A method, device and storage medium for searching automobile insurance cases

Abstract

The present invention discloses a method, device and medium for searching auto insurance cases, which relates to the technical field of case search, including: obtaining the target ID of the target auto insurance case through an ID generation module, and obtaining the target physical address; using the transmission module as the first operating thread to store the data on a disk to obtain a relational database; at the same time, using the retrieval module as the second operating thread to obtain a keyword library according to a keyword decomposition mechanism; adding the keyword library to the relational database to form a target routing table; searching for auto insurance cases according to the routing table set. The present invention solves the technical problems in the prior art that auto insurance case retrieval is difficult, inefficient and inaccurate, and it is difficult to balance data storage and retrieval efficiency, and achieves the technical effect of accurately finding similar cases, comprehensively building a keyword library to accurately describe case characteristics, and realizing efficient and accurate retrieval of auto insurance cases.

Description

Technical Field

[0001] The present invention relates to the technical field of case retrieval, and in particular to a method, device and storage medium for automobile insurance case retrieval. Background Art

[0002] In the field of auto insurance business, with the continuous growth of vehicle ownership and the increasing complexity and diversity of auto insurance cases, traditional auto insurance case retrieval methods face many challenges. At present, existing auto insurance case retrieval technologies usually rely only on simple case numbers or basic information for retrieval, which makes it difficult to fully and accurately describe case characteristics. This results in the inability to accurately locate truly relevant auto insurance cases when searching for similar cases, and the accuracy of similar case searches is low. At the same time, due to the lack of an effective multi-link keyword analysis mechanism, the processing of key links such as report information, investigation information, and damage assessment information of the target auto insurance case is relatively simple, and the key information of each link cannot be fully mined to build a comprehensive keyword library, resulting in an incomplete description of case characteristics. This inefficient retrieval method not only wastes a lot of time and labor costs, but also affects the speed and accuracy of auto insurance claims, and cannot meet the growing needs of auto insurance business.

[0003] The existing technology has technical problems such as difficulty in retrieving auto insurance cases, low efficiency and inaccuracy, and difficulty in balancing data preservation and retrieval efficiency. Summary of the invention

[0004] The present application provides a method, device and storage medium for retrieving auto insurance cases, which are used to solve the technical problems in the prior art that auto insurance case retrieval is difficult, inefficient and inaccurate, and it is difficult to balance data preservation and retrieval efficiency.

[0005] In view of the above problems, the present application provides a method, device and storage medium for retrieving auto insurance cases.

[0006] In a first aspect of the present application, a method for retrieving a motor vehicle insurance case is provided, the method comprising:

[0007] The target ID of the target auto insurance case is obtained through the ID generation module, and the target physical address of the target ID is obtained through the routing module; the transmission module is used as the first operating thread to save the mapping relationship between the target ID and the target physical address to a disk to obtain a relational database; at the same time, the retrieval module is used as the second operating thread to perform multi-link keyword decomposition on the target auto insurance case according to the keyword decomposition mechanism to obtain a keyword library; the keyword library is added to the relational database to form a target routing table, and a routing table set is formed through the target routing table; and auto insurance cases are retrieved according to the routing table set.

[0008] A second aspect of an embodiment of the present application provides an electronic device, comprising: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement a method for retrieving auto insurance cases as described in any one of claims 1 to 6.

[0009] A third aspect of an embodiment of the present application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements a method for retrieving auto insurance cases provided by the present application.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] The target ID of the target auto insurance case is obtained through the ID generation module, and the target physical address of the target ID is obtained through the routing module; the transmission module is used as the first operating thread to store the data on the disk and obtain a relational database; at the same time, the retrieval module is used as the second operating thread to perform multi-link keyword decomposition on the target auto insurance case to obtain a keyword library; the keyword library is added to the relational database to form a target routing table; the auto insurance case is retrieved according to the routing table set. The technical effect of accurately finding similar cases, comprehensively building a keyword library to accurately describe case characteristics, and realizing efficient and accurate retrieval of auto insurance cases is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0013] Figure 1 A flowchart of a method for retrieving auto insurance cases provided in an embodiment of the present application;

[0014] Figure 2 This is a schematic diagram of the structure of an exemplary electronic device of the present application.

[0015] Description of the reference numerals: bus 300 , receiver 301 , processor 302 , transmitter 303 , memory 304 , bus interface 305 . DETAILED DESCRIPTION

[0016] The present application provides a method, device and storage medium for retrieving auto insurance cases, which are used to solve the technical problems in the prior art that auto insurance case retrieval is difficult, inefficient and inaccurate, and it is difficult to balance data preservation and retrieval efficiency.

[0017] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0018] Embodiment 1, as Figure 1 As shown, the present application provides a method for retrieving auto insurance cases, the method comprising:

[0019] Step S100: obtaining a target ID of a target auto insurance case through the ID generation module, and obtaining a target physical address of the target ID through the routing module.

[0020] Specifically, first, the ID generation module starts working and is responsible for generating a globally unique ID. For the target auto insurance case, the target ID is generated by comprehensively considering multiple key information of the case, such as the reporting time, vehicle information (such as license plate number, vehicle type, etc.), accident location, and some key features related to the accident. By processing this information, a unique identifier is generated as the target ID to ensure that each auto insurance case has its own unique identification, which is convenient for subsequent storage and retrieval operations. Then, the routing module comes into play. It is responsible for routing the ID to a real physical address. After obtaining the target ID, the routing module will find the corresponding target physical address based on the internal mapping mechanism. This mapping mechanism is based on a pre-established routing table, which records the correspondence between different IDs and corresponding physical addresses. By performing a search operation in this routing table, the routing module can determine the target physical address pointed to by the target ID. This address is the specific location where the data related to the target auto insurance case is stored, such as a specific path on a storage device or a specific storage area on a server node.

[0021] Step S200: Using the transmission module as a first operation thread, the mapping relationship between the target ID and the target physical address is saved on a disk to obtain a relational database.

[0022] Specifically, the transmission module is set as the first job thread, and its core operation is to save the mapping relationship between the target ID and the target physical address. When the insurance company's case is sent over, it first generates a unique digital ID (i.e., the target ID) through the ID generation module, and calculates the storage physical location of this ID (i.e., the target physical address) through the routing module. In this step, the transmission module will obtain the mapping relationship between the target ID and the corresponding target physical address, and store this information persistently on a storage medium such as a disk. Through this disk saving process, a relational database is finally formed. This relational database is crucial for subsequent case retrieval operations, and provides a basis for quickly locating case data, so that the corresponding physical address can be quickly found according to the target ID, and then the case-related information can be obtained.

[0023] Step S300: At the same time, the search module is used as the second operation thread to perform multi-link keyword decomposition on the target auto insurance case according to the keyword decomposition mechanism to obtain a keyword library.

[0024] Specifically, the retrieval module is used as the second operation thread and the operation of the transmission module is carried out simultaneously. The retrieval module processes the target auto insurance case according to the keyword decomposition mechanism. The auto insurance case is mainly divided into multiple links such as reporting, investigation, and damage assessment. The keyword decomposition module presets a large number of keywords for these links. For the target auto insurance case, first, in the reporting link, the information in the case is compared with the preset keywords in the reporting link. For example, the preset reporting keywords include collision, scratch, animal, fire and many other accident-related words. If the information in the case contains these keywords or their variant words (such as the deformation comparison of the center grille decorative strip and the air intake grille decorative strip), the generated target ID will be written into the dictionary table of the corresponding keywords. Similarly, similar operations are performed in the investigation link and the damage assessment link. The preset keywords in the investigation link are such as non-motor vehicle, motor vehicle, unilateral, etc., and the preset keywords in the damage assessment link are such as comprehensive repair shop, front bumper lower trim, etc. The information of the case in these links is compared with the preset keywords in each link, and the case ID is also written into the corresponding keyword dictionary table if it meets the conditions. After comparing and processing keywords in multiple links such as reporting, investigation, and damage assessment, a keyword library containing relevant keywords in each link and corresponding case IDs is finally formed. This keyword library provides an important basis for subsequent case retrieval, enabling the rapid location of relevant cases based on the keywords entered by the user. By running the keyword and data storage actions in parallel and at the same time, it not only ensures the storage efficiency, but also facilitates the subsequent light and fast retrieval. Compared with traditional full-text retrieval, it has better user experience and efficiency advantages.

[0025] Step S400: adding the keyword library to the relational database to form a target routing table, and forming a routing table set through the target routing table.

[0026] Specifically, the keyword library obtained in step S300 is added to the relational database formed in step S200. Each keyword in the keyword library has corresponding case ID information, which is integrated into the relational database to form a new structure, namely the target routing table. The target routing table not only contains the mapping relationship between the target ID and the target physical address, but also incorporates the corresponding relationship between the keyword and the case ID. Then, a routing table set is formed through this target routing table. This routing table set is a collection of multiple routing tables (including target routing tables formed by different case types or different processing stages), which provides a more comprehensive and richer information basis for the retrieval of subsequent motor insurance cases. Based on this routing table set, according to the keywords or other search conditions entered by the user, the physical address and corresponding case ID of the relevant case can be quickly located, and the detailed information of the case can be obtained.

[0027] Step S500: Retrieve auto insurance cases according to the routing table set.

[0028] Specifically, when a user initiates a search request, according to the search conditions entered by the user (such as keywords and the logical relationship between keywords, such as AND, OR, NOT, etc.), first, the routing table related to these keywords is searched in the routing table set. For each keyword, the corresponding keyword dictionary table will be found (these dictionary tables are formed when the keyword library is constructed, and have been added to the relational database to form the target routing table, and are included in the routing table set), and then the corresponding case ID set is obtained. Then, according to the logical relationship between the keywords specified by the user and the keywords, the corresponding set operations (such as intersection, union, difference, etc.) are performed on each case ID set. For example, if the user requires to retrieve cases that meet multiple keyword conditions at the same time, the intersection of the relevant case ID sets is taken; if the case that meets one or more of the conditions is required to be retrieved, the union of the relevant case ID sets is taken, etc. After the final case ID set is obtained through operation, these case IDs are handed over to the routing module. The routing module will query the physical address of each case through these case IDs based on the target routing table in the routing table set. Finally, these physical addresses are handed over to the transmission module, which extracts case data based on the physical addresses and presents the cases to the user, thus completing the retrieval process of the auto insurance cases.

[0029] In a possible manner, step S400 further includes:

[0030] Step S410: Obtaining an arbitrary case search instruction, wherein the arbitrary case search instruction refers to an instruction for searching any motor vehicle insurance case.

[0031] Step S420: Based on the arbitrary case search instruction, the keyword decomposition mechanism is retrieved, and an arbitrary keyword library of the arbitrary motor vehicle insurance case is obtained.

[0032] Step S430: traverse the arbitrary keyword library in the routing table set to obtain similar car insurance cases corresponding to similar keyword libraries.

[0033] Step S440: Send the similar auto insurance cases as the search results of the arbitrary case search instruction to the client.

[0034] Specifically, S410 is the trigger point of the entire search process. When the user needs to find a specific auto insurance case, a search instruction will be sent through the client. This search instruction is received and identified as an arbitrary case search instruction, which clearly indicates that any auto insurance case is to be searched, that is, it is not limited to a specific known case, but to find auto insurance cases that meet the requirements based on various conditions provided by the user. This search instruction is designed to meet the diverse search needs of users. Auto insurance cases contain rich information, such as reporting information, investigation information, damage assessment information and other links. Users initiate searches based on any one or more keywords or conditions in these links. This instruction serves as the starting signal of the search operation and triggers a series of subsequent search processes.

[0035] Once any case search instruction is received, the keyword decomposition mechanism will be called according to the instruction. The keyword decomposition mechanism is a processing method that presets a large number of keywords for multiple links such as reporting, investigation, and damage assessment of auto insurance cases. This mechanism is also followed for any auto insurance case. First, in the reporting stage, it will be checked whether the case information contains preset keywords such as collision, scratch, animal, fire, etc. If it contains or its variants meet the requirements, the case-related identification will be written into the dictionary table of the corresponding keywords. Then, in the investigation stage, it is compared whether there are keywords such as motor vehicles, and the identification writing operation is also performed if it meets the conditions. Finally, in the damage assessment stage, check whether there are keywords such as comprehensive repair shop and front bumper lower trim. If they meet the requirements, corresponding processing is performed. By comparing and processing the keywords of various links such as reporting, investigation, and damage assessment, the arbitrary keyword library of the arbitrary auto insurance case is finally obtained. This keyword library contains the keywords related to each link and the corresponding case identification information, which provides an important basis for subsequent retrieval operations.

[0036] After obtaining any keyword library, it is traversed in the routing table set. The routing table set is carefully constructed in the previous data processing process. It integrates the mapping relationship between case ID and physical address and the corresponding relationship between keywords and case ID. During the traversal process, the keywords in any keyword library are compared with the keyword dictionary table contained in each routing table in the routing table set. This comparison involves the complete match and partial match of keywords, such as considering synonyms, near synonyms or related concept words of keywords. When the keywords in any keyword library have a high degree of similarity with the keywords in the keyword dictionary table in a routing table (which may be based on a pre-set similarity standard, such as a certain number of keywords are the same or key information matches, etc.), the keyword library corresponding to the routing table is identified as a similar keyword library. The auto insurance cases corresponding to the similar keyword library are similar auto insurance cases. These similar auto insurance cases are found through the similarity association between their corresponding keyword library and any keyword library, which to a certain extent meets the search needs initiated by users and lays the foundation for providing accurate search results in the future.

[0037] At this time, these similar auto insurance cases are used as the response results to any case search command, with the purpose of providing users with fast and accurate auto insurance case search services. These similar auto insurance cases contain rich information, such as report information, investigation information, damage assessment information and other detailed contents of each link. These case information is sorted and packaged, and then sent to the client through network communication and other means. After receiving these search results, the client will display them to the user. The user can further understand the situation of related cases based on the information of these similar auto insurance cases, so as to meet his needs for auto insurance case search. This step completes the entire search process and realizes the complete process from the user initiating the search command to obtaining the search results.

[0038] In a possible manner, step S430 further includes:

[0039] S431: Step a: extracting a first routing table from the routing table set, wherein the first routing table corresponds to a first relational database, and the first relational database has a mapping relationship between a first keyword library, a first ID and a first physical address.

[0040] S432: Step b: Obtain a first similarity ratio, where the first similarity ratio refers to a similarity ratio between the arbitrary keyword library and the first keyword library.

[0041] S433: Step c: Repeat steps a to b, and form a similarity ratio set based on the first similarity ratio obtained.

[0042] S434: Step d: Filter the keyword library corresponding to the target similarity ratio in the similarity ratio set, record it as the similar keyword library, and use the auto insurance case corresponding to the similar keyword library as the similar auto insurance case.

[0043] Specifically, when processing the routing table set, the first thing to do is to perform an extraction operation, that is, to select one of the routing tables from the routing table set, which is defined as the first routing table here. The first routing table has a corresponding relationship with a specific first relational database, and this corresponding relationship is established during the processing and storage of the motor vehicle insurance case data. The first relational database contains multiple key elements, including the mapping relationship between the first keyword library, the first ID, and the first physical address. The first keyword library is formed by decomposing the motor vehicle insurance case by keywords, and storing the relevant keywords and the case ID in association with each link of the case (such as reporting, investigation, damage assessment, etc.), which is an important basis for subsequent similarity judgment and retrieval. The first ID is a unique identifier for a specific motor vehicle insurance case in the database, which is associated with various information of the case, and the specific location of the case data in the storage device is determined through the mapping relationship of the first physical address. This mapping relationship between the first keyword library, the first ID, and the first physical address possessed by the first relational database corresponding to the first routing table.

[0044] In step b, in order to obtain the first similarity ratio, the matching and weight of the keywords need to be considered comprehensively. First, the keywords in the two keyword libraries are comprehensively compared. For each keyword in any keyword library, check whether it has a completely matched keyword in the first keyword library. If there is a completely matched keyword, it is recorded as a complete match. At the same time, the matching of related concepts of keywords should also be considered. If a keyword in any keyword library does not have a complete match in the first keyword library, but there is a related concept match (such as synonyms, near synonyms, variant words, etc.), it is also regarded as a match. Then, a weight is set for each keyword, and the weight is set according to its importance in the retrieval of auto insurance cases. For example, the key action keywords in the reporting link (such as "collision") may have a higher weight, while some relatively less critical descriptive keywords may have a lower weight. When calculating the first similarity ratio, the following formula is used: First Similarity Ratio = (Number of Exact Matches × Average Weight of Exact Match Keywords + Number of Related Concept Matches × Average Weight of Related Concept Match Keywords) ÷ (Total Number of Keywords in Any Keyword Library × Average Weight of All Keywords) Among them, the number of exact matches is the number of exact matches in the two keyword libraries; the average weight of exact matches is the average of the weights of all exact matches; the number of related concept matches is the number of related concept matches; the average weight of related concept match keywords is the average of the weights of all related concept match keywords; the total number of keywords in any keyword library is the total number of keywords in any keyword library; the average weight of all keywords is the average of the weights of all keywords in the two keyword libraries.

[0045] In step c, it is necessary to repeat the operations of step a and step b to obtain multiple similarity ratios and form a similarity ratio set. First, repeating step a means extracting a routing table from the routing table set again as a new "first routing table", which corresponds to a new relational database, which also has a corresponding mapping relationship between a keyword library, an ID and a physical address. Then, repeat step b, for the keyword library corresponding to the newly extracted routing table, calculate its similarity ratio with any keyword library, that is, obtain a new "first similarity ratio". This process will take into account factors such as keyword matching and weight, and by comparing the keywords in the two keyword libraries, the similarity ratio is obtained according to the established algorithm. By continuously repeating steps a and b, a similarity ratio with any keyword library is calculated for each routing table in the routing table set. Then, these obtained similarity ratios are added to the similarity ratio set in sequence. The formation of the similarity ratio set is to comprehensively reflect the similarity between any keyword library and the keyword library corresponding to each routing table in the routing table set, and provide a data basis for subsequent screening of similarity ratios that meet the requirements in the set, so as to find similar auto insurance cases that better match the user's search needs.

[0046] Each similarity ratio in the similarity ratio set is analyzed, and the target similarity ratio is defined as the value with the largest similarity ratio, which means that the keyword library that is most similar to any keyword library is being found. By comparing all the ratio values ​​in the similarity ratio set, the largest ratio is determined. This largest similarity ratio represents the highest degree of similarity between the two keyword libraries. When the target similarity ratio is determined, the corresponding keyword library is recorded as a similar keyword library, which indicates that the keyword library is closest in content to any keyword library input by the user. Because there is a corresponding relationship between the keyword library and the auto insurance case, the auto insurance case corresponding to this similar keyword library is determined as a similar auto insurance case. This screening method is intended to provide users with the most relevant auto insurance case results that best meet their search intentions. By finding the keyword library with the largest similarity ratio, the accuracy and relevance of the search results can be maximized, allowing users to obtain the required auto insurance case information more quickly and efficiently.

[0047] In one possible manner, step S432 further includes:

[0048] Step S4321: extract the first arbitrary word in the arbitrary keyword library.

[0049] Step S4322: Determine whether the first arbitrary word exists in the first keyword library.

[0050] Step S4323: If it exists, add the first arbitrary word to the inner point set; if it does not exist, retrieve the first arbitrary deformation word set of the first arbitrary word, wherein the first arbitrary deformation word set includes the first arbitrary deformation word.

[0051] Step S4324: Determine whether the first arbitrary variant word exists in the first keyword library.

[0052] Step S4325: If it exists, add the first arbitrary word to the inner point set; if it does not exist, add the first arbitrary word to the outer point set.

[0053] Step S4326: Obtain the total number of words in the inner point set and the outer point set.

[0054] Step S4327: record the ratio of the number of words in the inner point set to the total number of words as the first similarity ratio.

[0055] Specifically, the first arbitrary word is extracted from the arbitrary keyword library. This operation is the basis for subsequent comparison. First, the extraction of the first arbitrary word starts the entire comparison process. The arbitrary keyword library is generated to meet the user's search needs for specific auto insurance cases. It contains various keywords related to auto insurance cases. The extracted first arbitrary word will be the first object to be compared. The basic comparison is to directly search and match the first arbitrary word in other keyword libraries (such as the first keyword library). If the exact same first arbitrary word is found in the first keyword library, it preliminarily indicates that the two keyword libraries have a certain similarity in this word. The variant comparison is performed when the basic comparison does not find a match. If the first arbitrary word does not exist in the first keyword library, the variant word set of the first arbitrary word will be retrieved, including synonyms, antonyms, variant words, etc. Then these transformed words are searched in the first keyword library. If the transformed words of the first arbitrary word are found, it can also indicate to a certain extent that there are similarities between the two keyword libraries. Although the forms of the words are different, the meanings expressed may be similar. Through such a comparison method, the similarity between the two keyword libraries can be judged more comprehensively and accurately, thereby providing a strong basis for finding similar auto insurance cases.

[0056] Determine whether the first arbitrary word extracted from the arbitrary keyword library exists in the first keyword library. The first keyword library comes from the relational database corresponding to the specific routing table in the routing table set, which represents a keyword set of a specific batch of auto insurance cases. If the first arbitrary word exists in the first keyword library, it means that there is a direct match between the two keyword libraries, indicating that the auto insurance cases corresponding to the two keyword libraries are similar in some aspects. For example, if the first arbitrary word is "collision" and there is also the word "collision" in the first keyword library, it means that the two auto insurance cases being compared both involve collision events. This judgment operation is one of the important steps in screening similar auto insurance cases. By judging whether each word in the arbitrary keyword library exists in the first keyword library one by one, the similarity between the two keyword libraries is preliminarily determined, providing a basis for subsequent further analysis and screening. If multiple common keywords are found, it is more likely to determine that the corresponding auto insurance cases are similar, thereby meeting the user's search needs.

[0057] If the first arbitrary word is found in the first keyword library, the first arbitrary word is added to the inner point set. The inner point set represents a vocabulary set with direct similarity between the two keyword libraries, which means that this word appears in the descriptions of two different auto insurance cases, further indicating that the corresponding auto insurance cases are similar in this aspect. If the first arbitrary word does not exist in the first keyword library, the first arbitrary variant word set of the first arbitrary word will be retrieved. This variant word set includes various variant words related to the first arbitrary word, such as synonyms, near synonyms, and variant words. In the description of auto insurance cases, different people will use different but similar words to express the same event or feature. By considering these variant words, the similarity between the two keyword libraries can be more comprehensively evaluated. For example, if the first arbitrary word is "car accident", its variant word set may include "traffic accident" and "car crash". In this way, even if the first arbitrary word itself is not in the first keyword library, if its variant words are in, it can also indicate the similarity between the two keyword libraries to a certain extent, providing more clues and basis for finding similar auto insurance cases in the future.

[0058] The inflected words of the first arbitrary word are checked. If the inflected word set of the first arbitrary word is retrieved because the first arbitrary word does not exist in the first keyword library in step S4323, it is now necessary to determine whether these inflected words exist in the first keyword library. If the first arbitrary inflected word exists in the first keyword library, it means that although the original first arbitrary word is not in the first keyword library, similarities are found through the inflected words, which is very important in the retrieval of auto insurance cases because different people may use different words to express the same concept when describing auto insurance cases. For example, the first arbitrary word is "scrape", and there is no such word in the first keyword library, but if its inflected word "scrape" exists in the first keyword library, it means that the two keyword libraries have certain similarities in this sense. This judgment helps to more comprehensively evaluate the similarity between the arbitrary keyword library and the first keyword library, thereby providing a stronger basis for accurately finding similar auto insurance cases. If similarities are found through inflected words, the impact of these similarities on auto insurance cases can be further analyzed later to better meet the user's retrieval needs.

[0059] If the first arbitrary word exists in the first keyword library, it means that similarity with the first arbitrary word is found through the word, and the first arbitrary word is added to the inner point set. The inner point set represents a vocabulary set with similarity between the two keyword libraries. This shows that although the first arbitrary word does not exist in the first keyword library, it is associated through its word, which further shows that the two keyword libraries have similarities to a certain extent, and the corresponding auto insurance cases may also have similarities. If the first arbitrary word does not exist in the first keyword library, the system adds the first arbitrary word to the outer point set, which represents a vocabulary set with no similarity between the two keyword libraries, which means that neither the first arbitrary word itself nor its word can find a match in the first keyword library, indicating that this word has no direct or indirect similarity with the content in the first keyword library, and the corresponding auto insurance case is unlikely to have similarity in the aspect represented by this word. Through such an operation, the first arbitrary word extracted from the arbitrary keyword library is clearly classified, providing a clear basis for the subsequent calculation of the similarity ratio, so as to more accurately determine the similarity between the arbitrary keyword library and the first keyword library, so as to find similar auto insurance cases that meet the user's search needs.

[0060] The total number of words in the inner point set and the outer point set is obtained. This total number of words reflects the total number of all words considered after judging the first arbitrary word and its variants. The inner point set and the outer point set represent words with similarity and no similarity between the two keyword libraries, respectively. By determining the total number of words, basic data is provided for calculating the similarity ratio.

[0061] The ratio of the number of words in the inner point set to the total number of words is recorded as the first similarity ratio. This ratio intuitively reflects the similarity between any keyword library and the first keyword library. If the number of words in the inner point set is large and the total number of words is relatively small, then the ratio will be high, indicating that the two keyword libraries are similar to a large extent; conversely, if the number of words in the inner point set is small, the ratio will be low, indicating that the similarity between the two keyword libraries is low. In the retrieval of auto insurance cases, this similarity ratio is very important, helping to determine the keyword library that is most similar to any keyword library entered by the user, so as to find the corresponding similar auto insurance cases. By repeating this process and comparing multiple keyword libraries, we can finally select the similar auto insurance case set that best meets the user's needs, providing users with accurate and efficient retrieval results.

[0062] In a possible manner, step S300 further includes:

[0063] Step S310: Obtain target information of the target auto insurance case, the target information including target report information, target investigation information and target damage assessment information; obtain a preset keyword library, the preset keyword library including a first preset word; determine whether the target report information contains the first preset word; if it does, add the first preset word to the target report keyword library; construct a primary double decomposition channel, the primary double decomposition channel includes a primary high-speed decomposition channel and a primary low-speed decomposition channel, and store the target report keyword library in the primary high-speed decomposition channel, and store the preset keyword library in the primary low-speed decomposition channel; respectively perform decomposition on the target investigation information through the primary high-speed decomposition channel and the primary low-speed decomposition channel. Decompose and judge to obtain a first judgment result; obtain a target investigation keyword library according to the first judgment result; construct an ultimate double decomposition channel, the ultimate double decomposition channel includes an ultimate high-speed decomposition channel and an ultimate low-speed decomposition channel, and store the target reporting keyword library and the target investigation keyword library in the ultimate high-speed decomposition channel, and store the preset keyword library in the ultimate low-speed decomposition channel; decompose and judge the target damage assessment information through the ultimate high-speed decomposition channel and the ultimate low-speed decomposition channel respectively to obtain a second judgment result; obtain a target damage assessment keyword library according to the second judgment result; the target reporting keyword library, the target investigation keyword library and the target damage assessment keyword library together constitute the keyword library.

[0064] Specifically, the target information includes target reporting information, target investigation information and target damage assessment information. The target reporting information is obtained by the following technical means: a data interface is established with the insurance company's business system, using a RESTful API or an interface based on the SOAP protocol. When a car insurance case occurs, the parties report the case to the insurance company, and the report information will be recorded in the insurance company's database system. By calling the data interface, the target reporting information is extracted from the database, including the time, location, vehicle model, contact information of the reporter and other details of the accident. This information is stored in a temporary data storage area, such as a data structure in memory or a temporary file, for subsequent rapid processing. When obtaining the target investigation information, surveyors are dispatched to the accident site for investigation. The surveyors will use mobile devices, such as tablets or smart phones, to record photos of the scene, measurement data, preliminary judgment of accident responsibility and other information. These devices upload the investigation information to the insurance company's server via a wireless network. By establishing a connection with the server, the target investigation information is obtained using the data interface. For the acquisition of target damage assessment information, the extent of the vehicle's loss is assessed based on the survey results and the damage to the vehicle, and the repair plan and cost estimate are determined. The damage assessment information will be recorded in the insurance company's damage assessment management system. Through the data interface with the damage assessment management system, the target damage assessment information is extracted, including repair items, a list of replacement parts, and repair cost estimates. This information is stored in a specific data storage area for subsequent comprehensive analysis with other target information. By establishing a data interface to interact with the insurance company's various business systems, the target information of the target auto insurance case is effectively obtained, providing basic data for subsequent analysis and processing.

[0065] Obtain a preset keyword library, which is pre-stored in a database, such as a relational database MySQL or a non-relational database MongoDB. Through a database query statement, such as an SQL query statement, retrieve the content of the preset keyword library from the database, including the first preset word. In addition, store the preset keyword library in a configuration file, such as an XML file. When the program starts, read the configuration file and load the keyword library into the memory for subsequent quick access.

[0066] Next, determine whether the target report information contains the first preset word, use a string matching algorithm to make the determination, and use an exact matching algorithm to directly compare the text in the target report information with the first preset word to see if they are completely consistent. If the target report information is complex, preprocess it first, such as performing a word segmentation operation to divide the text into individual words or phrases, and then perform an exact match.

[0067] If the judgment result is that the target report information contains the first preset word, the first preset word is added to the target report keyword library. The target report keyword library can be a data structure, such as a list, a set or a dictionary. If a list is used, the first preset word is directly added to the end of the list. If a set is used, the uniqueness of the keyword is ensured to avoid repeated addition. If a dictionary is used, the first preset word is used as a key, or stored together with some related attribute values, so as to facilitate more complex analysis and processing later.

[0068] The primary dual decomposition channel is constructed to process the information of the target auto insurance case more efficiently. First, the primary dual decomposition channel is constructed by designing a specific data structure. A channel class is created using an object-oriented programming method, which contains two properties: the primary high-speed decomposition channel and the primary low-speed decomposition channel. The primary high-speed decomposition channel is mainly used to quickly process information related to the target report keyword library. It is designed as a data structure with efficient search and matching functions. The hash table structure is used to quickly map keywords to corresponding storage locations through hash functions, thereby achieving fast search and matching. When the target report keyword library is stored in the primary high-speed decomposition channel, each keyword in the target report keyword library is traversed and stored in the corresponding position in the hash table. In this way, when the target investigation information is processed later, it is possible to quickly find out whether there are keywords related to the target investigation information in the primary high-speed decomposition channel. The primary low-speed decomposition channel is used to process the preset keyword library. Since the preset keyword library is usually relatively stable and universal, it does not need to pursue extreme speed like the high-speed decomposition channel. A simple array or linked list structure is used to store the preset keyword library. When the preset keyword library is stored in the primary low-speed decomposition channel, the keywords in the preset keyword library only need to be added to the array or linked list in sequence. When the target survey information is subsequently decomposed and judged, the primary low-speed decomposition channel can be used as a reference and work together with the primary high-speed decomposition channel to more comprehensively analyze the relationship between the target survey information and the keyword library.

[0069] Regarding the step of decomposing and judging the target survey information through the primary high-speed decomposition channel and the primary low-speed decomposition channel to obtain the first judgment result, first, for the primary high-speed decomposition channel, since it stores the target report keyword library, when processing the target survey information, a fast matching algorithm is adopted, using the fast search function of the hash table, traversing each word or phrase in the target survey information, and searching in the hash table whether there is a target report keyword that matches it. If a match is found, it is recorded as part of the judgment result. For the primary low-speed decomposition channel, because it stores the preset keyword library, the processing process can be relatively slow but more comprehensive. The words in the target survey information are compared with the preset keywords one by one, and a string matching algorithm is used, such as exact matching, fuzzy matching or regular expression matching. If matching keywords are found, they are also recorded. The matching results of the two channels are combined to obtain the first judgment result, which is a list of keywords related to the target survey information and the identification of which channel they are from.

[0070] According to the first judgment result, a target survey keyword library is obtained. First, an empty data structure is created to store the target survey keywords, such as using a list or a set, traversing the keyword list in the first judgment result, and adding the keywords therein to this data structure. If it is necessary to remove duplicate keywords, use the characteristics of the set to ensure uniqueness. At the same time, different weights or attributes are assigned to the keywords according to the channels from which they come. For example, if a keyword comes from a primary high-speed decomposition channel, it means that it has a strong association with the target report information, and a higher weight is assigned. If it comes from a primary low-speed decomposition channel, the association is relatively weak, and a lower weight is assigned. In this way, a target survey keyword library is constructed, providing a more targeted keyword set for subsequent processing.

[0071] When constructing the ultimate dual decomposition channel, for the ultimate high-speed decomposition channel, the red-black tree algorithm is selected to achieve efficient data storage and fast search. The red-black tree is a self-balancing binary search tree with the following characteristics and advantages: First, the time complexity of the insertion, deletion and search operations of the red-black tree is O(logn), where n is the number of nodes in the tree, which means that even when processing a large number of keywords, it can maintain a high efficiency. The process of storing the target report keyword library and the target investigation keyword library in the ultimate high-speed decomposition channel is as follows: First, create an empty red-black tree data structure. Then, traverse each keyword in the target report keyword library and insert it as a node into the red-black tree. The insertion operation will sort according to the characteristics of the keyword (such as the lexicographic order of the string) to ensure the balance of the tree. Then, traverse the keywords in the target investigation keyword library and insert them into the red-black tree. In this way, the ultimate high-speed decomposition channel contains all the keywords from the target report and target investigation, and stores them in an orderly and efficient way. For the ultimate low-speed decomposition channel, a relatively simple array structure is still used to store the preset keyword library. Just add the keywords in the preset keyword library to the array in sequence. When the target damage assessment information is subsequently decomposed and judged, the red-black tree in the ultimate high-speed decomposition channel quickly searches for words in the target damage assessment information that match the keywords of the target report or target investigation. If no match is found in the red-black tree, a comparison is performed one by one in the array of the ultimate low-speed decomposition channel to find out whether it matches the preset keywords. The ultimate dual decomposition channel constructed in this way can make full use of the efficiency of the red-black tree and the simplicity of the array, providing strong technical support for accurately decomposing and judging the target damage assessment information.

[0072] The target damage assessment information is decomposed and judged through the ultimate high-speed decomposition channel and the ultimate low-speed decomposition channel respectively to obtain the second judgment result. For the ultimate high-speed decomposition channel, since it is constructed using the red-black tree algorithm, it can efficiently search when processing the target damage assessment information. Traverse each word in the target damage assessment information and search in the red-black tree. If a matching keyword is found, it means that the target damage assessment information is associated with a keyword in the target report keyword library or the target investigation keyword library, and the match is recorded. For the ultimate low-speed decomposition channel, it is an array structure that stores a preset keyword library. The words in the target damage assessment information are compared with the preset keywords in the array one by one. If a match is found, it is also recorded. By processing the two channels separately, the second judgment result is obtained comprehensively. This result includes the words in the target damage assessment information that are associated with each keyword library and their source channel information.

[0073] According to the second judgment result, the target damage assessment keyword library is obtained, and a new data structure, such as a set or list, is created to store the target damage assessment keywords. The matching words in the second judgment result are traversed and added to the target damage assessment keyword library. If repeated keywords need to be removed, the characteristics of the set are used or judgment is performed before adding. At the same time, different attributes or weights are assigned to the words according to their source channels for subsequent analysis and processing.

[0074] Finally, the target reporting keyword library, the target investigation keyword library and the target damage assessment keyword library together constitute a complete keyword library. This keyword library integrates the key information of the target auto insurance case at each stage of reporting, investigation and damage assessment, and provides strong support for the comprehensive analysis and processing of auto insurance cases. Through this method of constructing the keyword library in stages and finally integrating it, we can more accurately grasp the characteristics and needs of auto insurance cases and improve the efficiency and accuracy of claims processing.

[0075] In a possible manner, step S310 further includes:

[0076] Step S311: The primary high-speed decomposition channel fully traverses the target case reporting keyword library in the target investigation information based on the first traversal speed to obtain a first traversal result.

[0077] Step S312: the primary high-speed decomposition channel locally traverses each keyword in the preset keyword library in the target survey information based on the second traversal speed to obtain a second traversal result, and the first traversal speed is much greater than the second traversal speed.

[0078] Step S313: Use the first traversal result and the second traversal result as the judgment basis for the first judgment result.

[0079] Specifically, the primary high-speed decomposition channel traverses the target report keyword library in the target investigation information at a faster first traversal speed, and checks each part of the target investigation information one by one to see whether there are keywords in the target report keyword library. Since the target report keyword library is directly related to the current target auto insurance case, the full traversal can ensure that no possible connection points are missed, and the inspection of the entire target investigation information can be completed in a shorter time to obtain the first traversal result.

[0080] The primary high-speed decomposition channel performs a partial traversal of each keyword in the preset keyword library in the target survey information at a relatively slow second traversal speed. The preset keyword library is a relatively general keyword set, which may not be completely consistent with the current target auto insurance case. Therefore, partial traversal can reduce unnecessary computing resource consumption while ensuring a certain coverage range. Moreover, the first traversal speed is much greater than the second traversal speed. This design is because the target reporting keyword library is more targeted to the current case and needs to be checked more quickly, while the preset keyword library is used as an auxiliary reference and the speed can be relatively slow.

[0081] The first traversal result and the second traversal result are used together as the judgment basis for the first judgment result. The first traversal result reflects the degree of correlation between the target survey information and the target report keyword library, while the second traversal result reflects the relationship between the target survey information and the preset keyword library. Combining these two results, we can have a more comprehensive understanding of the characteristics and potential problems of the target survey information, and provide an accurate basis for subsequent processing. For example, if the first traversal result shows that there are multiple words related to the target report keyword library in the target survey information, it means that the target survey information is highly consistent with the report situation; at the same time, if the second traversal result also finds some words related to the preset keyword library, it may imply some potential risks or special situations that require further attention.

[0082] Embodiment 2, as Figure 2 As shown, based on the inventive concept of a method for retrieving a motor vehicle insurance case in the aforementioned embodiment, the present application also provides an electronic device, comprising: at least one processor; a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the steps of any one of the methods described in the aforementioned embodiment one.

[0083] Figure 2 This is a schematic diagram of the structure of an exemplary electronic device of this application. Figure 2In the embodiment, the bus architecture is represented by bus 300, which may include any number of interconnected buses and bridges, and bus 300 connects various circuits including one or more processors represented by processor 302 and memory represented by memory 304. Bus 300 may also connect various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. Bus interface 305 provides an interface between bus 300 and receiver 301 and transmitter 303. Receiver 301 and transmitter 303 may be the same element, namely a transceiver, providing a unit for communicating with various other devices over a transmission medium. Processor 302 is responsible for managing bus 300 and general processing, while memory 304 may be used to store data used by processor 302 when performing operations.

[0084] Embodiment three, based on the same inventive concept as a method for retrieving auto insurance cases in the aforementioned embodiment, the present application also provides a computer-readable storage medium, on which a computer program is stored, and the computer program implements the steps of any one of the methods described in the aforementioned embodiment one when executed.

[0085] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0086] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0087] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A method for retrieving auto insurance cases, characterized in that: The method for retrieving a motor vehicle insurance case is applied to a device for retrieving a motor vehicle insurance case. The device for retrieving a motor vehicle insurance case includes an ID generation module, a routing module, a transmission module, and a retrieval module. The method for retrieving a motor vehicle insurance case includes: Obtaining a target ID of a target auto insurance case through the ID generation module, and obtaining a target physical address of the target ID through the routing module; The transmission module is used as a first operation thread to store the mapping relationship between the target ID and the target physical address in a disk to obtain a relational database; At the same time, the search module is used as a second operation thread to perform multi-link keyword decomposition on the target auto insurance case according to a keyword decomposition mechanism to obtain a keyword library; Adding the keyword library to the relational database to form a target routing table, and forming a routing table set through the target routing table; Car insurance cases are retrieved according to the routing table set.

2. A method for retrieving auto insurance cases according to claim 1, characterized in that: include: Obtaining an arbitrary case search instruction, wherein the arbitrary case search instruction refers to an instruction for searching any motor vehicle insurance case; Based on the arbitrary case search instruction, the keyword decomposition mechanism is retrieved, and an arbitrary keyword library of the arbitrary motor vehicle insurance case is obtained; Traversing the arbitrary keyword library in the routing table set to obtain similar auto insurance cases corresponding to similar keyword libraries; The similar auto insurance cases are sent to the client as the search results of the arbitrary case search instruction.

3. The method for retrieving auto insurance cases according to claim 2, characterized in that: include: Step a: extracting a first routing table from the routing table set, wherein the first routing table corresponds to a first relational database, and the first relational database has a mapping relationship between a first keyword library, a first ID, and a first physical address; Step b: obtaining a first similarity ratio, where the first similarity ratio refers to a similarity ratio between the arbitrary keyword library and the first keyword library; Step c: repeating step a to step b, and forming a similarity ratio set according to the first similarity ratio obtained; Step d: A keyword library corresponding to a target similarity ratio is obtained by screening the similarity ratio set, which is recorded as the similar keyword library, and the auto insurance case corresponding to the similar keyword library is taken as the similar auto insurance case.

4. A method for retrieving auto insurance cases according to claim 3, characterized in that: include: Extracting the first arbitrary word in the arbitrary keyword library; Determining whether the first arbitrary word exists in the first keyword library; If it exists, add the first arbitrary word to the inner point set; if it does not exist, retrieve a first arbitrary variant word set of the first arbitrary word, wherein the first arbitrary variant word set includes the first arbitrary variant word; determining whether the first arbitrary variant word exists in the first keyword library; If it exists, add the first arbitrary word to the inner point set; if it does not exist, add the first arbitrary word to the outer point set; Obtaining the total number of words in the inner point set and the outer point set; The ratio of the number of words in the internal point set to the total number of words is recorded as the first similarity ratio.

5. The method for retrieving auto insurance cases according to claim 1, characterized in that: include: Obtaining target information of the target auto insurance case, wherein the target information includes target case reporting information, target investigation information, and target damage assessment information; Acquire a preset keyword library, wherein the preset keyword library includes a first preset word; Determining whether the target reporting information contains the first preset word; If it is included, adding the first preset word to the target reporting keyword library; Constructing a primary dual decomposition channel, the primary dual decomposition channel comprising a primary high-speed decomposition channel and a primary low-speed decomposition channel, storing the target case reporting keyword in the primary high-speed decomposition channel, and storing the preset keyword in the primary low-speed decomposition channel; Decomposing and judging the target survey information through the primary high-speed decomposition channel and the primary low-speed decomposition channel respectively, and obtaining a first judgment result; Obtaining a target survey keyword library according to the first judgment result; Constructing an ultimate dual decomposition channel, the ultimate dual decomposition channel includes an ultimate high-speed decomposition channel and an ultimate low-speed decomposition channel, storing the target case reporting keyword library and the target investigation keyword library in the ultimate high-speed decomposition channel, and storing the preset keyword library in the ultimate low-speed decomposition channel; Decomposing and judging the target damage assessment information through the ultimate high-speed decomposition channel and the ultimate low-speed decomposition channel respectively, and obtaining a second judgment result; Obtaining a target damage assessment keyword library according to the second judgment result; The target reporting keyword library, the target investigation keyword library and the target damage assessment keyword library together constitute the keyword library.

6. A method for retrieving auto insurance cases according to claim 5, characterized in that: include: The primary high-speed decomposition channel fully traverses the target case reporting keyword library in the target investigation information based on the first traversal speed to obtain a first traversal result; The primary high-speed decomposition channel locally traverses each keyword in the preset keyword library in the target survey information based on a second traversal speed to obtain a second traversal result, and the first traversal speed is much greater than the second traversal speed; The first traversal result and the second traversal result are used as judgment basis for the first judgment result.

7. An electronic device, characterized in that: The electronic device comprises: A memory for storing executable instructions; The processor is used to implement a method for retrieving auto insurance cases according to any one of claims 1 to 6 when executing executable instructions stored in the memory.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, a method for retrieving automobile insurance cases as described in any one of claims 1 to 6 is implemented.

Images