Work problem library management method and system and program product
By using an automated problem database management method, the problems of inefficiency and inaccurate classification in traditional methods have been solved, achieving efficient and accurate problem database management, reducing labor costs and avoiding redundancy, and forming a knowledge system for continuous learning.
Patent Information
- Application Number
- CN202511665442.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-13
- Publication Date
- 2026-02-10
AI Technical Summary
Traditional problem database management methods rely on manual input and classification, which is inefficient, has poor classification accuracy, makes it difficult to identify duplicate problems, and leads to redundancy in database content, affecting usage efficiency.
An automated problem database management method is adopted, which involves obtaining questionnaires, extracting keyword combinations, performing hierarchical classification and similarity judgment, and realizing the automated collection, classification and deduplication of problems into the database.
It achieves full automation of the question database process, reduces labor costs, improves classification accuracy, avoids redundancy, and forms a knowledge system for continuous learning.
Smart Images
Figure CN121503476A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of data management technology, specifically relating to a method, system, and program product for managing a work problem database. Background Technology
[0002] In the management of modern enterprises and institutions, establishing and maintaining an efficient problem database is crucial for accumulating organizational knowledge and improving employees' problem-solving abilities. Traditional problem database management methods rely heavily on manual data entry and classification, which has the following significant drawbacks: 1. Inefficient and reliant on manual labor: From problem collection and classification to data entry, the entire process requires a large amount of manual intervention, which is time-consuming and labor-intensive, and is prone to inconsistencies in classification due to personal subjective judgment.
[0003] 2. Crude classification and poor accuracy: Inconsistent manual classification standards make it difficult to handle complex and vague problem descriptions, resulting in inaccurate problem classification and affecting subsequent search results.
[0004] 3. Difficulty in deduplication and redundancy: It's difficult to automatically identify whether newly submitted questions are duplicates or highly similar to existing questions in the database, leading to redundancy and bloat in the database, reducing its purity and usability. Summary of the Invention
[0005] The purpose of this invention is to provide a method, system, and program product for managing a work problem database, in order to solve the aforementioned problems existing in the prior art.
[0006] To achieve the above objectives, the present invention adopts the following technical solution: Firstly, a method for managing a work problem database is provided, including: Obtain a work-related problem questionnaire and extract a combination of work-related problem texts from the questionnaire, the combination of work-related problem texts including the question stem and the completion question text; Keyword extraction is performed on the question text and the fill-in-the-blank question text to obtain the question keyword combination of the work question text, and the question keyword combination of the work question text is used as the question keyword combination of the fill-in-the-blank question text. Based on the combination of question keywords, the text of the completed questions is hierarchically classified to determine the question type of the completed text. Retrieve a set of archived questions of the corresponding question type from a pre-set question database. The set of archived questions includes several archived question texts and combinations of question keywords associated with each archived question text. Based on the filled-in question text, the combination of question keywords in the filled-in question text, each archived question text, and the combination of question keywords in each archived question text, determine the mixed similarity between the filled-in question text and each archived question text; When the mixed similarity between the filled-in question text and each of the archived question texts meets the set question addition conditions, the filled-in question text will be used as the question text to be archived; Add the text of the question to be archived to the archived question set of the corresponding question type in the question library after numbering it, as a newly added archived question text, and simultaneously add the associated question keyword combination to the corresponding archived question set.
[0007] In one possible design, the acquisition of the job-related questionnaire includes: The system receives encrypted questionnaire packages uploaded by the information collection terminal and decrypts them using a pre-stored key to obtain a work-related question survey questionnaire.
[0008] In one possible design, the keyword extraction from the question stem and the fill-in-the-blank question text to obtain the question keyword combination of the work question text includes: The KeyBERT keyword extraction method was used to extract keywords from the question stem and the fill-in-the-blank questions, resulting in several keywords for each. By deduplicating and summarizing the keywords in the question text and the answer text, we obtain the keyword combination of the work question text.
[0009] In one possible design, the hierarchical classification of the completed question text based on question keyword combinations to determine the question type includes: The question keyword combination filled in is substituted into a preset question type mapping dictionary for matching, and the number of keyword overlaps between the question keyword combination and the keyword set corresponding to each question type in the question type mapping dictionary is determined. The question type mapping dictionary contains several question types and the keyword set corresponding to each question type. Determine the percentage of keyword overlap between the keyword combination of the question and the keyword set corresponding to each question type within the keyword set corresponding to each question type; The question type with the highest proportion and exceeding the set threshold will be selected as the question type for filling in the question text; When the highest percentage does not exceed the set threshold, the question keywords are preprocessed and then input into a pre-set question classification model for question classification to obtain the corresponding question classification result. The corresponding question classification result is used as the question type for filling in the question text. The question classification model adopts a pre-trained BERT model.
[0010] In one possible design, determining the mixed similarity between the filled-in question text and each archived question text, based on the filled-in question text, the combination of question keywords in the filled-in question text, each archived question text, and the combination of question keywords in each archived question text, includes: Perform text content matching between the entered question text and the corresponding archived question text to determine the degree of text matching between the entered question text and the corresponding archived question text; A pre-trained Word2vec word vector model is used to vectorize the keyword combinations of the fill-in question text and the keyword combinations of the saved question texts to obtain the keyword group vector of the fill-in question text and the keyword group vector of the saved question texts. The keyword group vector contains several keyword vector components. Calculate the cosine similarity between the keyword vector of the fill-in question text and the keyword vector of each archived question text, and use the calculated cosine similarity as the word vector similarity between the fill-in question text and the corresponding archived question text; Determine the intersection and union of the keyword combinations in the completed question text and the keyword combinations in each archived question text, and determine the number of keywords in the intersection and the number of keywords in the union; Divide the number of keywords in the intersection by the number of keywords in the union to obtain the keyword overlap between the completed question text and the corresponding archived question text; The weighted summation of text matching degree, word vector similarity and keyword overlap between the filled-in question text and the corresponding archived question text is used to obtain the mixed similarity between the filled-in question text and the corresponding archived question text.
[0011] In one possible design, the step of performing text content matching between the filled-in question text and the corresponding archived question text, and determining the text matching degree between the filled-in question text and the corresponding archived question text, includes: The text of the question to be filled in and the corresponding archived text of the question to be filled in are processed into text segments to obtain several first text blocks of the text of the question to be filled in and several second text blocks of the text of the archived text. The first and second text blocks are processed by character encoding to obtain the encoding sequence of each first text block and the encoding sequence of each second text block; Iterate through and determine the sequence overlap between the encoded sequence of each first text block and the encoded sequence of each second text block, and determine the number of first text blocks and second text blocks with a sequence overlap lower than a set overlap threshold, S1, and the number of first text blocks and second text blocks with a sequence overlap not lower than the set overlap threshold, S2. Based on the number of first and second text blocks with corresponding sequence overlap less than a set overlap threshold, S1, and the number of first and second text blocks with corresponding sequence overlap not less than a set overlap threshold, S2, calculate the text matching degree P between the filled-in question text and the corresponding archived question text, where P = S2 ÷ (S1 + S2).
[0012] In one possible design, the step of using the filled-in question text as the question text to be archived when the mixed similarity between the filled-in question text and each archived question text meets the set question addition conditions includes: If the mixed similarity between the entered question text and each of the archived question texts does not exceed the set similarity threshold, the entered question text will be used as the question text to be archived.
[0013] Secondly, a work-related question database management system is provided, including a questionnaire acquisition unit, a phrase extraction unit, a question classification unit, an archive retrieval unit, a text matching unit, an archive determination unit, and a question update unit, wherein: The questionnaire acquisition unit is used to acquire a work problem survey questionnaire and extract a work problem text combination from the work problem survey questionnaire, wherein the work problem text combination includes the question stem text and the completion question text. The phrase extraction unit is used to extract keywords from the question text and the fill-in-the-blank question text to obtain the question keyword combination of the work question text, and use the question keyword combination of the work question text as the question keyword combination of the fill-in-the-blank question text; The question classification unit is used to hierarchically classify the completed question text based on the combination of question keywords, and determine the question type of the completed question text; The archive retrieval unit is used to retrieve a set of archived questions of the corresponding question type from a preset question database. The set of archived questions includes several archived question texts and combinations of question keywords associated with each archived question text. The text matching unit is used to determine the mixed similarity between the filled-in question text and each archived question text based on the filled-in question text, the question keyword combination of the filled-in question text, each archived question text, and the question keyword combination of each archived question text. The archiving determination unit is used to select the filled-in question text as the question text to be archived when the mixed similarity between the filled-in question text and each archived question text meets the set question addition conditions. The issue update unit is used to add the text of the issue to be archived, after numbering it, to the set of archived issues of the corresponding issue type in the issue database as a newly added archived issue text, and to synchronously add the associated issue keyword combination to the corresponding set of archived issues.
[0014] Thirdly, a work problem database management system is provided, including: Memory, used to store instructions; A processor is configured to read instructions stored in the memory and execute any one of the working problem library management methods described in the first aspect above, according to the instructions.
[0015] Fourthly, a computer-readable storage medium is provided, on which instructions are stored, which, when executed on a computer, cause the computer to perform any of the working problem database management methods described in the first aspect. Simultaneously, a computer program product is also provided, which, when executed on a computer, performs any of the working problem database management methods described in the first aspect.
[0016] Beneficial Effects: This invention extracts questions from work-related questionnaires, performs keyword extraction and question classification, and then uses a special similarity judgment method based on question keywords and question types to determine whether it is a new question text. When it is determined to be a new question text, it uses the new question text to achieve accurate and efficient updates to the question database. This invention automates the entire process from question collection, extraction, classification to deduplication and database entry, greatly reducing labor costs; it can combine rule matching and deep semantic understanding of questions to achieve fast and high-precision question classification; it effectively avoids redundancy in the question database while automatically updating it; and it can automatically identify and absorb new questions, enabling the work-related question database to continuously learn and evolve, forming a knowledge system that integrates various question types. Attached Figure Description
[0017] 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.
[0018] Figure 1 This is a flowchart illustrating the method in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the system configuration in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the system configuration in Embodiment 3 of the present invention. Detailed Implementation
[0019] It should be noted that the descriptions of these embodiments are intended to aid in understanding the invention and do not constitute a limitation thereof. The specific structural and functional details disclosed herein are merely for describing exemplary embodiments of the invention. However, the invention may be embodied in many alternative forms and should not be construed as being limited to the embodiments described herein.
[0020] It should be understood that, unless otherwise explicitly specified and limited, the corresponding terms should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in the embodiments according to the specific circumstances.
[0021] Specific details are provided in the following description to provide a complete understanding of the exemplary embodiments. However, those skilled in the art will understand that the exemplary embodiments can be implemented without these specific details. For example, apparatus may be shown in block diagrams to avoid obscuring the examples with unnecessary details. In other embodiments, well-known processes, structures, and techniques may be omitted with non-essential details to avoid obscuring the embodiments.
[0022] Example 1: This embodiment provides a method for managing a work problem database, which can be applied to a corresponding data management server, such as... Figure 1 As shown, the method includes the following steps: S1. Obtain the work problem questionnaire and extract the work problem text combination from the work problem questionnaire, wherein the work problem text combination includes the question stem text and the fill-in-the-blank question text.
[0023] In practice, the server receives the encrypted questionnaire package uploaded by the information collection terminal, and decrypts the encrypted questionnaire package using a pre-stored key to obtain the work problem survey questionnaire. Then, the server extracts the work problem text combination from the work problem survey questionnaire, which includes the question stem text and the fill-in-the-blank text.
[0024] S2. Extract keywords from the question text and the fill-in-the-blank question text to obtain the question keyword combination of the work question text, and use the question keyword combination of the work question text as the question keyword combination of the fill-in-the-blank question text.
[0025] In practice, the server can use KeyBERT (an unsupervised keyword extraction tool based on the BERT model that extracts the most representative keywords and phrases from text through semantic similarity calculation) to extract keywords from the question stem and fill-in-the-blank texts, obtaining several keywords for each. Then, these keywords are deduplicated and summarized to obtain a keyword combination for the work-related question text. Finally, this keyword combination is used as the keyword combination for the fill-in-the-blank question text.
[0026] S3. Based on the combination of question keywords, classify the completed question text hierarchically to determine the question type of the completed question text.
[0027] In practice, the server matches the keyword combinations of the submitted question text with a pre-set question type mapping dictionary to determine the overlap between the keyword combinations and the keyword sets corresponding to each question type in the dictionary. This question type mapping dictionary, pre-built or configured and uploaded to the server by administrators, contains several question types and their corresponding keyword sets. Then, the server determines the percentage of keyword overlap between the keyword combinations and the keyword sets corresponding to each question type within those sets. The question type with the highest percentage overlap exceeding a set threshold is selected as the question type for the submitted question text. If the highest percentage overlap does not exceed the set threshold, the pre-processed keyword combinations are input into a pre-set question classification model for question classification. The corresponding classification result is then used as the question type for the submitted question text. The question classification model uses a pre-trained BERT model.
[0028] S4. Retrieve the archived question set of the corresponding question type from the preset question database. The archived question set contains several archived question texts and question keyword combinations associated with each archived question text.
[0029] In practice, after determining the question type for the question text to be filled in, the server can retrieve the corresponding set of archived questions from the pre-set question database. The set of archived questions contains several numbered archived question texts and combinations of question keywords associated with each archived question text.
[0030] S5. Based on the filled-in question text, the combination of question keywords in the filled-in question text, each archived question text, and the combination of question keywords in each archived question text, determine the mixed similarity between the filled-in question text and each archived question text.
[0031] In specific implementation, the server can perform text content matching between the filled-in question text and the corresponding archived question text to determine the text matching degree between the filled-in question text and the corresponding archived question text. Specifically, this includes: (1) performing text segmentation processing on the filled-in question text and the corresponding archived question text respectively to obtain several first text blocks of the filled-in question text and several second text blocks of the archived question text; (2) performing character encoding processing on each first text block and second text block to obtain the encoding sequence of each first text block and the encoding sequence of each second text block; (3) traversing to determine the encoding sequence of each first text block and the encoding sequence of each second text block. (3) The sequence overlap between the encoded sequences of the two text blocks is determined, and the number of first and second text blocks with corresponding sequence overlap lower than the set overlap threshold S1 and the number of first and second text blocks with corresponding sequence overlap not lower than the set overlap threshold S2 are determined; (4) Based on the number of first and second text blocks with corresponding sequence overlap lower than the set overlap threshold S1 and the number of first and second text blocks with corresponding sequence overlap not lower than the set overlap threshold S2, the text matching degree P between the filled question text and the corresponding archived question text is calculated, where P = S2 ÷ (S1 + S2).
[0032] The server can use a pre-trained Word2vec word vector model to vectorize the keyword combinations of the fill-in-the-question text and the keyword combinations of each archived question text, resulting in keyword group vectors for both the fill-in-the-question text and the archived question texts. Each keyword group vector contains several keyword vector components. Then, the cosine similarity between the keyword group vectors of the fill-in-the-question text and the keyword group vectors of each archived question text is calculated, and this calculated cosine similarity is used as the word vector similarity between the fill-in-the-question text and the corresponding archived question text.
[0033] At the same time, the server can determine the intersection and union of the keyword combinations in the question text and the keyword combinations in each of the archived question texts, and determine the number of keywords in the intersection and the number of keywords in the union; then, by dividing the number of keywords in the intersection by the number of keywords in the union, the keyword overlap between the question text and the corresponding archived question text is obtained.
[0034] Finally, the server performs a weighted summation of the text matching degree, word vector similarity, and keyword overlap between the filled-in question text and the corresponding archived question text to obtain the mixed similarity between the filled-in question text and the corresponding archived question text.
[0035] S6. When the mixed similarity between the filled-in question text and each of the archived question texts meets the set question addition conditions, the filled-in question text will be used as the question text to be archived.
[0036] In practice, when the mixed similarity between the submitted question text and any of the archived question texts does not exceed the set similarity threshold, the server will treat the submitted question text as a question text to be archived. However, when the mixed similarity between the submitted question text and a certain archived question text exceeds the set similarity threshold, it indicates that the submitted question text is highly similar to that archived question text, and the submitted question text does not need to be updated as a new question text.
[0037] S7. Add the text of the question to be archived to the archived question set of the corresponding question type in the question library after numbering it, as a newly added archived question text, and synchronously add the associated question keyword combination to the corresponding archived question set.
[0038] In practice, after determining that the question text to be filled in is the question text to be archived, the server adds the question text to be archived with a number and adds it to the archived question set of the corresponding question type in the question database as a new archived question text. The associated question keyword combination is also added to the corresponding archived question set simultaneously to update the question text in the question database.
[0039] This method automates the entire process from problem collection, extraction, classification to deduplication and database entry, greatly reducing labor costs. It can combine rule matching and deep semantic understanding of problems to achieve fast and high-precision problem classification. While automatically updating the problem database, it effectively avoids redundancy. Furthermore, it can automatically identify and incorporate new problems, enabling the working problem database to continuously learn and evolve, forming a knowledge system that aggregates various problem types.
[0040] Example 2: This embodiment provides a work problem database management system, such as... Figure 2 As shown, it includes a questionnaire acquisition unit, a phrase extraction unit, a question classification unit, an archive retrieval unit, a text matching unit, an archive determination unit, and a question update unit, wherein: The questionnaire acquisition unit is used to acquire a work problem survey questionnaire and extract a work problem text combination from the work problem survey questionnaire, wherein the work problem text combination includes the question stem text and the completion question text. The phrase extraction unit is used to extract keywords from the question text and the fill-in-the-blank question text to obtain the question keyword combination of the work question text, and use the question keyword combination of the work question text as the question keyword combination of the fill-in-the-blank question text; The question classification unit is used to hierarchically classify the completed question text based on the combination of question keywords, and determine the question type of the completed question text; The archive retrieval unit is used to retrieve a set of archived questions of the corresponding question type from a preset question database. The set of archived questions includes several archived question texts and combinations of question keywords associated with each archived question text. The text matching unit is used to determine the mixed similarity between the filled-in question text and each archived question text based on the filled-in question text, the question keyword combination of the filled-in question text, each archived question text, and the question keyword combination of each archived question text. The archiving determination unit is used to select the filled-in question text as the question text to be archived when the mixed similarity between the filled-in question text and each archived question text meets the set question addition conditions. The issue update unit is used to add the text of the issue to be archived, after numbering it, to the set of archived issues of the corresponding issue type in the issue database as a newly added archived issue text, and to synchronously add the associated issue keyword combination to the corresponding set of archived issues.
[0041] Example 3: This embodiment provides a work problem database management system, such as... Figure 3 As shown, at the hardware level, it includes: The data interface is used to establish data communication between the processor and external data terminals; Memory, used to store instructions; The processor is used to read instructions stored in the memory and execute the working problem library management method in Embodiment 1 according to the instructions.
[0042] Optionally, the system also includes an internal bus, through which the processor, memory, and data interface can be interconnected. This internal bus can be a PCIe (Peripheral Component Interconnect Eexpress) bus, which can be divided into an address bus, a data bus, a control bus, etc. The memory can include, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Flash Memory, First Input First Output (FIFO), and / or First In Last Out (FILO). The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.
[0043] Example 4: This embodiment provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the working problem library management method of Embodiment 1. 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.
[0044] This embodiment also provides a computer program product that, when run on a computer, executes the job problem database management method of Embodiment 1. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.
[0045] 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 managing a work problem database, characterized in that, include: Obtain a work-related problem questionnaire and extract a combination of work-related problem texts from the questionnaire, the combination of work-related problem texts including the question stem and the completion question text; Keyword extraction is performed on the question text and the fill-in-the-blank question text to obtain the question keyword combination of the work question text, and the question keyword combination of the work question text is used as the question keyword combination of the fill-in-the-blank question text. Based on the combination of question keywords, the text of the completed questions is hierarchically classified to determine the question type of the completed text. Retrieve a set of archived questions of the corresponding question type from a pre-set question database. The set of archived questions includes several archived question texts and combinations of question keywords associated with each archived question text. Based on the filled-in question text, the combination of question keywords in the filled-in question text, each archived question text, and the combination of question keywords in each archived question text, determine the mixed similarity between the filled-in question text and each archived question text; When the mixed similarity between the filled-in question text and each of the archived question texts meets the set question addition conditions, the filled-in question text will be used as the question text to be archived; Add the text of the question to be archived to the archived question set of the corresponding question type in the question library as a new archived question text, and simultaneously add the associated question keyword combination to the corresponding archived question set.
2. The method for managing a work problem database according to claim 1, characterized in that, The questionnaire for obtaining work-related questions includes: The system receives encrypted questionnaire packages uploaded by the information collection terminal and decrypts them using a pre-stored key to obtain a work-related question survey questionnaire.
3. The method for managing a work problem database according to claim 1, characterized in that, The keyword extraction from the question stem and the fill-in-the-blank question text yields a combination of keywords for the work-related question text, including: The KeyBERT keyword extraction method was used to extract keywords from the question stem and the fill-in-the-blank questions, resulting in several keywords for each. By deduplicating and summarizing the keywords in the question text and the answer text, we obtain the keyword combination of the work question text.
4. The method for managing a work problem database according to claim 1, characterized in that, The hierarchical classification of completed question texts based on question keyword combinations to determine the question type includes: The question keyword combination filled in is substituted into a preset question type mapping dictionary for matching, and the number of keyword overlaps between the question keyword combination and the keyword set corresponding to each question type in the question type mapping dictionary is determined. The question type mapping dictionary contains several question types and the keyword set corresponding to each question type. Determine the percentage of keyword overlap between the keyword combination of the question and the keyword set corresponding to each question type within the keyword set corresponding to each question type; The question type with the highest proportion and exceeding the set threshold will be selected as the question type for filling in the question text; When the highest percentage does not exceed the set threshold, the question keywords are preprocessed and then input into a pre-set question classification model for question classification to obtain the corresponding question classification result. The corresponding question classification result is used as the question type for filling in the question text. The question classification model adopts a pre-trained BERT model.
5. The method for managing a work problem database according to claim 1, characterized in that, The determination of the mixed similarity between the filled-in question text and each archived question text, based on the filled-in question text, the combination of question keywords in the filled-in question text, each archived question text, and the combination of question keywords in each archived question text, includes: Perform text content matching between the entered question text and the corresponding archived question text to determine the degree of text matching between the entered question text and the corresponding archived question text; A pre-trained Word2vec word vector model is used to vectorize the keyword combinations of the fill-in question text and the keyword combinations of the saved question texts to obtain the keyword group vector of the fill-in question text and the keyword group vector of the saved question texts. The keyword group vector contains several keyword vector components. Calculate the cosine similarity between the keyword vector of the fill-in question text and the keyword vector of each archived question text, and use the calculated cosine similarity as the word vector similarity between the fill-in question text and the corresponding archived question text; Determine the intersection and union of the keyword combinations in the completed question text and the keyword combinations in each archived question text, and determine the number of keywords in the intersection and the number of keywords in the union; Divide the number of keywords in the intersection by the number of keywords in the union to obtain the keyword overlap between the completed question text and the corresponding archived question text; The weighted summation of text matching degree, word vector similarity and keyword overlap between the filled-in question text and the corresponding archived question text is used to obtain the mixed similarity between the filled-in question text and the corresponding archived question text.
6. The method for managing a work problem database according to claim 5, characterized in that, The step of matching the text content of the entered question text with the corresponding archived question text to determine the text matching degree between the entered question text and the corresponding archived question text includes: The text of the question to be filled in and the corresponding archived text of the question to be filled in are processed into text segments to obtain several first text blocks of the text of the question to be filled in and several second text blocks of the text of the archived text. The first and second text blocks are processed by character encoding to obtain the encoding sequence of each first text block and the encoding sequence of each second text block; Iterate through and determine the sequence overlap between the encoded sequence of each first text block and the encoded sequence of each second text block, and determine the number of first text blocks and second text blocks with a sequence overlap lower than a set overlap threshold, S1, and the number of first text blocks and second text blocks with a sequence overlap not lower than the set overlap threshold, S2. Based on the number of first and second text blocks with corresponding sequence overlap less than a set overlap threshold, S1, and the number of first and second text blocks with corresponding sequence overlap not less than a set overlap threshold, S2, calculate the text matching degree P between the filled-in question text and the corresponding archived question text, where P = S2 ÷ (S1 + S2).
7. The method for managing a work problem database according to claim 1, characterized in that, When the mixed similarity between the filled-in question text and each of the archived question texts meets the set question addition conditions, the filled-in question text is used as the question text to be archived, including: If the mixed similarity between the entered question text and each of the archived question texts does not exceed the set similarity threshold, the entered question text will be used as the question text to be archived.
8. A work problem database management system, characterized in that, It includes a questionnaire acquisition unit, a phrase extraction unit, a question classification unit, an archive retrieval unit, a text matching unit, an archive determination unit, and a question update unit, among which: The questionnaire acquisition unit is used to acquire a work problem survey questionnaire and extract a work problem text combination from the work problem survey questionnaire, wherein the work problem text combination includes the question stem text and the completion question text. The phrase extraction unit is used to extract keywords from the question text and the fill-in-the-blank question text to obtain the question keyword combination of the work question text, and use the question keyword combination of the work question text as the question keyword combination of the fill-in-the-blank question text; The question classification unit is used to hierarchically classify the completed question text based on the combination of question keywords, and determine the question type of the completed question text; The archive retrieval unit is used to retrieve a set of archived questions of the corresponding question type from a preset question database. The set of archived questions includes several archived question texts and combinations of question keywords associated with each archived question text. The text matching unit is used to determine the mixed similarity between the filled-in question text and each archived question text based on the filled-in question text, the question keyword combination of the filled-in question text, each archived question text, and the question keyword combination of each archived question text. The archiving determination unit is used to select the filled-in question text as the question text to be archived when the mixed similarity between the filled-in question text and each archived question text meets the set question addition conditions. The issue update unit is used to add the text of the issue to be archived, after numbering it, to the set of archived issues of the corresponding issue type in the issue database as a newly added archived issue text, and to synchronously add the associated issue keyword combination to the corresponding set of archived issues.
9. A work problem database management system, characterized in that, include: Memory, used to store instructions; A processor is configured to read instructions stored in the memory and execute the working problem library management method according to any one of claims 1-7.
10. A computer program product, characterized in that, When the computer program product is run on a computer, the working problem library management method according to any one of claims 1-7 is executed.