After-sales data analysis method based on electronic product sales

Abstract

The invention discloses an after-sales data analysis method based on electronic product sales, and relates to the technical field of data analysis, and the method comprises the following steps: carrying out the after-sales feature construction of a product based on multi-source data, and obtaining an after-sales feature set; performing fault assessment according to the after-sales feature set to obtain a fault assessment result; performing risk feature extraction according to a fault assessment result to obtain high-risk features; performing defect correlation analysis according to the high-risk features to obtain a defect correlation result; performing after-sales risk assessment according to the defect association result to obtain a risk assessment result; and performing high-risk product identification according to the risk assessment result to obtain an identification result. According to the method, the comprehensiveness of risk scoring is improved, the interpretability of a risk assessment result is enhanced, and technical and business departments are helped to quickly identify risk reasons.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis, in particular to a post-sale data analysis method based on electronic product sales. BACKGROUND

[0002] With the increasing competition in the electronic product market, enterprises are increasingly relying on post-sale data analysis methods to identify potential high-risk product models, so as to conduct early investigation, recall or process optimization, and reduce post-sale costs and brand risks.

[0003] In the prior art, there are deficiencies in risk assessment: the existing risk assessment method only uses a single index (such as sales multiplied by failure rate) to perform a relatively simple linear weighting method, which is difficult to truly reflect the comprehensive risk of different models under multi-dimensional characteristics such as defect association and risk concentration, and also lacks sufficient interpretability. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides a post-sale data analysis method based on electronic product sales to solve the problems raised in the background art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0006] In a first aspect, the present application provides a post-sale data analysis method based on electronic product sales, comprising the following steps:

[0007] S1, product post-sale feature construction based on multi-source data, to obtain a post-sale feature set;

[0008] S2, failure evaluation according to the post-sale feature set, to obtain a failure evaluation result;

[0009] S3, risk feature extraction according to the failure evaluation result, to obtain a high-risk feature;

[0010] S4, defect association analysis according to the high-risk feature, to obtain a defect association result;

[0011] S5, post-sale risk evaluation according to the defect association result, to obtain a risk evaluation result;

[0012] S6, high-risk product identification according to the risk evaluation result, to obtain an identification result.

[0013] Further optimization of the technical solution, the failure evaluation in step S2 comprises:

[0014] Based on the obtained set of after-sales characteristics of product models, a weighted failure model is used to quantify the business importance and economic impact of different types of failures. A quantifiable and comparable failure rate index is constructed to comprehensively assess the after-sales risk of product models and obtain failure assessment results.

[0015] To further optimize this technical solution, the weighted fault model includes:

[0016]

[0017] in:

[0018] : No. Weighted failure rate of product model;

[0019] The number of different fault types;

[0020] : No. Weighting coefficients for different types of faults;

[0021] : No. Model product number Number of occurrences of this type of failure;

[0022] : No. Cumulative sales volume of the product model.

[0023] To further optimize this technical solution, the weighting coefficients include:

[0024]

[0025] in:

[0026] : No. Model product number The average repair cost for this type of fault;

[0027] The maximum average repair cost across all fault types for all product models;

[0028] By normalizing the average repair costs of various types of faults for different product models, weighting coefficients for various types of faults for each product model are calculated.

[0029] To further optimize this technical solution, the risk feature extraction in step S3 includes:

[0030] According to the failure evaluation result, risk feature extraction is performed, and through difference analysis, a core index of weighted failure rate is used to find the association between the after-sales features of the product model and the high failure rate, identify high-risk features that have a significant contribution to the high failure rate, and provide a basis for subsequent steps.

[0031] Further optimization of the technical solution, the defect correlation analysis in step S4 includes:

[0032] Through defect correlation analysis, the obtained high-risk features are associated with defect causes, potential root causes are found out, defect cause hypotheses are obtained, and verification is performed based on historical defect knowledge and business processes, and defect correlation results are output, providing a basis for subsequent steps.

[0033] Further optimization of the technical solution, the after-sales risk assessment in step S5 includes:

[0034] According to the obtained defect correlation results, the after-sales feature set of the product model and the weighted failure rate are combined, a risk assessment model is used, the objective historical risk indicators of the product are fused with the defect level interpretation indicators, the product risk score that comprehensively reflects the quantitative risk, economic risk and defect risk is obtained, and the risk assessment result is used as a quantitative basis for subsequent steps.

[0035] Further optimization of the technical solution, the risk assessment model includes:

[0036]

[0037] Among them:

[0038] : the risk score of the first model product;

[0039] : the normalized value of the average time to first failure of the first model product;

[0040] : the defect correlation result indication item of the first model product;

[0041] : the weight coefficient of the weighted failure rate;

[0042] : the weight coefficient of the average time to first failure;

[0043] : the weight coefficient of the defect correlation result.

[0044] Further optimization of the technical solution, the normalized value of the average time to first failure includes:

[0045]

[0046] Wherein:

[0047] : the first failure time of the product of the model; : the minimum average first failure time of all products of the model;

[0048] : the maximum average first failure time of all products of the model;

[0049] : the maximum average first failure time of all products of the model;

[0050] By inversely normalizing the maximum and minimum values of the average first failure time, the greater the normalized value, the greater the impact on the risk score.

[0051] Further optimization of the technical solution, the high-risk product identification in step S6 includes:

[0052] According to the obtained risk assessment result, high-risk product identification is performed to obtain an identification result, and a clear high-risk product set is output, thereby improving the pertinence of after-sales risk management, providing a target list for subsequent root cause analysis, defect improvement, and production adjustment, and supporting management guidance for subsequent troubleshooting, recall, and resource allocation.

[0053] In a second aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the computer program instructions are executed by the processor to implement the steps of the after-sales data analysis method based on electronic product sales according to the first aspect of the present application.

[0054] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program instructions are executed by the processor to implement the steps of the after-sales data analysis method based on electronic product sales according to the first aspect of the present application.

[0055] Compared with the prior art, the present application provides an after-sales data analysis method based on electronic product sales, which has the following beneficial effects:

[0056] The after-sales data analysis method based on electronic product sales, through the risk assessment model, is no longer limited to a single index, combines multiple indexes, improves the comprehensiveness of the risk score, enhances the interpretability of the risk assessment result, and helps the technical and business departments to quickly identify the risk causes. BRIEF DESCRIPTION OF DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0058] Fig. 1 A flowchart of a post-sale data analysis method based on electronic product sales according to the present application;

[0059] Fig. 2 A flowchart of a weighted failure model of a post-sale data analysis method based on electronic product sales according to the present application;

[0060] Fig. 3 A flowchart of a risk assessment model of a post-sale data analysis method based on electronic product sales according to the present application. DETAILED DESCRIPTION

[0061] In order to make the above objectives, features and advantages of the present application more apparent and comprehensible, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0062] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be practiced in other ways different from those described herein without departing from the scope of the present application, and those skilled in the art can make similar extensions without departing from the scope of the present application, so the present application is not limited to the specific embodiments disclosed below.

[0063] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor does it mean an embodiment that is separate or selectively excluded from other embodiments.

[0064] Embodiment one:

[0065] Reference Figs. 1-3 For the first embodiment of the present application, the embodiment provides a post-sale data analysis method based on electronic product sales, including the following steps:

[0066] S1, product post-sale feature construction based on multi-source data is performed to obtain a post-sale feature set.

[0067] In the present embodiment, the product post-sale feature construction includes:

[0068] Electronic products are usually sold through multiple channels, such as online e-commerce, offline distributors, and self-owned stores, and the corresponding after-sales data is scattered in multiple systems, such as after-sales service work order system, customer follow-up system, maintenance center management system, etc. Due to differences in design background, purpose and structure, these data systems often have inconsistent data field naming and types, non-uniform time formats, inability to directly associate records of the same product, missing values and dirty data, etc. Moreover, single-source data can only reflect a partial business perspective and is difficult to fully reflect the reliability, customer perception and cost performance of products in the entire actual use cycle, resulting in distorted analysis results and inaccurate identification of defects and improvement direction. Therefore, it is necessary to integrate multi-source data and build product after-sales features to provide high consistency and high completeness of input data for subsequent steps.

[0069] The purpose of this step is to integrate multi-source data by unifying data structure, data cleaning and data matching, and build a standardized and structured product after-sales feature set, so as to ensure that different source data has a unified meaning in subsequent analysis, provide consistent input for subsequent steps, avoid calculation errors caused by field differences or missing, ensure the comparability of cross-time and cross-channel analysis, and improve the accuracy and stability of subsequent steps.

[0070] The implementation method of this step includes:

[0071] Feature parameter determination: Determine the parameters included in the after-sales feature set, including product model, fault type, cumulative sales volume, fault occurrence frequency, average maintenance cost of each type of fault, etc. Record the time length from the sale of each product to the first after-sales, and obtain the average fault first occurrence time length of each product model to analyze the early failure tendency and design defect risk.

[0072] Unified data structure: Analyze the data fields of each data source (sales system, after-sales work order system, follow-up system, etc.), map and rename the fields, for example, unify the fields representing product model in different systems to model number, unify the sales area to area number, etc.

[0073] Cleaning and standardization: Handle missing values and outliers, unify time format (such as all converted to standard timestamp), so that subsequent time series calculation and cross-system data matching can be performed;

[0074] Matching and integration: Associate sales records with after-sales records one by one through product unique identifier (such as SN number or order number), and aggregate each type of data under the same product model;

[0075] Feature set generation: After unifying data structure, cleaning and standardizing data, and matching and integrating data, the after-sales feature set of each product model with integrity and uniformity is obtained.

[0076] S2. Perform fault assessment based on the after-sales feature set to obtain the fault assessment results.

[0077] In this embodiment, the fault assessment includes:

[0078] In step S1, the after-sales characteristic set of each product model was obtained. Simply relying on the number of occurrences of various types of faults cannot reflect the differences in business value of different faults. This step uses a weighted fault model to quantify the business importance and economic impact of different types of faults, and constructs a quantifiable and comparable fault rate index that is oriented towards actual business. By comprehensively considering the after-sales risks of each product model, the fault assessment results are obtained, providing a data foundation for subsequent steps.

[0079] Furthermore, the weighted fault model includes:

[0080]

[0081] in:

[0082] : No. The weighted failure rate of each product model reflects the failure risk of each product model and provides a numerical indicator for subsequent steps.

[0083] The number of different fault types;

[0084] : No. The weighting coefficient for each type of failure is used to adjust the impact of different failure types, amplifying the proportion of failures with a greater impact. It is set according to the impact of each type of failure on economic losses, and ranges from 0 to 1.

[0085] : No. Model product number The number of occurrences of each type of failure quantifies the frequency of occurrence of each type of failure;

[0086] : No. The cumulative sales volume of a product model is used to normalize the number of failures to obtain the failure rate, thus avoiding distortion caused by direct comparisons due to differences in sales volume of different models.

[0087] Furthermore, the weighting coefficients include:

[0088]

[0089] in:

[0090] : No. Model product number Average repair cost of each fault type of each product model

[0091] Maximum value of average repair cost of all fault types of all product models, used to realize normalization of average repair cost to obtain a weighted coefficient

[0092] By normalizing the average repair cost of each fault type of each product model, the weighted coefficient of each fault type of each product model is calculated.

[0093] The model describes how to construct a product failure rate indicator through the weighted coefficient.

[0094] The traditional failure assessment method usually only uses a simple product failure rate calculation, i.e., only the sum of the occurrence probabilities of each fault type is calculated, without distinguishing the influence degree of different fault types. However, the present model introduces a weighted coefficient that can reflect the influence of each fault type on actual business, and the occurrence probabilities of each fault type are weighted and summed, so that the result can more truly reflect the comprehensive influence on customer experience and cost, improve the business sensitivity to high-risk faults, and enhance the risk identification capability.

[0095] The steps for using the model include:

[0096] Data acquisition: In step S1, the cumulative sales of each product model is obtained , the occurrence number of each fault type of each product model , the average repair cost of each fault type of each product model , and the maximum value of average repair cost ;

[0097] Coefficient calculation: According to the obtained average repair cost of each fault type of each product model and the maximum value of average repair cost , the weighted coefficient of each fault type is calculated to obtain the weighted coefficient corresponding to each fault type ;

[0098] Failure assessment: According to the obtained cumulative sales of each product model , the occurrence number of each fault type of each product model , and in combination with the weighted coefficient , the weighted failure rate of each product model is calculated .

[0099] S3, according to the failure assessment result, risk features are extracted to obtain high-risk features.

[0100] In the present embodiment, the risk feature extraction includes:

[0101] In step S2, the weighted failure rate of each product model is calculated, and the failure evaluation result is obtained. However, the failure rate of a single product can only describe the result and find out which product model has a higher failure rate, but cannot explain the reason for the higher failure rate. The high failure rate of different product models is often not random, but is strongly related to specific risk characteristics. Enterprises need to know the risk characteristics that lead to high failure rates in order to fundamentally improve the design, production or after-sales strategy.

[0102] The purpose of this step is to extract risk characteristics according to the failure evaluation result. Through difference analysis, the characteristics of each product model that are significantly associated with high failure rates are found out, such as production batch, part supplier, production date, etc. The high-risk characteristics that significantly contribute to high failure rates are identified, so that the source of the problem can be identified more precisely, facilitating subsequent defect cause investigation, quality improvement or recall decision-making, and providing a basis for subsequent steps.

[0103] The implementation method of this step includes:

[0104] Analysis object construction: establish a mapping relationship between the after-sales characteristics of each product model obtained in step S1 and the weighted failure rate of the corresponding product model obtained in step S2 as the analysis object;

[0105] Difference analysis: for each characteristic, analyze the mean and variance of the weighted failure rate corresponding to different values. For example, compare the mean of the weighted failure rate of all product models with mainboards of type ZB1 with the product models with mainboards of type ZB2 to see if there is a significant difference. If there is a significant difference between different mainboard types (judged according to actual needs and experience), it means that the mainboard type is a key risk characteristic. The analysis is not limited to a single characteristic, but can also be done for a combination of characteristics (such as "a certain production batch + a certain factory"). After identifying the key characteristics, compare the mean of the weighted failure rate of product models with a specific characteristic value (such as the same mainboard type or the same production batch) with the average weighted failure rate of all product models. If the mean of the weighted failure rate of product models with a specific characteristic value is significantly higher (judged according to actual needs and experience) than the average weighted failure rate of all product models, it means that the characteristic value may be directly related to high risk, thereby identifying which characteristic leads to higher risk for the product model. If the difference only appears in a small number of models, it needs to be carefully confirmed.

[0106] High-risk characteristic output: After difference analysis, the characteristics in the after-sales characteristics of each product model that are determined to have a significant impact on high failure rates are extracted to form a high-risk characteristic set.

[0107] S4, defect correlation analysis is performed according to the high-risk characteristics, and a defect correlation result is obtained.

[0108] In this embodiment, the defect correlation analysis includes:

[0109] In step S3, the high-risk features of each product model are obtained, but the high-risk features themselves are only statistically significant results and have not yet been explained why the high risk is generated, and cannot guide technical or production decisions. The enterprise needs to know the causal or strong correlation between these features and specific defects in order to implement targeted improvements.

[0110] The purpose of this step is to establish a connection between the obtained high-risk features and possible defect causes through defect correlation analysis, find out potential root causes, obtain defect cause hypotheses, and verify them based on historical defect knowledge and business processes, output defect correlation results, and provide a basis for subsequent steps.

[0111] The implementation method of this step includes:

[0112] Business data backtracking analysis: backtrack the business records directly related to the high-risk features from the historical data, including quality inspection reports, non-conforming product records, repair reason descriptions, production line detection records, etc. For example, for the high-risk feature "ZB1 model motherboard", the historical defect descriptions and statistical information of this motherboard in the production and after-sales links are summarized. For the high-risk feature "PC1 product batch", the process parameter fluctuations during the production of this batch, supplier delivery records, etc. are checked, so as to convert the features in the statistical perspective into specific traceable business events, ensure that each high-risk feature can be corresponded to the business process and physical batch, and facilitate subsequent confirmation of defect nature;

[0113] Historical defect knowledge correlation: compare the business information obtained by backtracking with the defect knowledge base inside or outside the enterprise, the defect knowledge base includes typical defect types (such as virtual welding of welding points, circuit short circuit, material aging, etc.), cause modes and typical features, analyze which defect modes have been associated with similar production conditions, components or batches in the past, for example, it is found through comparison that the ZB1 model motherboard has caused overheating due to power chip process defects in the past, and it occurs in the same supplier or production line;

[0114] Defect Association Result Verification and Output: The results of backtracking analysis and knowledge comparison are combined. If there is no obvious match, the output is a weak association result (meaning that the historical defect records support limited information or only a few cases have occurred, with no repetition, indicating that the feature has low risk or high uncertainty in the short term), which requires further technical investigation. If the high-risk feature is highly consistent with the historical defect pattern, or the same defect description appears repeatedly in multiple places, the output is a strong association result (meaning that the historical defect pattern and the current high-risk feature are highly consistent, multiple pieces of evidence or historical records support this association, the failure mechanism is clear, the probability and severity of occurrence are high, and it is likely to lead to large-scale after-sales risks in the short term), which is the defect association result. The defect association result clearly points out the relationship between the specific high-risk feature value of each product model (such as the ZB1 model motherboard) and the specific defect cause (such as the power chip cold solder joint), and gives information such as the frequency of occurrence and the scope of impact, so as to facilitate the use of the technical, quality and supply chain teams and provide direct basis for subsequent steps.

[0115] S5. Conduct a post-sales risk assessment based on the defect correlation results to obtain the risk assessment results.

[0116] In this embodiment, the after-sales risk assessment includes:

[0117] Based on the defect association results obtained in step S4, and combined with the after-sales feature set and weighted failure rate of each product model, a risk assessment model is used to integrate the product's objective historical risk indicators with defect-level explanatory indicators to obtain a product risk score that comprehensively reflects quantity risk, economic risk, and defect risk. This score serves as a quantitative basis for subsequent steps and enhances the interpretability and credibility of after-sales data analysis.

[0118] Furthermore, the risk assessment model includes:

[0119]

[0120] in:

[0121] : No. The risk score for each product model; a higher value indicates a higher risk.

[0122] : No. The normalized value of the mean time to first failure for product model, ranging from 0 to 1;

[0123] : No. The defect association result indicator for a product model is used to indicate the impact of the defect association result on the risk. A strong association result indicates a high risk and the indicator is 1, while a weak association result indicates a low risk and the indicator is 0, in order to improve the risk score of the strong association result.

[0124] : weight coefficient of weighted failure rate, range from 0 to 1, sum of weight coefficients is 1, set according to enterprise's focus on risk assessment, if enterprise pays more attention to historical statistical risk (e.g. high rate of return, great impact on brand reputation), increase , for example, if an enterprise pays most attention to historical statistical risk, second attention to early risk, and least attention to defect explanation and root cause for new product after-sales risk, then is 0.5, is 0.3, is 0.2, so that the influence of historical statistical risk and early risk on risk score is greater, if an enterprise has many product models, but it is difficult to locate defects, and it wants to prioritize potential defect models, then appropriately reduce the weight of historical statistical risk and early risk, and increase the weight of defect explanation and root cause, is 0.4, is 0.2, is 0.4, so that if a product model has strong defect association results, its risk score will be significantly pulled up, which is more conducive to production investigation;

[0125] : weight coefficient of average time to first failure, range from 0 to 1, if enterprise pays more attention to early concentrated outbreak risk (e.g. high risk of a large number of returns in the early stage of listing), increase ;

[0126] : weight coefficient of defect association result, range from 0 to 1, if enterprise pays more attention to defect explanation and root cause (e.g. needs to cooperate with production or quality control department to quickly investigate defects), increase .

[0127] Further, the normalized value of the average time to first failure includes:

[0128]

[0129] wherein:

[0130] : average time to first failure of the model product, the smaller the value, the higher the risk, so it needs to be normalized so that the larger the normalized value, the higher the risk;

[0131] : minimum average time to first failure of all model products, used to normalize the average time to first failure to the range of 0 to 1;

[0132] : the maximum average time to first failure of all product models, used to normalize the average time to first failure to the range of 0 to 1;

[0133] The normalized value is inversely normalized by the maximum and minimum value of the average time to first failure, and the greater the normalized value, the greater the impact on the risk score.

[0134] The model describes how to calculate the risk score of each product model based on the weighted failure rate, average time to first failure and defect correlation result of each product model.

[0135] Traditional risk assessment methods usually only focus on failure rate or sales weighting, lack comprehensive consideration of defect information, and lack of interpretability. The present model introduces defect correlation result indicator and first failure time combined with weighted failure rate for comprehensive risk assessment, improves the comprehensiveness, interpretability and adaptability of risk assessment, makes the results more in line with business decision needs, helps technical and business departments to quickly identify risk causes, and supports more detailed after-sales management and resource allocation.

[0136] The use steps of the above model include:

[0137] Data acquisition: obtain the average time to first failure of each product model from step S1 , obtain the weighted failure rate of each product model from step S2 , obtain the defect correlation result of each product model from step S4 ;

[0138] Parameter calculation: according to the obtained data, the average time to first failure is inversely normalized to obtain the normalized value of the average time to first failure ;

[0139] Risk assessment: according to the obtained data and calculated parameters, the risk score is calculated combined with the weight coefficient , the higher the risk score, the greater the risk of the product model, and the risk score can be used for high-risk model warning, key investigation or resource scheduling.

[0140] S6, according to the risk assessment result, high-risk product identification is carried out to obtain the identification result.

[0141] In this embodiment, the high-risk product identification includes:

[0142] In step S5, the risk assessment result is obtained, which shows the risk size of each product model, but only the risk score value itself is not enough to meet the business needs, and it is also necessary to combine the actual threshold judgment or ranking identification to convert the risk assessment result into a product set, so as to facilitate decision makers to directly lock the key product model, guide subsequent investigation, recall or resource allocation.

[0143] The purpose of this step is to identify high-risk products according to the obtained risk assessment results, obtain an identification result, and output a clear high-risk product set, so as to improve the pertinence of after-sales risk management, concentrate resources on high after-sales or brand risk product models, provide a target list for subsequent root cause analysis, defect improvement, and production adjustment, and support management to make measures such as recall, technical investigation, or market intervention.

[0144] The implementation method of this step includes:

[0145] Identification strategy setting: aggregate the risk scores of all product models obtained in step S5, determine the identification rule according to the enterprise historical experience or risk preference, for example, sort the product models according to the risk scores from high to low, and take a certain percentage of product models in the front (for example, the enterprise historical data indicates that the product models with risk scores in the top 10% to 20% are more likely to cause after-sales risks, so take 15%), or set a fixed risk score threshold (for example, through analysis of the enterprise historical data, the product models with risk scores greater than 0.6 are more likely to cause after-sales problems in the future, so the threshold is taken as 0.6), and the product models greater than the threshold are regarded as high-risk;

[0146] Risk sorting and screening: sort the risk scores, and screen out high-risk models according to the selected rule to form a high-risk product set;

[0147] Device result output: output the high-risk product set, including the risk score and associated core explanation information (such as defect association state, average first failure time, etc.) of each model, and deliver the result to the decision-making department, technical quality department, production or market team for subsequent analysis and disposal.

[0148] Embodiment two:

[0149] The embodiment also provides a computer device suitable for the case of the after-sales data analysis method based on electronic product sales, including a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to realize the after-sales data analysis method based on electronic product sales as proposed in the above embodiment.

[0150] The embodiment also provides a storage medium having a computer program stored thereon, and the program is executed by a processor to realize the after-sales data analysis method based on electronic product sales as proposed in the above embodiment.

[0151] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved by WIFI, an operator network, NFC (Near Field Communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or can be a key, a trackball or a touchpad arranged on the shell of the computer device, or can be an external keyboard, a touchpad or a mouse, etc.

[0152] If the functions are implemented in the form of software function units and sold or used as independent products, the functions can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the parts of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the embodiments of the present application. The aforementioned storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0153] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a list of executable instructions for implementing logic functions, which can be specifically embodied in any computer readable medium for use by or in conjunction with an instruction execution system, device or apparatus, such as a computer-based system, a system including a processor or other system that can fetch and execute instructions from the instruction execution system, device or apparatus. For the purpose of this specification, the "computer readable medium" can be any device that can contain, store, communicate, propagate or transport programs for use by or in conjunction with the instruction execution system, device or apparatus, or in conjunction with these instruction execution systems, devices or apparatus.

[0154] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can also be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example, via an optical scanner, then compiled, interpreted, or otherwise processed, as necessary, and stored in a computer memory.

[0155] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the embodiments described above, various steps or methods can be implemented, for example, in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and in another embodiment, any of the following technologies, known in the art, or combinations thereof, can be used: discrete logic circuitry having logic gates for implementing logic functions upon an application of data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and the like.

[0156] It should be noted that the above examples are merely intended to illustrate the technical solutions of the present application and not to limit the same. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, and all such modifications or replacements should be included in the scope of the claims of the present application.

Claims

1. An after-sales data analysis method based on electronic product sales, characterized by, The method comprises the following steps: S1, constructing product after-sales features based on multi-source data to obtain an after-sales feature set; S2, performing fault evaluation according to the after-sales feature set to obtain a fault evaluation result; S3, extracting risk features according to the fault evaluation result to obtain high-risk features; S4, performing defect correlation analysis according to the high-risk features to obtain a defect correlation result; S5, performing after-sales risk evaluation according to the defect correlation result to obtain a risk evaluation result; S6, identifying high-risk products according to the risk evaluation result to obtain an identification result.

2. The method for analyzing after-sales data based on electronic product sales according to claim 1, characterized in that, The fault evaluation in the step S2 comprises: According to the obtained after-sales feature set of the product model, a quantifiable and comparable fault rate index is constructed by using a weighted fault model to quantify the business importance and economic impact of type faults, and the after-sales risk of the product model is comprehensively evaluated to obtain the fault evaluation result.

3. The method for analyzing after-sales data based on electronic product sales according to claim 2, characterized in that, The weighted fault model comprises: The weighted coefficient comprises: : the first weighted failure rate of the product of the model; : number of categories of fault types; : the first weighting coefficient of the class failure; : 1st : 2nd : 3rd : 1st Cumulative sales of a product model.

4. The method for analyzing after-sales data based on electronic product sales according to claim 3, characterized in that, The weighted coefficient is calculated by normalizing the average maintenance cost of various types of faults of various product models. The risk feature extraction in the step S3 comprises: : the first : the first : the average repair cost of the first : maximum of average repair cost over all product models for all failure types; According to the fault evaluation result, risk features are extracted, and through difference analysis, the association between the after-sales features of the product model and the high fault rate is found out, and the high-risk features that significantly contribute to the high fault rate are identified, thereby providing a basis for the subsequent steps.

5. The method for analyzing after-sales data based on electronic product sales according to claim 1, characterized in that, The defect correlation analysis in the step S4 comprises: Through defect correlation analysis, the obtained high-risk features are associated with defect causes, potential root causes are found out, defect cause hypotheses are obtained, and the defect correlation result is output based on historical defect knowledge and business processes, thereby providing a basis for the subsequent steps.

6. The method for analyzing after-sales data based on electronic product sales according to claim 1, wherein, The after-sales risk evaluation in the step S5 comprises: According to the obtained defect correlation result, the after-sales feature set of the product model and the weighted fault rate are combined, and a risk evaluation model is used to fuse the objective historical risk index of the product and the defect level interpretation index, thereby obtaining a product risk score that comprehensively reflects the quantity risk, economic risk and defect risk, as a quantitative basis for the subsequent steps.

7. The method for analyzing after-sales data based on electronic product sales according to claim 1, wherein, The risk evaluation model comprises: The normalized value of the average first failure time comprises:

8. The method for analyzing after-sales data based on electronic product sales according to claim 7, characterized in that, The normalized value is inversely normalized by the maximum and minimum values of the average first failure time, so that the greater the normalized value, the greater the impact on the risk score. The high-risk product identification in the step S6 comprises: : 1st Risk score for the model product; : the normalized value of the average time to first failure of the product of the model; : the normalized value of the average time to first failure of the product of the model; : the first defect association result indication item of the product of the model : weight coefficient of weighted failure rate; : weight coefficient of the average first failure duration; : weight coefficient of defect association result.

9. The method of claim 8, wherein, According to the obtained risk evaluation result, high-risk products are identified to obtain an identification result, and a clear high-risk product set is output, thereby improving the pertinence of after-sales risk management, providing a target list for subsequent root cause analysis, defect improvement and production adjustment, and supporting management to guide subsequent troubleshooting, recall and resource allocation. ​ : 1st average time to first failure for a model product; : minimum average time to first failure for all models; : maximum average time to first failure for all models; ​ 10. The method for analyzing after-sales data based on electronic product sales according to claim 1, wherein, ​ ​

Images