A correlation analysis system for automobile parts price fluctuations

Abstract

The present invention discloses a system for analyzing the price fluctuation correlation of automobile parts, which relates to the field of information management technology. The system comprises: a data preparation module collects and generates a standard claim data set for automobile insurance damage assessment. An entity extraction module performs entity recognition on the standard claim data set to obtain a repair shop characteristic entity set. An attribute clustering module extracts the attributes of the repair shop characteristic entity set, generates a repair shop accessory entity attribute set, sorts out data dimensions, and generates a repair shop accessory aggregate data set. A relationship confirmation module obtains a repair shop location relationship library based on the repair shop accessory aggregate data set to form a repair shop entity location relationship set. A graph construction module constructs a repair shop knowledge graph network. A fluctuation propagation module uses a water wave algorithm to perform price fluctuation correlation calculations on the repair shop knowledge graph network to obtain the results of the correlation impact of accessory price fluctuations. This achieves the technical effect of improving the ability to respond to price fluctuations and enhancing the transparency of accessory prices.

Description

Technical Field

[0001] The invention relates to the technical field of information management, and in particular to a system for analyzing the price fluctuation correlation of automobile parts. Background Art

[0002] As the number of cars in China continues to rise, the auto parts market is becoming increasingly large and complex. As an important channel for the sale of auto parts, the pricing strategy of repair shops directly affects the cost expenditure of consumers and insurance companies. Most of the existing auto parts price data collection systems and methods rely on historical price data and market conditions, lacking analysis and prediction of price change trends. At the same time, the mutual influence between repair shops is ignored, and there are technical problems such as low responsiveness to price fluctuations and affecting the transparency of parts selection. Summary of the invention

[0003] The present invention provides an automobile parts price fluctuation correlation analysis system to solve the technical problems in the prior art of low price fluctuation responsiveness and affecting the transparency of parts selection, and achieve the technical effects of improving price fluctuation responsiveness and enhancing parts price transparency.

[0004] The present invention provides an automobile parts price fluctuation correlation analysis system, comprising:

[0005] The data preparation module is used to collect and obtain a motor vehicle insurance damage assessment and claims data set, perform data cleaning and encryption preprocessing on the motor vehicle insurance damage assessment and claims data set, and obtain a motor vehicle insurance damage assessment and claims standard data set.

[0006] An entity extraction module is used to define entity category information, perform entity recognition on the standard claim data set for motor vehicle insurance damage assessment based on the entity category information, and obtain a repair shop feature entity set.

[0007] The attribute clustering module is used to extract attributes from the repair shop feature entity set, obtain a repair shop parts entity attribute set, and organize data dimensions based on the repair shop parts entity attribute set to generate a repair shop parts aggregate data set.

[0008] A relationship confirmation module is used to obtain a repair shop location relationship library based on the repair shop parts aggregation data set, extract relationships from the repair shop feature entity set based on the repair shop location relationship library, and obtain a repair shop entity location relationship set.

[0009] A graph construction module is used to construct a repair shop knowledge graph network based on the repair shop feature entity set, the repair shop parts entity attribute set and the repair shop entity position relationship set.

[0010] The fluctuation propagation module is used to use the water wave algorithm to perform price fluctuation correlation calculations on the repair shop knowledge graph network, obtain the results of the correlation impact of price fluctuations of repair shop accessories, and perform visual display and strategy-assisted decision-making based on the results of the correlation impact of price fluctuations of repair shop accessories.

[0011] In a feasible implementation, the steps of obtaining a standard claim data set for motor vehicle damage assessment include:

[0012] Build a data cleaning program that includes removing duplicate data, filling missing values, handling outliers, and formatting data.

[0013] The motor vehicle insurance damage assessment and claims data set is subjected to data cleaning preprocessing based on the data cleaning program to obtain a usable motor vehicle insurance damage assessment and claims data set.

[0014] The available motor vehicle insurance damage assessment and claim settlement data set is semantically disambiguated to obtain a consistent motor vehicle insurance damage assessment and claim settlement data set.

[0015] An encryption algorithm is configured to encrypt the consistent motor vehicle insurance damage assessment and claim data set to obtain the motor vehicle insurance damage assessment and claim data set.

[0016] In a feasible implementation, the steps of constructing a repair shop knowledge graph network include:

[0017] Graph node definition information and graph edge definition information are obtained, and based on the graph node definition information, the repair shop feature entity set is abstracted into an entity graph feature node set.

[0018] Based on the graph edge definition information, attribute definition is performed on the repair shop entity position relationship set to determine the entity graph feature edge set.

[0019] The entity graph feature node set is attributed through the repair shop parts entity attribute set to obtain an entity attribute feature node set.

[0020] The entity attribute feature node set is topologically connected based on the entity graph feature edge set to construct the repair shop knowledge graph network.

[0021] In a feasible implementation, the steps of obtaining the correlation effect result of the price fluctuation of the repair shop parts include:

[0022] The price fluctuation of each type of accessories is calculated based on the repair shop accessories aggregation data set to generate a set of price fluctuation factors of the accessory types of the target repair shop.

[0023] The price impact propagation calculation is performed on the repair shop knowledge graph network through the set of price fluctuation factors of the accessory type to generate a set of price fluctuation influence relationships of the repair shop.

[0024] Based on the repair shop price fluctuation influence relationship set, refraction and breaking wave influence analysis is performed to obtain the repair shop parts price fluctuation correlation influence result.

[0025] In a feasible implementation, the steps of generating a repair shop price fluctuation impact relationship set include:

[0026] The maximum extension length and the maximum extension number of nodes are set, and each price fluctuation factor in the set of price fluctuation factors of the accessory type is taken as a starting point. Based on the out-degree relationship in the repair shop knowledge graph network, the maximum extension length and the maximum extension number of nodes are expanded outward to construct a first price fluctuation propagation ripple set.

[0027] Based on the first price fluctuation propagation ripple set, similarity calculation and weighted summation of price fluctuation factors are performed to obtain the first layer output result.

[0028] Iteratively repeat expansion calculations are performed based on the first layer output results to obtain a multi-layer output calculation result set, and the first layer output results and the multi-layer output calculation result set are accumulated to obtain a final representation vector of the repair shop.

[0029] The repair shop fluctuation impact is reconstructed based on the final representation vector of the repair shop to generate the repair shop price fluctuation impact relationship set.

[0030] In a feasible implementation, the steps of obtaining the correlation effect result of the price fluctuation of the repair shop parts include:

[0031] Refraction influencing factors are obtained, where the refraction influencing factors include entity relationship weights, entity attributes, and entity criticality.

[0032] The repair shop characteristic entity set is evaluated and calculated based on the refraction influencing factors to obtain a repair shop entity refraction coefficient set.

[0033] A preset price breaking wave threshold is obtained, and a breaking wave effect is triggered when the preset price breaking wave threshold is exceeded, and price fluctuation influencing parameters are obtained based on the breaking wave effect analysis.

[0034] Based on the repair shop entity refraction coefficient set and the price fluctuation impact parameter, an impact update analysis is performed on the repair shop price fluctuation impact relationship set to obtain the repair shop parts price fluctuation associated impact result.

[0035] In a feasible implementation, the steps of performing visual display and strategy-assisted decision-making based on the correlation and impact results of the price fluctuation of the repair shop parts include:

[0036] Based on the repair shop parts price fluctuation correlation impact results, parameter extraction and analysis are performed to obtain a repair shop impact list, repair shop price change parameters, and repair shop impact degree indicators.

[0037] According to the data visualization measure set, the repair shop impact list, the repair shop price change parameters and the repair shop impact degree index are respectively visualized to obtain the visualization information of the price fluctuation of the repair shop parts.

[0038] Based on the visualized information of price fluctuations of repair shop parts, risk management and control analysis is performed to obtain parts price risk management and control strategy parameters, and risk auxiliary decision-making is performed using the parts price risk management and control strategy parameters.

[0039] In a feasible implementation, the steps of obtaining the accessories price risk management and control strategy parameters include:

[0040] A parts price risk control strategy library is constructed, and strategy matching is performed on the repair shop parts price fluctuation visualization information based on the parts price risk control strategy library to obtain a price risk control matching strategy.

[0041] Based on the visualized information on price fluctuations of repair shop parts, parameter analysis is performed on the price risk management and control matching strategy to determine a threshold for selecting strategy parameters.

[0042] A global evaluation is performed within the strategy parameter selection threshold to obtain the accessories price risk management and control strategy parameters.

[0043] The present invention discloses a system for analyzing the price fluctuation of automobile parts, including: a data preparation module collects and obtains a data set of automobile insurance damage assessment and claim settlement, performs data cleaning and encryption preprocessing on the data set of automobile insurance damage assessment and claim settlement, and obtains a standard data set of automobile insurance damage assessment and claim settlement; an entity extraction module defines entity category information, performs entity recognition on the standard data set of automobile insurance damage assessment and claim settlement based on the entity category information, and obtains a repair shop characteristic entity set; an attribute clustering module performs attribute extraction on the repair shop characteristic entity set, obtains a repair shop parts entity attribute set, and performs data dimension sorting based on the repair shop parts entity attribute set to generate a repair shop parts aggregated data set ; The relationship confirmation module obtains the repair shop location relationship library based on the repair shop parts aggregation data set, extracts relationships from the repair shop feature entity set based on the repair shop location relationship library, and obtains the repair shop entity location relationship set; the graph construction module constructs the repair shop knowledge graph network based on the repair shop feature entity set, the repair shop parts entity attribute set and the repair shop entity location relationship set; the fluctuation propagation module uses the water wave algorithm to calculate the price fluctuation correlation of the repair shop knowledge graph network, obtains the repair shop parts price fluctuation correlation impact results, and performs visualization and strategy-assisted decision-making based on the repair shop parts price fluctuation correlation impact results. The automobile parts price fluctuation correlation analysis system disclosed in the present invention solves the technical problems of low price fluctuation responsiveness and affecting the transparency of parts selection, and achieves the technical effects of improving price fluctuation responsiveness and enhancing parts price transparency. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a structural schematic diagram of a system for analyzing the price fluctuation correlation of automobile parts according to the present invention;

[0045] Figure 2 The present invention is a flowchart of obtaining a standard claim data set for automobile insurance damage assessment in an automobile parts price fluctuation correlation analysis system.

[0046] Explanation of the accompanying drawings: data preparation module 11, entity extraction module 12, attribute clustering module 13, relationship confirmation module 14, graph construction module 15, wave propagation module 16. DETAILED DESCRIPTION

[0047] The technical solution provided in the embodiments of the present invention is to solve the technical problems of low price fluctuation responsiveness and affecting the transparency of accessory selection in the prior art. The overall idea adopted is as follows:

[0048] First, the data preparation module collects and obtains the motor vehicle insurance damage assessment and claims data set, performs data cleaning and encryption preprocessing on the motor vehicle insurance damage assessment and claims data set, and obtains the motor vehicle insurance damage assessment and claims data set. Then, the entity extraction module defines the entity category information, performs entity recognition on the motor vehicle insurance damage assessment and claims data set based on the entity category information, and obtains the repair shop feature entity set. Then, the attribute clustering module extracts attributes from the repair shop feature entity set, obtains the repair shop accessories entity attribute set, and organizes the data dimension based on the repair shop accessories entity attribute set to generate the repair shop accessories aggregated data set. Next, the relationship confirmation module obtains the repair shop location relationship library based on the repair shop accessories aggregated data set, extracts the relationship of the repair shop feature entity set based on the repair shop location relationship library, and obtains the repair shop entity location relationship set. Furthermore, the graph construction module constructs the repair shop knowledge graph network based on the repair shop feature entity set, the repair shop accessories entity attribute set, and the repair shop entity location relationship set. Finally, the wave propagation module uses the water wave algorithm to calculate the price fluctuation correlation of the repair shop knowledge graph network, obtains the correlation impact results of the price fluctuation of repair shop accessories, and performs visualization and strategy-assisted decision-making based on the correlation impact results of the price fluctuation of repair shop accessories.

[0049] The above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods of the specification to better understand the above technical solution. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments used only to explain the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, it should be noted that, for the convenience of description, only the parts related to the present invention are shown in the drawings, rather than all of them. Example

[0050] Figure 1 The present invention is a schematic diagram of the structure of an automobile parts price fluctuation correlation analysis system, wherein the system comprises:

[0051] The data preparation module 11 is used to collect and obtain a motor vehicle insurance damage assessment and claim data set, perform data cleaning and encryption preprocessing on the motor vehicle insurance damage assessment and claim data set, and obtain a motor vehicle insurance damage assessment and claim standard data set.

[0052] Specifically, first, obtain access permissions and data extraction interfaces, interact with motor insurance damage assessment and claims data sources through API or batch export functions, and extract motor insurance damage assessment and claims data sets. Motor insurance damage assessment and claims data sources include motor insurance claims management systems, third-party accident reporting and assessment platforms, and traffic accident databases.

[0053] Specifically, the data in the motor vehicle insurance damage assessment and claims data source includes policy information, insurance company information, policy information, basic accident information, accident participant information, accident description, vehicle loss information, personal injury information, claims processing information, etc.

[0054] Specifically, the motor vehicle insurance damage assessment and claim data set is part of the data in the above-mentioned motor vehicle insurance damage assessment and claim data source, including vehicle information (model, type, frame number, etc.), vehicle loss information (damaged parts name, part number, cost, quantity, etc.), maintenance unit information (maintenance plant name, contact information), etc.

[0055] In some embodiments, Figure 2 As shown, the steps of obtaining the standard claim data set for motor vehicle damage assessment include:

[0056] Construct a data cleaning program, which includes removing duplicate data, filling missing values, processing outliers, and formatting data. Perform data cleaning preprocessing on the motor vehicle insurance damage assessment and claim data set based on the data cleaning program to obtain an available motor vehicle insurance damage assessment and claim data set. Perform semantic disambiguation processing on the available motor vehicle insurance damage assessment and claim data set to obtain a consistent motor vehicle insurance damage assessment and claim data set. Configure an encryption algorithm to encrypt the consistent motor vehicle insurance damage assessment and claim data set to obtain the motor vehicle insurance damage assessment and claim standard data set.

[0057] Specifically, a data cleaning program is constructed, in which duplicate data is removed by identifying unique fields (such as policy numbers, accident numbers, etc.), detecting and deleting duplicate data records. Filling missing values ​​is used to process missing data using methods such as mean filling, interpolation, and deleting records with missing values. Statistical analysis methods (such as box plots) can be used to identify outliers, and outliers in the data can be processed according to business rules. Formatting data is used to unify the format of data, including converting data to the same unit of measurement, etc., to ensure consistency in data format. Among them, the baseline format of formatted data is determined based on the analysis requirements of the analysis scenario to ensure that subsequent analysis and extraction can understand the content of the data set.

[0058] Furthermore, based on the data cleaning program constructed above, the auto insurance damage assessment and claims dataset is preprocessed to remove duplicate data, fill missing values, process outliers and format data. A dataset with high data quality and consistent format is obtained and stored as an available auto insurance damage assessment and claims dataset.

[0059] Specifically, semantic disambiguation is used to solve the problem that the same word has different meanings in different contexts. Specifically, it includes the problem of homonymy where the same word has different meanings in different contexts. For example, "amount" may refer to different types of amounts in different business scenarios, which can be distinguished through context analysis. The problem of synonyms where different words express the same meaning, for example, "payment amount" and "claim amount" may represent the same concept, which can be unified through semantic analysis. Exemplarily, natural language processing technology (such as word embedding models) is used to perform semantic analysis and comparison on data from different sources to ensure the consistency of data meaning.

[0060] Specifically, first, select a suitable encryption algorithm, such as AES (Advanced Encryption Standard) or RSA (asymmetric encryption algorithm). Then, encrypt the consistent auto insurance loss assessment and claims data set to protect data privacy and security and prevent data theft and tampering. The encrypted consistent auto insurance loss assessment and claims data set is the auto insurance loss assessment standard claims data set.

[0061] The entity extraction module 12 is used to define entity category information, perform entity recognition on the motor vehicle insurance damage assessment standard claim data set based on the entity category information, and obtain a repair shop feature entity set.

[0062] Specifically, an entity refers to a node in a knowledge graph, and entity category information refers to classification information of vehicle accessories. Exemplarily, the vehicle accessories categories in the entity category information include: power accessories, transmission accessories, suspension accessories, brake accessories, electrical accessories, body accessories, air conditioning and refrigeration accessories, emission control accessories, interior accessories, etc.

[0063] Specifically, based on the defined entity category information, entity recognition is performed on the standard claim data set for automobile insurance damage assessment, which means that repair shops are divided into multiple groups of repair shops corresponding to multiple entity categories according to the included entity categories, and multiple groups of repair shops constitute a repair shop feature entity set. Among them, repair shops in the same group have usage records of automobile accessories of the entity category.

[0064] Specifically, entity recognition is performed based on entity category information. First, the standard claims data set for auto insurance damage assessment is parsed to identify the vehicle accessories and prices, repair shops, and vehicle models involved in the data to obtain multiple entities. Then, according to the categories of vehicle accessories, the repair shop entities in the claims data are divided into multiple groups to form multiple entity category groups. The same repair shop can appear repeatedly in different entity category groups. For example, if repair shop A uses power accessories and suspension accessories in its records, then repair shop A belongs to both the power accessories group and the suspension accessories group. Finally, the repair shops in each entity category group are marked with the entity category as the tag content, and the marked multiple entity category groups form a repair shop feature entity set.

[0065] Through the above steps, the standard claims data set for motor vehicle damage assessment is systematically classified and identified, generating a structured repair shop feature entity set. This not only improves the organization and usability of the data, but also provides important support for subsequent data analysis and decision-making.

[0066] The attribute clustering module 13 is used to extract attributes from the repair shop feature entity set to obtain a repair shop parts entity attribute set, and to organize data dimensions based on the repair shop parts entity attribute set to generate a repair shop parts aggregated data set.

[0067] Specifically, the repair shop feature entity set is traversed, and all entity categories marked by multiple repair shops are counted respectively, and stored as the repair shop parts entity attribute set. The repair shop parts entity attribute set defines the categories of auto parts processed by each group of multiple repair shops. In other words, the parts entity attribute set of each repair shop records all the parts categories involved and their corresponding attributes. The repair shop parts entity attribute set helps to understand which auto parts price fluctuations affect different repair shops.

[0068] Specifically, the entity category information is used as multiple data dimensions to classify and organize the repair shop parts entity attribute set. For example, corresponding repair shop index lists or forms are established for power parts, transmission parts, suspension parts, etc. Then, the organized repair shop parts entity attribute set is aggregated to generate a repair shop parts aggregate data set. This data set contains the parts usage and related attributes of all repair shops, which is helpful for the two-way positioning between the parts category and the repair shop.

[0069] The relationship confirmation module 14 is used to obtain a repair shop location relationship library based on the repair shop parts aggregation data set, extract relationships from the repair shop feature entity set based on the repair shop location relationship library, and obtain a repair shop entity location relationship set.

[0070] Specifically, based on the aggregated data set of repair shop parts, the Geographic Information System (GIS) technology is used to convert the addresses into geographic coordinates (latitude and longitude) through geocoding, extract the geographic location information of multiple repair shops, and establish the geographic location of each repair shop and their relationship with each other.

[0071] Specifically, the distances between different repair shop locations are used to determine the relationships between the entities in the repair shop feature entity set, and to determine the repair shop entity location relationship set. The distance between the repair shop locations includes a straight-line distance or a path distance. Exemplarily, the straight-line distance is determined using the Haversine formula, which is a straight-line distance between two points calculated using the longitude and latitude of the repair shop. The path distance uses road network data and a path planning algorithm (such as the Dijkstra algorithm and the A* algorithm) to calculate the actual path distance between the repair shops.

[0072] Specifically, the relationship extraction results are stored in a database to form a repair shop entity location relationship set. Exemplarily, these relationship data are stored and queried through a graph database (such as Neo4j).

[0073] Furthermore, the entity relationships in the repair shop entity position relationship set are determined according to the entity category to which the repair shop belongs, that is, the entity relationships in the repair shop entity position relationship set have corresponding entity category tags, which are used to determine the relationship between repair shops under different parts categories. In other words, two repair shops that have an entity relationship in the first category may not have an entity relationship in the second category.

[0074] The above steps form a repair shop entity position relationship set, record the spatial relationship and characteristics between the repair shop entities, and provide comprehensive data support for subsequent analysis and decision-making.

[0075] The graph construction module 15 is used to construct a repair shop knowledge graph network based on the repair shop feature entity set, the repair shop parts entity attribute set and the repair shop entity position relationship set.

[0076] In some embodiments, the steps of constructing a repair shop knowledge graph network include:

[0077] Obtain graph node definition information and graph edge definition information, and abstract the repair shop feature entity set into an entity graph feature node set based on the graph node definition information. Define the attributes of the repair shop entity position relationship set based on the graph edge definition information, and determine the entity graph feature edge set. Assign attributes to the entity graph feature node set through the repair shop accessory entity attribute set to obtain an entity attribute feature node set. Topologically connect the entity attribute feature node set based on the entity graph feature edge set to construct the repair shop knowledge graph network.

[0078] Specifically, first, define the types and attributes of nodes and edges in the knowledge graph, including graph node definition information and graph edge definition information. The graph node definition information includes the information of each entity node in the repair shop feature entity set, such as the repair shop, accessories, service type and other node types and their attributes (such as name, location, category, etc.). The graph edge definition information includes defining the relationship between nodes, such as distance, cooperation relationship, supply relationship, etc.

[0079] Specifically, according to the graph node definition information, a node set is created using a programming language (such as Python) and a graph database (such as Neo4j), and each entity in the repair shop feature entity set is abstracted as a node in the knowledge graph.

[0080] Specifically, according to the graph edge definition information, the attributes of each relationship in the repair shop entity location relationship set are defined, the distance between the repair shops, the cooperative relationship, etc. are abstracted as graph edges, and each edge is assigned corresponding attributes.

[0081] Specifically, the attributes in the repair shop parts entity attribute set are assigned to the graph nodes to form a node set with attributes. For example, the node attributes are updated using the graph database API or query language (such as Cypher). The attributes of the repair shop, parts, etc. (such as name, category, value) are assigned to the corresponding graph nodes.

[0082] Furthermore, based on the relationships in the entity graph feature edge set, the entity attribute feature node set is topologically connected to construct a repair shop knowledge graph network. The above steps systematically represent the repair shop and its related entities and relationships by constructing a repair shop knowledge graph network, thereby improving the structuring and visualization of the data. By managing nodes and edges through a graph database, complex relationships between repair shops can be efficiently stored and queried, supporting complex relationship queries and analysis.

[0083] The fluctuation propagation module 16 is used to use the water wave algorithm to perform price fluctuation correlation calculation on the repair shop knowledge graph network, obtain the price fluctuation correlation impact results of the repair shop accessories, and perform visualization and strategy-assisted decision-making based on the price fluctuation correlation impact results of the repair shop accessories.

[0084] In some embodiments, the steps of obtaining the associated impact result of the price fluctuation of the repair shop parts include:

[0085] Based on the repair shop parts aggregation data set, price fluctuation calculations are performed for each type of parts to generate a set of price fluctuation factors for the parts type of the target repair shop. Price impact propagation calculations are performed on the repair shop knowledge graph network through the set of price fluctuation factors for the parts type to generate a set of price fluctuation influence relationships for the repair shop. Refraction and breaking wave impact analysis is performed based on the set of price fluctuation influence relationships for the repair shop to obtain the associated impact results of the price fluctuation of the parts for the repair shop.

[0086] Specifically, first, based on the repair shop parts aggregation data set, typical price values ​​of multiple types of auto parts are obtained, such as average, weighted average, median price, etc. Then, the deviation degree between the price of each type of auto parts of multiple repair shops corresponding to multiple nodes and the typical price value is calculated respectively, and multiple price fluctuation factors are obtained, which are then stored as a set of price fluctuation factors of accessory types.

[0087] In some implementations, the step of generating a repair shop price fluctuation impact relationship set includes:

[0088] Set the maximum extension length and the maximum extension number of nodes, take each price fluctuation factor in the set of price fluctuation factors of the accessory type as the starting point, expand outward according to the maximum extension length and the maximum extension number of nodes based on the out-degree relationship in the repair shop knowledge graph network, and construct a first price fluctuation propagation ripple set. Perform similarity calculation and weighted summation of price fluctuation factors based on the first price fluctuation propagation ripple set to obtain the first layer output result. Perform iterative repeated extension calculation based on the first layer output result to obtain a multi-layer output calculation result set, accumulate the first layer output result and the multi-layer output calculation result set to obtain the final representation vector of the repair shop. Reconstruct the repair shop fluctuation influence based on the final representation vector of the repair shop to generate the repair shop price fluctuation influence relationship set.

[0089] Specifically, first, define the propagation rules of the water wave algorithm in the knowledge graph, including propagation range, attenuation coefficient, etc. Among them, the propagation range is used to define the maximum range of water wave propagation (such as how many layers of nodes it propagates to). The attenuation coefficient is used to define the attenuation coefficient of price fluctuations during water wave propagation, simulating the weakening of the impact of price fluctuations.

[0090] Specifically, the propagation range is defined as the maximum extension length and the maximum extension number of nodes, where the maximum extension length refers to the maximum number of layers of water wave propagation (such as 3 layers), and the maximum extension number of nodes refers to the maximum number of nodes that each node can affect in a single water wave propagation.

[0091] Specifically, each price fluctuation factor in the set of price fluctuation factors of accessory types is selected as the starting point, and the out-degree relationship in the repair shop knowledge graph network is traversed using a programming language (such as Python), and expanded outward according to the set maximum expansion length and the maximum number of node expansions. The nodes and their price fluctuation factors are recorded in each expansion to form the first price fluctuation propagation ripple set.

[0092] Specifically, similarity calculation methods (such as cosine similarity, Euclidean distance, etc.) are used to calculate the similarity of price fluctuation factors, and the price fluctuation factors are weighted and summed according to the similarity to form the first-layer output results. Then, the first-layer output results are used as a new starting point for iterative expansion calculations, and the output results of each iteration are recorded to form a multi-layer output calculation result set.

[0093] Furthermore, the first layer output results and the multi-layer output calculation result sets are accumulated to obtain the final representation vector of the repair shop. That is to say, each repair shop is affected by the price fluctuations of multiple related repair shops, and the changes and fluctuations in the prices of repair shop parts are caused by the joint action of the multi-layer output calculation results. Among them, the length of the repair shop's final representation vector represents the magnitude of the price fluctuation of the parts, and the direction of the repair shop's final representation vector represents the direction of the price fluctuation of the parts.

[0094] In some implementations, the step of obtaining the result of the price fluctuation correlation impact of the repair shop parts includes:

[0095] Obtain refraction influencing factors, which include entity relationship weights, entity attributes, and entity criticality. Evaluate and calculate the repair shop feature entity set based on the refraction influencing factors to obtain a repair shop entity refraction coefficient set. Obtain a preset price breaking wave threshold, trigger a breaking wave effect when the preset price breaking wave threshold is exceeded, and obtain price fluctuation influencing parameters based on the breaking wave effect analysis. Based on the repair shop entity refraction coefficient set and the price fluctuation influencing parameters, perform an impact update analysis on the repair shop price fluctuation influencing relationship set to obtain the repair shop parts price fluctuation associated impact results.

[0096] Specifically, first, the nodes are classified and labeled using the attribute data of the repair shop (such as scale, geographical location, etc.). Specific repair shop nodes that may cause price fluctuation refraction are identified, such as large chain repair shops or central repair shops in a specific area. Then, based on the type and attributes of the repair shop, a refraction coefficient is defined for each specific repair shop node to represent the change in the direction and intensity of the price fluctuation propagation at the node. For example, the refraction coefficient of a large chain repair shop may be higher, while the refraction coefficient of a small repair shop may be lower.

[0097] Furthermore, during the propagation of price fluctuations, when the fluctuations affect a specific repair shop node, the refraction coefficient is applied to adjust the propagation path of the price fluctuations so that it is refracted at the specific node, and the propagation intensity of the price fluctuations is adjusted according to the refraction coefficient to simulate the refraction phenomenon in the actual market.

[0098] Specifically, the preset price breaking wave threshold is set according to historical data and market analysis. For example, when the price fluctuation of a certain accessory exceeds 10%, the breaking wave effect is triggered. Among them, the preset price breaking wave threshold is adjusted dynamically according to market changes and demand. Then, the intensity of price fluctuations is monitored in real time to determine whether it exceeds the preset price breaking wave threshold. When the intensity of price fluctuations exceeds the threshold, the breaking wave effect is triggered, and the trigger point and related data are recorded. According to the trigger point and intensity of the breaking wave effect, the influencing parameters of price fluctuations, such as fluctuation amplitude, propagation speed, and fluctuation acceleration, are calculated.

[0099] Furthermore, based on the repair shop entity refraction coefficient set and price fluctuation impact parameters, the impact update analysis of the repair shop price fluctuation impact relationship set is carried out, including changing the direction and intensity of price fluctuation propagation, expanding the price fluctuation impact relationship (such as price surges, so that originally unrelated nodes establish new associations with a certain type of accessories, and the original weak association is transformed and enhanced into a strong association, etc.), adjusting the maximum extension length and the maximum number of node extensions, etc.

[0100] In some embodiments, the steps of performing visual display and strategy-assisted decision-making based on the results of the price fluctuation correlation impact of the repair shop parts include:

[0101] Based on the results of the correlation impact of the price fluctuation of the repair shop parts, parameter extraction and analysis are performed to obtain a repair shop impact list, a repair shop price change parameter, and a repair shop impact degree index. According to the data visualization measure set, the repair shop impact list, the repair shop price change parameter, and the repair shop impact degree index are respectively visualized to obtain visualization information of the price fluctuation of the repair shop parts. Based on the visualization information of the price fluctuation of the repair shop parts, risk control analysis is performed to obtain the risk control strategy parameters of the parts price, and risk auxiliary decision-making is performed through the risk control strategy parameters of the parts price.

[0102] Specifically, based on the results of the impact of price fluctuations on repair shop parts, the affected repair shops are analyzed and obtained, and the corresponding price change parameters and impact indexes are extracted. For example, the price change parameters include the absolute value of price change and the relative value of price change, and the impact index includes the proportion of affected parts categories (the proportion of different parts categories affected by price fluctuations to the total parts categories), the overall cost fluctuation ratio (calculating the overall cost change ratio caused by price fluctuations), etc.

[0103] Specifically, data visualization tools (such as Tableau, Power BI, D3.js, etc.) are used to visualize the repair shop impact list, repair shop price change parameters, and repair shop impact degree indicators according to the data visualization measure set, so as to obtain visualization information on price fluctuations of repair shop parts. The data visualization measure set includes optional data visualization paths, including visualization on the user side, visualization on the business side, and visualization on the information management platform side.

[0104] Optionally, select appropriate chart types according to different data types. For example, the impact list uses a network diagram or relationship diagram to show the impact relationship between repair shops. The price change parameter uses a line chart, bar chart, or heat map to show price fluctuations. The impact degree indicator uses a radar chart, bar chart, etc. to show the degree of impact on the repair shop in different dimensions.

[0105] Optionally, potential risk points in price fluctuations can be identified through visual information, such as accessories with drastic price fluctuations and repair shops with greater impact. Then, based on the identified risk points, corresponding risk management strategies can be formulated, such as inventory adjustment, supply chain optimization, price adjustment strategy, etc. Among them, risk management strategy parameters include the number of response measures, execution time, expected effect, etc.

[0106] In some implementations, the step of obtaining the accessory price risk management and control strategy parameters includes:

[0107] Construct a parts price risk control strategy library, and perform strategy matching on the visualization information of the price fluctuation of parts in the repair shop based on the parts price risk control strategy library to obtain a price risk control matching strategy. Perform parameter analysis on the price risk control matching strategy based on the visualization information of the price fluctuation of parts in the repair shop to determine a strategy parameter selection threshold. Perform global evaluation and optimization within the strategy parameter selection threshold to obtain the parts price risk control strategy parameters.

[0108] Specifically, a parts price risk management strategy library is constructed. The parts price risk management strategy library is a structured strategy library that includes multiple risk management strategies, wherein each strategy is associated with a corresponding application feature indicator.

[0109] Specifically, based on the visualization information of price fluctuations of parts in repair shops, price fluctuation characteristic indicators are extracted, including, for example, the amplitude of price fluctuations, fluctuation rate, information entropy of fluctuation distribution, etc. Then, based on the price fluctuation characteristic indicators and application characteristic indicators, traversal matching is performed to obtain the price risk control matching strategy.

[0110] Furthermore, based on the degree and distribution of price fluctuations, the matching strategy is analyzed by parameters, including fluctuation degree analysis: analyzing the degree of fluctuation according to the absolute and relative values ​​of price fluctuations. Fluctuation distribution analysis: analyzing the distribution of price fluctuations and calculating the uniformity of fluctuations. Then, based on the parameter analysis results, the fluctuation level of the current repair shop parts price fluctuation visualization information is defined, where the higher the fluctuation level, the higher the price fluctuation risk.

[0111] Specifically, the threshold for selecting strategy parameters is determined based on the degree and distribution of fluctuations. The more severe the fluctuations and the more uneven the distribution, the higher the level of risk management strategy parameters required. For example, this includes selecting a stricter premium penalty amount or providing a higher number of accessories.

[0112] In summary, the automobile parts price fluctuation correlation analysis system provided by the present invention has the following technical effects:

[0113] The data preparation module collects and obtains the data set of motor insurance damage assessment and claims, performs data cleaning and encryption preprocessing on the data set of motor insurance damage assessment and claims, and obtains the standard data set of motor insurance damage assessment and claims. The entity extraction module defines entity category information, performs entity recognition on the standard data set of motor insurance damage assessment and claims based on the entity category information, and obtains the characteristic entity set of the repair shop. The attribute clustering module extracts attributes from the characteristic entity set of the repair shop, obtains the entity attribute set of the repair shop accessories, and organizes the data dimension based on the entity attribute set of the repair shop accessories to generate the aggregated data set of the repair shop accessories. The relationship confirmation module obtains the location relationship library of the repair shop based on the aggregated data set of the repair shop accessories, extracts the relationship of the characteristic entity set of the repair shop based on the location relationship library of the repair shop, and obtains the entity location relationship set of the repair shop. The graph construction module constructs the knowledge graph network of the repair shop based on the characteristic entity set of the repair shop, the entity attribute set of the repair shop accessories, and the entity location relationship set of the repair shop. The wave propagation module uses the water wave algorithm to calculate the price fluctuation correlation of the knowledge graph network of the repair shop, obtains the price fluctuation correlation impact results of the repair shop accessories, and performs visual display and strategy-assisted decision-making based on the price fluctuation correlation impact results of the repair shop accessories. This will achieve the technical effect of improving the ability to respond to price fluctuations and increasing the transparency of accessory prices.

[0114] It should be understood that the embodiments disclosed in the present invention and the above description can enable those skilled in the art to use the present invention to implement the present invention. At the same time, the present invention is not limited to the above-mentioned embodiments. It should be understood that those skilled in the art can still modify the technical solutions recorded in the above-mentioned embodiments, or replace some of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A system for analyzing the price fluctuation of automobile parts, characterized in that: The system comprises: A data preparation module, the data preparation module is used to collect and obtain a motor vehicle insurance damage assessment and claim data set, perform data cleaning and encryption preprocessing on the motor vehicle insurance damage assessment and claim data set, and obtain a motor vehicle insurance damage assessment and claim standard data set; An entity extraction module, the entity extraction module is used to define entity category information, perform entity recognition on the standard claim data set for motor vehicle insurance damage assessment based on the entity category information, and obtain a repair shop feature entity set; An attribute clustering module, the attribute clustering module is used to extract attributes from the repair shop feature entity set, obtain a repair shop parts entity attribute set, and perform data dimension sorting based on the repair shop parts entity attribute set to generate a repair shop parts aggregated data set; A relationship confirmation module, the relationship confirmation module is used to obtain a repair shop location relationship library based on the repair shop parts aggregation data set, and extract relationships from the repair shop feature entity set based on the repair shop location relationship library to obtain a repair shop entity location relationship set; A graph construction module, the graph construction module is used to construct a repair shop knowledge graph network based on the repair shop feature entity set, the repair shop parts entity attribute set and the repair shop entity position relationship set; A fluctuation propagation module, the fluctuation propagation module is used to perform price fluctuation correlation calculation on the repair shop knowledge graph network using a water wave algorithm, obtain the correlation impact results of the price fluctuation of the repair shop parts, and perform visual display and strategy auxiliary decision-making based on the correlation impact results of the price fluctuation of the repair shop parts; The steps of constructing a repair shop knowledge graph network include: Obtaining graph node definition information and graph edge definition information, and abstracting the repair shop feature entity set into an entity graph feature node set based on the graph node definition information; Based on the graph edge definition information, attribute definition is performed on the repair shop entity position relationship set to determine the entity graph feature edge set; Assigning attributes to the entity graph feature node set through the repair shop parts entity attribute set to obtain an entity attribute feature node set; The entity attribute feature node set is topologically connected based on the entity graph feature edge set to construct the repair shop knowledge graph network.

2. The system according to claim 1, characterized in that The steps of obtaining the standard claim data set for motor vehicle damage assessment include: Building a data cleaning program, which includes removing duplicate data, filling missing values, handling outliers, and formatting data; Performing data cleaning preprocessing on the motor vehicle insurance damage assessment and claims data set based on the data cleaning program to obtain an available motor vehicle insurance damage assessment and claims data set; Performing semantic disambiguation processing on the available motor vehicle insurance damage assessment and claims data set to obtain a consistent motor vehicle insurance damage assessment and claims data set; An encryption algorithm is configured to encrypt the consistent motor vehicle insurance damage assessment and claim data set to obtain the motor vehicle insurance damage assessment and claim data set.

3. The system according to claim 1, characterized in that The steps of obtaining the correlation and impact results of the price fluctuation of the repair shop parts include: Calculate the price fluctuation of each type of parts based on the repair shop parts aggregation data set to generate a set of price fluctuation factors of the parts types of the target repair shop; Performing price impact propagation calculation on the repair shop knowledge graph network through the set of price fluctuation factors of the accessory type to generate a set of price fluctuation influence relationships of the repair shop; Based on the repair shop price fluctuation influence relationship set, refraction and breaking wave influence analysis is performed to obtain the repair shop parts price fluctuation correlation influence result.

4. The system according to claim 3, characterized in that The steps of generating a repair shop price fluctuation influence relationship set include: Set a maximum extension length and a maximum number of node extensions, take each price fluctuation factor in the accessory type price fluctuation factor set as a starting point, expand outward according to the maximum extension length and the maximum number of node extensions based on the out-degree relationship in the repair shop knowledge graph network, and construct a first price fluctuation propagation ripple set; Based on the first price fluctuation propagation ripple set, similarity calculation and weighted summation of price fluctuation factors are performed to obtain a first-layer output result; Iteratively repeat expansion calculation based on the first layer output result to obtain a multi-layer output calculation result set, and accumulate the first layer output result and the multi-layer output calculation result set to obtain a final representation vector of the repair shop; The repair shop fluctuation impact is reconstructed based on the final representation vector of the repair shop to generate the repair shop price fluctuation impact relationship set.

5. The system according to claim 3, characterized in that The steps of obtaining the correlation and impact result of the price fluctuation of the repair shop parts include: Obtaining refraction influencing factors, wherein the refraction influencing factors include entity relationship weight, entity attribute, and entity criticality; Based on the refraction influencing factors, the repair shop characteristic entity set is evaluated and calculated to obtain a repair shop entity refraction coefficient set, where the repair shop entity refraction coefficient is used to represent the change in the propagation direction and intensity of price fluctuations at the repair shop node; Obtaining a preset price breaking wave threshold, triggering a breaking wave effect when the price fluctuation intensity exceeds the preset price breaking wave threshold, and obtaining a price fluctuation influence parameter based on the breaking wave effect analysis; Based on the repair shop entity refraction coefficient set and the price fluctuation impact parameter, an impact update analysis is performed on the repair shop price fluctuation impact relationship set to obtain the repair shop parts price fluctuation associated impact result.

6. The system according to claim 5, characterized in that The steps of performing visual display and strategy-assisted decision-making based on the correlation and impact results of the price fluctuation of the repair shop parts include: Perform parameter extraction and analysis based on the repair shop parts price fluctuation correlation impact results to obtain a repair shop impact list, repair shop price change parameters, and repair shop impact degree indicators; According to the data visualization measure set, the repair shop impact list, the repair shop price change parameters and the repair shop impact degree index are respectively visualized to obtain the visualization information of the price fluctuation of the repair shop parts; Based on the visualized information of price fluctuations of repair shop parts, risk management and control analysis is performed to obtain parts price risk management and control strategy parameters, and risk auxiliary decision-making is performed using the parts price risk management and control strategy parameters.

7. The system according to claim 6, characterized in that The steps of obtaining the accessories price risk control strategy parameters include: Constructing a parts price risk control strategy library, and performing strategy matching on the repair shop parts price fluctuation visualization information based on the parts price risk control strategy library to obtain a price risk control matching strategy; Perform parameter analysis on the price risk control matching strategy based on the visualized information of price fluctuations of the repair shop parts to determine the threshold for selecting strategy parameters; A global evaluation is performed within the strategy parameter selection threshold to obtain the accessory price risk management and control strategy parameters.

Images