Product quality evaluation method and device based on artificial intelligence attention mechanism

Abstract

The application discloses a product quality evaluation method and device based on an artificial intelligence attention mechanism and belongs to the technical field of artificial intelligence. The method comprises the following steps: constructing a product quality evaluation index system; acquiring a plurality of training samples based on the quality evaluation index system; the training samples comprise input information corresponding to each evaluation index acquired for a sample product and a score of the sample product; the input information corresponding to each evaluation index in the training samples is taken as input, and the score is taken as output, so as to train a neural network model based on an attention mechanism by using the plurality of training samples, so that the neural network model automatically identifies more focused evaluation indexes and adjusts the attention weights of different evaluation indexes; and the trained neural network model is used for quality evaluation of a product to be evaluated. The application can automatically adjust the attention weights of different evaluation indexes and realize the objectivity and consistency in the evaluation process of the product to be evaluated.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to a product quality evaluation method and device based on an artificial intelligence attention mechanism. BACKGROUND

[0002] In the medical, rehabilitation and elderly care fields, the application of intelligent devices is in a rapid development stage. From wearable health monitoring devices, to large-scale rehabilitation training equipment, to intelligent elderly care service robots, these devices play a key role in improving medical efficiency, improving rehabilitation effect, and improving the quality of life of the elderly. However, the quality of these devices is directly related to the health and safety of the user, so it is crucial to scientifically and comprehensively evaluate the quality of these devices.

[0003] Traditional product quality evaluation methods mainly rely on manual experience and simple detection means, and mainly test the functionality of the product to determine whether the product functionality meets the established requirements, and calculate the evaluation result based on the assigned weights to achieve product quality evaluation.

[0004] However, the weight distribution is greatly influenced by subjective factors, making it difficult to ensure the objectivity and consistency of the evaluation results. Therefore, a more objective and accurate evaluation method is needed. SUMMARY

[0005] The present application provides a product quality evaluation method and device based on an artificial intelligence attention mechanism. The technical solution is as follows:

[0006] On the one hand, a product quality evaluation method based on an artificial intelligence attention mechanism is provided, the method comprising:

[0007] constructing a quality evaluation index system of a product; the product has a specified function; the quality evaluation index system includes a plurality of evaluation indexes;

[0008] obtaining a plurality of training samples based on the quality evaluation index system; the training samples include input information corresponding to each evaluation index obtained for a sample product, and a score for the sample product; the plurality of training samples are obtained from at least a plurality of different types of sample products, and the plurality of different types of sample products all have the specified function; the score is calculated from a first score and at least one second score; the first score is a score calculated using the input information according to a set scoring rule; and the evaluation index in the quality evaluation index system is a physical performance index; the second score is a score obtained by evaluating a non-performance index;

[0009] The input information corresponding to each evaluation index in the training sample is taken as input, and the score is taken as output, so as to train the neural network model based on the attention mechanism by using the plurality of training samples, so that the neural network model automatically identifies the evaluation index that pays more attention to, and adjusts the attention weight of different evaluation indexes;

[0010] The trained neural network model is used to perform quality evaluation on the product to be evaluated.

[0011] In a possible implementation, the sample product is at least one of a medical device, a rehabilitation device, and an elderly care device; and the specified function is at least one of the functions possessed by the medical device, the rehabilitation device, and the elderly care device.

[0012] In a possible implementation, the quality evaluation index system is constructed based on an analytic hierarchy process, and the quality evaluation index system includes at least two levels of evaluation indexes.

[0013] In a possible implementation, the use of the trained neural network model to perform quality evaluation on the product to be evaluated includes:

[0014] Based on the quality evaluation index system, input information corresponding to each evaluation index of the product to be evaluated is obtained;

[0015] The input information corresponding to each evaluation index of the product to be evaluated is input into the trained neural network model to obtain an output score, and the score is taken as a quality evaluation result of the product to be evaluated.

[0016] In a possible implementation, the method further includes: using the test sample and the trained neural network model to determine the size relationship of the weights allocated to the plurality of evaluation indexes in the quality evaluation index system by the neural network model; wherein the test sample includes input information corresponding to each evaluation index obtained for a test product;

[0017] After the use of the trained neural network model to perform quality evaluation on the product to be evaluated, the method further includes: if the score output by the neural network model for the product to be evaluated is less than a set score, determining at least one target evaluation index affecting the score based on the size relationship of the weights allocated to the plurality of evaluation indexes in the quality evaluation index system and the input information of each evaluation index of the product to be evaluated, and outputting improvement suggestions for the product to be evaluated according to the at least one target evaluation index.

[0018] On the other hand, a product quality evaluation device based on an artificial intelligence attention mechanism is provided, and the device includes:

[0019] A constructing unit is configured to construct a quality evaluation index system of a product; the product has a specified function; and the quality evaluation index system includes multiple evaluation indexes;

[0020] An obtaining unit is configured to obtain multiple training samples based on the quality evaluation index system; the training samples include input information corresponding to each evaluation index obtained for a sample product and a score of the sample product; the multiple training samples are obtained from multiple sample products of different types, and the multiple sample products of different types all have the specified function; the score is calculated from a first score value and at least one second score value; the first score value is a score value calculated from the input information according to a set scoring rule; and the evaluation index in the quality evaluation index system is a performance index that can be physically measured and evaluated; and the second score value is a score value calculated from a non-performance index;

[0021] A training unit is configured to use the input information corresponding to each evaluation index in the training samples as input and the score as output to train a neural network model based on an attention mechanism using the multiple training samples, so that the neural network model automatically identifies more focused evaluation indexes and adjusts attention weights of different evaluation indexes.

[0022] An evaluation unit is configured to use the trained neural network model to perform quality evaluation on a product to be evaluated.

[0023] In a possible implementation, the method further includes:

[0024] A first determining unit is configured to determine a size relationship of weights allocated to multiple evaluation indexes in the quality evaluation index system by using a test sample and the trained neural network model; and the test sample includes input information corresponding to each evaluation index obtained for a test product.

[0025] A second determining unit is configured to, after the evaluation unit is executed, determine at least one target evaluation index affecting the score based on the size relationship of the weights allocated to the multiple evaluation indexes in the quality evaluation index system and the input information of each evaluation index of the product to be evaluated, if the score output by the neural network model for the product to be evaluated is less than a set score value, and output improvement suggestions for the product to be evaluated according to the at least one target evaluation index.

[0026] On the other hand, a computer device is provided, which includes a memory and a processor; the memory is configured to store a computer program; and the processor is configured to execute the computer program stored in the memory to implement the steps of the product quality evaluation method based on the artificial intelligence attention mechanism.

[0027] In another aspect, a computer readable storage medium is provided, and the storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the product quality evaluation method based on the artificial intelligence attention mechanism.

[0028] In another aspect, a computer program product is provided, and the computer program product comprises a computer program, and the computer program is executed by a processor to implement the steps of the product quality evaluation method based on the artificial intelligence attention mechanism.

[0029] The technical solutions provided by the present application can bring at least the following beneficial effects:

[0030] By constructing the product quality evaluation index system, the training sample products used for training the neural network model and the products to be evaluated are evaluated according to the same quality evaluation index system, and the input information of each evaluation index in the quality evaluation index system can be obtained by physical evaluation, thereby ensuring the objectivity and consistency of the data source. The training sample is obtained from at least one sample product of different types to ensure that the trained neural network model can evaluate the quality of products with specified functions and different types. The score output in the training sample is obtained by the first score and at least one second score. The first score is calculated by the input information according to the set scoring rule, and the second score is calculated by the non-performance index. That is, the second score is not calculated by the input information, but is scored by external factors. When the neural network model is used to evaluate the quality of new products, the performance index and the non-performance index can be evaluated based on the input information of the performance index. In the process of training the neural network model, the neural network model can map the input information and the output score to automatically identify the evaluation index that needs to be paid more attention to, so as to automatically adjust the attention weight of different evaluation indexes, and further realize the objectivity and consistency in the evaluation process of the product to be evaluated. BRIEF DESCRIPTION OF DRAWINGS

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0032] Figure 1 is a product quality evaluation method based on an artificial intelligence attention mechanism provided by an embodiment of the present application;

[0033] Figure 2It is an embodiment of the present application to provide a product quality evaluation device structure diagram based on an artificial intelligence attention mechanism.

[0034] Figure 3 It is an embodiment of the present application to provide a hardware architecture diagram of a computer device. DETAILED DESCRIPTION

[0035] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0036] Please refer to Figure 1 The embodiment of the present application provides a product quality evaluation method based on an artificial intelligence attention mechanism, which comprises the following steps:

[0037] Step 100, a quality evaluation index system of a product is constructed; the product has a specified function; the quality evaluation index system comprises a plurality of evaluation indexes;

[0038] Step 102, a plurality of training samples are obtained based on the quality evaluation index system; the training sample comprises input information corresponding to each evaluation index obtained for a sample product, and a score of the sample product; the plurality of training samples are obtained from at least a plurality of different types of sample products, and the plurality of different types of sample products all have the specified function; the score is calculated from a first score value and at least one second score value; the first score value is a score value calculated by using the input information according to a set scoring rule; and the evaluation index in the quality evaluation index system is a physical performance index;

[0039] Step 104, the input information corresponding to each evaluation index in the training sample is taken as the input, and the score is taken as the output, so as to train a neural network model based on an attention mechanism by using the plurality of training samples, so as to make the neural network model automatically identify more attention evaluation indexes and adjust the attention weight of different evaluation indexes;

[0040] Step 106, the trained neural network model is used to evaluate the quality of a product to be evaluated.

[0041] In the embodiment of the present application, by constructing the product quality evaluation index system, the training sample products used for training the neural network model and the products to be evaluated are evaluated according to the same quality evaluation index system, and the input information of each evaluation index in the quality evaluation index system can be obtained by physical evaluation, thereby ensuring the objectivity and consistency of the data source. The plurality of training samples are obtained from at least a plurality of different types of sample products, so as to ensure that the trained neural network model can evaluate the quality of products with specified functions and different types. The score output in the training sample is obtained by the first score and at least one second score. The first score is calculated by the input information according to the set scoring rule, and the second score is calculated by the non-performance index. That is, the second score is not calculated by the input information, but is scored by external factors. When the neural network model is used to evaluate the quality of new products, the performance index and the non-performance index can be evaluated based on the input information of the performance index. In the process of training the neural network model, the neural network model can map the input information and the output score to automatically identify the evaluation index that needs to be paid more attention to, so as to automatically adjust the attention weight of different evaluation indexes, and further realize the objectivity and consistency in the process of evaluating the products to be evaluated.

[0042] The following describes Figure 1 The execution mode of each step shown.

[0043] First, for step 100, a product quality evaluation index system is constructed. The product has a specified function. The quality evaluation index system includes a plurality of evaluation indexes.

[0044] In the embodiment of the present application, in order to ensure the consistency in the process of product evaluation, a product quality evaluation index system can be constructed. The quality evaluation index system is constructed for a product with a specified function, and the quality evaluation index system includes a plurality of evaluation indexes.

[0045] The product quality evaluation method of the embodiment of the present application is used to evaluate the quality of products, especially the quality of medical devices, rehabilitation devices and elderly care devices. Therefore, the specified function of the product can be at least one of the functions of the medical devices, rehabilitation devices and elderly care devices. For example, the specified function can be blood oxygen measurement function, blood pressure measurement function, heart rate measurement function, etc.

[0046] When constructing the quality evaluation index system, industry indexes, standard indexes, national indexes, industry indexes, etc. of the product can be used to construct.

[0047] In an implementation manner, the quality evaluation index system can be multiple evaluation indexes in a parallel relationship, that is, multiple evaluation indexes in the quality evaluation index system are located at the same level.

[0048] In another implementation manner, the quality evaluation index system can be constructed based on an analytic hierarchy process, and the quality evaluation index system includes evaluation indexes in at least two levels.

[0049] In the medical device quality evaluation, the quality evaluation target of the device is decomposed into criteria layers such as functionality, safety, and ease of use through the construction of a hierarchical structure model, and is further subdivided into specific index layers, so that the evaluator can compare each level of factors with each other to determine the relative importance weight, and then obtain the evaluation result. However, in the traditional way, the analytic hierarchy process also has obvious shortcomings: on the one hand, the construction of the judgment matrix depends on the subjective judgment of experts, and the knowledge structure, experience level, and personal preference of the experts will affect the accuracy of the judgment matrix, resulting in strong subjectivity in weight distribution; for example, when determining the relative importance of the functionality and safety of the medical device, different experts may give different judgments, making the evaluation result lack sufficient objectivity and reliability; on the other hand, the analytic hierarchy process has limited data processing capability and is difficult to deal with massive and complex multi-source data; in the case of explosive growth of intelligent device data, a large amount of sensor data and user feedback data generated during device operation, the analytic hierarchy process cannot fully mine the potential information in the data, and it is difficult to comprehensively and accurately reflect the quality status of the device.

[0050] Therefore, in the embodiment of the present application, the neural network model is used to automatically identify the weight of the evaluation index in combination with the artificial intelligence attention mechanism, which not only reduces the calculation amount but also reduces the participation of artificial subjectivity, and improves the objectivity and consistency of product quality evaluation.

[0051] For example, the quality evaluation index system includes three levels of evaluation indexes, that is, the quality evaluation index system includes first-level indexes, second-level indexes, and third-level indexes. In an implementation manner, the first-level indexes at least include one or more evaluation indexes in the following: appearance structure, aging suitability, functionality, safety, signal quality, environmental adaptability, and user experience. Among them, the appearance structure focuses on the overall design and material selection of the device; the aging suitability focuses on the adaptation of the device to the use habits and physical characteristics of the elderly; the functionality is used to measure the implementation and effect of each function of the device; the safety is used to ensure that the device does not harm the user during use; the environmental adaptability is used to evaluate the stable operation ability of the device under different environmental conditions; and the user experience is used to reflect the subjective feeling of the user to the device.

[0052] Preferably, the constructed quality evaluation index system at least includes an evaluation index of an aging suitability level. The aging suitability level, as the name implies, is a level of whether the product is suitable for use by the elderly. By using the evaluation index, the degree of adaptation of the product to the use habits and physical characteristics of the elderly can be focused on, not only the safety of the product, but also the practical value of the product can be evaluated from the actual needs of the elderly, so as to meet the actual needs of quality evaluation of medical devices, rehabilitation devices and elderly care devices.

[0053] For example, the secondary indicators of the appearance structure in the above-mentioned primary indicators can at least include one or more of the surface, the display screen, the watchband, the switch or the key and the logo; the secondary indicators of the aging suitability level in the above-mentioned primary indicators can at least include one or more of the perceptibility, the operability, the understandability, the compatibility, the safety, the remote assistance and the emergency application; the tertiary indicators of the perceptibility in the secondary indicators can at least include one or more of the font size adjustment, the line spacing, the contrast, the color and the verification code; the tertiary indicators of the operability in the secondary indicators can at least include one or more of the screen display magnification, the input keyboard, the screen touch interaction, the voice broadcast, the component focus size, the gesture operation, the operation time and the floating window.

[0054] Then, the step 102 of obtaining a plurality of training samples based on the quality evaluation index system and the step 104 of taking the input information corresponding to each evaluation index in the training samples as input and the score as output to train the neural network model based on the attention mechanism with the plurality of training samples to make the neural network model automatically identify the evaluation index that needs to be paid more attention to and adjust the attention weight of different evaluation indexes are simultaneously explained.

[0055] In the embodiment of the present application, the product quality is evaluated by the neural network model in the manner of scoring and outputting the input information. In order to ensure the objectivity and consistency of the trained neural network model in evaluating the product quality, the training samples used for training the neural network model are also obtained based on the quality evaluation index system.

[0056] The training samples include the input information corresponding to each evaluation index obtained for the sample product and the score of the sample product, and the plurality of training samples are obtained from a plurality of different types of sample products, and the plurality of different types of sample products all have the specified function.

[0057] In one embodiment of the present application, the score of the sample product in the training sample is calculated from a first score and at least one second score.

[0058] The first score is a score calculated according to a set scoring rule using the input information; and the evaluation indexes in the quality evaluation index system are performance indexes that can be physically measured.

[0059] The second score is a score obtained by using non-performance indexes for evaluation.

[0060] The first score and the second score are described below.

[0061] First, the first score is described.

[0062] Since the evaluation indexes in the quality evaluation index system are all performance indexes that can be physically measured, the input information corresponding to the evaluation indexes is all objective information that can be directly measured, thereby ensuring the objectivity in the evaluation process.

[0063] In the embodiment of the application, the scoring rule can be set in advance, and weights are configured for different evaluation indexes based on expert experience, so as to calculate the first score using the input information.

[0064] The input information can be collected through channels such as sensors, user feedback, and device operation logs. In terms of sensor data collection, high-precision and stable sensors can be selected to ensure the accuracy and reliability of the collected data; in terms of user feedback collection, an online and offline combination method can be used to expand the feedback collection range and improve the representativeness of the feedback data; and in terms of device operation logs, professional data analysis tools can be used to deeply mine the log data and extract valuable information.

[0065] Taking the aging level and its secondary index "perceptibility" as an example, the weight configured for the aging level is 20%, and the scoring method of each tertiary index of "perceptibility" can include:

[0066] Font size adjustment: according to the scoring rule of "interface font size can be set, adjusted with system settings, or mobile application has font size setting options", score for this evaluation index from the optional scoring range (3-5 points);

[0067] Line spacing: according to the scoring rule of "the line spacing of the text in the paragraph is at least 1.3 times, and the paragraph spacing is at least 1.3 times larger than the line spacing, while considering the mobile application application scenarios and display effects", score for this evaluation index from the optional scoring range (3-5 points);

[0068] Contrast: according to the scoring rule of "the contrast between text / text image presentation, icons and other elements is at least 4.5:1", score for this evaluation index from the optional scoring range (1-2 points);

[0069] Color: According to the scoring rule of "text color is not the only means to distinguish visual elements in terms of conveying information, indicating actions, etc.", score this evaluation indicator from the optional scoring range (1-2 points);

[0070] Verification code: According to the scoring rule of "if there is a non-text verification code in the mobile application that is not easy for the elderly to understand, a text verification code that can be understood by the elderly should be provided", score this evaluation indicator from the optional scoring range (1-2 points).

[0071] To further improve the objectivity of the input information, the optional scoring range of each evaluation indicator can be two scores, when the scoring rule of the evaluation indicator is not met, the lowest score is directly taken as the score of the evaluation indicator, when the scoring rule of the evaluation indicator is met, the highest score is directly taken as the score of the evaluation indicator. For example, the optional scoring range of the contrast evaluation indicator is 1 point and 2 points, if the contrast meets its scoring rule, the score is 2 points, otherwise, the score is 1 point.

[0072] When each item of the three-level indicator is scored based on the corresponding input information, in an implementation, the calculation method of the first score value can include: adding the scores of the three-level indicators to obtain the score of the corresponding two-level indicator, then adding the scores of each two-level indicator to obtain the score of the corresponding one-level indicator, using the score of each one-level indicator and its weight, the weighted sum of multiple one-level indicators can be obtained, and the weighted sum is taken as the first score value.

[0073] Then the second score value is described.

[0074] In order to give a multi-dimensional evaluation of the product, in addition to the performance indicators, non-performance indicators are also needed for evaluation. This is because, in addition to product performance, market feedback is more important to how good the product quality is, so the non-performance indicators at least include market feedback evaluation indicators to use market feedback to comprehensively evaluate the product. In addition, the non-performance indicators can also include expert evaluation indicators, etc.

[0075] For the market feedback evaluation indicators, at least one of the market selling price, market sales volume, and market score can be included. Based on these non-performance indicators, the second score value of the corresponding non-performance indicator can be calculated according to the set calculation rule.

[0076] It can be seen that, in order to obtain the score as the output in the training sample, not only the input information needs to be combined, but also the data other than the input information needs to be combined to obtain the score, so as to realize the quality evaluation for the subsequent product without the support of the data other than the input information. For example, for a new product before entering the market, since there is no market feedback evaluation index, the performance index that can be physically evaluated can be used as the input information to output the score of the comprehensive quality evaluation.

[0077] Then, based on the description of the above embodiment, the training sample can be obtained for the sample product.

[0078] In order to ensure that the trained neural network model can evaluate the quality of different types of products, the plurality of training samples used to train the neural network model are at least obtained from a plurality of sample products of different types, and the plurality of sample products of different types all have the specified function.

[0079] When the product quality evaluation method provided by the embodiment of the application is applied to medical equipment, rehabilitation equipment or elderly care equipment, the sample product is at least one of medical equipment, rehabilitation equipment and elderly care equipment; and the specified function of the sample product is at least one of the functions possessed by the medical equipment, the rehabilitation equipment and the elderly care equipment.

[0080] It should be noted that, if the specified function is a plurality of functions, in one implementation, the sample product can be a product having at least one of the plurality of functions. For example, the specified function is a blood pressure measurement function and a blood oxygen measurement function, and then the sample product can be a product only having the blood pressure measurement function or a product only having the blood oxygen measurement function.

[0081] The types of the sample product can include smart watches, wristbands, sports earphones, smart phones, finger clip blood oxygen meters, medical blood oxygen monitors, etc. Thus, the sample product covers more types, so that the trained neural network model can evaluate the quality of products of different types.

[0082] After obtaining the plurality of training samples, the input information corresponding to each evaluation index in the training sample is taken as the input, and the score is taken as the output to train the neural network model based on the attention mechanism. The neural network model can deeply analyze and extract features for the input information of different evaluation indexes, learn the mapping relationship between the evaluation index input information and the score, automatically identify the evaluation index that pays more attention, and then automatically adjust the attention weight of different evaluation indexes. That is, although the weights of the evaluation indexes are artificially set based on expert experience when the first score is calculated, the weights of the evaluation indexes change after the score obtained by combining the first score with the second score, and the change can better reflect the real weight of each evaluation index. The adjusted weight is more objective and accurate.

[0083] In addition, when the quality evaluation index system is constructed based on the analytic hierarchy process, there are numerous evaluation indexes at each level involved in the quality evaluation index system, and there are complex correlation relationships between evaluation indexes at different levels. According to the description of the above embodiment, it can be known that the traditional analytic hierarchy process not only lacks sufficient objectivity and reliability, but also cannot accurately mine which information in the data is more important in the case of large data volume, resulting in poor accuracy of the product quality evaluation result. In the embodiment of the present application, the artificial intelligence attention mechanism is combined, and the neural network model is used to automatically identify the weight of each evaluation index, which not only reduces the calculation amount and reduces the participation of artificial subjectivity, but also improves the objectivity and consistency of product quality evaluation.

[0084] In the training process of the neural network model, the back propagation algorithm can be used to accurately adjust the model parameters, and various optimization algorithms such as stochastic gradient descent, Adagrad, Adadelta, etc. can be flexibly used. According to the training effect and convergence of the model, the learning rate is dynamically adjusted, the model performance is continuously optimized, and the accuracy, stability and generalization ability of the model are ensured.

[0085] Finally, for step 106, the trained neural network model is used to evaluate the quality of the product to be evaluated.

[0086] The product to be evaluated has a specified function.

[0087] Specifically, based on the quality evaluation index system, the input information corresponding to each evaluation index of the product to be evaluated can be obtained, and the input information corresponding to each evaluation index can be input into the trained neural network model to obtain the output score. The score is taken as the quality evaluation result of the product to be evaluated.

[0088] It can be seen that even if the product to be evaluated only has performance index input information, the neural network model can also be used to obtain a comprehensive score with non-performance index.

[0089] Further, for the trained neural network model, the test sample and the trained neural network model can be used to determine the size relationship of the weight allocated to the multiple evaluation indexes in the quality evaluation index system by the neural network model. The test sample includes the input information corresponding to each evaluation index obtained for the test product. Specifically, the position disturbance method, SHAP (machine learning model interpretability) method, gradient change and other methods can be used to determine a number of evaluation indexes with the highest allocated weight.

[0090] The determination process is explained by using the position disturbance method. The input information corresponding to each evaluation index in the test sample is input into the trained neural network model to obtain the output score; the input information of two evaluation indexes in the test sample is exchanged, and then the exchanged input information is input into the trained neural network model to obtain the score after information exchange; if the difference between the score after information exchange and the score before information exchange exceeds a set threshold, it is determined that at least one of the two evaluation indexes after information exchange has a higher allocated weight. Based on this, the position disturbance method can be used to determine the size relationship of the allocated weights of the multiple evaluation indexes in the quality evaluation index system.

[0091] Further, after determining the neural network model as the size relationship of the allocated weights of the multiple evaluation indexes in the quality evaluation index system, if the score output by the neural network model for the product to be evaluated is less than a set score, it indicates that the product needs to be improved in some performance. Based on the size relationship of the allocated weights of the multiple evaluation indexes in the quality evaluation index system and the input information of each evaluation index, at least one target evaluation index affecting the score is determined, and improvement suggestions are output for the product according to the at least one target evaluation index.

[0092] The embodiments of the present application are described below with actual cases.

[0093] Smart elderly bed mattress case.

[0094] A quality evaluation index system is constructed for the sleep monitoring function, and the construction method is realized based on the analytic hierarchy process. The first-level indexes include appearance structure, aging level, functionality, safety, environmental reliability, and user experience; the second-level indexes are subdivided into sleep monitoring function, health warning function, etc. under functionality; the aging level is subdivided into whether the bed mattress height is convenient for the elderly to get on and off the bed, whether the operation is simple and easy to understand, etc.; the third-level indexes are subdivided into sleep stage recognition accuracy, heart rate monitoring error, etc. under sleep monitoring function.

[0095] Multiple training samples are obtained for different models of smart elderly bed mattresses. For the input information of each evaluation index, the acquisition method can include: collecting data such as user's sleep posture and turning over times through the pressure sensor on the bed mattress; collecting information such as comfort and operation convenience of the bed mattress through user feedback; obtaining data such as bed mattress fault records and running time from device running logs; and calculating a first score using the obtained input information, and then calculating a second score using the performance index, and taking the sum of the first score and the second score as the score of the training sample.

[0096] The neural network model based on the artificial intelligence attention mechanism is trained, and then the input information of the to-be-evaluated intelligent pension mattress is input into the trained neural network model to obtain a comprehensive score, the comprehensive score corresponds to a quality rating of excellent, and an improvement suggestion for optimizing the edge design of the mattress is given.

[0097] Please refer to Figure 2 The embodiment of the present application provides a product quality evaluation device based on an artificial intelligence attention mechanism, which comprises:

[0098] The construction unit 200 is configured to construct a quality evaluation index system of a product; the product has a specified function; and the quality evaluation index system comprises a plurality of evaluation indexes.

[0099] The acquisition unit 202 is configured to acquire a plurality of training samples based on the quality evaluation index system; the training samples comprise input information corresponding to each evaluation index acquired for a sample product, and a score obtained for the sample product; the plurality of training samples are obtained from at least a plurality of different types of sample products, and the plurality of different types of sample products all have the specified function; the score is calculated from a first score value and at least one second score value; the first score value is a score value calculated by using the input information according to a set scoring rule; and the evaluation index in the quality evaluation index system is a physical performance index; and the second score value is a score value obtained by evaluating a non-performance index;

[0100] The training unit 204 is configured to use the input information corresponding to each evaluation index in the training samples as input and the score as output to train a neural network model based on the attention mechanism by using the plurality of training samples, so that the neural network model automatically identifies more attention evaluation indexes and adjusts the attention weights of different evaluation indexes.

[0101] The evaluation unit 206 is configured to perform quality evaluation on a to-be-evaluated product by using the trained neural network model.

[0102] In an embodiment of the present application, the sample product is at least one of a medical device, a rehabilitation device and an elderly care device; and the specified function is at least one of the functions possessed by the medical device, the rehabilitation device and the elderly care device.

[0103] In an embodiment of the present application, the quality evaluation index system is constructed based on the analytic hierarchy process, and the quality evaluation index system comprises at least two levels of evaluation indexes.

[0104] In an embodiment of the present application, the evaluation unit is specifically configured to: based on the quality evaluation index system, obtain input information corresponding to each evaluation index for the product to be evaluated; and input the input information corresponding to each evaluation index for the product to be evaluated into the trained neural network model to obtain an output score, and take the score as the quality evaluation result of the product to be evaluated.

[0105] In an embodiment of the present application, the device can further include:

[0106] The first determination unit is configured to determine the size relationship of the weights allocated to the multiple evaluation indexes in the quality evaluation index system by using the test sample and the trained neural network model; wherein the test sample includes input information corresponding to each evaluation index for a test product.

[0107] The second determination unit is configured to, when the score output by the neural network model for the product to be evaluated is less than a set score, determine at least one target evaluation index affecting the score based on the size relationship of the weights allocated to the multiple evaluation indexes in the quality evaluation index system and the input information of each evaluation index for the product to be evaluated, and output improvement suggestions for the product to be evaluated according to the at least one target evaluation index.

[0108] It should be noted that the product quality evaluation device based on the artificial intelligence attention mechanism provided in the above embodiments is only exemplified by the division of the above functional modules, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the product quality evaluation device based on the artificial intelligence attention mechanism provided in the above embodiments and the product quality evaluation method based on the artificial intelligence attention mechanism embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0109] Embodiments of the present application also provide a computer device, which refers to Figure 3 The computer device includes a processor and a memory, and the memory stores at least one instruction, at least one program, a code set or an instruction set. The at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the product quality evaluation method based on the artificial intelligence attention mechanism provided by each method embodiment.

[0110] The embodiment of the present application further provides a computer readable storage medium, and at least one instruction, at least one program, a code set or an instruction set are stored on the computer readable storage medium, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by a processor to implement the product quality evaluation method based on the artificial intelligence attention mechanism provided in each method embodiment.

[0111] The embodiment of the present application further provides a computer program product, and the computer program product comprises a computer program, and a processor of a computer device reads the computer program from a computer readable storage medium, and the processor executes the computer program, so that the computer device executes the product quality evaluation method based on the artificial intelligence attention mechanism described in any one of the above embodiments.

[0112] For the convenience of description, the above system or device is described as various modules or units in terms of functions respectively. Of course, functions of each unit can be implemented in one or more software and / or hardware in the implementation of the present application.

[0113] From the above description of the embodiments, those skilled in the art can clearly understand that the present application can be implemented by means of software and necessary universal hardware platforms. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which can be stored in a storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments of the present application.

[0114] Finally, it should be noted that, in this document, relational terms such as first and second, and the like can only be used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that these entities or operations exist in any actual relationship or order. Moreover, the terms "comprise", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a list of elements does not exclude other elements not explicitly listed, or other elements inherent to such process, method, article or device. Without more limitations, an element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article or device including the element.

[0115] The above merely describes the preferred embodiments of the present application, and it should be pointed out that, for those skilled in the art, some improvements and refinements can be made without departing from the principles of the present application, and these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A product quality evaluation method based on artificial intelligence attention mechanism, characterized in that, The method is used for comprehensive quality evaluation of new products before they are launched on the market, when there are no market feedback evaluation indicators. Construct a quality evaluation index system for the product; the product has specified functions; the quality evaluation index system includes multiple evaluation indicators; the constructed quality evaluation index system includes at least the evaluation indicator of age-friendliness level, in order to focus on the degree to which the product is adapted to the usage habits and physical characteristics of the elderly. Based on the aforementioned quality evaluation index system, multiple training samples are obtained. Each training sample includes input information corresponding to each evaluation index for a sample product, and a score for that sample product. The multiple training samples are obtained from at least multiple sample products of different types, and all of these sample products possess the specified function. The score is calculated from a first score and at least one second score. The first score is calculated using the input information according to a set scoring rule. The evaluation indexes in the quality evaluation index system are physically measurable performance indicators. The second score is obtained using non-performance indicators. These non-performance indicators include at least market feedback evaluation indicators. Market feedback evaluation indicators include at least one of market selling price, market sales volume, and market rating. The sample product is at least one of medical equipment, rehabilitation equipment, and elderly care equipment. The specified function is at least one of the functions found in medical equipment, rehabilitation equipment, and elderly care equipment. The input information corresponding to each evaluation indicator in the training samples is used as input and the score is used as output. The neural network model based on the attention mechanism is trained using the multiple training samples so that the neural network model can automatically identify the evaluation indicators that are of more interest and adjust the attention weights of different evaluation indicators. Based on the aforementioned quality evaluation index system, input information corresponding to each physically measurable performance index is obtained for the product to be evaluated; the input information corresponding to each physically measurable performance index obtained for the product to be evaluated is input into the trained neural network model to obtain the output score, and the score is used as the quality evaluation result of the product to be evaluated; the product to be evaluated has a specified function.

2. The method according to claim 1, characterized in that, The quality evaluation index system is constructed based on the analytic hierarchy process (AHP) and includes evaluation indicators at least two levels.

3. The method according to any one of claims 1-2, characterized in that, It also includes: using test samples and a trained neural network model to determine the relative weights of multiple evaluation indicators in the quality evaluation indicator system assigned by the neural network model; wherein, the test samples include input information corresponding to each evaluation indicator obtained for the test product; After the trained neural network model is used to evaluate the quality of the product to be evaluated, the method further includes: if the score output by the neural network model for the product to be evaluated is less than a set score, then based on the weight relationship of multiple evaluation indicators in the quality evaluation index system and the input information of each evaluation indicator for the product to be evaluated, at least one target evaluation indicator that affects the score is determined, and improvement suggestions are output for the product to be evaluated based on the at least one target evaluation indicator.

4. A product quality evaluation device based on an artificial intelligence attention mechanism, characterized in that, The apparatus is used for comprehensive quality evaluation of new products before they are launched on the market, in situations where there are no market feedback evaluation indicators. A construction unit is used to construct a quality evaluation index system for a product; the product has specified functions; the quality evaluation index system includes multiple evaluation indicators; the constructed quality evaluation index system includes at least one evaluation indicator for age-friendliness, in order to focus on the degree to which the product is adapted to the usage habits and physical characteristics of the elderly. The acquisition unit is used to acquire multiple training samples based on the quality evaluation index system. The training samples include input information corresponding to each evaluation index for each sample product, and a score for the sample product. The multiple training samples are obtained from at least multiple sample products of different types, and all sample products of different types have the specified function. The score is calculated from a first score and at least one second score. The first score is calculated using the input information according to a set scoring rule. The evaluation indexes in the quality evaluation index system are physically measurable performance indicators. The second score is obtained using non-performance indicators. These non-performance indicators include at least market feedback evaluation indicators. Market feedback evaluation indicators include at least one of market selling price, market sales volume, and market rating. The sample product is at least one of medical equipment, rehabilitation equipment, and elderly care equipment. The specified function is at least one of the functions found in medical equipment, rehabilitation equipment, and elderly care equipment. The training unit is used to take the input information corresponding to each evaluation index in the training samples as input and the score as output, so as to train the attention-based neural network model using the multiple training samples, so that the neural network model can automatically identify the evaluation index that is more important and adjust the attention weight of different evaluation indexes. The evaluation unit is used to obtain input information corresponding to each physically measurable performance indicator for the product to be evaluated based on the quality evaluation index system; input the input information corresponding to each physically measurable performance indicator for the product to be evaluated into the trained neural network model to obtain the output score, and use the score as the quality evaluation result of the product to be evaluated; the product to be evaluated and the specified function.

5. The apparatus according to claim 4, characterized in that, Also includes: The first determining unit is used to determine the relative weights of multiple evaluation indicators in the quality evaluation indicator system assigned by the neural network model to the test samples and the trained neural network model; wherein, the test samples include input information corresponding to each evaluation indicator obtained for the test product; The second determining unit is used to determine at least one target evaluation indicator that affects the score if the score output by the neural network model for the product to be evaluated is less than a set score after the evaluation unit has been executed. This is based on the weight relationship of multiple evaluation indicators in the quality evaluation index system and the input information of each evaluation indicator for the product to be evaluated. The unit then outputs improvement suggestions for the product to be evaluated based on the at least one target evaluation indicator.

6. A computer device, characterized in that, The computer device includes a memory and a processor. The memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory to implement the steps of the method according to any one of claims 1-3.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method described in any one of claims 1-3.

8. A computer program product, characterized in that, Includes a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1-3.

Images