtaining the result, thereby implementing the above-mentioned method for automatically labeling or modifying patent-related text.

[0184] In one embodiment, the automatic labeling or modification method for the patent-related text is executed by program instructions, all of which are located locally on the user's computer and are executed by the user's computer to implement the above-mentioned automatic labeling or modification method for the patent-related text.

[0185] In one embodiment, the automatic labeling or modification method for the text related to this patent application is entirely run on the server, and the execution results are displayed through the user's computer, i.e., the client.

[0186] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating the number of technical features indicated. Therefore, a feature defined with "first," "second," or "third" may explicitly or implicitly include at least one of those features. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationships and movements between components in a specific orientation (as shown in the figures). If the specific orientation changes, the directional indication will also change accordingly. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. A process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0187] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for automatically numbering patent-related text, characterized in that, including: Determine the scope of the claims or the specification in the patent-related text; Define the structural names in the scope of the claims or the specification; Perform hierarchical division on at least two of the defined structural names; Number the at least two structural names according to the hierarchical division.

2. The method according to claim 1, wherein the defining of the structural names in the scope of the claims includes: Using the first word introducing the structural name and the trailing symbol after the structural name in the claims, search for the first content between the first word and the trailing symbol or the second content between adjacent first words in the scope of the claims, and use the first content and the second content as the structural name, or further determine the structural name from the first content and the second content.

3. The method according to claim 2, wherein the first word includes any one of the following: containing, having, provided with, including, and the trailing symbol includes punctuation marks.

4. The method according to claim 2, wherein the further determination of the structural name from the first content and the second content includes: Finding the content between the last "de" and before the trailing symbol from the first content and the second content as the structural name.

5. The method according to claim 2, wherein the defining of the structural names in the scope of the claims includes: Determining at least two identical words in the scope of the claims, and at least one of the identical words is preceded by the definite article "the" or "this", and the other is not preceded by the definite article "the" or "this", and using the identical words as the structural nouns.

6. The method according to claim 5, wherein the determination of at least two identical words in the scope of the claims includes: Determining a word of two characters in the scope of the claims, and traversing all words of two characters identical to the word in the scope of the claims; In response to finding more than one of the identical words, adding one character to the left or right of the word as another word of three characters, and continuing to refer to the step of traversing all words of two characters identical to the word in the scope of the claims, performing traversal and subsequent steps for the other word, and continuing to loop; In response to not finding more than one of the identical words, determining another word of two characters in the scope of the claims, and continuing to refer to the step of traversing all words of two characters identical to the word in the scope of the claims, performing traversal and subsequent steps for the other word, and continuing to loop until the scope of the claims is traversed.

7. The method according to claim 1, wherein the hierarchical division of at least two of the defined structural names includes: Dividing two of the structural names immediately adjacent to the first word into upper and lower levels; The numbering of the at least two structural names according to the hierarchical division includes: The two structural names at the upper and lower levels are assigned corresponding labels according to their hierarchical designations.

8. The method according to claim 1, characterized in that, Determining the scope of the claims in the patent-related text includes: Locate the "claims" in the patent-related texts; In response to finding the "claims", the found result is used as the first boundary, and the "technical field", "abstract", "abstract drawing", "drawings", "description" or document end mark is searched after the "claims". In response to finding "technical field", "specification abstract", "abstract drawing", "specification drawing" or "specification", the found result is taken as the second boundary, and the area between the first boundary and the second boundary is taken as the scope of the claims.

9. The method according to claim 1, characterized in that, The labeling of at least two structure names includes: Receive a first instruction to insert a label into the patent-related text; In response to the first instruction, the reference numeral is inserted into the patent-related text; Using the aforementioned label as the third boundary, determine the name of a structure immediately preceding the third boundary; Search for the structure name in the patent-related text, and insert the number after the search results, or The labeling of at least two structure names includes: Receive a second instruction to delete or replace a label; In response to the second instruction, the label is deleted or replaced; By deleting or replacing the label as the fourth boundary, a structure name immediately preceding the fourth boundary is determined; Searching for the structure name in the patent-related text, deleting or replacing the label of the found content, or The definition of the structural name in the scope of the claims includes: Receive a third instruction to revise the patent-related text; In response to the third instruction, determine whether the revision is directed at one of the structure names or adjacent to one of the structure names; In response to the revision concerning one of the structure names, or immediately adjacent to one of the structure names, the same word in at least one of the other structure names in the patent-related text is replaced with the revised structure name.

10. A method for automatically numbering patent-related text, characterized in that, include: Define the scope of the claims or the structural names in the specification; The at least two defined structure names are labeled using random or predefined labeling methods.

11. A method for automatically modifying patent-related text, characterized in that, include: Receive a manual revision completion instruction for the patent-related text; In response to the manual revision end instruction, or in response to the fulfillment of preset conditions, determine the pre-revision content in the patent-related text and define the scope of the claims or the structural names in the specification; In response to the matching of the structure name with the pre-revision content, the remaining or selected portion of the patent-related text is matched with the structure name or the pre-revision content; Based on the revised content, revise all or part of the matched target content.

12. The method according to claim 11, characterized in that, The definition of the structural names in the scope of the claims includes: Using the first word that introduces the structure name in the claims and the last word that follows the structure name, search for the first content between the first word and the last word, or the second content between adjacent first words, within the scope of the claims. Use the first content and the second content as the structure name, or further determine the structure name from the first content and the second content.

13. A computer device, characterized in that, The computer device includes: processor; A memory, connected to the processor, is used to store program instructions that can be executed by the processor to implement the automatic labeling method or automatic modification method for patent-related text as described in any one of claims 1-12.

Images