A form field annotation automatic generation method, device and equipment and storage medium

Abstract

The application discloses a form field annotation automatic generation method and device, equipment and storage medium, and relates to the artificial intelligence auxiliary technical field.The method is to firstly perform word segmentation processing on field description information of a certain form field in a target form to obtain at least one field description word, then for each field description word, it is judged whether the corresponding word needs to be annotated, if yes, at least one other entity having a meaning understanding relationship with the corresponding entity is found from a knowledge graph having a triple structure, then a natural language processing technology is used to convert the at least one other entity into user understandable text, and finally when a user fills in the certain form field, the user understandable text is output and displayed, so that professional terms or complex concepts in the form can be automatically understood for the target form, and corresponding prompts or annotations can be generated, and thus the understanding degree, form filling efficiency and accuracy when the user fills in the form can be improved.

Description

Technical Field

[0001] This invention belongs to the field of artificial intelligence-assisted technology, specifically relating to a method, apparatus, device, and storage medium for automatically generating annotations for form fields. Background Technology

[0002] Forms are digital media used to store data. They contain various components, and users can design and generate forms according to their needs. In modern enterprise information systems, forms are widely used for data input and management, and have become an important tool for executing business processes in large organizations.

[0003] Currently, many online forms contain complex or specialized field descriptions, which can be difficult for users to understand, affecting the efficiency and accuracy of form completion. Although some systems utilize NLP (Natural Language Processing) technology to assist with form completion, they primarily rely on preset rules or respond to user input, rather than automatically providing domain-specific understanding support. Furthermore, these systems cannot generate targeted hints or annotations based on specific form fields. Therefore, how to automatically understand the technical terms or complex concepts in a target form and generate corresponding hints or annotations to improve user comprehension and accuracy when filling out forms is a crucial research topic for those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a method, apparatus, computer device, and computer-readable storage medium for automatically generating annotations for form fields, in order to solve the problem that existing form filling assistance systems can only provide help for general questions, but cannot automatically generate accurate and easy-to-understand prompts or annotations for form fields with professional or complex concepts.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: Firstly, a method for automatically generating form field annotations is provided, including: Obtain the field description information of a certain form field in the target form, wherein the field description information includes the field name, field type and / or field filling instructions; The field description information is segmented to obtain at least one field description term; For each field description term in the at least one field description term, determine whether the corresponding term needs to be annotated; if so, mark the corresponding term as a term to be annotated. For a given term to be annotated, at least one other entity with a semantic understanding relationship with the corresponding entity is found from a knowledge graph with a triple structure, wherein the triple structure is composed of <entity, relation, entity>. Natural language processing techniques are used to convert the at least one other entity into user-understandable text for the word to be annotated. When a user fills in a form field, the user-understandable text of the term to be annotated is output and displayed.

[0006] Based on the above-mentioned invention, a novel solution is provided for automatically generating and displaying form field annotations using knowledge graphs and natural language processing (NLP) technologies. First, the field description information of a form field in the target form is segmented to obtain at least one field description term. Then, for each field description term among these at least one field description term, it is determined whether the corresponding term requires annotation. If so, at least one other entity with a semantic understanding relationship to the corresponding entity is found from a knowledge graph with a triple structure. Next, NLP technologies are used to convert this at least one other entity into user-understandable text. Finally, when a user fills in a form field, the user-understandable text is output and displayed. This method can automatically understand technical terms or complex concepts in the target form and generate corresponding prompts or annotations, thereby improving user comprehension, form filling efficiency, and accuracy, facilitating practical application and promotion.

[0007] In one possible design, for a certain field description term among the at least one field description terms, it is determined whether the corresponding term needs annotation. If so, the corresponding term is marked as a term to be annotated, including: Search a preset dictionary containing common, non-technical terms to see if a descriptive term for a particular field exists. If not, then it is determined that the descriptive term of a certain field needs to be annotated, and the descriptive term of that field is marked as a term to be annotated.

[0008] In one possible design, for a certain field description term among the at least one field description terms, it is determined whether the corresponding term needs annotation. If so, the corresponding term is marked as a term to be annotated, including: The descriptive words of a certain field are subjected to word feature extraction processing to obtain multiple word features; The multiple word features are imported into a pre-trained word comprehension difficulty level recognition model based on machine learning algorithms, and the word comprehension difficulty level recognition result of the descriptive words of a certain field is output. When the word comprehension difficulty level indicated by the word comprehension difficulty level recognition result exceeds the preset difficulty level, it is determined that the descriptive words of a certain field need to be annotated, and the descriptive words of the certain field are marked as words to be annotated.

[0009] In one possible design, for a given term to be annotated, at least one other entity with a semantic understanding relationship to the corresponding entity is found from a knowledge graph with a triple structure, including the following steps S41 to S44: S41. For a given term to be annotated, find at least one other initial entity that has a direct semantic relationship with the corresponding entity from a knowledge graph with a triple structure, and then execute step S42, wherein the triple structure is composed of <entity, relation, entity>. S42. For each of the at least one initial other entity, perform step S43; S43. Determine whether the corresponding entity needs annotation. If so, continue to search for at least one new other entity in the knowledge graph that has a direct relationship with the meaning understanding of the corresponding entity. Then, for each of the new other entities in the at least one new other entity, return to step S43. S44. All other entities found that do not require annotation are grouped together as at least one other entity that has a semantic understanding relationship with the entity of the term to be annotated.

[0010] In one possible design, natural language processing techniques are used to convert the at least one other entity into user-understandable text for the term to be annotated, including: The at least one other entity is converted into user-understandable text for the term to be annotated using text generation or automatic summarization methods in natural language processing.

[0011] In one possible design, when a user fills in a form field, the user-understandable text displaying the term to be annotated is output, including: When a user fills in a form field, if the mouse cursor moves over the input control of the form field or the word to be annotated, the user-understandable text of the word to be annotated will be displayed in a temporary floating window. Alternatively, when a user fills in a form field, the user-understandable text of the term to be annotated can be temporarily displayed below the input control of that form field.

[0012] In one possible design, if no other entity with a semantic understanding relationship to the entity of a certain term to be annotated can be found in the knowledge graph, the method further includes: The term to be annotated is imported into the AI ​​large language model, and the user-understandable text of the term to be annotated is output.

[0013] Secondly, an automatic form field annotation generation device is provided, which includes an information acquisition module, a word segmentation processing module, an annotation judgment module, an entity lookup module, a text conversion module, and a text display module connected in sequence. The information acquisition module is used to acquire field description information of a certain form field in the target form, wherein the field description information includes field name, field type and / or field filling instructions; The word segmentation module is used to segment the field description information to obtain at least one field description word; The annotation judgment module is used to determine whether the corresponding words need to be annotated for each field description word in the at least one field description word; if so, the corresponding words are marked as words to be annotated. The entity lookup module is used to find at least one other entity with a semantic understanding relationship with the corresponding entity from a knowledge graph with a triple structure for a given term to be annotated, wherein the triple structure is composed of <entity, relation, entity>. The text conversion module is used to convert the at least one other entity into user-understandable text of the word to be annotated using natural language processing technology. The text display module is used to output and display the user-understandable text of the term to be annotated when the user fills in a certain form field.

[0014] Thirdly, the present invention provides a computer device comprising a memory, a processor, and a transceiver connected in sequence for communication, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the automatic generation method for form field annotations as described in the first aspect or any possible design in the first aspect.

[0015] Fourthly, the present invention provides a computer-readable storage medium storing instructions that, when executed on a computer, perform the automatic generation method for form field annotations as described in the first aspect or any possible design within the first aspect.

[0016] Fifthly, the present invention provides a computer program product containing instructions that, when executed on a computer, cause the computer to perform the automatic generation method for form field annotations as described in the first aspect or any possible design in the first aspect.

[0017] The beneficial effects of the above scheme are: (1) This invention creatively provides a new solution for automatically generating and displaying form field annotations based on knowledge graph and natural language processing technology. First, the field description information of a certain form field in the target form is segmented to obtain at least one field description word. Then, for each field description word in the at least one field description word, it is determined whether the corresponding word needs to be annotated. If so, at least one other entity with a meaning understanding relationship with the corresponding entity is found from the knowledge graph with a triple structure. Then, natural language processing technology is used to convert the at least one other entity into user-understandable text. Finally, when the user fills in the form field, the user-understandable text is output and displayed. In this way, the professional terms or complex concepts in the form can be automatically understood for the target form, and corresponding prompts or annotations can be generated, thereby improving the user's understanding, form filling efficiency and accuracy when filling in the form, which is convenient for practical application and promotion. (2) Due to the combination of knowledge graph and NLP technology, this invention is able to handle forms of various fields and complexities, and has strong applicability and flexibility. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the method for automatically generating form field annotations provided in an embodiment of this application.

[0020] Figure 2 A schematic diagram of the structure of the form field annotation automatic generation device provided in the embodiments of this application.

[0021] Figure 3 A schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in conjunction with the accompanying drawings and descriptions of the embodiments or the prior art. Obviously, the following description of the structure of the accompanying drawings is only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0023] It should be understood that although the terms "first" and "second", etc., may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, the first object may be referred to as the second object, and similarly, the second object may be referred to as the first object, without departing from the scope of the exemplary embodiments of the invention.

[0024] It should be understood that the term "and / or" that may appear in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, B exists alone, or A and B exist simultaneously. Another example is A, B and / or C, which can mean that any one of A, B, and C or any combination thereof exists. The term " / and" that may appear in this document describes another relationship between related objects, indicating that two relationships can exist. For example, A / and B can mean: A exists alone or A and B exist simultaneously. In addition, the character " / " that may appear in this document generally indicates that the related objects before and after it are in an "or" relationship.

[0025] Example: like Figure 1 As shown, the automatic generation method for form field annotations provided in the first aspect of this embodiment can be executed, but is not limited to, by a computer device with certain computing resources, such as a platform server, a personal computer (PC, referring to a multi-purpose computer of a size, price, and performance suitable for personal use; desktop computers, laptops, mini-laptops, tablets, and ultrabooks are all personal computers), a smartphone, a personal digital assistant (PDA), or a wearable device. Figure 1 As shown, the method for automatically generating form field annotations may include, but is not limited to, the following steps S1 to S6.

[0026] S1. Obtain the field description information of a certain form field in the target form, wherein the field description information includes, but is not limited to, the field name, field type and / or field filling instructions.

[0027] In step S1, the target form is an online form to be filled out by the user, such as a medical insurance form or a bank loan application form, which can be submitted by the user filling out the form or the form designer. Furthermore, a specific form field refers to any field to be filled in the target form, such as a field named "Reimbursement Ratio" in the medical insurance form, or a field named "Annualized Interest Rate" in the bank loan application form.

[0028] S2. Perform word segmentation on the field description information to obtain at least one field description term.

[0029] In step S2, specifically, the word segmentation process can, but is not limited to, using the jieba word segmentation tool. Furthermore, the field description terms can be in the form of words or phrases.

[0030] S3. For each field description term in the at least one field description term, determine whether the corresponding term needs to be annotated. If so, mark the corresponding term as a term to be annotated.

[0031] In step S3, the specific method for determining whether field description terms need annotation can be based on dictionary lookup or artificial intelligence technology. For the former, specifically, for a particular field description term among the at least one field description terms, it is determined whether the corresponding word needs annotation. If so, the corresponding word is marked as a word to be annotated. This includes, but is not limited to: first searching in a preset dictionary containing common non-professional terms to see if the particular field description term exists; if not, it is determined that the particular field description term needs annotation, and it is marked as a word to be annotated. For the latter, specifically, for a particular field description term among the at least one field description terms, it is determined whether the corresponding word needs annotation. If so, the corresponding word is marked as a word to be annotated. This includes, but is not limited to, the following steps S321 to S323.

[0032] S321. Perform word feature extraction processing on the words used to describe a certain field to obtain multiple word features.

[0033] In step S321, the multiple word features include, but are not limited to, the number of characters, string length, the average occurrence rate of characters in a preset dictionary containing common non-professional vocabulary, and / or the total number of strokes. These word features can be extracted using conventional statistical analysis methods.

[0034] S322. Import the multiple word features into a pre-trained word comprehension difficulty level recognition model based on machine learning algorithms, and output the word comprehension difficulty level recognition result of the descriptive words of a certain field.

[0035] In step S322, the machine learning algorithm is an algorithm that learns the structure of data and improves the estimation of a certain target from a large amount of data and historical experience without human intervention. Its learning tasks include: learning a function from input to output; learning the latent structure of unlabeled data; and instance-based learning, such as classifying new entities and determining their categories based on training data. Therefore, the word comprehension difficulty level recognition model can be trained through a conventional model training process based on the multiple word features of multiple sample words and the manually labeled word comprehension difficulty level labels on each sample word, and applied to this embodiment. Furthermore, the machine learning algorithm may, but is not limited to, adopt existing algorithms based on LightGBM. LightGBM (Light Gradient Boosting Machine) is a framework for implementing the GBDT (Gradient Boosting Decision Tree) algorithm, which supports efficient parallel training and has advantages such as faster training speed, lower memory consumption, better accuracy, support for distributed processing, and the ability to quickly process massive amounts of data. Therefore, the word comprehension difficulty level recognition model can be built and trained based on existing knowledge.

[0036] S323. When the word comprehension difficulty level indicated by the word comprehension difficulty level recognition result exceeds the preset difficulty level, it is determined that the descriptive words of a certain field need to be annotated, and the descriptive words of the certain field are marked as words to be annotated.

[0037] S4. For a given term to be annotated, find at least one other entity in a knowledge graph with a triple structure that has a meaning-understanding relationship with the corresponding entity, wherein the triple structure is composed of <entity, relation, entity>.

[0038] In step S4, the knowledge graph (KG) is an important branch of knowledge engineering. It describes the concepts and their relationships in the physical world in a symbolic structure. The basic form of a knowledge graph is a triple of <entity, relation, entity>, where entities are interconnected through relations, forming a complex network of knowledge. For example, if the term to be annotated is "compensation ratio," and the knowledge graph contains three such <entity, relation, entity> triples: "compensation ratio" - meaning understanding relation - "insurance company," "compensation ratio" - meaning understanding relation - "accepting claims," ​​and "compensation ratio" - meaning understanding relation - "paying fees proportionally," then "insurance company," "accepting claims," ​​and "paying fees proportionally" can be considered as at least one other entity that has a meaning understanding relation with the entity of the term to be annotated. Considering that there may be entities that require further explanation among the at least one other entity, it is preferable that, for a certain term to be annotated, at least one other entity with a meaning understanding relationship with the corresponding entity is found from a knowledge graph with a triple structure, including but not limited to the following steps S41 to S44.

[0039] S41. For a given term to be annotated, find at least one other initial entity in the knowledge graph with a triple structure that has a direct relation to the meaning of the corresponding entity, and then execute step S42, wherein the triple structure is composed of <entity, relation, entity>.

[0040] S42. For each of the at least one initial other entity, perform step S43.

[0041] S43. Determine whether the corresponding entity needs annotation. If so, continue to search for at least one new other entity in the knowledge graph that has a direct meaning understanding relationship with the corresponding entity. Then, for each of the new other entities in the at least one new other entity, return to step S43.

[0042] In step S43, the specific means of determining whether an entity needs annotation can be, but are not limited to, referring to the aforementioned step S3, and will not be repeated here. For example, if "paying fees proportionally" requires further annotation, then it is necessary to continue searching the knowledge graph for at least one new entity that has a direct semantic relationship with the corresponding entity.

[0043] S44. All other entities found that do not require annotation are grouped together as at least one other entity that has a semantic understanding relationship with the entity of the term to be annotated.

[0044] S5. Use natural language processing technology to convert the at least one other entity into user-understandable text for the word to be annotated.

[0045] In step S5, the natural language processing (NLP) technology is a type of artificial intelligence that specializes in analyzing human language. Its working principle is as follows: receiving natural language (a language evolved from natural human use and used daily for communication); translating natural language (usually through probabilistic algorithms); analyzing natural language and outputting results; thus, by employing existing NLP technology, specialized terminology or complex concepts (i.e., the at least one other entity) obtained from a knowledge graph can be converted into language that the user can understand (i.e., user-understandable text). For example, "insurance company," "accepting claims," ​​and "proportional payment" can be converted into "the reimbursement ratio refers to the proportion of fees paid by the insurance company when accepting a claim," which explains the basic concept of "reimbursement ratio" and also mentions its role, helping users better understand and fill out the medical insurance form. Specifically, using NLP technology to convert the at least one other entity into user-understandable text for a specific term to be annotated includes, but is not limited to, using text generation or automatic summarization methods within NLP technology to convert the at least one other entity into user-understandable text for the specific term to be annotated.

[0046] Following step S4, considering the finite number of entities and relationships in the knowledge graph, to ensure the acquisition of user-understandable text for a given term, preferably, if no other entity with a semantic understanding relationship to the given term cannot be found in the knowledge graph, the method further includes: importing the given term into an AI large-scale language model and outputting user-understandable text for the given term. The AI ​​large-scale language model refers to a deep learning method based on artificial intelligence technology, used for reasoning and generating new language expressions with certain semantics and coherence. These models can automatically understand and process natural language, and even engage in dialogue, create various text content, and answer questions. Specifically, the AI ​​large-scale language model can be divided into two main categories: language generation models and language understanding models. The language generation model refers to training a neural network through learning from large-scale data, which automatically generates an article, story, or other type of text content consistent with the original input logic after inputting a sentence. The AI ​​large language model preferably adopts the GPT-3 language generation model so that it can be used as a chatbot tool for professional communication and response to obtain user-understandable text for a certain word to be annotated.

[0047] S6. When a user fills in a form field, output the user-understandable text of the term to be annotated.

[0048] In step S6, specifically, when a user fills in a form field, the user-understandable text of the term to be annotated is output and displayed, including but not limited to: when the user fills in a form field, if the mouse cursor moves over the input control of the form field or the term to be annotated, the user-understandable text of the term to be annotated is output and displayed in a temporary floating window; or, when the user fills in a form field, the user-understandable text of the term to be annotated is temporarily output and displayed directly below the input control of the form field.

[0049] Therefore, based on the automatic generation method for form field annotations described in steps S1 to S6 above, a new scheme for automatically generating and displaying form field annotations based on knowledge graphs and natural language processing technology is provided. First, the field description information of a certain form field in the target form is segmented to obtain at least one field description term. Then, for each field description term in the at least one field description term, it is determined whether the corresponding term needs annotation. If so, at least one other entity with a semantic understanding relationship with the corresponding entity is found from a knowledge graph with a triple structure. Then, natural language processing technology is used to convert the at least one other entity into user-understandable text. Finally, when the user fills in a certain form field, the user-understandable text is output and displayed. This can automatically understand the technical terms or complex concepts in the target form and generate corresponding prompts or annotations, thereby improving the user's understanding, efficiency, and accuracy when filling in the form, facilitating practical application and promotion.

[0050] like Figure 2 As shown, the second aspect of this embodiment provides a virtual device for implementing the automatic generation method of form field annotations described in the first aspect, including an information acquisition module, a word segmentation processing module, an annotation judgment module, an entity search module, a text conversion module, and a text display module that are sequentially connected in communication. The information acquisition module is used to acquire field description information of a certain form field in the target form, wherein the field description information includes field name, field type and / or field filling instructions; The word segmentation module is used to segment the field description information to obtain at least one field description word; The annotation judgment module is used to determine whether the corresponding words need to be annotated for each field description word in the at least one field description word; if so, the corresponding words are marked as words to be annotated. The entity lookup module is used to find at least one other entity with a semantic understanding relationship with the corresponding entity from a knowledge graph with a triple structure for a given term to be annotated, wherein the triple structure is composed of <entity, relation, entity>. The text conversion module is used to convert the at least one other entity into user-understandable text of the word to be annotated using natural language processing technology. The text display module is used to output and display the user-understandable text of the term to be annotated when the user fills in a certain form field.

[0051] The working process, working details and technical effects of the aforementioned device provided in the second aspect of this embodiment can be found in the automatic generation method of form field annotations described in the first aspect, and will not be repeated here.

[0052] like Figure 3 As shown, the third aspect of this embodiment provides a computer device for executing the automatic form field annotation generation method as described in the first aspect. The device includes a memory, a processor, and a transceiver connected in sequence. The memory stores a computer program, the transceiver sends and receives messages, and the processor reads the computer program and executes the automatic form field annotation generation method as described in the first aspect. Specifically, the memory may include, but is not limited to, random-access memory (RAM), read-only memory (ROM), flash memory, first-in-first-out (FIFO) memory, and / or first-in-last-out (FILO) memory, etc.; the processor may include, but is not limited to, a microprocessor of the STM32F105 series. Furthermore, the computer device may also include, but is not limited to, a power supply module, a display screen, and other necessary components.

[0053] The working process, working details and technical effects of the aforementioned computer device provided in the third aspect of this embodiment can be found in the automatic generation method of form field annotations described in the first aspect, and will not be repeated here.

[0054] This fourth aspect of the embodiment provides a computer-readable storage medium storing instructions comprising the automatic generation method for form field annotations as described in the first aspect. Specifically, the computer-readable storage medium stores instructions that, when executed on a computer, perform the automatic generation method for form field annotations as described in the first aspect. The computer-readable storage medium refers to a data storage medium, which may include, but is not limited to, floppy disks, optical disks, hard disks, flash memory, USB flash drives, and / or Memory Sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0055] The working process, working details and technical effects of the aforementioned computer-readable storage medium provided in the fourth aspect of this embodiment can be found in the automatic generation method of form field annotations as described in the first aspect, and will not be repeated here.

[0056] This fifth aspect of the embodiment provides a computer program product containing instructions that, when executed on a computer, cause the computer to perform the automatic generation method for form field annotations as described in the first aspect. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.

[0057] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for automatically generating annotations for form fields, characterized in that, include: Obtain the field description information of a certain form field in the target form, wherein the field description information includes the field name, field type and / or field filling instructions; The field description information is segmented to obtain at least one field description term; For each field description term in the at least one field description terminology, determine whether the corresponding word needs annotation. If so, mark the corresponding word as a word to be annotated. Specifically, this includes: performing word feature extraction processing on a certain field description term in the at least one field description terminology to obtain multiple word features; importing the multiple word features into a pre-trained word comprehension difficulty level recognition model based on machine learning algorithms to output the word comprehension difficulty level recognition result of the certain field description term; when the word comprehension difficulty level indicated by the word comprehension difficulty level recognition result exceeds a preset difficulty level, determine that the certain field description term needs annotation and mark the certain field description term as a word to be annotated. For a given term to be annotated, at least one other entity with a semantic understanding relationship to the corresponding entity is found from a knowledge graph with a triple structure. Specifically, this includes the following steps S41-S44: S41. For a given term to be annotated, at least one initial other entity with a direct semantic understanding relationship to the corresponding entity is found from a knowledge graph with a triple structure. Then, step S42 is executed, where the triple structure is in the form of <entity, relation, entity>. S42. For each of the initial other entities among the at least one initial other entity, step S43 is executed. S43. It is determined whether the corresponding entity needs annotation. If so, at least one new other entity with a direct semantic understanding relationship to the corresponding entity is found from the knowledge graph. Then, for each of the new other entities among the at least one new other entity, step S43 is returned. S44. All other entities found that do not need annotation are summarized as at least one other entity with a semantic understanding relationship to the entity of the given term to be annotated. Natural language processing techniques are used to convert the at least one other entity into user-understandable text for the word to be annotated. When a user fills in a form field, the user-understandable text of the term to be annotated is output and displayed.

2. The method for automatically generating form field annotations according to claim 1, characterized in that, For a specific field description term among the at least one field description terms, determine whether the corresponding term needs annotation. If so, mark the corresponding term as a term to be annotated, including: Search a preset dictionary containing common, non-technical terms to see if a descriptive term for a particular field exists. If not, then it is determined that the descriptive term of a certain field needs to be annotated, and the descriptive term of that field is marked as a term to be annotated.

3. The method for automatically generating form field annotations according to claim 1, characterized in that, Using natural language processing techniques to convert the at least one other entity into user-understandable text for the term to be annotated, including: The at least one other entity is converted into user-understandable text for the term to be annotated using text generation or automatic summarization methods in natural language processing.

4. The method for automatically generating form field annotations according to claim 1, characterized in that, When a user fills in a form field, the system outputs and displays the user-understandable text for the term to be annotated, including: When a user fills in a form field, if the mouse cursor moves over the input control of the form field or the word to be annotated, the user-understandable text of the word to be annotated will be displayed in a temporary floating window. Alternatively, when a user fills in a form field, the user-understandable text of the term to be annotated can be temporarily displayed below the input control of that form field.

5. The method for automatically generating form field annotations according to claim 1, characterized in that, If no other entity with a semantic understanding relationship to the entity of a certain term to be annotated can be found in the knowledge graph, the method further includes: The term to be annotated is imported into the AI ​​large language model, and the user-understandable text of the term to be annotated is output.

6. A device for automatically generating annotations for form fields, characterized in that, It includes an information acquisition module with sequential communication connections, a word segmentation processing module, an annotation judgment module, an entity lookup module, a text conversion module, and a text display module; The information acquisition module is used to acquire field description information of a certain form field in the target form, wherein the field description information includes field name, field type and / or field filling instructions; The word segmentation module is used to segment the field description information to obtain at least one field description word; The annotation judgment module is used to determine whether a corresponding word needs to be annotated for each field description word in the at least one field description word. If so, the corresponding word is marked as a word to be annotated. Specifically, it includes: performing word feature extraction processing on a certain field description word in the at least one field description word to obtain multiple word features; importing the multiple word features into a pre-trained word comprehension difficulty level recognition model based on machine learning algorithms, and outputting a word comprehension difficulty level recognition result for the certain field description word; when the word comprehension difficulty level indicated by the word comprehension difficulty level recognition result exceeds a preset difficulty level, it is determined that the certain field description word needs to be annotated, and the certain field description word is marked as a word to be annotated. The entity lookup module is used to find at least one other entity with a meaning-understanding relationship to the corresponding entity from a knowledge graph with a triple structure for a given term to be annotated. Specifically, it includes the following steps S41-S44: S41. For a given term to be annotated, find at least one initial other entity with a direct meaning-understanding relationship to the corresponding entity from the knowledge graph with a triple structure, and then execute step S42, wherein the triple structure is composed of <entity, relation, entity>; S42. For each of the initial other entities in the at least one initial other entity, execute step S43; S43. Determine whether the corresponding entity needs annotation. If so, continue to search the knowledge graph for at least one new other entity with a direct meaning-understanding relationship to the corresponding entity, and then return to execute step S43 for each of the at least one new other entity; S44. Summarize all the other entities found that do not need annotation as at least one other entity with a meaning-understanding relationship to the entity of the given term to be annotated. The text conversion module is used to convert the at least one other entity into user-understandable text of the word to be annotated using natural language processing technology. The text display module is used to output and display the user-understandable text of the term to be annotated when the user fills in a certain form field.

7. A computer device, characterized in that, The device includes a memory, a processor, and a transceiver connected in sequence, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the form field annotation automatic generation method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that... The computer-readable storage medium stores instructions that, when executed on a computer, perform the automatic generation method for form field annotations as described in any one of claims 1 to 5.

Images