Data processing method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of data processing, and discloses a data processing method and device, electronic equipment and a storage medium, and the method comprises the steps: respectively determining an analysis sub-object in each feedback statement, and an initial score corresponding to each analysis sub-object; according to the initial score corresponding to each analysis sub-object and the comprehensive score corresponding to the feedback statement to which each analysis sub-object belongs, calculating to obtain a target score corresponding to each analysis sub-object in each feedback statement; and according to the target scores corresponding to all the analysis sub-objects, calculating to obtain a net recommendation value of the target object. According to the method, the object keyword in the user feedback statement is analyzed to obtain the multiple analysis sub-objects, the initial scores for different analysis sub-objects in the complex feedback statement are obtained, the initial score of each analysis sub-object is corrected, the target score of each analysis sub-object is obtained, and the calculation basis of the net recommendation value is more accurate.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and in particular to a data processing method, apparatus, electronic device, and storage medium. Background Art

[0002] With the complex development of Internet business systems, feedback analysis technology has become an important basis for object optimization. In related technologies, the Net Promoter Score (NPS) is generally used as an important parameter for determining whether to recommend a target object, that is, based on a comprehensive score corresponding to the target object, the net promoter score of the target object is directly determined.

[0003] However, in the process of assigning a comprehensive score to the target object, there are many interference factors, which easily lead to errors in the comprehensive score obtained by scoring, resulting in inaccurate net promoter scores of the determined target objects and unable to objectively and accurately characterize the recommendation degree for the target objects. Summary of the Invention

[0004] In view of the above problems, this application provides a data processing method, apparatus, electronic device, and storage medium, which are used to solve the problem of inaccurate calculation of net promoter scores in the prior art.

[0005] According to one aspect of this application, a data processing method is provided. The method includes: for each object keyword included in each feedback statement, respectively determining the analysis sub-object in each feedback statement, and determining the initial score corresponding to each analysis sub-object according to the description keywords associated with the analysis sub-object in the feedback statement to which it belongs; where each feedback statement is a description statement for the same target object, and each feedback statement corresponds to a comprehensive score for the target object; calculating the target score corresponding to each analysis sub-object in each feedback statement according to the initial score corresponding to each analysis sub-object and the comprehensive score corresponding to the feedback statement to which each analysis sub-object belongs; calculating the net promoter score of the target object according to the target scores corresponding to all analysis sub-objects.

[0006] In an optional manner, the method further includes: traversing multiple feedback statements, and using the traversed feedback statement as the target feedback statement; if the matching coefficient between the target candidate word segment and the preset noun is greater than the first preset matching threshold, determining that the target candidate word segment is the target object keyword of the target feedback statement; where the target candidate word segment is any candidate word segment determined from the target feedback statement; if the adjacent candidate word segment of the target object keyword has a matching coefficient greater than the second preset matching threshold with the preset description word, determining that the adjacent candidate word segment is the description keyword associated with the analysis sub-object corresponding to the target object keyword.

[0007] In an alternative manner, determining the initial score corresponding to each analysis sub-object according to the description keywords associated with each analysis sub-object in the corresponding feedback statement includes: traversing each analysis sub-object, and taking the traversed analysis sub-object as the target analysis sub-object. The description keyword associated with the target analysis sub-object in the corresponding feedback statement is the target description keyword, and the comprehensive score corresponding to the feedback statement to which the target analysis sub-object belongs is the target comprehensive score; if the attribute value of the target description keyword is greater than the preset threshold, the initial score corresponding to the target analysis sub-object is: the score with the smallest difference from the target comprehensive score in the first score interval; if the attribute value of the target description keyword is equal to the threshold, the initial score corresponding to the target analysis sub-object is: the score with the smallest difference from the target comprehensive score in the second score interval; if the attribute value of the target description keyword is less than the threshold, the initial score corresponding to the target analysis sub-object is: the score with the smallest difference from the target comprehensive score in the third score interval; wherein, the minimum value of the first score interval is greater than the maximum value of the second score interval, and the minimum value of the second score interval is greater than the maximum value of the third score interval.

[0008] In an alternative manner, calculating the target score corresponding to each analysis sub-object in each feedback statement according to the initial score corresponding to each analysis sub-object and the comprehensive score corresponding to the feedback statement to which each analysis sub-object belongs includes: for multiple analysis sub-objects belonging to the same feedback statement, calculating the average initial score to determine the average initial score corresponding to each feedback statement and the target score interval to which each average initial score belongs. The target score interval is any one of the first score interval, the second score interval, and the third score interval; if the difference between the average initial score and the comprehensive score corresponding to the focus feedback statement is greater than the preset error threshold, adjusting the initial score corresponding to the focus analysis sub-object based on the target score interval to determine the target score corresponding to each analysis sub-object in each feedback statement; wherein, the focus analysis sub-object is the analysis sub-object in the focus feedback statement and whose initial score is within the target score interval.

[0009] In an alternative manner, calculating the net recommendation value of the target object according to the target scores corresponding to all analysis sub-objects includes: calculating the first proportion of the number of positive analysis sub-objects to the number of all analysis sub-objects, and the second proportion of the number of negative analysis sub-objects to the number of all analysis sub-objects; wherein, the positive analysis sub-object is the analysis sub-object whose target score is within the first score interval, and the negative analysis sub-object is the analysis sub-object whose target score is within the third score interval; determining the net recommendation value of the target object based on the difference between the first proportion and the second proportion.

[0010] In an alternative manner, each analysis sub-object corresponds to a belonging analysis set, and the method further includes: determining a net recommendation value of each target analysis set based on a difference between a third ratio and a fourth ratio corresponding to each analysis set; wherein, the third ratio is a ratio of the number of all positive analysis sub-objects to the number of all analysis sub-objects in each analysis set, and the fourth ratio is a ratio of the number of all negative analysis sub-objects to the number of all analysis sub-objects in each analysis set.

[0011] In an alternative manner, each analysis sub-object corresponds to a belonging analysis set, and the target scores corresponding to the respective analysis sub-objects are in any one of a first score interval, a second score interval, and a third score interval, where the minimum value of the first score interval is greater than the maximum value of the second score interval, and the minimum value of the second score interval is greater than the maximum value of the third score interval; the method further includes: using the analysis sub-objects with target scores in the first score interval in the target analysis set as first score-losing objects, the analysis sub-objects with target scores in the second score interval as second score-losing objects, and the analysis sub-objects with target scores in the third score interval as third score-losing objects; wherein, the target analysis set is any one of the analysis sets; calculating a score-losing weight of the target analysis set based on the first score-losing objects, the second score-losing objects, the third score-losing objects, and their respective corresponding score-losing weights in the target analysis set; obtaining a score-losing value of the target analysis set based on the net recommendation value of the target object and the score-losing weight of the target analysis set, so as to determine the optimization degree of the target analysis set.

[0012] According to another aspect of the embodiments of the present application, there is provided a data processing device, including: a determining module, configured to respectively determine analysis sub-objects in each feedback statement for each object keyword included in each feedback statement, and determine initial scores corresponding to the respective analysis sub-objects according to the description keywords associated with the respective analysis sub-objects in the belonging feedback statements; wherein each feedback statement is a description statement for the same target object, and each feedback statement corresponds to a comprehensive score for the target object; an adjusting module, configured to calculate target scores corresponding to the respective analysis sub-objects in each feedback statement according to the initial scores corresponding to the respective analysis sub-objects and the comprehensive scores corresponding to the belonging feedback statements of the respective analysis sub-objects; a calculating module, configured to calculate a net recommendation value of the target object according to the target scores corresponding to all the analysis sub-objects.

[0013] According to one aspect of the embodiments of the present application, there is provided an electronic device, including: a controller; a memory, configured to store one or more programs, and when the one or more programs are executed by the controller, to execute the above output method.

[0014] According to one aspect of the embodiments of the present application, a storage medium is further provided, on which computer-readable instructions are stored. When the computer-readable instructions are executed by a processor of a computer, the computer is caused to execute the above output method.

[0015] According to one aspect of the embodiments of the present application, a computer program product or a computer program is further provided. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the above output method.

[0016] The present application calculates the Net Promoter Score (NPS) of a target object based on multiple feedback statements and their corresponding comprehensive scores. Compared with directly determining the NPS according to a single comprehensive score corresponding to the target object, the problem that the NPS is inaccurate due to a large scoring error of a single comprehensive score is greatly reduced. Among them, multiple analysis sub-objects are parsed from the object keywords in each feedback statement, and in combination with the description keywords associated with the analysis sub-objects in the feedback statement, the initial scores of each analysis sub-object are determined, which can split the scores for different analysis sub-objects in a complex feedback statement, avoiding the influence of a large error in the local evaluation of the target object caused by relying only on a single comprehensive score; according to the comprehensive score corresponding to the feedback statement, the initial scores of each analysis sub-object are corrected to obtain the target scores of each analysis sub-object, ensuring that the scoring of each analysis sub-object does not deviate from the overall comprehensive score, further reducing the scoring error and making the calculation basis of the NPS more accurate. In addition, the more the number of feedback statements, the more categories and the larger the number of analysis sub-objects included, making the calculation samples of the NPS more, and the error smaller.

[0017] The above description is only an overview of the technical solutions of the embodiments of the present application. In order to be able to understand the technical means of the embodiments of the present application more clearly, it can be implemented according to the content of the description. And in order to make the above and other purposes, features and advantages of the embodiments of the present application more obvious and understandable, the specific embodiments of the present application are specifically exemplified below. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings here are incorporated into the description and form a part of this description, showing embodiments consistent with the present application, and are used together with the description to explain the principles of the present application. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1It is a schematic flowchart of a data processing method shown in an exemplary embodiment of the present application.

[0020] Figure 2 It is based on Figure 1 The shown exemplary embodiment shows a schematic flowchart of a method for calculating the target scores of each analysis sub-object.

[0021] Figure 3 It is based on Figure 1 The shown exemplary embodiment shows a schematic flowchart of a net promoter score calculation method.

[0022] Figure 4 It is a schematic flowchart of a method for calculating deduction scores shown in an exemplary embodiment of the present application.

[0023] Figure 5 It is a schematic flowchart of another data processing method shown in an exemplary embodiment of the present application.

[0024] Figure 6 It is a schematic diagram of an application scenario of the data processing method of the present application.

[0025] Figure 7 It is a schematic structural diagram of a data processing device shown in an exemplary embodiment of the present application.

[0026] Figure 8 It is a schematic structural diagram of a computer system of an electronic device shown in an exemplary embodiment of the present application. Detailed implementation manners

[0027] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all the implementation manners consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0028] The block diagrams shown in the drawings are only functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0029] The flowcharts shown in the drawings are only exemplary descriptions and do not necessarily include all the contents and operations / steps, nor do they necessarily need to be executed in the described order. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined. Therefore, the actual execution order may change according to the actual situation.

[0030] As used in this application, "a plurality of" means two or more. "And / or" describes the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0031] In the related art, when calculating the net promoter score based on the comprehensive score of the target object, due to many interference factors in the process of assigning comprehensive scores to the target object, the obtained comprehensive scores are prone to errors, resulting in inaccurate net promoter scores of the determined target objects and being unable to objectively and accurately characterize the recommendation degree for the target objects.

[0032] For this reason, one aspect of this application provides a data processing method. For details, please refer to Figure 1 , Figure 1 is a schematic flowchart of a data processing method shown in an exemplary embodiment of this application. This method at least includes S110 to S130, which are introduced in detail as follows:

[0033] S110: For each object keyword included in each feedback statement, respectively determine the analysis sub-objects in each feedback statement, and determine the initial scores corresponding to the respective analysis sub-objects according to the description keywords associated with the analysis sub-objects in the respective feedback statements.

[0034] Among them, each feedback statement is a description statement for the same target object, and each feedback statement corresponds to a comprehensive score for the target object.

[0035] The target object refers to the product to be evaluated, including tangible products such as hardware and process materials, as well as intangible products such as software and services. The feedback statement is a description statement for the target object, which can be text or voice. The comprehensive score represents the overall score of a certain physical evaluation system or physical object for the target object.

[0036] Exemplarily, the target object can be a car, a mobile phone, a consulting service, or an application software, etc. The feedback statement is a text message, such as "The system performance is excellent but the battery life is poor". The comprehensive score corresponding to the feedback statement can be obtained by a pre-trained evaluation model or computer software scoring the target object, or can be obtained by a user within different time periods, or multiple users scoring the target object at the same time.

[0037] In this application, the object keywords are generally nouns, that is, words representing the referred objects; the description keywords are generally adjectives or adverbs of degree, which are used to represent words such as the advantages and disadvantages, high and low of the corresponding attributes of the referred objects, or words representing the degree of the action performance of the referred objects. At least one analysis sub-object can be determined according to the object keywords in each feedback statement, for example, a certain component of a tangible product; combined with the description keywords associated with each analysis sub-object in the corresponding feedback statement, such as "good", the initial scores corresponding to each analysis sub-object can be determined.

[0038] S120: Calculate the target scores corresponding to each analysis sub-object in each feedback statement according to the initial scores corresponding to each analysis sub-object and the comprehensive scores corresponding to the feedback statements to which each analysis sub-object belongs.

[0039] In this application, the initial scores corresponding to each analysis sub-object are determined according to the description keywords in the feedback statement. Therefore, the initial scores have a high correlation with the feedback statement. According to the comprehensive scores corresponding to the feedback statements to which each analysis sub-object belongs, the initial scores of each analysis sub-object are adjusted to obtain the target scores corresponding to each analysis sub-object in each feedback statement, which improves the accuracy of the basis for calculating the net promoter score.

[0040] S130: Calculate the net promoter score of the target object according to the target scores corresponding to all analysis sub-objects respectively.

[0041] Exemplarily, assume that there are a total of feedback statement 1 and feedback statement 2, and feedback statement 1 contains analysis sub-object A and analysis sub-object B, and feedback statement 2 contains analysis sub-object A and analysis sub-object C. Then, according to the target score a1 of analysis sub-object A in feedback statement 1, the target score b of analysis sub-object B in feedback statement 1, the target score a2 of analysis sub-object A in feedback statement 2, and the target score c of analysis sub-object C in feedback statement 2, calculate the net promoter score of the target object.

[0042] This embodiment calculates the net recommendation value of the target object based on multiple feedback statements and their corresponding comprehensive scores. Compared with directly determining the net recommendation value based on a comprehensive score corresponding to the target object, it greatly reduces the problem of inaccurate net recommendation value caused by large errors in assigning a single comprehensive score. Specifically, based on the object keyword analysis in each feedback statement, multiple analysis sub-objects are obtained, and the initial score of each analysis sub-object is determined in combination with the descriptive keywords associated with the analysis sub-object in the feedback statement. This can split the scores for different analysis sub-objects in a complex feedback statement, avoiding the influence of large errors in the local evaluation of the target object caused by relying solely on a single comprehensive score; based on the comprehensive score corresponding to the feedback statement, the initial score of each analysis sub-object is corrected to obtain the target score of each analysis sub-object, ensuring that the score of each analysis sub-object does not deviate from the overall comprehensive score when determining the score of each analysis sub-object, further reducing the scoring error, and making the calculation basis of the net recommendation value more accurate.

[0043] In another exemplary embodiment of the present application, how to parse the object keywords and description keywords in the feedback sentence is described in detail. Based on S110 to S130, the specific steps further include S210 to S230, which are described in detail as follows:

[0044] S210: traverse multiple feedback sentences, and use the traversed feedback sentences as target feedback sentences.

[0045] This embodiment ensures that each feedback statement is processed without omission by means of traversal.

[0046] S220: If the matching coefficient between the target candidate segmentation word and the preset noun is greater than a first preset matching threshold, the target candidate segmentation word is determined to be the target object keyword of the target feedback sentence.

[0047] The target candidate segmentation word is any candidate segmentation word determined from the target feedback sentence.

[0048] In this application, the StructBERT model is used to perform text preprocessing on the target feedback sentence, including removing special symbols, correcting typos, and word segmentation, so as to split the target feedback sentence into multiple candidate word segmentations, so as to facilitate matching each candidate word segmentation with the preset noun.

[0049] For example, taking the target object as a car, a preset noun library containing components such as doors, steering wheels, seats, and air conditioners is set up. If the matching coefficient between the target candidate segmentation word and any preset noun in the preset noun library is greater than a first preset matching threshold, it indicates that the target candidate segmentation word is a target object keyword in the target feedback sentence. The matching coefficient is the similarity between the target candidate segmentation word and the preset noun. The higher the similarity, the larger the matching coefficient. The matching threshold is determined based on the actual training results of the model.

[0050] S230: If the matching coefficient between the adjacent candidate word segmentations of the target object keyword and the preset description word is greater than the second preset matching threshold, then determine the adjacent candidate word segmentation as the description keyword associated with the analysis sub-object corresponding to the target object keyword.

[0051] In this application, the adjacent candidate word segmentations of the target object keyword can be the pre-candidate word segmentations of the target object keyword or the post-candidate word segmentations of the target object keyword. During the process of matching the description keyword, wildcards can be set at the beginning and end positions of the target object keyword to represent strings of any length, implementing an open matching mode and simplifying the matching operation.

[0052] Exemplarily, use * to represent any character between two words in a feedback statement. For example: [audio * good], where "audio" is the determined target object keyword, and the matching coefficient between the word "good" and the preset description words in the preset description word library is greater than the second preset matching threshold, then determine it as the description keyword associated with "audio" in the target feedback statement to which it belongs. And "*" replaces other content in the target feedback statement that fails to match the target object keyword and the description keyword, such as the word "effect" in [audio effect good].

[0053] As shown in Table 1, Table 1 is a corresponding relationship table after some feedback statements are successfully matched with the preset nouns and preset description words.

[0054] Target feedback statement Target object keyword Description keyword The in-vehicle camera recording time is short Camera Short The sound effect is good Sound Good The first row of seats is hard Seat Hard

[0055] Table 1

[0056] Perform a matching operation on the feedback statement with multiple preset nouns and preset description words, and use the preset nouns that are successfully matched as the target object keywords, and use the preset description words that are successfully matched as the description keywords.

[0057] In addition, in practical applications, the values of the first preset matching threshold and the second matching threshold can be the same or different, and this application does not limit this.

[0058] This embodiment provides a method for parsing the object keyword and the description keyword in the feedback statement. By pre-saving the nouns and description words related to the target object for keyword recognition of the feedback statement, complex natural language processing algorithms are not required, and the corresponding vocabulary can be directly searched to complete the matching, which can improve the efficiency and accuracy of keyword matching.

[0059] In another exemplary embodiment of this application, based on the above Figure 1 shown method, it illustrates how to determine the initial scores of each analysis sub-object. The specific steps further include S310 - S340, which are introduced in detail as follows:

[0060] S310: Traverse each analysis sub-object, and use the traversed analysis sub-object as the target analysis sub-object. The description keyword associated with the target analysis sub-object in the corresponding feedback statement is the target description keyword, and the comprehensive score corresponding to the feedback statement to which the target analysis sub-object belongs is the target comprehensive score.

[0061] When determining the initial score of each analysis sub-object, the feedback statement can be used as the basic operation unit, that is, in one operation process, the processing of the analysis sub-object in one feedback statement is completed, and then in the next operation process, the analysis sub-objects in other feedback statements are processed. Or the analysis sub-object can be directly used as the basic operation unit, that is, in one operation process, the processing of one analysis sub-object in any feedback statement is completed.

[0062] This embodiment uses the analysis sub-object as the basic operation unit, and ensures that each analysis sub-object is processed without omission through traversal. If there are the same analysis sub-objects in different feedback statements, they should be regarded as different operation units.

[0063] S320: If the attribute value of the target description keyword is greater than the preset threshold, the initial score corresponding to the target analysis sub-object is: the score in the first score interval with the smallest difference from the target comprehensive score.

[0064] In this application, the target description keyword is classified based on a dictionary or machine learning, specifically including positive reviews, medium reviews, and negative reviews. The intervals where the initial scores of the analysis sub-objects under different categories are located are different.

[0065] Exemplarily, if the attribute value of the target description keyword is greater than the preset threshold, and it is determined that the classification of the target description keyword is a positive review, then the initial score of the target analysis sub-object is in the first score interval corresponding to the positive review, such as 9 - 10 points on a ten-point scale, that is, the first score interval is [9, 10]. Assuming the target comprehensive score is 6 points, the initial score of the target analysis sub-object is 9 points, which is the score with the smallest difference from the target comprehensive score.

[0066] In some alternative embodiments, if the preset score interval is an open interval, for example, (9, 10), the endpoint values 9 and 10 can still be used for difference calculation with the target comprehensive score and be used as the final target score when meeting the conditions. The same applies hereinafter.

[0067] Among them, the attribute value of the target description keyword can be the similarity with the preset word in the dictionary or the score value of the target description keyword in machine learning. The preset thresholds corresponding to different classification methods are different, and this application does not limit this.

[0068] S330: If the attribute value of the target description keyword is equal to the preset threshold, the initial score corresponding to the target analysis sub-object is: the score in the second score interval with the smallest difference from the target comprehensive score.

[0069] Exemplarily, if the attribute value of the target description keyword is equal to the preset threshold, it is determined that the classification of the target description keyword is a medium review. Then, the initial score of the target analysis sub-object is in the second score interval corresponding to the medium review, such as 7 - 8 points in a ten-point system, that is, the second score interval is [7, 8]. Suppose the target comprehensive score is 6 points, then the initial score of the target analysis sub-object is 7 points, which is the score with the smallest difference from the target comprehensive score.

[0070] S340: If the attribute value of the target description keyword is less than the preset threshold, the initial score corresponding to the target analysis sub-object is: the score in the third score interval with the smallest difference from the target comprehensive score.

[0071] Among them, the minimum value of the first score interval is greater than or equal to the maximum value of the second score interval, and the minimum value of the second score interval is greater than or equal to the maximum value of the third score interval.

[0072] Exemplarily, if the attribute value of the target description keyword is less than the preset threshold, it is determined that the classification of the target description keyword is a bad review. Then, the initial score of the target analysis sub-object is in the third score interval corresponding to the bad review, such as 0 - 6 points in a ten-point system, that is, the third score interval is [0, 6]. Suppose the target comprehensive score is 6 points, then the initial score of the target analysis sub-object is 6 points, which is the score with the smallest difference from the target comprehensive score.

[0073] The score interval corresponding to a good review is greater than the score interval corresponding to a medium review, and the score interval corresponding to a medium review is greater than the score interval corresponding to a bad review.

[0074] In another alternative embodiment, if there is no associated description keyword for the object keyword, or the final classification cannot be confirmed directly through the description keyword, algorithms such as the bidirectional attention mechanism can be used to capture the context dependence relationship in the feedback statement to complete the classification determination of the analysis sub-object in the specific semantics, and then determine the initial score of the analysis sub-object.

[0075] This embodiment provides a method for determining the initial score of each analysis sub-object. By determining the score interval to which the initial score corresponding to each analysis sub-object belongs through the classification of the description keyword, compared with directly assigning the initial score between zero and full marks, dividing the score interval can ensure that there is no large error between the initial score and the actual semantics of the feedback statement. Then, in combination with the comprehensive score, the final initial score is determined in the corresponding score interval, reducing the error from the overall score of the feedback statement.

[0076] In another exemplary embodiment of the present application, based on the aboveFigure 1 The method shown provides a method for calculating the target scores of each analysis sub-object. For details, please refer to Figure 2 , Figure 2 which is based on Figure 1 the schematic flowchart of the method for calculating the target scores of each analysis sub-object shown in the exemplary embodiment.

[0077] This method includes at least S410 to S420 in step S120 as shown in Figure 1 and is introduced in detail as follows:

[0078] S410: For multiple analysis sub-objects belonging to the same feedback statement, calculate the average initial score to determine the average initial score corresponding to each feedback statement and the target score interval to which each average initial score belongs.

[0079] The target score interval is any one of the first score interval, the second score interval, and the third score interval.

[0080] Exemplarily, assume that feedback statement 3 contains five analysis sub-objects A, B, C, D, and E, and their respective corresponding initial scores are 6, 6, 6, 6, and 9. Then, calculate the average initial score of feedback statement 3 as (6 + 6 + 6 + 6 + 9) / 5 = 6.6 points, and determine that the target score interval to which the average initial score belongs is the third score interval [0, 6].

[0081] S420: If the difference between the average initial score corresponding to the focus feedback statement and the comprehensive score is greater than the preset error threshold, adjust the initial score corresponding to the focus analysis sub-object based on the target score interval to determine the target scores corresponding to each analysis sub-object in each feedback statement.

[0082] Exemplarily, assume that the comprehensive score of feedback statement 3 is 6 points, the preset error threshold is 0.5 points, and the difference between the average initial score of feedback statement 3 and the comprehensive score is 0.6 points, that is, the difference is greater than the preset error threshold. Then, it indicates that feedback statement 3 is a focus feedback statement, and the initial scores of the focus analysis sub-objects therein need to be adjusted.

[0083] In this application, the focus analysis sub-object is the analysis sub-object in the focus feedback statement and whose initial score is within the target score interval. That is, the focus analysis sub-objects in feedback statement 3 are analysis sub-objects A, B, C, and D whose initial scores are within the third score interval.

[0084] The specific adjustment method is as follows: If the average initial score is greater than the comprehensive score, subtract 1 from the initial score of the focus analysis sub-object; if the average initial score is less than the comprehensive score, add 1 to the initial score of the focus analysis sub-object until the difference between the adjusted average initial score and the comprehensive score is less than the preset error threshold.

[0085] For example, after subtracting 1 from the initial scores of analysis sub-objects A, B, C, and D respectively, the average initial score of feedback statement 3 recalculated is (5 + 5 + 5 + 5 + 9) / 5 = 5.8, and the error between 5.8 and the comprehensive score of 6 is less than 0.5. Therefore, the target scores corresponding to each analysis sub-object in feedback statement 3 are 5, 5, 5, 5, and 9 respectively.

[0086] Among them, the error threshold can be adjusted according to the actual situation, and the present application does not limit this.

[0087] Calculate the target scores of each analysis sub-object in all feedback statements based on the above method.

[0088] This embodiment provides a method for determining the target scores of each analysis sub-object. By using the comprehensive score and the average initial score of the feedback statement, the initial scores of the analysis sub-objects are fine-tuned, so that the final average initial score is closer to the comprehensive score. When splitting and determining the scores of different analysis sub-objects, it can ensure that the overall comprehensive score of the target object is not deviated, thereby ensuring the effectiveness of the target scores.

[0089] In another exemplary embodiment of the present application, based on the above Figure 1 shown method, a method for calculating the net promoter score of the target object is provided. For details, refer to Figure 3 , Figure 3 is a schematic flowchart of a method for calculating the net promoter score shown in the Figure 1 shown exemplary embodiment. This method further includes S510 to S520 in S130 as shown in Figure 1 shown, and the details are introduced as follows:

[0090] S510: Calculate the first proportion of the number of positive analysis sub-objects in the total number of all analysis sub-objects, and the second proportion of the number of negative analysis sub-objects in the total number of all analysis sub-objects.

[0091] Among them, the positive analysis sub-object is an analysis sub-object whose target score is in the first score interval, and the negative analysis sub-object is an analysis sub-object whose target score is in the third score interval.

[0092] In the present application, the physical evaluation system or physical object corresponding to the feedback statement to which the positive analysis sub-object belongs is the recommender of the positive analysis sub-object, and the physical evaluation system or physical object corresponding to the feedback statement to which the negative analysis sub-object belongs is the detractor of the negative analysis sub-object. The first proportion of the number of positive analysis sub-objects in the total number of all analysis sub-objects is the proportion of the recommenders, and the second proportion of the number of negative analysis sub-objects in the total number of all analysis sub-objects is the proportion of the detractors.

[0093] S520: Determine the Net Promoter Score (NPS) of the target object based on the difference between the first proportion and the second proportion.

[0094] According to the NPS calculation formula NPS = % of promoters - % of detractors, the NPS of the target object is calculated as the difference between the first proportion and the second proportion, where the value range of NPS is [-100, 100].

[0095] This embodiment introduces how to calculate the NPS of the target object. Using each analysis sub-object as the basis for calculating the NPS of the target object can improve the accuracy of the NPS.

[0096] In some alternative embodiments, the analysis sub-object can also be the after-sales service for the target object. Based on the target scores of the after-sales service in multiple feedback statements, the NPS of the after-sales service can be calculated, thereby reflecting the user's satisfaction with the after-sales service.

[0097] In another exemplary embodiment of the present application, each analysis sub-object corresponds to an analysis set to which it belongs. Then, according to the target scores of multiple analysis sub-objects in different analysis sets, the NPS of each analysis set can be calculated, specifically including: determining the NPS of each target analysis set based on the difference between the third proportion and the fourth proportion corresponding to each analysis set. Among them, the third proportion is the ratio of the number of all positive analysis sub-objects to the number of all analysis sub-objects in each analysis set, and the fourth proportion is the ratio of the number of all negative analysis sub-objects to the number of all analysis sub-objects in each analysis set.

[0098] Exemplarily, if the target object is a car, the analysis sub-objects are each component of the car, such as doors, seats, audio systems, cameras, entertainment systems, etc. Among them, the audio system, camera, and entertainment system all belong to the in-vehicle entertainment system, that is, they belong to the same analysis set. Then, according to the target scores of the audio system, camera, and entertainment system in each feedback statement, the proportion of positive analysis sub-objects and the proportion of negative analysis sub-objects within the analysis set of the in-vehicle entertainment system are determined, and using the NPS calculation formula, the NPS of the analysis set of the in-vehicle entertainment system can be calculated, thereby quantifying the user's satisfaction with the in-vehicle entertainment system.

[0099] In addition, the doors, seats, audio system, camera, and entertainment system all belong to the intelligent cockpit, that is, the intelligent cockpit can be used as an analysis set at a higher level than the in-vehicle entertainment system. Therefore, the above method can be used to calculate the NPS of the intelligent cockpit, thereby determining the recommendation degree for the intelligent cockpit.

[0100] This embodiment classifies each analysis sub-object to obtain multiple first-level or second-level analysis sets and calculates the NPS of the analysis sets, which can determine the recommendation degree for each system of the target object.

[0101] On the other hand, in the related art, only the possibility of a user recommending a target object to others is measured by the Net Promoter Score, and the lack of a strategy design for optimizing the target object is provided.

[0102] To this end, in order to better identify the defects of the target object and provide a direction and basis for the subsequent rectification work for the target object, in another exemplary embodiment of the present application, a method for calculating the score loss of the target object is provided. For details, please refer to Figure 4 , Figure 4 FIG. is a schematic flowchart of a method for calculating the score loss shown in an exemplary embodiment of the present application. The method at least includes S610 to S630, which are introduced in detail as follows:

[0103] S610: In the target analysis set, the analysis sub-objects with target scores in the first score interval are used as the first score-loss objects, the analysis sub-objects with target scores in the second score interval are used as the second score-loss objects, and the analysis sub-objects with target scores in the third score interval are used as the third score-loss objects.

[0104] Among them, the target analysis set is any one of the analysis sets.

[0105] In this embodiment, according to the target scores, each analysis sub-object in the target analysis set is divided into different score-loss objects to determine their respective corresponding weights.

[0106] Exemplarily, in the traditional NPS calculation process, according to the comprehensive score for the target object, the physical evaluation system or physical object with a good review is determined as the recommender, the physical evaluation system or physical object with a medium review is determined as the passive person, and the physical evaluation system or physical object with a bad review is determined as the detractor. Based on the Net Promoter Score formula NPS = recommender% - detractor%, it can be inferred that:

[0107] NPS = (100% - detractor% - passive%) - detractor% = 100% - passive% - 2 * detractor%

[0108] Thus, the influence weight of the recommender is determined to be 0, the influence weight of the passive person is 1, and the influence weight of the detractor is 2.

[0109] Since the first score interval in the present application corresponds to a good review, the second score interval corresponds to a medium review, and the third score interval corresponds to a bad review, it can be analogously inferred that the score-loss weight of the first score-loss object is 0, the score-loss weight of the second score-loss object is 1, and the weight of the third score-loss object is 2.

[0110] S620: Calculate the loss weight of the target analysis set based on the first, second, and third score-losing objects and their respective loss weights in the target analysis set.

[0111] Exemplarily, if the number of the first score-losing objects in the target analysis set I is 6, the number of the second score-losing objects is 4, and the number of the third score-losing objects is 17, then the loss weight of the target analysis set I is 6*0 + 17*1 + 4*2 = 25.

[0112] S630: Obtain the loss score of the target analysis set based on the net promoter score of the target object and the loss weight of the target analysis set, so as to determine the optimization degree of the target analysis set.

[0113] In this application, according to the net promoter score NPS of the target object 总 , calculate the total loss score of the target object, and then calculate the loss score (Promoter Loss, PL) of the target analysis set according to the total loss score and the loss weight of the target analysis set. The calculation formula is:

[0114]

[0115] Among them, the total loss weight is the sum of the loss weights of all analysis sub-objects.

[0116] Exemplarily, assume that the net promoter score NPS of the target object 总 = 75.3, then the total loss score PL of the target object 总 = 100 - 75.3 = 24.7, and the loss score of the target analysis set I is

[0117]

[0118] Among them, 127 is the sum of the loss weights of all analysis sub-objects.

[0119] Similarly, the loss score corresponding to a certain analysis sub-object can be calculated based on the above method.

[0120] In this embodiment, by assigning loss weights to the analysis sub-objects, the loss scores of a single analysis set or a single analysis sub-object are calculated. The analysis set with a higher loss score indicates more defects. At the same time, compared with directly using the total loss score of the target object, introducing the loss weight to calculate the loss score of a single analysis set can obtain the evaluation proportion of a certain part of the target object in the whole system, and better clarify the influence of the part on the whole. Therefore, this method can shorten the time for identifying problems in all aspects of the target object, provide a clear direction and basis for the subsequent rectification work, and significantly improve the rectification efficiency.

[0121] Refer to Figure 5, Figure 5 It is a schematic flowchart of another data processing method shown in an exemplary embodiment of the present application. The detailed description is as follows:

[0122] First, receive feedback statements from different users regarding a target object.

[0123] Then, use the StructBERT model to obtain multiple analysis sub-objects and their respective target scores based on multiple feedback statements.

[0124] Finally, use the NPS calculation model to calculate the net promoter score of the target object based on each analysis sub-object and its respective target score.

[0125] In the embodiment of the present application, multiple analysis sub-objects are parsed based on the object keywords in each feedback statement, and the initial scores of each analysis sub-object are determined by combining the description keywords associated with the analysis sub-objects in the feedback statement, which can split the scores for different analysis sub-objects in complex feedback statements, avoiding the influence of large errors in the local evaluation of the target object caused by relying only on a single comprehensive score; according to the comprehensive score corresponding to the feedback statement, the initial scores of each analysis sub-object are corrected to obtain the target scores of each analysis sub-object, ensuring that the scoring of each analysis sub-object does not deviate from the overall comprehensive score, further reducing the scoring error and making the calculation basis of the net promoter score more accurate. It can be widely applied to user feedback analysis in fields such as automobile manufacturing, electronic products, and service industries, thus helping enterprises accurately locate problems and optimize product quality.

[0126] In another exemplary embodiment of the present application, an exemplary description of the application scenarios of the above multiple data processing methods is given. For details, please refer to Figure 6 , Figure 6 It is a schematic diagram of the application scenario of the data processing method of the present application. Among them, it includes a client 100 and a server 200, which can be connected by wireless communication. The present application does not limit their connection method.

[0127] The server 200 can be used as an execution entity to receive the net promoter score calculation instruction of the input data sent by the client 100 and execute the data processing method shown in any of the above exemplary embodiments. The following is an exemplary description:

[0128] Client 100 sends the feedback statements collected by itself for the target object to Server 200, and instructs Server 200 to calculate the Net Promoter Score (NPS) of the target object. After receiving the feedback statements, Server 200, based on the StructBERT model, respectively determines the analysis sub-objects in each feedback statement for the object keywords included in each feedback statement, and determines the initial scores corresponding to each analysis sub-object according to the description keywords associated with the respective analysis sub-objects in the feedback statement to which they belong. Server 200 calculates the target scores corresponding to each analysis sub-object in each feedback statement according to the initial scores corresponding to each analysis sub-object and the comprehensive scores corresponding to the feedback statements to which each analysis sub-object belongs. Server 200 uses the NPS calculation model to calculate the Net Promoter Score of the target object based on each analysis sub-object and its respective corresponding target scores. Finally, Server 200 returns the calculation result of the Net Promoter Score to Client 100.

[0129] Optionally, Client 100 may be an electronic device such as a mobile phone, a computer, or a tablet. The StructBERT model and the NPS calculation model may be integrated within Server 200, or may be remotely called by Server 200 through a preset interface. This application places no restrictions on this.

[0130] Another aspect of this application also provides a data processing device, as Figure 7 shown Figure 7 is a schematic structural diagram of the data processing device shown in an exemplary embodiment of this application. Device 700 includes:

[0131] A determination module 710 that respectively determines the analysis sub-objects in each feedback statement for the object keywords included in each feedback statement, and determines the initial scores corresponding to each analysis sub-object according to the description keywords associated with the respective analysis sub-objects in the feedback statement to which they belong; wherein, each feedback statement is a description statement for the same target object, and each feedback statement corresponds to a comprehensive score for the target object;

[0132] An adjustment module 720 that calculates the target scores corresponding to each analysis sub-object in each feedback statement according to the initial scores corresponding to each analysis sub-object and the comprehensive scores corresponding to the feedback statements to which each analysis sub-object belongs;

[0133] A calculation module 730 that calculates the Net Promoter Score of the target object according to the target scores corresponding to all the analysis sub-objects respectively.

[0134] In an optional manner, device 700 further includes:

[0135] A first traversal module that traverses multiple feedback statements and uses the traversed feedback statement as the target feedback statement;

[0136] The first matching module determines the target object keyword of the target feedback statement if the matching coefficient between the target candidate word segment and the preset noun is greater than the first preset matching threshold; wherein, the target candidate word segment is any candidate word segment determined from the target feedback statement.

[0137] The second matching module determines the descriptive keyword associated with the analysis sub-object corresponding to the target object keyword if the matching coefficient between the adjacent candidate word segment of the target object keyword and the preset descriptive word is greater than the second preset matching threshold.

[0138] In an optional manner, the determination module 710 further includes:

[0139] The traversal unit traverses each analysis sub-object, and uses the traversed analysis sub-object as the target analysis sub-object. The descriptive keyword associated with the target analysis sub-object in its belonging feedback statement is the target descriptive keyword, and the comprehensive score corresponding to the feedback statement to which the target analysis sub-object belongs is the target comprehensive score.

[0140] The first judgment unit, if the attribute value of the target descriptive keyword is greater than the preset threshold, the initial score corresponding to the target analysis sub-object is: the score in the first score interval with the smallest difference from the target comprehensive score.

[0141] The second judgment unit, if the attribute value of the target descriptive keyword is equal to the threshold, the initial score corresponding to the target analysis sub-object is: the score in the second score interval with the smallest difference from the target comprehensive score.

[0142] The third judgment unit, if the attribute value of the target descriptive keyword is less than the threshold, the initial score corresponding to the target analysis sub-object is: the score in the third score interval with the smallest difference from the target comprehensive score; wherein, the minimum value of the first score interval is greater than the maximum value of the second score interval, and the minimum value of the second score interval is greater than the maximum value of the third score interval.

[0143] In an optional manner, the adjustment module 720 further includes:

[0144] The mean calculation unit calculates the average initial score for multiple analysis sub-objects belonging to the same feedback statement to determine the average initial score corresponding to each feedback statement and the target score interval to which each average initial score belongs. The target score interval is any one of the first score interval, the second score interval, and the third score interval.

[0145] Adjustment subunit: If the difference between the average initial score and the comprehensive score corresponding to the focus feedback statement is greater than the preset error threshold, then adjust the initial score corresponding to the focus analysis sub-object based on the target score range to determine the target score corresponding to each analysis sub-object in each feedback statement; wherein, the focus analysis sub-object is an analysis sub-object in the focus feedback statement and whose initial score is within the target score range.

[0146] In an alternative manner, the calculation module 730 further includes:

[0147] Ratio calculation unit: Calculate the first ratio of the number of positive analysis sub-objects to the number of all analysis sub-objects, and the second ratio of the number of negative analysis sub-objects to the number of all analysis sub-objects; wherein, the positive analysis sub-object is an analysis sub-object whose target score is within the first score range, and the negative analysis sub-object is an analysis sub-object whose target score is within the third score range.

[0148] Difference calculation unit: Determine the net recommendation value of the target object based on the difference between the first ratio and the second ratio.

[0149] In an alternative manner, the apparatus 700 further includes:

[0150] Difference calculation module: Determine the net recommendation value of each target analysis set based on the difference between the third ratio and the fourth ratio corresponding to each analysis set; wherein, the third ratio is the ratio of the number of all positive analysis sub-objects to the number of all analysis sub-objects in each analysis set, and the fourth ratio is the ratio of the number of all negative analysis sub-objects to the number of all analysis sub-objects in each analysis set.

[0151] In an alternative manner, the apparatus 700 further includes:

[0152] Classification module: Take the analysis sub-objects in the target analysis set whose target scores are within the first score range as the first score-losing objects, the analysis sub-objects whose target scores are within the second score range as the second score-losing objects, and the analysis sub-objects whose target scores are within the third score range as the third score-losing objects; wherein, the target analysis set is any one of the analysis sets.

[0153] Weighting module: Calculate the score-losing weight of the target analysis set based on the first score-losing objects, the second score-losing objects, the third score-losing objects in the target analysis set and their respective score-losing weights.

[0154] Optimization module: Obtain the score-losing value of the target analysis set based on the net recommendation value of the target object and the score-losing weight of the target analysis set, so as to be used to determine the optimization degree of the target analysis set.

[0155] It should be noted that the data processing device provided in the above embodiments and the data processing method provided in the foregoing embodiments belong to the same concept. The specific manners in which each module and unit perform operations have been described in detail in the method embodiments and will not be elaborated here.

[0156] On the other hand, the present application further provides an electronic device, including: a controller; a memory for storing one or more programs, which, when executed by the controller, are configured to execute the above management method.

[0157] Please refer to Figure 8 , Figure 8 which is a schematic structural diagram of a computer system of the electronic device shown in an exemplary embodiment of the present application, showing a schematic structural diagram of a computer system of the electronic device suitable for implementing the embodiments of the present application.

[0158] It should be noted that Figure 8 the computer system 800 of the electronic device shown is only an example and should not impose any limitation on the functions and usage scope of the embodiments of the present application.

[0159] As Figure 8 shown, the computer system 800 includes a central processing unit (CPU) 801, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 802 or the program loaded from the storage section 808 into the random access memory (RAM) 803, such as executing the method in the above embodiments. In the RAM 803, various programs and data required for system operation are also stored. The CPU 801, ROM 802, and RAM 803 are connected to each other through a bus 804. The input / output (I / O) interface 805 is also connected to the bus 804.

[0160] The following components are connected to the I / O interface 805: an input part 806 including a keyboard, a mouse, etc.; an output part 807 including, for example, a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage part 808 including a hard disk, etc.; and a communication part 809 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication part 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the I / O interface 805 as required. A removable medium 811, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 810 as required so that a computer program read therefrom is installed into the storage part 808 as required.

[0161] Specifically, according to an embodiment of the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication part 809, and / or installed from the removable medium 811. When the computer program is executed by a central processing unit (CPU) 801, various functions defined in the system of the present application are executed.

[0162] It should be noted that the computer-readable medium shown in the embodiments of the present application may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer-readable storage medium may be any tangible medium that contains or stores a program, and the program may be used by or in conjunction with an instruction execution system, apparatus, or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable computer program. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium may also be any computer-readable medium other than the computer-readable storage medium, and the computer-readable medium may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The computer program contained on the computer-readable medium may be transmitted by any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0163] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. Among them, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the above module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in an order different from that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, and the combination of blocks in the block diagram or flowchart, may be implemented by a dedicated hardware-based system for performing the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0164] The units involved in the embodiments of the present application can be implemented in software or in hardware, and the described units can also be provided in a processor. Among them, the names of these units do not constitute a limitation to the unit itself in some cases.

[0165] Another aspect of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the management method as described above is implemented. The computer-readable storage medium may be included in the electronic device described in the above embodiments, or may exist alone without being assembled into the electronic device.

[0166] Another aspect of the present application also provides a computer program product or a computer program. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the management methods provided in the above various embodiments.

[0167] According to one aspect of the embodiments of the present application, a computer system is also provided, including a Central Processing Unit (CPU). It can perform various appropriate actions and processes according to the program stored in a Read-Only Memory (ROM) or the program loaded from a storage section into a Random Access Memory (RAM), such as executing the method in the above embodiments. In the RAM, various programs and data required for system operations are also stored. The CPU, ROM, and RAM are connected to each other via a bus. An Input / Output (I / O) interface is also connected to the bus.

[0168] The following components are connected to the I / O interface: an input section including a keyboard, a mouse, etc.; an output section including, for example, a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), etc., and a speaker, etc.; a storage section including a hard disk, etc.; and a communication section including a network interface card such as a Local Area Network (LAN) card, a modem, etc. The communication section performs communication processing via a network such as the Internet. A drive is also connected to the I / O interface as required. A removable medium, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive as required, so that the computer program read from it can be installed into the storage section as required.

[0169] The above content is only a preferred exemplary embodiment of the present application and is not intended to limit the implementation of the present application. Those of ordinary skill in the art can easily make corresponding adaptations or modifications according to the main concept and spirit of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope required by the claims.

Claims

1. A data processing method, characterized in that, The method includes: For each object keyword included in each feedback statement, respectively determine the analysis sub-object in each feedback statement, and determine the initial score corresponding to each analysis sub-object according to the description keyword associated with the analysis sub-object in the feedback statement to which it belongs; wherein, each feedback statement is a description statement for the same target object, and each feedback statement corresponds to a comprehensive score for the target object. Calculate the target score corresponding to each analysis sub-object in each feedback statement according to the initial score corresponding to each analysis sub-object and the comprehensive score corresponding to the feedback statement to which each analysis sub-object belongs. Calculate the net promoter score of the target object according to the target scores corresponding to all analysis sub-objects respectively.

2. The method according to claim 1, characterized in that, The method further includes: Traverse multiple feedback statements, and use the traversed feedback statement as the target feedback statement. If the matching coefficient between the target candidate word segment and the preset noun is greater than the first preset matching threshold, determine that the target candidate word segment is the object keyword of the target feedback statement; wherein, the target candidate word segment is any candidate word segment determined from the target feedback statement. If the matching coefficient between the adjacent candidate word segment of the target object keyword and the preset description word is greater than the second preset matching threshold, determine that the adjacent candidate word segment is the description keyword associated with the analysis sub-object corresponding to the target object keyword.

3. The method according to claim 1, characterized in that The determining the initial score corresponding to each analysis sub-object according to the description keyword associated with the analysis sub-object in the feedback statement to which it belongs includes: Traverse each analysis sub-object, and use the traversed analysis sub-object as the target analysis sub-object. The description keyword associated with the target analysis sub-object in the feedback statement to which it belongs is the target description keyword, and the comprehensive score corresponding to the feedback statement to which the target analysis sub-object belongs is the target comprehensive score. If the attribute value of the target description keyword is greater than the preset threshold, the initial score corresponding to the target analysis sub-object is: the score in the first score interval with the smallest difference from the target comprehensive score. If the attribute value of the target description keyword is equal to the preset threshold, the initial score corresponding to the target analysis sub-object is: the score in the second score interval with the smallest difference from the target comprehensive score. If the attribute value of the target description keyword is less than the preset threshold, the initial score corresponding to the target analysis sub-object is: the score in the third score interval with the smallest difference from the target comprehensive score; wherein, the minimum value of the first score interval is greater than the maximum value of the second score interval, and the minimum value of the second score interval is greater than the maximum value of the third score interval.

4. The method according to claim 3, characterized in that, The calculating the target score corresponding to each analysis sub-object in each feedback statement according to the initial score corresponding to each analysis sub-object and the comprehensive score corresponding to the feedback statement to which each analysis sub-object belongs includes: For multiple analysis sub-objects belonging to the same feedback statement, calculate the average initial score to determine the average initial score corresponding to each feedback statement and the target score range to which each average initial score belongs, where the target score range is any one of the first score range, the second score range, and the third score range; If there is a difference between the average initial score and the comprehensive score corresponding to the focus feedback statement, which is greater than a preset error threshold, the initial score corresponding to the focus analysis sub-object is adjusted based on the target score range to determine the target score corresponding to each analysis sub-object in each feedback statement; wherein the focus analysis sub-object is the analysis sub-object in the focus feedback statement, and the initial score is within the target score range.

5. The method according to claim 3, characterized in that The calculating of the net recommendation value of the target object according to the target scores corresponding to all analyzed sub-objects includes: Calculating a first ratio of the number of positive analysis sub-objects to the number of all analysis sub-objects, and a second ratio of the number of negative analysis sub-objects to the number of all analysis sub-objects; wherein the positive analysis sub-objects are analysis sub-objects whose target scores are within the first score range, and the negative analysis sub-objects are analysis sub-objects whose target scores are within the third score range; A net recommendation value of the target object is determined based on a difference between the first proportion and the second proportion.

6. The method according to claim 5, wherein Each analysis sub-object corresponds to a corresponding analysis set, and the method further includes: Based on the difference between the third proportion and the fourth proportion corresponding to each analysis set, the net recommendation value of each target analysis set is determined; wherein the third proportion is the ratio of the number of all positive analysis sub-objects in each analysis set to the number of all analysis sub-objects in the respective analysis set, and the fourth proportion is the ratio of all negative analysis sub-objects in each analysis set to the number of all analysis sub-objects in the respective analysis set.

7. The method according to claim 1, wherein Each analysis sub-object corresponds to a corresponding analysis set, and the target score corresponding to each analysis sub-object is within any one of a first score interval, a second score interval, and a third score interval, wherein the minimum value of the first score interval is greater than the maximum value of the second score interval, and the minimum value of the second score interval is greater than the maximum value of the third score interval; the method further includes: In the target analysis set, the analysis sub-objects whose target scores are within the first score range are used as first point-dropping objects, the analysis sub-objects whose target scores are within the second score range are used as second point-dropping objects, and the analysis sub-objects whose target scores are within the third score range are used as third point-dropping objects; wherein the target analysis set is any one of the analysis sets; Calculating the point loss weight of the target analysis set based on the first point loss object, the second point loss object, the third point loss object and their corresponding point loss weights in the target analysis set; Based on the net recommendation value of the target object and the loss weight of the target analysis set, the loss value of the target analysis set is obtained to determine the degree of optimization of the target analysis set.

8. A data processing device, characterized in that, The device comprises: A determination module, for each object keyword included in each feedback statement, respectively determine the analysis sub-objects in each feedback statement, and determine the initial scores corresponding to each analysis sub-object according to the description keywords associated with the respective analysis sub-objects in the feedback statements to which they belong; wherein each feedback statement is a description statement for the same target object, and each feedback statement corresponds to a comprehensive score for the target object. An adjustment module, calculate the target scores corresponding to each analysis sub-object in each feedback statement according to the initial scores corresponding to each analysis sub-object and the comprehensive scores corresponding to the feedback statements to which the respective analysis sub-objects belong. A calculation module, calculate the net promoter score of the target object according to the target scores corresponding to all the analysis sub-objects respectively.

9. An electronic device, characterized in that, Including: A controller; A memory for storing one or more programs, which when executed by the controller, cause the controller to implement the method according to any one of claims 1 to 7.

10. A storage medium, characterized in that, A computer-readable instruction is stored thereon, which when executed by a processor of a computer, causes the computer to execute the method according to any one of claims 1 to 7.

Images