Full-text retrieval method and system for interactive electronic technical manual of ship equipment
By constructing a technical database of ship equipment and user behavior analysis, combined with multimodal preference level judgment, the problem of low efficiency in ship equipment information retrieval is solved, customized services and efficient maintenance are realized, and user satisfaction and equipment management are improved intelligently level.
Patent Information
- Application Number
- CN202411024588.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-29
AI Technical Summary
The prior art cannot effectively meet the personalized information needs of ship equipment users, resulting in low information retrieval efficiency and accuracy, especially when repairing and replacing equipment, it is difficult to provide targeted and customized services.
By establishing a technical database of ship equipment, using natural language processing and semantic analysis technology, the equipment association index and adaptation index are constructed, combined with user behavior analysis, multimodal preference level judgment is provided, interactive electronic technical manual full text search system is realized, and customized Recommended information and repair plans.
It improves the efficiency and accuracy of ship equipment management, meets users' personalized needs, improves user satisfaction and equipment reliability, and provides intelligent and convenient management support.
Smart Images

Figure CN119003750A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of ship equipment technical manual retrieval, and in particular to a full-text retrieval method and system for an interactive electronic technical manual of ship equipment. Background Art
[0002] In today's increasingly complex and competitive marine equipment field, a good technical manual is not only a guide for operation, but also a key support for ship operation and maintenance. Although existing technologies provide users with a basic framework for equipment information, there are still a series of challenges that need to be urgently addressed.
[0003] With the diversification of ship equipment types and sizes, users' demand for information is becoming increasingly diversified and personalized. However, existing technologies often fail to effectively understand and meet users' personalized needs, lacking targeted information recommendations and customized services. For example, when users consult about the maintenance of ship equipment, they will think about later repairs and replacement of similar products, and some users have different acceptance of text and video technical explanations, which makes it difficult for users to find specific information, reducing the efficiency and accuracy of information retrieval. Summary of the invention
[0004] 1. Technical issues to be resolved
[0005] In view of the deficiencies in the prior art, the present invention provides a full-text search method and system for an interactive electronic technical manual of ship equipment, which solves the problems mentioned in the background technology.
[0006] (II) Technical solution
[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: A full-text retrieval method for interactive electronic technical manuals of ship equipment, comprising:
[0008] The crawling technology is used to crawl different types of ship equipment information from the public database. The ship equipment information includes the function, technical parameters and operation guide of the equipment, and a ship technology database is established. After preprocessing the ship technology database, the ship equipment information in the ship technology database is entity recognized and the association relationship between the equipment is extracted through natural language processing technology, and the equipment association index Gl is constructed. The equipment with the equipment association index Gl higher than or equal to the association threshold Q is classified into one category to obtain several association groups.
[0009] Represent several association groups in the form of a graph, construct a ship equipment knowledge graph, identify the nodes and edges between the devices in each association group, and add the value of the association index Gl above the edge; collect the adaptation value of the associated device and the number of replaceable device models from the public database, construct the adaptation index Sp, and add the first replaceable device label to the device with the adaptation index Sp higher than or equal to the adaptation threshold X;
[0010] When the user extracts a query request, the user's ship equipment usage instructions are sent and collected, the query user request information and the actual usage information of the user's ship equipment are collected, and a query data set is established. Through semantic analysis technology, the first query word is identified and obtained, and the user's historical search records and query preferences are collected to construct a user behavior index Xw; the query data set is analyzed to obtain the user's ship equipment maintenance and troubleshooting records, and a maintenance index Wx is constructed. After evaluation based on the maintenance index Wx, a suitable search scheme is matched according to the evaluation result. In obtaining the suitable search scheme, the user behavior index Xw is judged to obtain the judgment result, and based on the multimodal preference level of the judgment result, the text, voice or video data is uniformly processed and sent to the user end.
[0011] Preferably, target databases are selected, including ship equipment manufacturer websites, technical forums, industry association websites, and public databases, including PubMed and IEEE's equivalent function Xplore;
[0012] Select Scrapy, Beautiful Soup and Selenium tools, analyze the structure and content of the target website, and formulate a crawling strategy, including determining the target page to be crawled, the frequency of crawling and the depth of crawling; determine the crawled ship keywords, tags and URL patterns, and locate the target information of ship equipment; write the crawling code, crawl the link request of ship equipment information, process the response of the ship equipment information page, extract the detailed information of the ship equipment, return the results, save them in the folder of each ship equipment, and establish a ship technology database.
[0013] Preferably, the raw data in the ship technical database is cleaned, deduplicated and formatted for pre-processing;
[0014] Using natural language processing technology, perform entity recognition and association extraction on the pre-processed data to identify entities in ship equipment information, including equipment names and technical parameters; extract associations between equipment, including functional associations, control associations, safety associations, environmental protection associations, and auxiliary equipment associations;
[0015] Based on the extracted association relationships and the statistical analysis method of the association relationships, the device association index Gl is generated by the following formula:
[0016]
[0017] Among them, W ik The weight of the association between device i and device j is determined based on the strength, frequency, and importance of the association, and is adjusted by the user. jk represents the association index between device i and device j, variable k represents the index or number of the device, which is used to represent the iteration in the calculation of the association relationship weight between devices. Variable k is used to cyclically calculate the contribution of each device to the association index; n represents the number of devices.
[0018] Preferably, an association threshold Q is set, and the association index Gl between the traversed devices is traversed, and the devices with the association index Gl higher than or equal to the association threshold Q are classified into the same category, and the association groups include a ship function association group, a ship control association group, a ship function association group, a ship safety association group, a ship environmental protection association group, and a ship auxiliary equipment association group;
[0019] Devices with a relevance index GI lower than the relevance threshold Q are classified as “unclassified”;
[0020] Several association groups are represented in the form of a graph to construct a ship equipment knowledge graph, and the nodes and edges between the equipment in each association group are identified, and the value of the association index Gl is added above the edge.
[0021] Preferably, the replaceable device information of each device is collected from a public database, the replaceable device information including the model, specification, brand and supplier of the replaceable device, and for each device, the adaptation value of the replaceable device of each device and the number N of replaceable devices are collected and obtained;
[0022] The adaptability value of the replaceable device includes the software adaptability value SAS and the hardware adaptability value HAS. After dimensionless processing, the adaptability index Sp is generated by the following formula:
[0023]
[0024] Wherein, the software adaptation value SAS represents the adaptability between software, including version compatibility and functional matching; the hardware adaptation value HAS represents the adaptability between hardware, including interface compatibility and specification matching; N represents the number of replaceable devices, that is, the number of spare devices with the same function; max(N) represents the maximum value of the number of replaceable devices with the same function, which is used to normalize the number of replaceable devices.
[0025] Preferably, the adaptation index Sp of the device is evaluated:
[0026] For devices with an adaptation index Sp higher than or equal to the adaptation threshold X, add the first replaceable device label to indicate that the device has high adaptability and is the preferred replacement device. This makes it convenient for these devices to be the first-priority preferred backup replacement options during device maintenance, replacement, or update.
[0027] For devices with an adaptation index Sp lower than the adaptation threshold X, add the second non-recommended replacement device label to prompt the user or system that these devices have low adaptability and are not suitable as the preferred replacement devices. If such a device is a customized device, add special processing marks, including "customized device" and "special requirements", and perform significant color marking.
[0028] Preferably, when the user submits a query request, the system sends an instruction to the marine device through the network or interface connected to the marine device to request the usage information of the device. The system collects the query request information submitted by the user, including keywords and query content; and synchronously collects the actual usage information of the marine device provided by the user, including device model, usage time, and maintenance records.
[0029] Combine the collected user request information and the actual usage information of the marine device to form a query dataset; extract the maintenance and fault handling records of the user's marine device in the query dataset, and generate a maintenance index Wx through the following formula:
[0030]
[0031] In the formula, MF represents the device maintenance frequency, MR represents the device usage years, FT represents the fault degree value, quantify the influence range of the fault on the operation of other marine devices into a value from 1 to 10, and the values from 1 to 10 increase the severity level value, MT represents the maintenance time, MC represents the maintenance cost value, a, b, c, d, and f respectively represent the weight values of the device maintenance frequency MF, the device usage years MR, the fault degree value FT, the maintenance time MT, and the maintenance cost value MC, 0 < a < 1, 0 < b < 1, 0 < c < 1, 0 < d < 1, 0 < f < 1, and a + b + c + d + f = 1, A represents the first correction constant.
[0032] Preferably, analyze the query dataset and collect historical user query information to obtain historical commonly used query keywords, query frequency cxpl, query time cxsj, browsing frequency llpl, number of times of browsing multi-modal videos cs1, number of times of browsing text cs2, number of times of listening to voice explanations cs3, and historical satisfaction number value cs4. Combine the keywords with the ship code to generate the first query sentence.
[0033] The user behavior index Xw is obtained by extracting the query frequency cxpl, query time cxsj, browsing frequency llpl, number of multimodal video browsing times cs1, number of text browsing times cs2, number of audio explanation listening times cs3 and historical satisfaction times cs4 from the historical user query information, and after dimensionless processing, the user behavior index Xw is calculated by the following formula:
[0034]
[0035] In the formula, the query frequency cxpl and the query time cxsj in the user query information are multiplied and divided by the browsing frequency llpl plus 1 to avoid division by 0; then, the squares of the number of times the multimodal video is browsed cs1, the number of times the text is browsed cs2, and the number of times the voice explanation is listened to cs3 are summed and divided by the historical satisfaction number cs4 plus 1, also to avoid division by 0, and B is the second correction constant.
[0036] Preferably, the maintenance index Wx is evaluated to obtain an evaluation result, including:
[0037] When the maintenance index Wx value is between 0 and 30, it means that the equipment maintenance demand is low, the equipment status is qualified, and the daily maintenance manual, operation guide or operation precautions are adapted;
[0038] When the maintenance index Wx value is between 31-70, it means that the equipment maintenance demand is intermediate and an emergency treatment plan is needed, and common troubleshooting methods and maintenance guidance for the equipment are adapted;
[0039] When the maintenance index Wx value is between 71 and 100, it means that the equipment has a high maintenance demand and requires an emergency maintenance manual and emergency treatment plan. The replacement equipment with a high adaptation index Sp will be pushed an ordering link and the maintenance personnel contact information will be provided;
[0040] The user behavior index Xw is judged, and the judgment results are as follows:
[0041] When the user behavior index Xw value is between 0 and 30, it means that the user has a high degree of acceptance of text data, so the multimodal preference level is "text preference";
[0042] When the user behavior index Xw value is between 31 and 70, it means that the user has a high degree of acceptance of voice data, so the multimodal preference level is "voice preference";
[0043] When the user behavior index Xw is between 71 and 100, it indicates that the user tends to have a high degree of reception of video data, and therefore, the multimodal preference level is “video preference”.
[0044] The full-text retrieval system of interactive electronic technical manuals for marine equipment includes a data acquisition module, a data preprocessing module, an association extraction module, an association analysis and classification module, a knowledge graph construction module, an equipment adaptation evaluation module, a user behavior analysis module, a maintenance evaluation module, a multimodal preference level judgment module, and a result matching and sending module;
[0045] The data acquisition module is used to capture different types of ship equipment information from public databases using crawling technology, including functions, technical parameters and operating instructions, establish a ship technology database, select target databases, including ship equipment manufacturer websites, technical forums, industry association websites and public databases, including PubMed and IEEE Xplore with the same function, and write crawling codes to capture ship equipment information and save it to the ship technology database;
[0046] The data preprocessing module is used to perform cleaning, deduplication and formatting preprocessing operations on the ship technical database;
[0047] The association relationship extraction module is used to use natural language processing technology to perform entity recognition and association relationship extraction on the data to construct a device association index Gl;
[0048] The association analysis classification module is used to classify devices whose device association index Gl is higher than or equal to the association threshold Q into one category to obtain a plurality of association groups;
[0049] The knowledge graph construction module represents the association group in the form of a graph, constructs a ship equipment knowledge graph, and displays the association relationship between the equipment and the value of the association index Gl in the graph;
[0050] The device adaptation evaluation module is used to collect the adaptation value of the associated device and the quantity value of the replaceable device model from the public database, construct the adaptation index Sp, and add the first replaceable device label or the second non-recommended replacement device label to the device according to the value of the adaptation index Sp;
[0051] The user behavior analysis module is used to collect user ship equipment usage instructions and actual usage information, establish a query data set, and analyze the query data set to obtain a user behavior index Xw;
[0052] The maintenance evaluation module is used to evaluate according to the maintenance index Wx, determine the maintenance requirement of the equipment, and match the corresponding adaptation retrieval scheme;
[0053] The multimodal preference level judgment module is used to judge the user's multimodal preference level according to the value of the user behavior index Xw, including text preference, voice preference or video preference;
[0054] The result matching and sending module is used to match the corresponding search scheme and multimodal data according to the evaluation results and the judgment of the user behavior index, and send the results to the user end after uniformly processing the text, voice or video data.
[0055] (III) Beneficial effects
[0056] The present invention provides a full-text search method and system for interactive electronic technical manuals of ship equipment. It has the following beneficial effects:
[0057] (1) Through data collection and preprocessing, the system can obtain and organize ship equipment information from multiple sources to ensure the integrity and accuracy of the data. Association extraction and knowledge graph construction enable users to clearly understand the association between equipment and provide guidance for equipment maintenance and management. The equipment adaptation evaluation module helps users quickly find replacement equipment and classify it according to the adaptation index, improving maintenance efficiency and accuracy. The user behavior analysis and multimodal preference level judgment module take into account the personalized needs of users and provide users with a customized service experience. The maintenance evaluation module provides maintenance solutions based on the equipment status, improving the reliability and stability of the equipment. Finally, the result matching and sending module sends the results of the comprehensive analysis to the user end, enabling users to obtain the required information in a timely manner, improving user satisfaction and usage efficiency. The combined effect of these functions makes ship equipment management more intelligent and convenient, providing strong support for ship operation and management.
[0058] (2) Through the construction and analysis of the user behavior index Xw, the user's personalized needs and preferences can be understood. The maintenance index Wx constructed based on the user's historical search records and query preferences, as well as the maintenance and troubleshooting records of ship equipment, can achieve understanding and response to the user's personalized needs. According to the evaluation results, the appropriate retrieval scheme is matched, and judgment is made based on the user behavior index Xw to provide targeted information recommendations and customized services to meet the user's personalized needs. Through the judgment of multimodal preference levels and the unified processing of text, voice or video data, information can be sent to the user end in a form that is more in line with the user's taste according to the user's preferences. This not only improves the user's experience, but also enhances the user's trust and reliance on the technical manual, thereby improving user satisfaction.
[0059] (3) The association group is represented in the form of a graph, a ship equipment knowledge graph is constructed, and the value of the association index Gl is added above the edge, which can intuitively display the association relationship between ship equipment. Such a graph can help users understand the relationship between equipment more intuitively and provide an intuitive reference for decision-making and problem solving.
[0060] (4) Calculate and evaluate the device's adaptation index Sp. If the device's adaptation index Sp is higher than or equal to the adaptation threshold X, add a first replaceable device label to indicate that the device has high adaptability and is used as the preferred replacement device. This facilitates the use of these devices as the first priority backup replacement options when the device is maintained, replaced, or updated.
[0061] If the device's adaptation index Sp is lower than the adaptation threshold X, a second not recommended replacement device label is added to remind the user or system that these devices have low adaptability and are not suitable as the first choice replacement devices; if this device is a customized device, a special processing mark is added, including "customized device" and "special needs" and marked with a prominent color.
[0062] Through such processing and marking, users can more clearly understand the compatibility of replaceable equipment during equipment maintenance and replacement, thereby effectively selecting appropriate backup replacement options, improving the efficiency and accuracy of maintenance and replacement, and reducing maintenance delays and failure risks caused by equipment incompatibility. BRIEF DESCRIPTION OF THE DRAWINGS
[0063] Figure 1 A schematic diagram of the steps of the method for full-text retrieval of the interactive electronic technical manual of ship equipment of the present invention;
[0064] Figure 2 The present invention is a schematic diagram of the flowchart of the full-text retrieval system for the interactive electronic technical manual of ship equipment. DETAILED DESCRIPTION
[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0066] Example 1
[0067] See also Figure 1 The present invention provides a full-text retrieval method for an interactive electronic technical manual of ship equipment.
[0068] The crawling technology is used to crawl different types of ship equipment information from the public database. The ship equipment information includes the function, technical parameters and operation guide of the equipment, and a ship technology database is established. After preprocessing the ship technology database, the ship equipment information in the ship technology database is entity recognized and the association relationship between the equipment is extracted through natural language processing technology, and the equipment association index Gl is constructed. The equipment with the equipment association index Gl higher than or equal to the association threshold Q is classified into one category to obtain several association groups.
[0069] Represent several association groups in the form of a graph, construct a ship equipment knowledge graph, identify the nodes and edges between the devices in each association group, and add the value of the association index Gl above the edge; collect the adaptation value of the associated device and the number of replaceable device models from the public database, construct the adaptation index Sp, and add the first replaceable device label to the device with the adaptation index Sp higher than or equal to the adaptation threshold X;
[0070] When the user extracts a query request, the user's ship equipment usage instructions are sent and collected, the query user request information and the actual usage information of the user's ship equipment are collected, and a query data set is established. Through semantic analysis technology, the first query word is identified and obtained, and the user's historical search records and query preferences are collected to construct a user behavior index Xw; the query data set is analyzed to obtain the user's ship equipment maintenance and troubleshooting records, and a maintenance index Wx is constructed. After evaluation based on the maintenance index Wx, a suitable search scheme is matched according to the evaluation result. In obtaining the suitable search scheme, the user behavior index Xw is judged to obtain the judgment result, and based on the multimodal preference level of the judgment result, the text, voice or video data is uniformly processed and sent to the user end.
[0071] In this embodiment, the crawling technology is used to obtain different types of ship equipment information from the public database, a ship technology database is established, and the entity recognition and association relationship extraction of the equipment information are realized through the natural language processing technology. This enables users to obtain the required content more quickly and accurately when retrieving specific information, and improves the efficiency and accuracy of information retrieval. Through the construction and analysis of the user behavior index Xw, the personalized needs and preferences of users can be understood. According to the maintenance index Wx constructed based on the user's historical search records and query preferences, as well as the maintenance and troubleshooting records of the ship equipment, the understanding and response to the personalized needs of users can be realized. According to the evaluation results, the matching retrieval scheme is matched, and the user behavior index Xw is judged to provide targeted information recommendations and customized services, so as to meet the personalized needs of users. Through the judgment of multimodal preference levels and the unified processing of text, voice or video data, the information can be sent to the user end in a form that is more in line with the user's taste according to the user's preference. This not only improves the user's experience, but also enhances the user's trust and dependence on the technical manual, thereby improving the user's satisfaction.
[0072] Example 2: This example is an explanation of Example 1. Specifically, target databases are selected, including ship equipment manufacturer websites, technical forums, industry association websites, and public databases. Public databases include PubMed and IEEE Xplore, which has the same function.
[0073] Select Scrapy, Beautiful Soup and Selenium tools, analyze the structure and content of the target website, and formulate a crawling strategy, including determining the target page to be crawled, the frequency of crawling and the depth of crawling; determine the crawled ship keywords, tags and URL patterns, and locate the target information of ship equipment; write the crawling code, crawl the link request of ship equipment information, process the response of the ship equipment information page, extract the detailed information of the ship equipment, return the results, save them in the folder of each ship equipment, and establish a ship technology database.
[0074] In this embodiment, selecting a target database and adopting appropriate crawling tools and strategies will help improve the information acquisition efficiency, accuracy and practicality of the full-text retrieval system for interactive electronic technical manuals of ship equipment and provide users with a better user experience.
[0075] Example 3: This example is an explanation of Example 1. Specifically, the original data in the ship technical database is cleaned, deduplicated and formatted.
[0076] Using natural language processing technology, perform entity recognition and association extraction on the pre-processed data to identify entities in ship equipment information, including equipment names and technical parameters; extract associations between equipment, including functional associations, control associations, safety associations, environmental protection associations, and auxiliary equipment associations;
[0077] Based on the extracted association relationships and the statistical analysis method of the association relationships, the device association index Gl is generated by the following formula:
[0078]
[0079] Among them, W ik The weight of the association between device i and device j is determined based on the strength, frequency, and importance of the association, and is adjusted by the user. jk represents the association index between device i and device j, variable k represents the index or number of the device, which is used to represent the iteration in the calculation of the association relationship weight between devices. Variable k is used to cyclically calculate the contribution of each device to the association index; n represents the number of devices.
[0080] Specifically, an association threshold Q is set, and the association index Gl between the traversed devices is traversed, and the devices with the association index Gl higher than or equal to the association threshold Q are classified into the same category, and the association groups include a ship function association group, a ship control association group, a ship function association group, a ship safety association group, a ship environmental protection association group, and a ship auxiliary equipment association group;
[0081] Devices with a relevance index GI lower than the relevance threshold Q are classified as “unclassified”;
[0082] Several association groups are represented in the form of a graph to construct a ship equipment knowledge graph, and the nodes and edges between the equipment in each association group are identified, and the value of the association index Gl is added above the edge.
[0083] In this embodiment, the association group is represented in the form of a graph, a ship equipment knowledge graph is constructed, and the value of the association index Gl is added above the edge, so that the association relationship between the ship equipment can be intuitively displayed. Such a graph can help users understand the relationship between the equipment more intuitively and provide an intuitive reference for decision-making and problem solving.
[0084] Embodiment 4: This embodiment is an explanation of Embodiment 1. Specifically, replaceable device information of each device is collected from a public database. The replaceable device information includes the model, specification, brand and supplier of the replaceable device. For each device, the adaptation value of the replaceable device of each device and the number of replaceable devices N are collected and obtained.
[0085] The adaptability value of the replaceable device includes the software adaptability value SAS and the hardware adaptability value HAS. After dimensionless processing, the adaptability index Sp is generated by the following formula:
[0086]
[0087] Wherein, the software adaptation value SAS represents the adaptability between software, including version compatibility and functional matching; the hardware adaptation value HAS represents the adaptability between hardware, including interface compatibility and specification matching; N represents the number of replaceable devices, that is, the number of spare devices with the same function; max(N) represents the maximum value of the number of replaceable devices with the same function, which is used to normalize the number of replaceable devices.
[0088] Evaluate the device's adaptation index Sp:
[0089] If the device's adaptation index Sp is higher than or equal to the adaptation threshold X, add the first replaceable device label, indicating that the device has high adaptability and is the preferred replacement device, so that when the device is maintained, replaced or updated, these devices are used as the first priority backup replacement options;
[0090] If the device's adaptation index Sp is lower than the adaptation threshold X, a second not recommended replacement device label is added to remind the user or system that these devices have low adaptability and are not suitable as the first choice replacement devices; if this device is a customized device, a special processing mark is added, including "customized device" and "special needs" and marked with a prominent color.
[0091] In this embodiment, through such processing and marking, users can more clearly understand the adaptation situation of replaceable devices during equipment maintenance and replacement, thereby effectively selecting appropriate spare replacement options, improving the efficiency and accuracy of maintenance and replacement, and reducing maintenance delays and failure risks caused by equipment incompatibility.
[0092] Embodiment 5. This embodiment is an explanatory description based on Embodiment 1. Specifically, when a user submits a query request, the system sends an instruction to the marine equipment through the network or interface connected to the marine equipment to request the acquisition of equipment usage information; the system collects the query request information submitted by the user, including keywords and query content; and synchronously collects the actual usage information of the marine equipment provided by the user, including equipment model, usage time, and maintenance records.
[0093] The user request information and the actual usage information of the marine equipment collected are combined to form a query data set; the maintenance and fault handling records of the user's marine equipment in the query data set are extracted, and the maintenance index Wx is generated through the following formula:
[0094]
[0095] In the formula, MF represents the equipment maintenance frequency, MR represents the equipment usage years, FT represents the fault degree value, the influence range of the fault on the operation of other marine equipment is quantified into a value from 1 to 10, and the value from 1 to 10 increases with the severity level value, MT represents the maintenance time, MC represents the maintenance cost value, a, b, c, d, and f respectively represent the weight values of the equipment maintenance frequency MF, equipment usage years MR, fault degree value FT, maintenance time MT, and maintenance cost value MC, 0 < a < 1, 0 < b < 1, 0 < c < 1, 0 < d < 1, 0 < f < 1, and a + b + c + d + f = 1, and A represents the first correction constant.
[0096] Through the calculation of the maintenance index Wx, the system can comprehensively evaluate the maintenance and fault handling situations of the user's equipment, provide more comprehensive and accurate information for the user, and support their decisions and actions in equipment maintenance and fault handling.
[0097] Specifically, analyze the query data set, collect historical user query information, obtain historical commonly used query keywords, query frequency cxpl, query time cxsj, browsing frequency llpl, number of times of browsing multimodal videos cs1, number of times of browsing text cs2, number of times of listening to voice explanations cs3, and historical satisfaction value cs4, combine the keywords with the ship code to generate the first query sentence.
[0098] The user behavior index Xw is obtained by extracting the query frequency cxpl, query time cxsj, browsing frequency llpl, number of multimodal video browsing times cs1, number of text browsing times cs2, number of audio explanation listening times cs3 and historical satisfaction times cs4 from the historical user query information, and after dimensionless processing, the user behavior index Xw is calculated by the following formula:
[0099]
[0100] In the formula, the query frequency cxpl and the query time cxsj in the user query information are multiplied and divided by the browsing frequency llpl plus 1 to avoid division by 0; then, the squares of the number of times the multimodal video is browsed cs1, the number of times the text is browsed cs2, and the number of times the voice explanation is listened to cs3 are summed and divided by the historical satisfaction number cs4 plus 1, also to avoid division by 0, and B is the second correction constant.
[0101] Specifically, the maintenance index Wx is evaluated to obtain evaluation results, including:
[0102] When the maintenance index Wx value is between 0 and 30, it means that the equipment maintenance demand is low, the equipment status is qualified, and the daily maintenance manual, operation guide or operation precautions are adapted;
[0103] When the maintenance index Wx value is between 31-70, it means that the equipment maintenance demand is intermediate and an emergency treatment plan is needed, and common troubleshooting methods and maintenance guidance for the equipment are adapted;
[0104] When the maintenance index Wx value is between 71 and 100, it means that the equipment has a high maintenance demand and requires an emergency maintenance manual and emergency treatment plan. The replacement equipment with a high adaptation index Sp will be pushed an ordering link and the maintenance personnel contact information will be provided;
[0105] The user behavior index Xw is judged, and the judgment results are as follows:
[0106] When the user behavior index Xw value is between 0 and 30, it means that the user has a high degree of acceptance of text data, so the multimodal preference level is "text preference";
[0107] When the user behavior index Xw value is between 31 and 70, it means that the user has a high degree of acceptance of voice data, so the multimodal preference level is "voice preference";
[0108] When the user behavior index Xw is between 71 and 100, it indicates that the user tends to have a high degree of reception of video data, and therefore, the multimodal preference level is “video preference”.
[0109] In this embodiment, by comprehensively analyzing user behavior and equipment maintenance needs, the system can more accurately understand user needs and provide personalized services and recommendations, thereby improving user experience and equipment maintenance efficiency.
[0110] Example 6, please refer to Figure 2 , a full-text retrieval system for interactive electronic technical manuals of marine equipment, including a data acquisition module, a data preprocessing module, an association extraction module, an association analysis classification module, a knowledge graph construction module, an equipment adaptation evaluation module, a user behavior analysis module, a maintenance evaluation module, a multimodal preference level judgment module, and a result matching and sending module;
[0111] The data acquisition module is used to capture different types of ship equipment information from public databases using crawling technology, including functions, technical parameters and operating instructions, establish a ship technology database, select target databases, including ship equipment manufacturer websites, technical forums, industry association websites and public databases, including PubMed and IEEE Xplore with the same function, and write crawling codes to capture ship equipment information and save it to the ship technology database;
[0112] The data preprocessing module is used to perform cleaning, deduplication and formatting preprocessing operations on the ship technical database;
[0113] The association relationship extraction module is used to use natural language processing technology to perform entity recognition and association relationship extraction on the data to construct a device association index Gl;
[0114] The association analysis classification module is used to classify devices whose device association index Gl is higher than or equal to the association threshold Q into one category to obtain a plurality of association groups;
[0115] The knowledge graph construction module represents the association group in the form of a graph, constructs a ship equipment knowledge graph, and displays the association relationship between the equipment and the value of the association index Gl in the graph;
[0116] The device adaptation evaluation module is used to collect the adaptation value of the associated device and the quantity value of the replaceable device model from the public database, construct the adaptation index Sp, and add the first replaceable device label or the second non-recommended replacement device label to the device according to the value of the adaptation index Sp;
[0117] The user behavior analysis module is used to collect user ship equipment usage instructions and actual usage information, establish a query data set, and analyze the query data set to obtain a user behavior index Xw;
[0118] The maintenance evaluation module is used to evaluate according to the maintenance index Wx, determine the maintenance requirement of the equipment, and match the corresponding adaptation retrieval scheme;
[0119] The multimodal preference level judgment module is used to judge the user's multimodal preference level according to the value of the user behavior index Xw, including text preference, voice preference or video preference;
[0120] The result matching and sending module is used to match the corresponding search scheme and multimodal data according to the evaluation results and the judgment of the user behavior index, and send the results to the user end after uniformly processing the text, voice or video data.
[0121] In this embodiment, through data collection and preprocessing, the system can obtain and organize ship equipment information from multiple sources to ensure the integrity and accuracy of the data. Association extraction and knowledge graph construction enable users to clearly understand the association relationship between devices and provide guidance for equipment maintenance and management. The equipment adaptation evaluation module helps users quickly find replacement equipment and classify it according to the adaptation index, improving maintenance efficiency and accuracy. The user behavior analysis and multimodal preference level judgment module take into account the personalized needs of users and provide users with a customized service experience. The maintenance evaluation module provides maintenance solutions based on the equipment status, improving the reliability and stability of the equipment. Finally, the result matching and sending module sends the results of the comprehensive analysis to the user end, enabling users to obtain the required information in a timely manner, improving user satisfaction and usage efficiency. The combined effect of these functions makes ship equipment management more intelligent and convenient, providing strong support for ship operation and management.
[0122] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A full-text retrieval method for interactive electronic technical manuals of marine equipment, characterized in that: include, The crawling technology is used to crawl different types of ship equipment information from the public database. The ship equipment information includes the function, technical parameters and operation guide of the equipment, and a ship technology database is established. After preprocessing the ship technology database, the ship equipment information in the ship technology database is entity recognized and the association relationship between the equipment is extracted through natural language processing technology, and the equipment association index Gl is constructed. The equipment with the equipment association index Gl higher than or equal to the association threshold Q is classified into one category to obtain several association groups. Represent several association groups in the form of a graph, construct a ship equipment knowledge graph, identify the nodes and edges between the equipment in each association group, and add the value of the association index Gl above the edge; and collecting the adaptation value of the associated device and the quantity value of the replaceable device model from the public database, constructing the adaptation index Sp, and adding the first replaceable device tag to the device whose adaptation index Sp is higher than or equal to the adaptation threshold X; When the user extracts a query request, the user's ship equipment usage instructions are sent and collected, the query user request information and the actual usage information of the user's ship equipment are collected, and a query data set is established. Through semantic analysis technology, the first query word is identified and obtained, and the user's historical search records and query preferences are collected to construct a user behavior index Xw; the query data set is analyzed to obtain the user's ship equipment maintenance and troubleshooting records, and a maintenance index Wx is constructed. After evaluation based on the maintenance index Wx, a suitable search scheme is matched according to the evaluation result. In obtaining the suitable search scheme, the user behavior index Xw is judged to obtain the judgment result, and based on the multimodal preference level of the judgment result, the text, voice or video data is uniformly processed and sent to the user end.
2. The method for full-text retrieval of interactive electronic technical manual of ship equipment according to claim 1, characterized in that: Select target databases, including ship equipment manufacturer websites, technical forums, industry association websites, and public databases, including PubMed and IEEE's equivalent function Xplore; Select Scrapy, Beautiful Soup and Selenium tools, analyze the structure and content of the target website, and formulate a crawling strategy, including determining the target page to be crawled, the frequency of crawling and the depth of crawling; determine the crawled ship keywords, tags and URL patterns, and locate the target information of ship equipment; write the crawling code, crawl the link request of ship equipment information, process the response of the ship equipment information page, extract the detailed information of the ship equipment, return the results, save them in the folder of each ship equipment, and establish a ship technology database.
3. The method for full-text retrieval of interactive electronic technical manual of ship equipment according to claim 1, characterized in that: Perform pre-processing operations such as cleaning, deduplication and formatting of raw data in the ship technical database; Using natural language processing technology, perform entity recognition and association extraction on the pre-processed data to identify entities in ship equipment information, including equipment names and technical parameters; extract associations between equipment, including functional associations, control associations, safety associations, environmental protection associations, and auxiliary equipment associations; Based on the extracted association relationships and the statistical analysis method of the association relationships, the device association index Gl is generated by the following formula: Among them, W ik The weight of the association between device i and device j is determined based on the strength, frequency, and importance of the association, and is adjusted by the user. jk represents the association index between device i and device j, variable k represents the index or number of the device, which is used to represent the iteration in the calculation of the association relationship weight between devices. Variable k is used to cyclically calculate the contribution of each device to the association index; n represents the number of devices.
4. The method for full-text retrieval of interactive electronic technical manual of ship equipment according to claim 3 is characterized by: Set an association threshold Q, and traverse the association index Gl between the devices, and classify the devices with an association index Gl higher than or equal to the association threshold Q into the same category, and the association groups include a ship function association group, a ship control association group, a ship function association group, a ship safety association group, a ship environmental protection association group, and a ship auxiliary equipment association group; Classify devices with a correlation index G1 lower than the correlation threshold Q into the "unclassified" category; Several association groups are represented in the form of a graph to construct a ship equipment knowledge graph, and the nodes and edges between the equipment in each association group are identified, and the value of the association index Gl is added above the edge.
5. The method for full-text retrieval of interactive electronic technical manual of ship equipment according to claim 1, characterized in that: Collect replaceable device information of each device from a public database, the replaceable device information includes the model, specification, brand and supplier of the replaceable device, and for each device, collect and obtain the adaptation value of the replaceable device of each device and the number of replaceable devices N; The adaptability value of the replaceable device includes the software adaptability value SAS and the hardware adaptability value HAS. After dimensionless processing, the adaptability index Sp is generated by the following formula: Wherein, the software adaptation value SAS represents the adaptability between software, including version compatibility and functional matching; the hardware adaptation value HAS represents the adaptability between hardware, including interface compatibility and specification matching; N represents the number of replaceable devices, that is, the number of spare devices with the same function; max(N) represents the maximum value of the number of replaceable devices with the same function, which is used to normalize the number of replaceable devices.
6. The method for full-text retrieval of interactive electronic technical manual of ship equipment according to claim 5, characterized in that: Evaluate the device's adaptation index Sp: If the device's adaptation index Sp is higher than or equal to the adaptation threshold X, add the first replaceable device label, indicating that the device has high adaptability and is the preferred replacement device, so that when the device is maintained, replaced or updated, these devices are used as the first priority backup replacement options; If the device's adaptation index Sp is lower than the adaptation threshold X, a second not recommended replacement device label is added to remind the user or system that these devices have low adaptability and are not suitable as the first choice replacement devices; if this device is a customized device, a special processing mark is added, including "customized device" and "special needs" and marked with a prominent color.
7. The method for full-text retrieval of interactive electronic technical manual of ship equipment according to claim 1, characterized in that: When a user makes a query request, the system sends a command to the ship equipment through the network or interface connected to the ship equipment, requesting to obtain the equipment usage information; the system collects the query request information proposed by the user, including keywords and query content; And simultaneously collect the actual usage information of ship equipment provided by users, including equipment model, usage time and maintenance records; Form a query dataset from the collected user request information and the actual usage information of the ship equipment; extract the maintenance and troubleshooting records of the user's communication equipment in the query dataset, and generate a maintenance index Wx through the following formula: In the formula, MF represents the equipment maintenance frequency, MR represents the equipment usage years, FT represents the fault degree value, the influence range of the fault affecting the operation of other ship equipment is quantified into a value from 1 to 10, and the value from 1 to 10 increases the severity level value, MT represents the maintenance time, MC represents the maintenance cost value, a, b, c, d, and f respectively represent the weight values of the equipment maintenance frequency MF, the equipment usage years MR, the fault degree value FT, the maintenance time MT, and the maintenance cost value MC, 0 < a < 1, 0 < b < 1, 0 < c < 1, 0 < d < 1, 0 < f < 1, and a + b + c + d + f = 1, A represents the first correction constant.
8. The method for full-text retrieval of interactive electronic technical manual of ship equipment according to claim 1, characterized in that: Analyze the query dataset, and collect historical user query information to obtain historical commonly used query keywords, query frequency cxpl, query time cxsj, browsing frequency llpl, number of times of browsing multi-modal videos cs1, number of times of browsing text cs2, number of times of listening to voice explanations cs3, and historical satisfaction number value cs4. Combine the keywords with the ship code to generate the first query sentence; The user behavior index Xw is obtained by extracting the query frequency cxpl, query time cxsj, browsing frequency llpl, number of times of browsing multi-modal videos cs1, number of times of browsing text cs2, number of times of listening to voice explanations cs3, and historical satisfaction number value cs4 in the historical user query information. After dimensionless processing, the user behavior index Xw is calculated through the following formula: In the formula, multiply the query frequency cxpl and the query time cxsj in the user query information, and divide by the browsing frequency llpl + 1 to avoid division by zero; then, sum the squares of the number of times of browsing multi-modal videos cs1, number of times of browsing text cs2, and number of times of listening to voice explanations cs3, and divide by the historical satisfaction number value cs4 + 1, which is also to avoid division by zero, and B is the second correction constant.
9. The method for full-text retrieval of interactive electronic technical manual of ship equipment according to claim 8, characterized in that: Evaluate the maintenance index Wx to obtain the evaluation result, including: When the value of the maintenance index Wx is between 0 and 30, it indicates that the equipment maintenance requirement is low, the equipment status is qualified, and it is suitable for the daily maintenance manual, operation guide, or operation precautions; When the value of the maintenance index Wx is between 31 and 70, it indicates that the equipment maintenance requirement is medium, and an emergency treatment plan is needed, which is suitable for the common fault troubleshooting methods and maintenance guides of the equipment; When the value of the maintenance index Wx is between 71 and 100, it indicates that the equipment maintenance requirement is high, and an emergency repair manual and an emergency treatment plan are needed. Push the order link for the replacement equipment with a high adaptation index Sp, and provide the contact information of the maintenance personnel; And judge the user behavior index Xw to obtain the judgment result including the following: When the value of the user behavior index Xw is between 0 and 30, it indicates that the user has a high acceptance of the text data. Therefore, the multi-modal preference level is "text preference"; When the user behavior index Xw value is between 31 and 70, it means that the user has a high degree of acceptance of voice data, so the multimodal preference level is "voice preference"; When the user behavior index Xw is between 71 and 100, it indicates that the user tends to have a high degree of reception of video data, and therefore, the multimodal preference level is "video preference".
10. A full-text retrieval system for interactive electronic technical manuals of ship equipment, applied to a full-text retrieval method for interactive electronic technical manuals of ship equipment according to any one of claims 1 to 9, characterized in that: It includes data collection module, data preprocessing module, association relationship extraction module, association analysis and classification module, knowledge graph construction module, equipment adaptation evaluation module, user behavior analysis module, maintenance evaluation module, multimodal preference level judgment module, and result matching and sending module; The data acquisition module is used to capture different types of ship equipment information from public databases using crawling technology, including functions, technical parameters and operating instructions, establish a ship technology database, select target databases, including ship equipment manufacturer websites, technical forums, industry association websites and public databases, including PubMed and IEEE Xplore with the same function, and write crawling codes to capture ship equipment information and save it to the ship technology database; The data preprocessing module is used to perform cleaning, deduplication and formatting preprocessing operations on the ship technical database; The association relationship extraction module is used to use natural language processing technology to perform entity recognition and association relationship extraction on the data to construct a device association index Gl; The association analysis classification module is used to classify devices whose device association index Gl is higher than or equal to the association threshold Q into one category to obtain a plurality of association groups; The knowledge graph construction module represents the association group in the form of a graph, constructs a ship equipment knowledge graph, and displays the association relationship between the equipment and the value of the association index Gl in the graph; The device adaptation evaluation module is used to collect the adaptation value of the associated device and the quantity value of the replaceable device model from the public database, construct the adaptation index Sp, and add the first replaceable device label or the second non-recommended replacement device label to the device according to the value of the adaptation index Sp; The user behavior analysis module is used to collect user ship equipment usage instructions and actual usage information, establish a query data set, and analyze the query data set to obtain a user behavior index Xw; The maintenance evaluation module is used to evaluate according to the maintenance index Wx, determine the maintenance requirement of the equipment, and match the corresponding adaptation retrieval scheme; The multimodal preference level judgment module is used to judge the user's multimodal preference level according to the value of the user behavior index Xw, including text preference, voice preference or video preference; The result matching and sending module is used to match the corresponding search scheme and multimodal data according to the evaluation results and the judgment of the user behavior index, and send the results to the user end after uniformly processing the text, voice or video data.
Citation Information
Patent Citations
Full-text retrieval device and method for interactive electronic technical manual of shipping equipment
CN105528411A
Ship technical condition comprehensive evaluation system based on state monitoring
CN112101764A
Aerospace measurement and control equipment state evaluation data-based maintenance scheme autonomous recommendation system
CN116862470A
Interactive electronic manual data conversion system
CN118069905A
Method and apparatus for recommending fault investigation plan and fault resolution plan, and storage medium
WO2023206520A1
Cited By
Auxiliary management method and system for design of fixing device of container ship
CN121581819A
Container ship fixture design assistance management method and system
CN121581819B
