Construction and evaluation method of vehicle competition network based on multi-source heterogeneous data, electronic equipment and medium

Abstract

The invention relates to the technical field of electric data processing, in particular to a method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data, electronic equipment and a medium. The method comprises the following steps: according to user behavior class data, obtaining vehicle type pairs with a competition relationship and a competition degree, and constructing a competition network; applying a community detection algorithm on the competitive network to obtain sub-competitive networks belonging to the same subdivision field; extracting a key topological attribute index for each sub-competition network; calculating a configuration competition index according to the vehicle type attribute and the configuration data; obtaining a price elasticity index and market sales expression according to the market expression data; obtaining a user comprehensive score according to the user behavior data; and according to the key topological attribute index, the configuration competition index, the price elasticity index, the market sales volume performance and the user comprehensive score, evaluating the vehicle type competition relationship. According to the invention, a competition relationship network between vehicle types can be constructed, and the competition relationship can be evaluated.

Description

Technical Field

[0001] The present application relates to the technical field of electrical data processing, and more particularly, to a method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data, an electronic device, and a medium. Background Art

[0002] In the automotive industry, the competition relationship between similar models usually unfolds based on multiple dimensions such as vehicle performance, price, brand value, and market positioning.

[0003] For electric vehicles, battery technology, driving range, and charging speed are key competitive points. The fuel consumption and emission levels of fuel vehicles are also important aspects of competition. Data provided by industry reports and market research institutions can help understand the market positioning and competition pattern of different models.

[0004] However, the existing competition relationships between models are mostly artificially defined, lacking authority and scientific nature, and there is no scientific and reasonable evaluation of the competition relationships between models. In view of this, the present application is proposed. Summary of the Invention

[0005] The purpose of the present application is to provide a method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data, an electronic device, and a medium, so as to construct a competition relationship network between models and evaluate the competition relationship.

[0006] To achieve the above purpose, the present application adopts the following technical solutions: In a first aspect, the present application provides a method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data, including: Collect user behavior data, vehicle model attributes and configuration data, and market performance data of multiple vehicle models; Based on the user behavior data, obtain vehicle model pairs with competition relationships and the competition degree of the vehicle model pairs; use the vehicle model pairs as two nodes and the competition degree as the weight of the edge to construct a competition network; Apply a community detection algorithm to the competition network to obtain sub-competition networks belonging to the same sub-field; the sub-competition networks include multiple vehicle model pairs with competition relationships; Extract key topological attribute indicators for each sub-competition network, and the key topological attribute indicators include degree centrality, weighted degree centrality, and betweenness centrality; Calculate a configuration competition index based on the vehicle model attributes and configuration data; obtain a price elasticity index and a market sales performance based on the market performance data; obtain a comprehensive user score based on the user behavior data; Evaluate the competitive relationship between vehicle models based on the key topological attribute indicators, configured competition index, price elasticity index, market sales performance, and comprehensive user ratings.

[0007] In a second aspect, the present application provides an electronic device, including: at least one processor, and 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 to enable the at least one processor to execute the above-mentioned method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data.

[0008] In a third aspect, the present application provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to cause a computer to execute the above-mentioned method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data.

[0009] Compared with the prior art, the beneficial effects of the present application are: By integrating data such as user evaluation texts, product configuration parameters, market price changes, and sales volume, the present application constructs a competitive network with visualization, community segmentation, and dynamics, and realizes the quantitative evaluation and trend analysis of the competitive relationship of automotive products in different market environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0011] Figure 1 is a flowchart of the method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data provided by the present application; Figure 2 is a diagram showing the number of vehicle models in different communities provided by an embodiment of the present application; Figure 3 is a schematic diagram of the competition network of different vehicle levels provided by an embodiment of the present application; Figure 4 is a heat map of the multi-dimensional competitive relationship provided by an embodiment of the present application; Figure 5 is a schematic diagram of the structure of the electronic device provided by the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0012] The following describes exemplary embodiments of the present application in conjunction with the accompanying drawings. Various details of the embodiments of the present application are included to assist understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, descriptions of well-known functions and structures are omitted in the following description for clarity and conciseness.

[0013] Figure 1 is a flowchart of a method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data provided by this embodiment. This method can be executed by an electronic device, and the electronic device can be integrated in the electronic device.

[0014] See Figure 1 , the method provided by this embodiment includes the following operations: S110. Collect user behavior data, vehicle model attributes and configuration data, and market performance data for multiple vehicle models.

[0015] The user behavior data mainly comes from user word-of-mouth information in automotive vertical media platforms, including user identity, purchase time, scoring results (such as scores on forums, comprehensive evaluations, etc.), user comment texts, etc. The vehicle model attributes and configuration data include: brand, energy type (fuel, pure electric, hybrid, etc.), vehicle model level (such as compact car, mid-size SUV, etc.), power system and other configuration parameters. The market performance data includes: market guiding price, actual transaction price, monthly sales volume, etc., which are used to assist in analyzing the market acceptance and competition performance of vehicle models.

[0016] Optionally, standardize the vehicle model names. Specifically, construct a vehicle model variant dictionary (including multiple variant names and standard names of vehicle models), standardize the variant names that appear in the user comment texts, merge synonymous variants, eliminate ambiguities, and improve the accuracy of subsequent vehicle model competitor extraction.

[0017] S120. Obtain vehicle model pairs with competitive relationships and the degree of competition of the vehicle model pairs according to the user behavior data; use the vehicle model pairs as two nodes and the degree of competition as the weight of the edge to construct a competition network.

[0018] S120 includes the following operations: S121. Process the user comment texts to obtain at least two vehicle models that appear in the same comment as potentially competitive vehicle models.

[0019] Using the dictionary matching method, batch identify the model of this vehicle and competing models in a user's review text to obtain the "this vehicle model - competing vehicle model" combination; at the same time, associate and bind the vehicle attributes, configuration data, and market performance data of the corresponding models collected in S110.

[0020] S122. Among the models that potentially have a competitive relationship, filter according to the price difference and vehicle class to obtain pairs of models with a competitive relationship.

[0021] Since the "this vehicle model - competing vehicle model" combinations that appear simultaneously in a user's review text do not necessarily have a direct competitive relationship, if the price difference between the two models is too large or the vehicle classes are different, it is considered that there is no direct competitive relationship.

[0022] Based on this, introduce a dual filtering mechanism of price difference and class. For all models that potentially have a competitive relationship, calculate the mean (μ) and standard deviation σ of the price difference distribution, and set an interval (μ ± 3σ) to determine whether the price difference is an outlier. If a certain price difference falls outside this interval, it is considered an outlier and the model pair is excluded. Vehicle class filtering is used to exclude model pairs that do not have a direct market competitive relationship. For example, a minicar and an MPV (Multi-Purpose Vehicle) have different vehicle classes and do not have a direct competitive relationship. The remaining model pairs after filtering perform the following operations.

[0023] S123. Construct a structured relationship data table with the model pairs having a competitive relationship and the co-occurrence frequency of the model pairs in the user's review text, providing a basis for constructing a weighted network in the future.

[0024] Among them, the structured relationship data table records: the model of this vehicle and competing models, and the co-occurrence frequency in the user's review text (for example, among a total of 1000 user review texts, the model of this vehicle and competing models appear simultaneously in 10 texts, then the co-occurrence frequency is 10). The structured relationship data table also records the user behavior data, vehicle attributes, configuration data, and market performance data of the model of this vehicle; it also records the user behavior data, vehicle attributes, configuration data, and market performance data of the competing models.

[0025] Optionally, obtain the competition degree of the model pairs having a competitive relationship according to the user behavior data, including: For the model of this vehicle i and the competing model j, use the following formula to calculate the competition degree of the competing model with respect to the model of this vehicle : ； ; Among them, is the probability that model i and competitor model j co - occur in all user review texts where the two models are in a competitive relationship (co - occur in the same user review text). That is, the number of user review texts in which model i and competitor model j co - occur divided by the number of user review texts that have a competitive relationship with model i or competitor model j. For example, if there are 1000 review texts for model i and competitor model j in total, and according to the structured relationship data table, the number of valid competitive relationship review texts is 100. Among them, the number of reviews in which model i and competitor model j co - occur is 10, the number of reviews in which model i co - occurs with other models is 50, and the number of reviews in which competitor model j co - occurs with other models is 40. Then the probability that model i and competitor model j co - occur is 10 / 100. is the probability that model i and other models (non - i models) appear in the review texts identified as having a valid competitive relationship according to the structured relationship data table. Continuing with the above example, the number of reviews in which model i co - occurs with other models is 60 (including the 10 reviews in which model i and competitor model j co - occur), divided by the number of valid competitive review texts 100, to get Pr ( i )。 is the probability that competitor model j and other models (non - j models) appear in the relevant user review texts. Continuing with the above example, the number of reviews in which competitor model j co - occurs with other models is 50 divided by 100, to get Pr ( j ); is the threshold, used to measure the degree to which model i and competitor model j are co - mentioned in the same user review text.

[0026] Considering that there are significant differences in the appearance frequencies of each model in all user review texts, affected by the actual sales volume scale, directly using lift value as the weight of the edge (i.e., the edge pointing from competitor model j to model i) may lead to distortion in representing the degree of competition. Therefore, normalization processing is introduced. After calculating the lift value, is normalized according to the number of purchase users of model i to obtain , that is divided by the number of user reviews of model i. This normalization operation uses the number of evaluation samples of the model as the normalization factor, thereby eliminating the offset caused by sales differences, making the degree of competition between different models in the network have a unified dimension basis. The formed complex network not only has good sparsity and layering in structure, but also reflects more objective and real market competition characteristics in the weights of the edges, and can support subsequent community identification and network index calculation.

[0027] If If it is greater than the threshold, the competition relationship between vehicle model i and competitor vehicle model j is relatively strong. The determination of the threshold is based on whether the formed network meets the research purpose, robustness analysis, and the convergence requirements of the model.

[0028] S130. Apply the community detection algorithm to the competition network to obtain sub-competition networks belonging to the same sub-segment.

[0029] Although in the foregoing solution, pairs of potentially competing vehicle models are filtered according to the price difference and vehicle model level, the sub-segment to which the vehicle model belongs is not considered. However, vehicle models in the same sub-segment have a stronger competition relationship, which is also the object to be studied in this embodiment.

[0030] This embodiment is different from the prior art solution that determines the sub-segment according to the vehicle model operation scenarios (commuting, logistics, transportation). Instead, it determines the vehicle models belonging to the same sub-segment according to the tight connection relationship between nodes, which has higher scientificity and rationality; and it can classify potentially cross-sub-market vehicle models into one sub-segment. On the basis of constructing the competition network, the Louvain community detection algorithm is used to refine the entire competition structure, so as to identify the potential competitor clustering relationships in different sub-segments. The Louvain algorithm is a hierarchical clustering algorithm based on modularity optimization, which has the efficiency and scalability to handle large-scale complex networks. Its core process includes the following iterative executions: S131. Initially, each node is regarded as a community.

[0031] S132. Determine whether merging the node into an adjacent community maximally improves the modularity; if it maximally improves the modularity, then merge the node into the adjacent community to form a preliminary community structure.

[0032] Before the merging, there are multiple communities in the competition network, and each community has at least one adjacent community. Try to move the node to its adjacent community to maximize the modularity. For each node, calculate the change in modularity (i.e., modularity increment) after it is moved to different communities, and select the community that maximizes the modularity increment for merging. The modularity Q of the competition network is calculated using the following formula: ; where M is the number of edges in the competition network, is an element in the adjacency matrix, indicating whether there is a connection between vehicle model i and competitor vehicle model j. If there is a connection, it is 1; if there is no connection, it is 0. and respectively represent the degrees of vehicle model i and competitor vehicle model j, is an indicator function, represents the community to which vehicle model i belongs, represents the community to which the competitor model j belongs. When the value is 1, it means that the current model i and the competing model j belong to the same community.

[0033] S133. Take each preliminary community structure as a supernode, and merge the supernodes into adjacent supernodes according to the modularity, and finally form a sub-competitive network belonging to the same segment.

[0034] Treat the supernode as a community and try to merge the supernode into the adjacent supernode to increase the modularity of the entire competitive network. Under the multi-layer iterative framework, the global modularity is continuously improved, and the nested community structure in the network is gradually mined. The modularity is calculated using the formula shown in S132, and i and j no longer represent single nodes, but supernodes.

[0035] After the Louvain algorithm, the competition network is divided into multiple communities, that is, multiple sub-competition networks, each of which includes multiple pairs of models with competitive relationships. Through a continuous modularity optimization process, the Louvain algorithm can automatically identify closely connected node groups in the network, thereby achieving accurate division of competitive model groups. It has higher computing efficiency and stronger structural analysis capabilities, and is suitable for automobile market competition networks with a large number of nodes and complex relationships. In practical applications, this method can effectively reveal the relationship structure of different competition levels. For example, the competition in the electric vehicle market between new power brands and traditional independent brands, and the confrontational relationship between independent brands and joint venture brands of the same level of models, can all be explored through this algorithm to explore the clustering characteristics and market distribution pattern behind them.

[0036] S140. Extract key topological attribute indicators for each sub-competition network, where the key topological attribute indicators include degree centrality, weighted degree centrality, and betweenness centrality.

[0037] Degree Centrality mainly measures the number of direct connections between a node and other nodes in the sub-competitive network, that is, the number of competing models that the current model directly competes with in the sub-competitive network, reflecting the breadth of the competitive relationship between the current model and the competing models. The higher the degree centrality, the stronger the market competitiveness of the model. It can be defined as the following formula, where Is a node The degree of the node The number of directly connected edges.

[0038] ; Weighted Degree Centrality takes into account both the number of connections of a node in the sub-competition network with other nodes and the weight of each edge, that is, the comparison frequency in the user review text, representing the degree of competition relationship, and can be defined by the following formula. Among them, represents the set of all nodes directly connected to node , is the weight of the edge between node and its neighbor node . A high weighted degree centrality of a vehicle model not only indicates that it has a competitive relationship with multiple vehicle models, but also means that the degree of these competitive relationships is large.

[0039] ; Betweenness Centrality is an index to measure the importance of nodes in the sub-competition network, indicating the frequency of a node appearing in the shortest paths between other nodes, revealing the role of a certain node as a "mediator" or "bridge" in the network. Each vehicle model is regarded as a node, and the higher the betweenness centrality, the more it represents the core node of the sub-competition network. Losing this vehicle model in the market may lead to changes in the overall competition structure. For each node, the betweenness centrality is defined by the following formula: ; Among them, represents the number of paths passing through vehicle model from vehicle model node to in all the shortest paths, is the total number of all shortest paths from vehicle model node to .

[0040] S150. Calculate the configuration competition index according to the vehicle model attributes and configuration data; obtain the price elasticity index and market sales performance according to the market performance data; obtain the user comprehensive score according to the user behavior data.

[0041] The configuration competition index comprehensively reflects the richness and difference of vehicle functional configurations, and is mainly composed of two parts: configuration richness, which measures the coverage of vehicle models in various functional configurations, and configuration difference, which measures the performance of vehicle models in innovative and forward-looking feature configurations, reflecting the differential competitiveness of vehicles in the market. Based on the vehicle model attributes and configuration data, a configuration item system is established, which includes multiple first-level indicators and multiple second-level indicators. According to the first-level indicators, it is divided into functional configurations and forward-looking configurations, as shown in the following table, and is used to calculate configuration richness and configuration difference respectively.

[0042]

[0043] Calculate the configuration competition index according to the following formula CCI : ; where is the weight, n is the total number of all functional configurations, m is the total number of all forward-looking configurations, is the indicator variable of the r th functional configuration. If this vehicle model has the r th functional configuration, the indicator variable is 1; if it does not have the r th functional configuration, the indicator variable is 0; is the indicator variable of the q th forward-looking configuration. If this vehicle model has the q th forward-looking configuration, the indicator variable is 1; if it does not have the q th forward-looking configuration, the indicator variable is 0.

[0044] In price-sensitive market segments, some companies will engage in price wars to enhance the competitiveness of their products. By monitoring the difference between the actual transaction price (TP) and the manufacturer's suggested retail price (MSRP) of different versions of vehicle models, calculate the price elasticity index (PEI) of the vehicle model to evaluate its competitive performance in the market. The calculation method can be expressed as the following formula: ; where w is the number of versions of this vehicle model, represents the number of actual selling prices of the vehicle model of the bth version (for example, if the selling prices are 100,000 and 120,000, then there are 2 actual selling price numbers), TP bc is the cth actual selling price value of the vehicle model of the bth version (for example, 100,000), represents the average actual selling price of the vehicle model of the bth version (for example, the average of 100,000 and 120,000 is 110,000). Finally, calculate the average actual selling price of the bth version relative to MSRP bThe deviation from the manufacturer's guide price of the b-th version (i.e., the manufacturer's guide price of the b-th version) has a directional result. A positive PEI result reflects that the model has a brand premium or high market recognition. The larger the PEI value, the stronger the competitiveness of the model, and consumers pay a price higher than the suggested retail price for this model. A negative PEI indicates that the model faces greater market pressure, and the smaller the value, the weaker the competitiveness of the model. When evaluating product competitiveness, in order to ensure the consistency of the direction of PEI with other indicators, it is necessary to calculate the standardized PEI index. As shown in the following formula, when PEI is negative, it is first reversed, and then the PEI result is standardized and mapped to the interval 0-1.

[0045] ; Among them, is the standardized price elasticity index, PEI is the price elasticity index of this model, is the minimum value of the positive PEIs of all models in the sub-competition network, is the maximum value of the positive PEIs of all models in the sub-competition network, is the maximum value of the negative PEIs of all models in the sub-competition network, is the minimum value of the negative PEIs of all models in the sub-competition network.

[0046] The market sales performance (Market Performance Score, MPS) is a key indicator mainly measuring the popularity of the model in the market and the actual market share, and is an important indicator for evaluating the overall competitiveness of the model. The market sales performance can be comprehensively quantified into two indicators: the monthly average sales volume of the model and the market share. At the same time, considering the market performance of other models in this sub-segment, the market sales performance and relative competitiveness are comprehensively evaluated. The calculation method of MPS is the ratio of the monthly average sales volume of this model to the monthly average sales volume of the largest model in the sub-competition network.

[0047] The user comprehensive score mainly evaluates the overall experience satisfaction of users with the model. The data comes from the comprehensive evaluation results of users in multiple dimensions, including space, power, fuel consumption or driving range, appearance, interior, cost performance, configuration, etc. The score of each dimension reflects the satisfaction of users in this aspect. Finally, the comprehensive score is obtained by calculating the average value, reflecting the user recognition of the vehicle in the market, and serving as an indicator in the evaluation system of automotive product competitiveness. This solution uses the comprehensive evaluation scores of car owners to obtain the user comprehensive scores of different models.

[0048] S160. Evaluate the competitive relationship of models based on key topological attribute indicators, configuration competition index, price elasticity index, market sales performance, and user comprehensive score.

[0049] The weighted sum of the key topological attribute indicators, configuration competition index, price elasticity index, market sales performance, and user comprehensive score for each vehicle model is calculated to obtain the competition relationship score between the vehicle model and other vehicle models.

[0050] Due to differences in consumer demands, market differences, product positioning, etc. in different sub - fields, the evaluation method of vehicle model competition relationship proposed in this application needs to be based on the division of different sub - fields to avoid biases caused by cross - market comparisons. For a specific sub - field, a competitiveness composite matrix is constructed by incorporating five main indicators: key topological attribute indicators, user comprehensive score, configuration competition index, price elasticity index, and market sales performance into the product competitiveness evaluation system. The performance of each vehicle model in different dimensions is quantified into a set of data. Among them, the sub - matrix of key topological attribute indicators needs to be obtained by calculating the data of three indicators: degree centrality, weighted degree centrality, and betweenness centrality. The composite matrix can be expressed by the following formula, where is the number of vehicle models in the sub - field,

[0051] ; In this application, the entropy weight method is used to objectively assign the weights of each indicator, and the indicator weights of the competition intensity sub - matrix and the competitiveness composite matrix are calculated respectively (i.e., the weights of degree centrality, weighted degree centrality, betweenness centrality, configuration competition index, price elasticity index, market sales performance, and user comprehensive score). The entropy weight method is a weight assignment method based on information entropy, which automatically adjusts the weights of each indicator by measuring the amount of information of each indicator, that is, the degree of data dispersion. First, the Min - Max normalization method is used to normalize the original indicator data to the interval to eliminate the influence of the difference in indicator dimensions, as shown in the following formula.

[0052] ; where and respectively represent the minimum and maximum values of the th indicator, is the value of the th vehicle model for the th indicator, is the normalized value of the th vehicle model for the th indicator. Then, according to the following formula, the proportion of the th indicator of vehicle model in all indicators is calculated to quantify the relative performance of the vehicle model in each indicator.

[0053] ; Based on the vehicle model of the proportion of the th index, the following formula is used to calculate the information entropy of the th index, which is a measure of the degree of dispersion of the index among different vehicle models. The smaller the entropy value, the higher the discrimination degree of the index and the greater the amount of information.

[0054] ; Finally, the weights of each index are calculated using information entropy , as shown in the following formula. According to the obtained index weights, the weighted average method is used to obtain the competition relationship scores of each vehicle model with other vehicle models. ;

[0055] ; The method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data proposed in this application integrates a result visualization and intelligent analysis module, supporting the full-process display from network structure interpretation to competitiveness evaluation output. Through a systematic graphical interface and multi-dimensional visualization means, users can intuitively obtain market structure insights and product competition performance evaluations, improving the efficiency and decision-making value of model applications.

[0056] First, based on the community detection results of the competition network, a structural visualization graph of the market competition groups is generated, as shown in Figure 2 . Different colors in the figure represent different communities, and the numbers represent the number of vehicle models in the community. Optionally, the representative vehicle models or brands of each community are automatically labeled to assist in understanding the composition and distribution of each competition group, thereby identifying the dominant forces and concentrated competition areas in the segmented fields.

[0057] Within the specified community, a competition network visualization solution at the vehicle model level is further provided. As shown in Figure 3 , Figure 3 different colors in

[0058] represent different vehicle model levels. The full name of SUV is Sport Utility Vehicle, representing a sport utility vehicle. Through node classification and relationship connection, the actual competition relationships between different vehicle model levels are displayed, revealing the cross-level alternative competition characteristics in the market and helping enterprises identify "asymmetric competition" situations, that is, the non-traditional path selection behaviors in users' car purchase decisions. Figure 4As shown, it intuitively presents the strong and weak performances of each vehicle model under various evaluation indicators. The heat map supports an interactive viewing function. Users can filter by indicators and focus on vehicle models to quickly locate products with obvious performance advantages or those in urgent need of optimization. At the same time, it provides ranking and weight details as a reference basis. It should be noted that Figure 4 the numerical values of each evaluation indicator in

[0059] are not protected. The above visualization and analysis functions are uniformly integrated by the front-end module, and users can obtain the Figures 2 to 4 results shown with one click through the graphical interface, significantly improving the analysis efficiency and output accuracy. This application also supports exporting the analysis results as a report to adapt to the actual application needs of enterprises in multiple scenarios such as market research and product planning.

[0060] The purpose of this application is to provide a method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data to solve the problems existing in the prior art, such as single evaluation dimension, inaccurate competitor identification, lack of structural modeling and intelligent scoring mechanism, etc. Especially in the context of the increasingly complex current market structure and diverse consumer behaviors, traditional competition relationship analysis methods have been difficult to meet the refined and systematic needs of enterprises for competitor monitoring and product decision-making.

[0061] Compared with the prior art, this application has the following beneficial effects: (1) Break through the limitation of data dimensions and integrate multi-source heterogeneous data. This application is based on multi-source heterogeneous information such as user evaluation texts, configuration parameters, market prices, and sales data, realizing a closed-loop modeling from user perception to market performance, significantly expanding the data basis and adaptability of competition analysis.

[0062] (2) Introduce complex network theory to achieve structured competition modeling. Compared with linear analysis methods, this application is based on a weighted complex network model, taking vehicle models as nodes and competition relationships as edges to construct a quantifiable structure map, and performing community identification through the modularity optimization method. This modeling method can reveal the small-world effect and competition cluster structure in the automotive market, improving the model's analytical ability for complex market competition patterns.

[0063] (3) Adopt a modular index system to improve the accuracy and dynamics of evaluation. This application designs a competitiveness evaluation matrix that integrates multi-dimensional indicators such as network centrality, configuration index, price elasticity, user score, and sales volume, supports objective weight assignment and adaptive update, and overcomes the limitations of subjective weight setting and static response in traditional models.

[0064] (4) System integration visualization and intelligent analysis enhance the applicability of the model. This application is equipped with a visualization analysis function, including a community structure map, a vehicle model competition relationship network, and a comprehensive score display in the form of a heat map, effectively supporting users in the intelligent identification of competitive group division, horizontal substitution relationships, and disadvantaged vehicle models, and enhancing the practicality and interpretability of the model.

[0065] As Figure 5 shown, this embodiment provides an electronic device, including: at least one processor; and a memory communicatively connected to at least one of the processors; wherein, the memory stores instructions executable by at least one of the processors, and the instructions are executed by at least one of the processors to enable at least one of the processors to execute the above method. At least one processor in this electronic device can execute the above method, and thus has at least the same advantages as the above method.

[0066] Optionally, the electronic device further includes interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Each component is interconnected using different buses and can be installed on a common motherboard or otherwise installed as needed. The processor can process instructions executed within the electronic device, including instructions for storing graphical information in the memory or on the memory to display a GUI (Graphical User Interface) on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors and multiple memories can be used together, and / or multiple buses and multiple memories can be used together. Similarly, multiple electronic devices can be connected (for example, as a server array, a set of blade servers, or a multi-processor system), and each device provides some necessary operations. Figure 5 Taking one processor 301 as an example.

[0067] The memory 302, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data in this embodiment of the application. The processor 301 executes various functional applications and data processing of the device by running the software programs, instructions, and modules stored in the memory 302, that is, implements the above method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data.

[0068] The memory 302 may mainly include a program storage area and a data storage area. Among them, the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 302 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some instances, the memory 302 may further include a memory remotely provided with respect to the processor 301, and these remote memories may be connected to the device through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0069] The electronic device may further include: an input device 303 and an output device 304. The processor 301, the memory 302, the input device 303, and the output device 304 may be connected through a bus or other means. Figure 5 Taking the connection through the bus as an example.

[0070] The input device 303 may receive input digital or character information, and the output device 304 may include a display device, an auxiliary lighting device (e.g., an LED), a tactile feedback device (e.g., a vibration motor), etc. The display device may include but is not limited to a liquid crystal display (LCD), a light-emitting diode (LED) display, and a plasma display. In some embodiments, the display device may be a touch screen.

[0071] This embodiment provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to cause a computer to execute the above method. The computer instructions on the computer-readable storage medium are used to cause a computer to execute the above method, and thus have at least the same advantages as the above method.

[0072] The medium in this application may adopt any combination of one or more computer-readable media. The medium may be a computer-readable signal medium or a computer-readable storage medium. The medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the medium (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the medium may be any tangible medium that contains or stores a program, and the program may be used by or in combination with an instruction execution system, apparatus, or device.

[0073] A computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which computer-readable program code is carried. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. The computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0074] The program code contained on a computer-readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wire, optical fiber cable, RF (Radio Frequency), etc., or any suitable combination of the foregoing.

[0075] The computer program code for performing the operations of this application may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., by using an Internet service provider to connect through the Internet).

[0076] It should be understood that the various forms of the processes shown above may be used, with steps reordered, added, or deleted. For example, the steps recited in this application may be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this application can be achieved. No limitation is made herein.

[0077] The above specific embodiments do not constitute a limitation on the protection scope of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application shall be included within the protection scope of this application.

Claims

1. A method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data, characterized in that: include: Collect user behavior data, vehicle attributes and configuration data, and market performance data for various vehicle models; According to the user behavior data, a pair of vehicle models having a competitive relationship and the degree of competition between the vehicle models are obtained; the vehicle model pairs are taken as two nodes, and the degree of competition is taken as the weight of the edge to construct a competition network; Applying a community detection algorithm on the competition network to obtain a sub-competition network belonging to the same segment; the sub-competition network includes a plurality of vehicle model pairs having a competition relationship; Extracting key topological attribute indicators for each sub-competitive network, wherein the key topological attribute indicators include degree centrality, weighted degree centrality and betweenness centrality; The configuration competition index is calculated based on the vehicle attributes and configuration data; the price elasticity index and market sales performance are obtained based on the market performance data; and the user comprehensive score is obtained based on the user behavior data; The competitive relationship between vehicle models is evaluated based on the key topological attribute indicators, configuration competition index, price elasticity index, market sales performance and user comprehensive scores.

2. The method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data according to claim 1 is characterized in that: After collecting user behavior data, vehicle attributes and configuration data, and market performance data for various vehicle models, it also includes: Standardize vehicle model names.

3. The method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data according to claim 2 is characterized in that: According to the user behavior data, a pair of vehicle models having a competitive relationship is obtained, including: Processing the user review text to obtain at least two car models that appear simultaneously in one review as potential competing car models; Among the potentially competitive models, filter them according to price difference and model level to obtain pairs of competing models; A structured relationship data table is constructed by combining the competing vehicle model pairs and the co-occurrence frequencies of the vehicle model pairs in the user review texts.

4. The method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data according to claim 3 is characterized in that: According to the user behavior data, the degree of competition between the competing vehicle models is obtained, including: For the current model i and the competing model j, the following formula is used to calculate the degree of competition of the competing model to the current model: : ； ； in, is the probability that the model i and the competitor model j appear together in the relevant user review text, is the probability of model i and other models appearing in the relevant user review text, is the probability of competitor model j and other models appearing in the relevant user review text; is the threshold value, Used to measure the degree to which model i and competitor model j are mentioned together in the same user review text; Will According to the number of user comments on this model i, we can get .

5. The method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data according to claim 4 is characterized in that: Applying a community detection algorithm on the competition network, a sub-competition network belonging to the same subdivision is obtained, including: Initially, each node is considered as a community; Determine whether merging nodes into adjacent communities can maximize modularity; if so, merge the nodes into adjacent communities to form a preliminary community structure. Each preliminary community structure is taken as a supernode, and the supernode is merged into the adjacent supernode according to the modularity, eventually forming a sub-competitive network belonging to the same segment.

6. The method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data according to claim 5 is characterized in that: The configuration competition index is calculated based on vehicle attributes and configuration data, including: The configuration competition index is calculated according to the following formula CCI : ； in, is the weight, n is the total number of all functional configurations, m is the total number of all forward-looking configurations, It is r The indicator variable of a functional configuration. If this vehicle model has r functional configuration, the indicator variable is 1, and there is no r If there is no functional configuration, the indicator variable is 0; It is q An indicator variable for a forward-looking configuration. If the vehicle model has q Forward-looking configuration, the indicator variable is 1, and there is no q If there is no forward-looking configuration, the indicator variable is 0.

7. The method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data according to claim 6 is characterized in that: The price elasticity index is obtained according to the market performance data, including: The price elasticity index PEI is calculated according to the following formula: ; Among them, w is the number of versions of this model, represents the actual selling price of the bth version of the car model, TP bc is the actual sales value of the cth model of the bth version, represents the average actual selling price of the bth version of the car model, and finally calculates the average actual selling price of each version relative to MSRP b The deviation, MSRP b is the manufacturer's suggested retail price of the bth version of the vehicle; The PEI results are normalized and mapped to the range of 0 to 1 according to the following formula: ； in, is the standardized price elasticity index, PEI is the price elasticity index of this model, is the minimum positive PEI of all models in the sub-competitive network, is the maximum value of the positive PEI of all models in the sub-competitive network, is the maximum value of negative PEI of all models in the sub-competition network, is the minimum negative PEI of all models in the sub-competition network.

8. The method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data according to claim 7, characterized in that: Based on the key topological attribute indicators, configuration competition index, price elasticity index, market sales performance and user comprehensive scores, the competitive relationship of vehicle models is evaluated, including: The key topological attribute indicators, configuration competition index, price elasticity index, market sales performance and user comprehensive score of each model are weighted and summed to obtain the competitive relationship score between the model and other models.

9. An electronic device, characterized in that: include: at least one processor, and a memory communicatively coupled to at least one of the processors; Wherein, the memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that: The medium stores computer instructions, which are used to enable a computer to execute the method for constructing and evaluating a vehicle competition network based on multi-source heterogeneous data according to any one of claims 1 to 8.

Images