Method, device, equipment and storage medium for generating index framework of automobile website
By crawling data from automotive websites and performing correlation calculations and cluster analysis, an indicator feature framework is automatically generated, solving the problems of low efficiency and poor rationality of manual definition, and realizing the generation of an efficient and scientific indicator framework.
Patent Information
- Application Number
- CN202210435497.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-04-24
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2042-04-24
AI Technical Summary
In existing technologies, the generation of indicator frameworks for automotive websites relies on manual definition, which is inefficient and lacks rationality, failing to meet the needs of report development and monitoring.
By crawling valuable data and behavioral indicators from automotive websites, we calculate the relevance level, perform hierarchical clustering and factor analysis, automatically generate an indicator feature framework, and determine its position and level within the framework.
It improves the efficiency and scientific rigor of indicator framework generation, ensures the framework's rationality, and meets the needs of report development and monitoring.
Smart Images

Figure CN114757709B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of computer and network technology, and in particular to a method, device, electronic device and storage medium for generating an indicator framework of an automobile website. Background Art
[0002] In the business activities of automobile companies, the brand's official website and dealer websites are the main channels for consumers to obtain information about the brand's automobiles. Therefore, the measurement of website performance occupies an important position in the online operations of automobile brands. Relevant departments need to monitor and report on the current operating status of automobile official websites.
[0003] In related technologies, the indicator framework of the automobile website is usually manually defined based on business common sense, and the conversion funnel is applied to the report.
[0004] However, in actual operations, consumers vary greatly from one individual to another and website data changes a lot. Manual definition based on intuition is highly subjective and the definition results are easily inconsistent with the actual situation, which cannot meet the monitoring needs in actual website operations. Summary of the invention
[0005] The present application provides a method, device, electronic device and storage medium for generating an indicator framework for an automobile website, so as to solve the problems that the indicator framework is usually generated by manual definition, which is inefficient and the finally generated indicator framework is not rational enough and cannot meet the needs of report development and monitoring.
[0006] A first aspect of the present application provides an indicator framework generation method for an automobile website, comprising the following steps: crawling value data and multiple website behavior indicator features of a target automobile from relevant automobile websites; calculating the correlation level between each indicator feature and the value based on the value data and the multiple website behavior indicator features, and matching the actual position of each indicator feature in an indicator feature framework based on the correlation level; clustering the multiple website behavior indicator features to obtain clustering results, determining the actual number of layers of the indicator feature framework based on the clustering results, and generating the indicator feature framework of the target automobile in the automobile website based on the actual position of each indicator feature in the indicator feature framework and the actual number of layers.
[0007] Optionally, clustering the multiple website behavior indicator features to obtain clustering results includes: performing hierarchical clustering on the multiple website behavior indicator features to obtain clustering results; performing factor analysis on the multiple website behavior indicator features to obtain analysis results; and clustering the indicator features according to the clustering results and the analysis results to obtain the clustering results.
[0008] Optionally, the hierarchical clustering of the multiple website behavior indicator features to obtain a clustering result includes: calculating the Euclidean distance between any indicator features among the multiple website behavior indicator features; combining the indicator features whose Euclidean distance is less than a preset distance to obtain a first new indicator feature; calculating the Euclidean distance between the first new indicator feature and the remaining features among the multiple website behavior indicator features until the Euclidean distance between all features is greater than or equal to the preset distance, thereby generating a tree structure of features.
[0009] Optionally, the factor analysis is performed on the multiple website behavior indicator features to obtain analysis results, including: calculating the load coefficient of each indicator feature in the multiple website behavior indicator features on multiple target factors; taking the load coefficient on each target factor that is greater than a preset coefficient as the target load coefficient of each target factor; and generating a factor load matrix based on the target load coefficient of each target factor.
[0010] Optionally, before calculating the correlation level between each indicator feature and the value based on the value data and the multiple website behavior indicator features, it also includes: performing data cleaning processing on the value data and the multiple website behavior indicator features to obtain processed indicator features and value data.
[0011] Optionally, before calculating the correlation level between each indicator feature and the value according to the value data and the multiple website behavior indicator features, it also includes: generating multiple second new indicator features according to each processed indicator feature and the delay operator.
[0012] Optionally, the correlation level between each indicator feature and the value is calculated based on the value data and the multiple website behavior indicator features, including: calculating the Pearson correlation coefficient between each second new indicator feature and the value; generating a correlation coefficient matrix based on the Pearson correlation coefficient, and rendering the values in the correlation coefficient matrix in a preset order to generate a correlation coefficient heat map; identifying the correlation coefficient heat map to obtain the correlation level between each indicator feature and the value.
[0013] The second aspect of the present application provides an indicator framework generation device for an automobile website, including: a crawling module, used to crawl value data and multiple website behavior indicator features of a target automobile from the automobile website; a matching module, used to calculate the correlation level between each indicator feature and the value based on the value data and the multiple website behavior indicator features, and match the actual position of each indicator feature in the indicator feature framework based on the correlation level; a generation module, used to cluster the multiple website behavior indicator features to obtain clustering results, determine the actual number of layers of the indicator feature framework based on the clustering results, and generate the indicator feature framework of the target automobile in the automobile website based on the actual position of each indicator feature in the indicator feature framework and the actual number of layers.
[0014] The third aspect of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for generating an indicator framework for an automobile website as described in the above embodiment.
[0015] A fourth aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the method for generating an indicator framework for an automobile website as described in the above embodiment.
[0016] Therefore, this application has at least the following beneficial effects:
[0017] The indicator framework of automobiles in automobile websites can be automatically generated based on data statistics, which improves the efficiency of framework generation, avoids the influence of human factors on the definition of indicator framework, improves the scientific nature of framework definition, makes the structure of indicator framework more reasonable, and meets the needs of report development and monitoring. Therefore, it solves the technical problems that the indicator framework is usually generated by manual definition, which is inefficient, and the indicator framework generated is not reasonable and cannot meet the needs of report development and monitoring.
[0018] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:
[0020] Figure 1 A flowchart of a method for generating an indicator framework for an automobile website provided according to an embodiment of the present application;
[0021] Figure 2 An output example diagram provided according to an embodiment of the present application;
[0022] Figure 3 This is an example diagram of a thermal map output provided according to an embodiment of the present application;
[0023] Figure 4 An example diagram of a Euclidean distance matrix for hierarchical clustering provided according to an embodiment of the present application;
[0024] Figure 5 Another example diagram of a Euclidean distance matrix for hierarchical clustering provided according to an embodiment of the present application;
[0025] Figure 6 This is an example diagram of hierarchical clustering output provided according to an embodiment of the present application;
[0026] Figure 7 A flowchart of a method for generating an indicator framework for an automobile website according to an embodiment of the present application;
[0027] Figure 8 This is an example diagram of factor analysis output provided according to an embodiment of the present application;
[0028] Fig. 9 This is an example diagram of an indicator framework generating device for an automobile website provided according to an embodiment of the present application;
[0029] Fig.10 It is a schematic diagram of the structure of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION
[0030] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.
[0031] The following describes the indicator framework generation method, device, electronic device and storage medium of the automobile website of the embodiment of the present application with reference to the accompanying drawings. In view of the problem mentioned in the above background technology that the indicator framework is usually generated by manual definition, which is inefficient, and the rationality of the indicator framework finally generated is poor, and it cannot meet the needs of report development and monitoring, the present application provides an indicator framework generation method for an automobile website. In this method, the indicator framework of the automobile in the automobile website can be automatically generated based on data statistics, the efficiency of framework generation is improved, the influence of human factors on the definition of the indicator framework is avoided, the scientific nature of the framework definition is improved, and the composition of the indicator framework is more reasonable to meet the needs of report development and monitoring. As a result, the problem that the indicator framework is usually generated by manual definition, which is inefficient, and the rationality of the indicator framework finally generated is poor, and it cannot meet the needs of report development and monitoring is solved.
[0032] Specifically, Figure 1 A flowchart of a method for generating an indicator framework for an automobile provided in an embodiment of the present application.
[0033] like Figure 1 As shown, the indicator framework generation method of the automobile website includes the following steps:
[0034] In step S101, the value data of the target car and multiple website behavior indicator features are crawled from relevant automobile websites.
[0035] The relevant automobile website may be the brand official website of the target automobile or a third-party website, etc.; the value data refers to the specific value of the vehicle, such as the sales volume of the vehicle, etc.; the website behavior indicator feature refers to the relevant features used to guide users entering the website to browse, such as Figure 2 Each layer is a website behavior indicator feature; the target car can be specifically set or selected according to the actual situation, and there is no specific limitation on this.
[0036] It is understandable that when obtaining data for generating an indicator architecture, the embodiment of the present application can obtain relevant data from the automobile website by crawling, and the crawling operation can be implemented in a variety of ways, such as Excel, Python, etc.
[0037] In a specific application, the embodiment of the present application can obtain the weekly session level statistical data indicators x1, x2..., x n And the sales data y of the brand, the final data is x1, x2…, x n ,y, that is, there are m rows (n+1) columns of data, one row is one week of data, and one column is one indicator feature.
[0038] In step S102, the correlation level between each indicator feature and the value is calculated based on the value data and multiple website behavior indicator features, and the actual position of each indicator feature in the indicator feature framework is matched based on the correlation level.
[0039] Among them, the correlation level can refer to the degree of correlation between each indicator feature and the value; the higher the correlation level, the stronger the correlation between the indicator feature and the value; the lower the correlation level, the weaker the correlation between the indicator feature and the value.
[0040] It can be understood that the embodiments of the present application can determine the degree of correlation between the indicator feature and the value according to the correlation level between each indicator feature and the value, and determine the actual position of each indicator feature according to the magnitude of the correlation degree.
[0041] In an embodiment of the present application, the correlation level between each indicator feature and the value is calculated based on the value data and multiple website behavior indicator features, including: calculating the Pearson correlation coefficient between each second new indicator feature and the value; generating a correlation coefficient matrix based on the Pearson correlation coefficient, and rendering the values in the correlation coefficient matrix in a preset order to generate a correlation coefficient heat map; identifying the correlation coefficient heat map to obtain the correlation level between each indicator feature and the value.
[0042] The preset order may be an order of values from high to low or from low to high, etc., and may be set according to actual conditions, and is not specifically limited thereto.
[0043] It can be understood that the embodiments of the present application can obtain the Pearson correlation coefficient and generate a coefficient matrix by calculating the correlation between the indicator characteristics and the value, and obtain a correlation coefficient heat map by rendering the matrix related numerical values, so as to determine the correlation level between each indicator characteristic and the value by reading the map.
[0044] For example, taking "dealer query" as an example, adding a 1st-order delay operator can represent one week, and create a new column: dealer query_lagl. The processed data is used to calculate the heat map correlation coefficient to generate a correlation coefficient matrix, with columns corresponding to 1st to 13th-order delay operators x i_lag1 ,x i_lag2 ,x i_lag3 …,x i_lag13 For example, in the row of "dealer query", the third column is the Pearson correlation coefficient between the feature and the sales data after applying the third-order delay operator; after obtaining the correlation coefficient matrix, color it according to the value, and finally obtain the correlation coefficient heat map. The heat map example can be as follows Figure 3 shown.
[0045] It should be noted that when calculating the Pearson correlation coefficient, not only can multiple website behavior indicator features obtained by crawling be directly used, but also multiple website behavior indicator features obtained by crawling can be processed to obtain a second new indicator feature, which is not specifically limited. Among them, the second new indicator feature will be described in detail in subsequent embodiments, and will not be described in detail here.
[0046] When calculating the Pearson correlation coefficient, the second new indicator feature is used as an example to illustrate the details as follows:
[0047] Let each second new index feature be x i , the value is y, and the calculation formula of Pearson correlation coefficient is:
[0048]
[0049] in, The coefficient value range can be [-1,1]. If the coefficient value is closer to 1, the positive correlation between the second new indicator feature and the value is stronger; the closer it is to -1, the stronger the negative correlation between the two; the closer it is to 0, the weaker the correlation between the two.
[0050] In an embodiment of the present application, before calculating the correlation level between each indicator feature and the value based on the value data and multiple website behavior indicator features, the embodiment of the present application can process the multiple website behavior indicator features obtained by crawling in a variety of ways to generate a second new indicator feature, and no specific limitation is made to this.
[0051] As a possible implementation method, multiple second new indicator features are generated based on multiple website behavior indicator features and delay operators obtained by crawling. It can be understood that the embodiment of the present application can generate multiple second new indicator features for the input original features, namely website behavior indicators {X1, X2, ..., X n} combined with the delay operator Β to create multiple feature combinations {X i_lag1 ,X i_lag2 ,X i_lag3 ,…,X i_lagg}, where X i_lagj Represents X i New features obtained after applying the m-th order delay operator;
[0052] As another possible implementation method, data cleaning processing is performed on the value data and multiple website behavior indicator features to obtain processed indicator features and value data; and multiple second new indicator features are generated according to each processed indicator feature and delay operator.
[0053] The delay operator may be a delay operation performed on a time function.
[0054] Among them, data cleaning processing can refer to the extraction and processing of the disordered original data obtained by crawling, including the elimination of abnormal values and null values, such as data anomalies caused by some buried points; the embodiment of the present application can achieve data cleaning through at least one method such as floor&capping, and there is no specific limitation on this.
[0055] It is understandable that the embodiments of the present application can first clean the original value data and website behavior indicator characteristics obtained by crawling, obtain the value data and indicator characteristics after eliminating outliers, and generate a second new indicator characteristic based on the processed indicator characteristics and delay operator.
[0056] When performing data cleaning, the floor&capping method is used as an example to illustrate the details as follows:
[0057] For any indicator feature xi , assuming that 95% of the data points should be in In the range, They are indicator features x i The mean and standard deviation of , for any j∈[1,m] are:
[0058]
[0059] For x1, x2, ..., x n ,y performs the above floor & capping operations to complete data cleaning, and for each feature x other than the valuable data i Add delay operator B k Create new features.
[0060] In step S103, the indicator features of multiple website behavior indicators are clustered to obtain clustering results, the actual number of layers of the indicator feature framework is determined according to the clustering results, and the indicator feature framework of the target car in the car website is generated according to the actual position and actual number of layers of each indicator feature in the indicator feature framework.
[0061] It can be understood that the embodiment of the present application requires a detailed division of the indicator framework, and the actual number of layers of the indicator feature framework can be determined by clustering the original indicator features of the website; the indicator features are specifically located in the indicator feature framework with the actual position and actual number of layers of each indicator feature as coordinates to generate the automobile website indicator feature framework of the target automobile.
[0062] In an embodiment of the present application, indicator feature clustering is performed on multiple website behavior indicator features to obtain clustering results, including: hierarchical clustering is performed on multiple website behavior indicator features to obtain clustering results; factor analysis is performed on multiple website behavior indicator features to obtain analysis results; indicator feature clustering is performed according to the clustering results and the analysis results to obtain clustering results.
[0063] Among them, clustering can refer to the process of dividing a collection of physical or abstract objects into multiple classes consisting of similar objects; factor analysis can refer to the statistical technique of studying the extraction of common factors from a group of variables.
[0064] It can be understood that when the embodiment of the present application needs to cluster multiple website behavior feature indicator features, in order to further cluster, hierarchical clustering and factor analysis can be performed on the multiple website behavior indicator features respectively, and the results obtained after hierarchical clustering and factor analysis can be integrated and clustered again to obtain the final clustering result, so as to improve the reliability and scientificity of the generated indicator framework.
[0065] In an embodiment of the present application, hierarchical clustering is performed on multiple website behavior indicator features to obtain a clustering result, including: calculating the Euclidean distance between any indicator features in the multiple website behavior indicator features; combining indicator features whose Euclidean distance is less than a preset distance to obtain a first new indicator feature; calculating the Euclidean distance between the first new indicator feature and the remaining features in the multiple website behavior indicator features, until the Euclidean distance between all features is greater than or equal to the preset distance, thereby generating a tree structure of features.
[0066] Among them, for some two features x i With x j , Euclidean distance
[0067] Among them, the preset distance can be used to divide the Euclidean distance between any characteristic values in the website behavior index. The preset distance can be specifically set according to actual conditions, and no specific limitation is made to this.
[0068] It can be understood that the embodiments of the present application can divide multiple indicator features into different categories according to similar features by performing hierarchical clustering processing on the indicator features, that is, assigning different indicator features to different clusters according to the specified clustering level, and obtaining the clustering results of hierarchical clustering of multiple website behavior indicator features.
[0069] For example, taking "dealer query" as an example, adding a 1st-order delay operator can represent one week, and create a new column: dealer query_lagl. Perform hierarchical clustering on the data, calculate the Euclidean distance between each indicator feature, and obtain the similarity matrix, such as Figure 4 As shown; the two features with the closest Euclidean distance are combined into a new feature, namely the first new indicator feature, and the remaining features and the first new indicator feature are continued to be calculated to obtain a similarity matrix, as shown Figure 5 As shown; from bottom to bottom, a tree structure is formed. This step can be completed by at least one statistical software such as Excel, and there is no specific limitation on this. The example diagram can be as shown in Figure 6 shown.
[0070] In an embodiment of the present application, factor analysis is performed on multiple website behavior indicator features to obtain analysis results, including: calculating the load coefficient of each indicator feature in the multiple website behavior indicator features on multiple target factors; taking the load coefficient on each target factor that is greater than a preset coefficient as the target load coefficient of each target factor; and generating a factor load matrix based on the target load coefficient of each target factor.
[0071] Among them, factor analysis can find hidden representative factors among many variables, and classifying variables of the same nature into one factor can reduce the variable items. The loading coefficient can be a coefficient introduced to facilitate the calculation of multiple target factors in actual factor analysis; the preset coefficient can be specifically set according to actual conditions, etc., and there is no specific limitation on this.
[0072] It can be understood that the embodiments of the present application can classify multiple indicator features that are essentially the same or have common factors into one factor through factor analysis, test the hypothesis of the relationship between variables, and obtain the clustering results of factor analysis of multiple website behavior feature indicator features.
[0073] For example, taking "dealer query" as an example, adding a first-order delay operator can represent one week, and create a new column: dealer query_lagl. Perform factor analysis on the data, and use statistical software to perform factor analysis on the original quantity that does not apply the delay operator feature except sales, that is, x1, x2..., x n Perform factor analysis; calculate the factor loading matrix, such as Figure 5 , each value in the matrix is the corresponding x i Load coefficient on factor i; find the variable with the largest load coefficient for each factor, such as Figure 4 In the figure, for factor 1, variables Q7, Q8, and Q9 are considered to have larger loading coefficients on factor 1, that is, factor 1 is the main factor behind these three variables.
[0074] In summary, after obtaining the value data of the target car and the characteristics of multiple website behavior indicators, the embodiment of the present application can apply the delay operator and the correlation coefficient to help determine the different positions of different indicators in the framework; cluster the indicators based on factor analysis and hierarchical clustering to find the indicator clusters with similar positions in the framework; and combine the analysis to generate the indicator framework of the automobile website.
[0075] The following will describe the index framework generation method of the automobile website of the embodiment of the present application through a specific embodiment. Figure 7 As shown, the specific steps are as follows:
[0076] 1. Get the weekly session-level statistics indicators x1, x2, ..., x n And the sales data y of the brand, the final data is x1, x2…, x n ,y, that is, there are m rows (n+1) columns of data, one row is a week of data, and one column is an indicator feature. 15 For example, this number is the 5th week data of indicator feature x1.
[0077] 2. Data cleaning, and analysis of each feature x except sales volume iAdd delay operator B k Create new features.
[0078] Among them, data cleaning can use the floor & capping method: that is, for any indicator feature x i , assuming that 95% of the data points should be in In the range, They are indicator features x i The mean and standard deviation of , for any j∈[1,m] are:
[0079]
[0080] For x1, x2, ..., x n ,y performs the above floor & capping operations to complete data cleaning. Wherein, k is the order of the delay operator, and the embodiment of the present application can add 1 to 13 order delay operators.
[0081] 3. For "dealer query", add a first-order lag operator, that is, add the corresponding time of this feature every week to the data interval of the data set, which is one week in the embodiment of the present application, and create a new column: dealer query_lag1; and perform heat map correlation coefficient calculation and drawing operations, hierarchical clustering operations and factor analysis operations on the processed data.
[0082] When calculating the heat map correlation coefficient and plotting:
[0083] (1) Use statistical software to calculate the characteristics x under each order of delay operator i The Pearson correlation coefficient between y and sales volume is calculated as:
[0084]
[0085] in, The coefficient value range is [-1,1]. If the coefficient value is closer to 1, the positive correlation between the feature and sales volume is stronger. If it is closer to -1, the negative correlation between the feature and sales volume is stronger. If it is closer to 0, the correlation between the two is weaker.
[0086] (2) Generate a correlation coefficient matrix, that is, the rows are the characteristics, and the columns are the corresponding 1st to 13th order delay operators x i_lag1 ,x i_lag2 ,x i_lag3 …,x i_lag13 For example, for the "dealer query" row, the third column is the Pearson correlation coefficient between the feature and the sales data after applying the third-order delay operator;
[0087] (3) After obtaining the correlation coefficient matrix, color it according to the value, and finally obtain the correlation coefficient heat map, such as Figure 3 As shown, the details are as follows:
[0088] When performing hierarchical clustering:
[0089] (1) For two features x i With x j , Euclidean distance Calculate the Euclidean distance between each feature and get the similarity matrix, such as Figure 4 As shown;
[0090] (2) Combine the two features with the closest Euclidean distance into a new feature, and continue the operation in (1) for the remaining features and this feature, such as Figure 5 As shown;
[0091] (3) The statistical software finally combines the data from bottom to bottom into a tree structure, such as Figure 6 shown.
[0092] When performing factor analysis:
[0093] (1) Use statistical software to perform factor analysis on the original quantities other than sales volume that do not use the delay operator feature, that is, x1, x2, ..., x n Conduct factor analysis;
[0094] (2) Calculate the factor loading matrix, such as Figure 5 As shown, each value in the matrix is the corresponding x i Loading coefficient on factor i;
[0095] (3) For each factor, find the variable with the largest load coefficient, such as Figure 5 In the figure, for factor 1, variables Q7, Q8, and Q9 are considered to have larger loading coefficients on factor 1, that is, factor 1 is the main factor behind these three variables.
[0096] 4. Analyze the results of correlation coefficient heat map, factor loading matrix, and hierarchical clustering.
[0097] (1) Figure 3 As shown in the correlation coefficient heat map, x1, x2, and x4 have the highest correlation coefficient with sales at the first-order delay operator, while x5 has the highest correlation coefficient with sales at the third order; that is, for x1, x2, and x4, last week's values have a strong correlation coefficient with this week's sales, so these three variables are more suitable to be placed at the bottom of the conversion funnel; and for x5, the value three weeks ago has a strong correlation with this week's sales, so this variable is more suitable to be placed in the upper part of the conversion funnel;
[0098] (2) Figure 6 As shown, variables A and F, B and C are close in the tree diagram, so it can be considered that variables A and F, B and C are in the same stage or a relatively close stage of the conversion funnel;
[0099] (3) Figure 8 As shown, variables Q7, Q8, and Q9 have larger loading coefficients on factor 1, and variables Q1, Q2, and Q3 have larger loading coefficients on factor 4, so Q7, Q8, and Q9 are considered to be at the same stage or a relatively close stage of the conversion funnel, while variables Q1, Q2, and Q3 are considered to be at the same stage or a relatively close stage of the conversion funnel;
[0100] (4) Based on common sense, if Figure 2 As shown in the figure, taking the automobile website as an example, the test drive booking stage is basically in the stage of intention before purchase, so it is placed at the bottom of the conversion funnel. As for the website visit indicator, it is more of a rough browse before purchase, so it can be placed in the upper part of the conversion funnel.
[0101] 5. Based on the results of the above steps, generate the indicator framework of the final automobile website.
[0102] According to the indicator framework generation method for an automobile website proposed in the embodiment of the present application, the indicator framework of automobiles in the automobile website can be automatically generated based on data statistics, thereby improving the efficiency of framework generation, avoiding the influence of human factors on the definition of the indicator framework, improving the scientific nature of the framework definition, and making the structure of the indicator framework more reasonable to meet report development and monitoring needs.
[0103] Next, the indicator framework generating device for the automobile website proposed in the embodiment of the present application is described with reference to the accompanying drawings.
[0104] Fig. 9 It is a block diagram of an indicator framework generating device for an automobile website according to an embodiment of the present application.
[0105] like Fig. 9 As shown, the indicator framework generating device 10 of the automobile website includes: a crawling module 100, a matching module 200 and a generating module 300.
[0106] In an embodiment of the present application, a crawling module 100 is used to crawl value data and multiple website behavior indicator features of a target car from a car website; a matching module 200 is used to calculate the correlation level between each indicator feature and the value based on the value data and multiple website behavior indicator features, and match the actual position of each indicator feature in the indicator feature framework based on the correlation level; a generation module 300 is used to perform indicator feature clustering on multiple website behavior indicator features to obtain clustering results, determine the actual number of layers of the indicator feature framework based on the clustering results, and generate the indicator feature framework of the target car in the car website based on the actual position and actual number of layers of each indicator feature in the indicator feature framework.
[0107] It should be noted that the above explanation of the embodiment of the method for generating an indicator framework for an automobile website is also applicable to the device for generating an indicator framework for an automobile website of this embodiment, and will not be elaborated herein.
[0108] According to the indicator framework generation device for the automobile website proposed in the embodiment of the present application, the indicator framework of the automobile in the automobile website can be automatically generated based on data statistics, thereby improving the efficiency of framework generation, avoiding the influence of human factors on the definition of the indicator framework, improving the scientific nature of the framework definition, and making the structure of the indicator framework more reasonable to meet the needs of report development and monitoring.
[0109] Fig.10 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include:
[0110] A memory 1001 , a processor 1002 , and a computer program stored in the memory 1001 and executable on the processor 1002 .
[0111] When the processor 1002 executes the program, the indicator framework generation method for the automobile website provided in the above embodiment is implemented.
[0112] Furthermore, the electronic device further comprises:
[0113] The communication interface 1003 is used for communication between the memory 1001 and the processor 1002 .
[0114] The memory 1001 is used to store computer programs that can be executed on the processor 1002 .
[0115] The memory 1001 may include a high-speed RAM (Random Access Memory) memory, and may also include a non-volatile memory, such as at least one disk memory.
[0116] If the memory 1001, the processor 1002 and the communication interface 1003 are implemented independently, the communication interface 1003, the memory 1001 and the processor 1002 can be connected to each other through a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig.10 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.
[0117] Optionally, in a specific implementation, if the memory 1001, the processor 1002 and the communication interface 1003 are integrated on a chip, the memory 1001, the processor 1002 and the communication interface 1003 can communicate with each other through an internal interface.
[0118] The processor 1002 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.
[0119] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for generating an indicator framework for an automobile website.
[0120] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms are not necessarily directed to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.
[0121] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.
[0122] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.
[0123] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array, a field programmable gate array, etc.
[0124] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.
Claims
1. A method for generating an indicator framework for an automobile website, characterized in that: The following steps are involved: Crawl the target car’s value data and multiple website behavior indicator features from relevant car websites; Calculate the correlation level between each indicator feature and the value according to the value data and the plurality of website behavior indicator features, and match the actual position of each indicator feature in the indicator feature framework based on the correlation level; Performing indicator feature clustering on the multiple website behavior indicator features to obtain clustering results, determining the actual number of layers of the indicator feature framework according to the clustering results, and generating the indicator feature framework of the target car in the car website according to the actual position and the actual number of layers of each indicator feature in the indicator feature framework; The clustering of the plurality of website behavior indicator features to obtain a clustering result includes: performing hierarchical clustering on the plurality of website behavior indicator features to obtain a clustering result; performing factor analysis on the plurality of website behavior indicator features to obtain an analysis result; clustering the indicator features according to the clustering result and the analysis result to obtain the clustering result; The hierarchical clustering of the multiple website behavior indicator features to obtain a clustering result includes: calculating the Euclidean distance between any indicator features in the multiple website behavior indicator features; combining the indicator features whose Euclidean distance is less than a preset distance to obtain a first new indicator feature; calculating the Euclidean distance between the first new indicator feature and the remaining features in the multiple website behavior indicator features until the Euclidean distance between all features is greater than or equal to the preset distance, thereby generating a tree structure of features; The performing of factor analysis on the plurality of website behavior indicator characteristics to obtain analysis results includes: calculating the load coefficient of each indicator characteristic of the plurality of website behavior indicator characteristics on a plurality of target factors; taking the load coefficient on each target factor that is greater than a preset coefficient as the target load coefficient of each target factor; and generating a factor load matrix according to the target load coefficient of each target factor; The method of calculating the correlation level between each indicator feature and the value based on the value data and the multiple website behavior indicator features includes: calculating the Pearson correlation coefficient between each second new indicator feature and the value; generating a correlation coefficient matrix based on the Pearson correlation coefficient, and rendering the values in the correlation coefficient matrix in a preset order to generate a correlation coefficient heat map; identifying the correlation coefficient heat map to obtain the correlation level between each indicator feature and the value.
2. The method according to claim 1, characterized in that Before calculating the correlation level between each indicator feature and the value according to the value data and the plurality of website behavior indicator features, the method further includes: The value data and the plurality of website behavior indicator characteristics are cleansed to obtain processed indicator characteristics and value data.
3. The method according to claim 2, characterized in that Before calculating the correlation level between each indicator feature and the value according to the value data and the plurality of website behavior indicator features, the method further includes: A plurality of second new indicator features are generated according to each processed indicator feature and the delay operator.
4. A device for generating an indicator framework for an automobile website, characterized in that: include: The crawling module is used to crawl the value data of the target car and multiple website behavior indicator characteristics from the car website; A matching module, used to calculate the correlation level between each indicator feature and the value according to the value data and the plurality of website behavior indicator features, and match the actual position of each indicator feature in the indicator feature framework based on the correlation level; A generating module, configured to cluster the indicator features of the plurality of website behavior indicator features to obtain a clustering result, determine the actual number of layers of the indicator feature framework according to the clustering result, and generate the indicator feature framework of the target car in the car website according to the actual position of each indicator feature in the indicator feature framework and the actual number of layers; The generating module is further used to: perform hierarchical clustering on the plurality of website behavior indicator features to obtain clustering results; Performing factor analysis on the plurality of website behavior indicator characteristics to obtain analysis results; clustering the indicator characteristics according to the clustering results and the analysis results to obtain the clustering results; The generating module is further used to: calculate the Euclidean distance between any indicator features in the multiple website behavior indicator features; combine the indicator features whose Euclidean distance is less than a preset distance to obtain a first new indicator feature; calculate the Euclidean distance between the first new indicator feature and the remaining features in the multiple website behavior indicator features until the Euclidean distance between all features is greater than or equal to the preset distance, thereby generating a tree structure of features; The generating module is further used to: calculate the load coefficient of each indicator feature of the multiple website behavior indicator features on multiple target factors; use the load coefficient on each target factor that is greater than the preset coefficient as the target load coefficient of each target factor; and generate a factor load matrix according to the target load coefficient of each target factor; The matching module is further used to calculate the Pearson correlation coefficient between each second new indicator feature and the value; generate a correlation coefficient matrix based on the Pearson correlation coefficient, and render the values in the correlation coefficient matrix in a preset order to generate a correlation coefficient heat map; identify the correlation coefficient heat map to obtain the correlation level between each indicator feature and the value.
5. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for generating an indicator framework for an automobile website as described in any one of claims 1 to 3.
6. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the indicator framework generation method for an automobile website as described in any one of claims 1 to 3.
