Order processing method and apparatus, electronic device, medium, chip and product

Abstract

This disclosure provides an order processing method, apparatus, electronic device, medium, chip, and product. The method includes: responding to a user's refund request by obtaining the user's first feedback tone and first feedback text; determining a first risk value corresponding to the user based on the first feedback tone and first feedback text using a pre-trained model; and performing a first processing on the user based on the first risk value. This method, by utilizing the user's first feedback tone and first feedback text to determine the user's risk value, thereby judging whether the user's refund request is reasonable, improves the reasonableness of processing customer refund requests and protects the rights and interests of both customers and merchants.

Description

Technical Field

[0001] This disclosure relates to the field of data processing, and more particularly to an order processing method, apparatus, electronic device, medium, chip, and product. Background Technology

[0002] In handling customer complaints, we often encounter legitimate malfunctions such as product damage that require compensation to users. However, some users may fabricate facts to demand a full refund. Customer service representatives, lacking the skills to verify the authenticity of the user's claims, may directly grant a refund, causing unnecessary losses. Similarly, even if a user has a refund record, if the current complaint reflects a genuine situation, customer service may be unable to determine its veracity and thus refuse to grant a refund, potentially leading to a negative user experience. Summary of the Invention

[0003] This disclosure provides an order processing method, apparatus, electronic device, medium, chip, and product to address problems in the related art and improve the reasonableness of processing customer refund requests.

[0004] A first aspect of this disclosure provides an order processing method, the method comprising: in response to a user's refund request, obtaining the user's first feedback tone and first feedback text; based on the first feedback tone and first feedback text, using a pre-trained model, determining a first risk value corresponding to the user; and performing a first processing on the user based on the first risk value.

[0005] In some embodiments of this disclosure, in response to a user's refund request, obtaining the user's feedback data includes: obtaining the user's first feedback audio based on the refund request; converting the first feedback audio into first feedback text using a speech recognition algorithm based on the first feedback audio; and determining the user's first feedback tone in the first feedback audio using a feature extraction algorithm based on the first feedback audio.

[0006] In some embodiments of this disclosure, determining the user's first feedback tone in the first feedback audio using a feature extraction algorithm includes: extracting fundamental frequency parameters, sound intensity parameters, and prosodic feature parameters from the first feedback audio using a feature extraction algorithm; and determining the first feedback tone based on at least one of the pitch information indicated by the fundamental frequency parameters, the speech loudness information indicated by the sound intensity parameters, and the rhythm and intonation information indicated by the prosodic feature parameters.

[0007] In some embodiments of this disclosure, determining the first risk value corresponding to a user based on feedback data and using a pre-trained model includes: semantically splitting the first feedback text based on the word segmentation module in the pre-trained model to obtain the first word segment; determining the feature label set of the first word segment using the encoder in the pre-trained model based on the first word segment, wherein the feature label set includes at least one of the following: similarity feature, time feature, spatial feature, sentiment feature, and context feature; and determining the user's first risk value based on the tone of the first feedback and the feature label set.

[0008] In some embodiments of this disclosure, determining a user's first risk value based on the first feedback tone and feature tag set includes: determining the credibility of a second risk value based on the first feedback tone and feature tag set; when the credibility is greater than or equal to a preset credibility, determining the second risk value as the first risk value; when the credibility is less than the preset credibility, obtaining the user's second feedback audio based on the first feedback tone and first feedback text to determine the first risk value.

[0009] In some embodiments of this disclosure, performing a first process on a user based on a first risk value includes: refunding the user when the first risk value is less than a preset risk value and the user has no historical refund record; obtaining on-site video or images corresponding to the refund request when the first risk value is less than the preset risk value and the user has a historical refund record to determine whether to refund the user; and refusing to refund the user when the first risk value is greater than or equal to the preset risk value.

[0010] In some embodiments of this disclosure, the method further includes: obtaining a training dataset, which includes at least user problem description data, customer service feedback data, and user risk probability; constructing a model to be trained based on a semantic model and a long short-term memory network; determining the problem description data as the input sequence dataset of the model to be trained, and determining the customer service feedback data and user risk probability as the output label dataset of the model to be trained; and training the model to be trained based on the input sequence data and the output label dataset to obtain a pre-trained model.

[0011] A second aspect of this disclosure provides an order processing apparatus, comprising: an acquisition unit, configured to acquire a user's first feedback tone and first feedback text in response to a user's refund request; a determination unit, configured to determine a first risk value corresponding to the user based on the first feedback tone and first feedback text using a pre-trained model; and a processing unit, configured to perform a first processing on the user based on the first risk value.

[0012] A third aspect of this disclosure provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the methods described in the first aspect of this disclosure.

[0013] A fourth aspect of this disclosure provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to perform the methods described in the first aspect of this disclosure.

[0014] A fifth aspect of this disclosure provides a chip characterized by including at least one processor and a communication interface; the communication interface is used to receive signals input to the chip or signals output from the chip, and the processor communicates with the communication interface and implements the method described in the first aspect of this disclosure through logic circuits or executing code instructions.

[0015] A sixth aspect of this disclosure provides a computer program product including a computing program stored on a computer-readable storage medium. The computer program includes program instructions that, when executed by a computer, cause the computer to perform the method in any implementation of the first aspect described above.

[0016] In summary, the order processing method proposed in this disclosure includes: responding to a user's refund request by obtaining the user's initial feedback tone and initial feedback text; determining the user's corresponding initial risk value based on the initial feedback tone and initial feedback text using a pre-trained model; and performing initial processing on the user based on the initial risk value. This method, by utilizing the user's initial feedback tone and initial feedback text to determine the user's risk value, thereby judging the reasonableness of the user's refund request, improves the reasonableness of processing customer refund requests and protects the rights and interests of both customers and merchants.

[0017] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure, and are not intended to unduly limit this disclosure.

[0019] Figure 1 A flowchart illustrating an order processing method provided in an embodiment of this disclosure;

[0020] Figure 2 A flowchart illustrating an order processing method provided in an embodiment of this disclosure;

[0021] Figure 3 A flowchart illustrating an order processing method provided in an embodiment of this disclosure;

[0022] Figure 4 This is a schematic diagram of the structure of an order processing device provided in an embodiment of the present disclosure;

[0023] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure;

[0024] Figure 6 This is a schematic diagram of the chip structure provided in an embodiment of this disclosure. Detailed Implementation

[0025] Embodiments of this disclosure are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this disclosure, and should not be construed as limiting this disclosure.

[0026] In handling customer complaints, we often encounter legitimate issues such as product damage that require compensation to users. However, many users fabricate facts to demand full refunds. Customer service representatives, lacking the skills to verify the authenticity of user claims, may directly grant refunds, causing unnecessary losses. Similarly, even if a user has a refund record, if the current complaint reflects a genuine situation, customer service may be unable to determine its veracity and thus refuse to grant a refund, potentially leading to a negative user experience.

[0027] Therefore, in order to solve the above problems, this disclosure proposes an order processing method. This method determines the user's risk value by obtaining the user's initial feedback tone and feedback text when submitting order feedback, thereby judging whether the user's refund request is reasonable, improving the reasonableness of processing customer refund requests, and protecting the rights and interests of customers and merchants.

[0028] The specific details of this method are as follows.

[0029] Figure 1 This is a flowchart illustrating an order processing method provided in an embodiment of this disclosure. Figure 1 As shown, the method may include the following steps.

[0030] Step 101: In response to the user's refund request, obtain the user's initial feedback tone and initial feedback text.

[0031] In some embodiments, when a refund request is received from a user, customer service can ask the user questions to allow the user to describe the reason for the refund, product details, on-site situation, and other order-related information via voice, thereby obtaining the user's initial feedback tone and initial feedback text.

[0032] The aforementioned customer service representatives can be human or intelligent models; this disclosure does not specify which.

[0033] In other words, based on the refund request, the user's first feedback audio can be obtained, and based on the first feedback audio, the user's first feedback tone and first feedback text can be obtained.

[0034] In some embodiments, a speech recognition algorithm can be used to recognize the speech description in the first feedback audio as a text description, thereby obtaining the first feedback text; however, it is not limited thereto. When the user also directly provides text feedback in addition to the first audio, the text description and the text feedback from the user can be combined to obtain the first feedback text. This disclosure does not limit this.

[0035] In some instances, feature extraction algorithms can be used to extract features from the first feedback audio to determine the user's initial tone of voice based on the extracted audio parameters.

[0036] Step 102: Based on the tone and text of the first feedback, use a pre-trained model to determine the first risk value corresponding to the user.

[0037] In some embodiments, a pre-trained model can be used to semantically segment the first feedback text to break down the complex text into one or more first words, so that the pre-trained model can understand the semantics of the first feedback text.

[0038] Then, the first segmented word is encoded to determine one or more feature table labels of the first segmented word.

[0039] In some embodiments, a pre-trained model can be used to combine feature labels and first feedback tone to determine the user's first risk value.

[0040] In some embodiments, a first risk value is used to indicate the veracity of the description of the refund request in the first feedback audio.

[0041] Step 103: Based on the first risk value, perform the first processing on the user.

[0042] In some embodiments, the same or different first treatment methods may be applied to users based on different first risk values.

[0043] In some optional embodiments, when the first risk value is less than the preset risk value and the user has no history of refunds, it indicates that the user has good credit, that is, the probability of the user maliciously requesting a refund is low, and the user can be refunded directly in this case.

[0044] In some optional embodiments, when the first risk value is less than the preset risk value and the user has a history of refunds, it indicates that the probability of the user maliciously refunding this order is low. However, since the user has a previous refund record, in order to further verify whether the user's description is true, the user can be asked to provide on-site videos or images to further determine the user's behavior.

[0045] Furthermore, when the video or image provided by the user matches the description in the first feedback text, a refund will be issued to the user; when the video or image provided by the user differs from the description in the first feedback text, a refund can be refused to the user using pre-set scripts or other methods.

[0046] In some optional embodiments, when the first risk value is greater than or equal to the preset risk value, it indicates that the probability of the user maliciously requesting a refund is high. In this case, the user can be refused a refund by using preset wording or other methods.

[0047] In other words, the first step may include any one of the following three approaches: refunding the user; obtaining on-site video or images corresponding to the refund request to determine whether to refund the user; or refusing to refund the user.

[0048] In summary, the order processing method disclosed herein includes: responding to a user's refund request by obtaining the user's initial feedback tone and initial feedback text; determining the user's corresponding initial risk value using a pre-trained model based on the initial feedback tone and initial feedback text; and performing initial processing on the user based on the initial risk value. This method, by utilizing the user's initial feedback tone and initial feedback text to determine the user's risk value, thereby judging the reasonableness of the user's refund request, improves the reasonableness of processing customer refund requests and protects the rights and interests of both customers and merchants.

[0049] Figure 2 This is a flowchart illustrating an order processing method provided in an embodiment of this disclosure. Figure 2 As shown, based on Figure 1 The illustrated embodiment shows that the method includes the following steps.

[0050] Step 201: Based on the refund request, obtain the user's first feedback audio.

[0051] In some embodiments, when a refund request is received from a user, customer service can ask the user questions to allow the user to describe the reason for the refund, product details, on-site situation, and other order-related information via voice, thereby obtaining the user's first audio recording.

[0052] Step 202: Based on the first feedback audio, use a speech recognition algorithm to convert the first feedback audio into first feedback text.

[0053] In some embodiments, speech recognition algorithms such as Hidden Markov Models and Vector Quantization can be used to recognize the speech in the first feedback audio, so as to convert the recognized language into text and obtain the first feedback text. This disclosure does not limit the speech recognition algorithms used.

[0054] Step 203: Based on the first feedback audio, use a feature extraction algorithm to determine the user's first feedback tone in the first feedback audio.

[0055] In some embodiments, feature extraction algorithms such as line spectrum frequency algorithm and perceptual linear prediction algorithm can be used to extract features from the speech in the first feedback audio to obtain the fundamental frequency parameter, sound intensity parameter and prosodic feature parameter in the first feedback audio.

[0056] Then, based on at least one of the pitch information indicated by the fundamental frequency parameter, the speech loudness information indicated by the sound intensity parameter, and the rhythm and intonation information indicated by the prosodic feature parameter, the first feedback tone is determined.

[0057] For example, if the user's voice is determined to be low based on the fundamental frequency parameter and intermittent based on the prosodic feature parameter, then the first feedback tone can be determined to be a guilty tone.

[0058] For example, if the user's voice is determined to be steady based on the loudness information and continuous based on the prosodic feature parameters, then the first feedback tone can be determined as a declarative tone.

[0059] Step 204: Based on the word segmentation module in the pre-trained model, perform semantic segmentation on the first feedback text to obtain the first word segment.

[0060] In some embodiments, semantic segmentation can be performed on the first feedback text based on the word segmentation module to obtain the first segmented word.

[0061] For example, if the first feedback text is "vehicle brake failure", then the first participles are "vehicle", "brake", and "failure".

[0062] Step 205: Based on the first word segmentation, use the encoder in the pre-trained model to determine the feature label set of the first word segmentation.

[0063] In some embodiments, feature labels may include at least one of the following: similarity features, temporal features, spatial features, sentiment features, and contextual features.

[0064] In some embodiments, an encoder may be used to encode the first word segment to convert the first word segment into a set of feature labels that can be recognized by a pre-trained model.

[0065] Step 206: Determine the user's first risk value based on the first feedback tone and feature tag set.

[0066] In some embodiments, the pre-trained model can be made to understand the semantics in the first feedback text based on the feature label set.

[0067] Furthermore, by combining the tone of the feedback with the semantic understanding, the user's initial risk level can be determined.

[0068] In some embodiments, a first risk value is used to indicate the veracity of the description of the refund request in the first feedback audio.

[0069] In some embodiments, the credibility of the second risk value can be determined based on the first feedback tone and feature tag set; when the credibility is greater than or equal to a preset credibility, the second risk value is determined as the first risk value; when the credibility is less than the preset credibility, the user's second feedback audio is obtained based on the first feedback tone and the first feedback text to determine the first risk value.

[0070] For example, when the tone of the first feedback is declarative, but the pre-trained model determines that the description in the first feedback text is contradictory based on the feature label set, the pre-trained model cannot determine whether the user is lying. That is, the credibility of the second risk value determined at this time is low. Therefore, based on the tone and text of the user's first feedback, the model can continue to ask questions to obtain the user's second feedback audio. Then, the first audio and the second audio can be combined to determine the first risk value.

[0071] For example, if the first feedback tone is guilty and the pre-trained model determines, based on the feature label set, that the description in the first feedback text is contradictory, then the pre-trained model determines that the user is more likely to be lying. In this case, the user's second risk value can be determined to be a larger value, and the credibility of the second risk value can be determined to be higher. At this point, the second risk value can be determined as the first risk value.

[0072] For example, when the tone of the first feedback is declarative, and the pre-trained model determines that the description in the first feedback text is consistent based on the feature label set, the pre-trained model determines that the probability of the user lying is small. Therefore, the user's second risk value can be determined to be a small value, and the credibility of the second risk value can be determined to be high. At this time, the second risk value can be determined as the first risk value.

[0073] Step 207: Based on the first risk value, perform the first processing on the user.

[0074] In some embodiments, the same or different first treatment methods may be applied to users based on different first risk values.

[0075] In some optional embodiments, when the first risk value is less than the preset risk value and the user has no history of refunds, it indicates that the user has good credit, that is, the probability of the user maliciously requesting a refund is low, and the user can be refunded directly in this case.

[0076] In some optional embodiments, when the first risk value is less than the preset risk value and the user has a history of refunds, it indicates that the probability of the user maliciously refunding this order is low. However, since the user has a previous refund record, in order to further verify whether the user's description is true, the user can be asked to provide on-site videos or images to further determine the user's behavior.

[0077] Furthermore, when the video or image provided by the user matches the description in the first feedback text, a refund will be issued to the user; when the video or image provided by the user differs from the description in the first feedback text, a refund can be refused to the user using pre-set scripts or other methods.

[0078] In some optional embodiments, when the first risk value is greater than or equal to the preset risk value, it indicates that the probability of the user maliciously requesting a refund is high. In this case, the user can be refused a refund by using preset wording or other methods.

[0079] In other words, the first step may include any one of the following three approaches: refunding the user; obtaining on-site video or images corresponding to the refund request to determine whether to refund the user; or refusing to refund the user.

[0080] In summary, the order processing method disclosed herein includes: obtaining the user's first feedback audio based on the refund request; converting the first feedback audio into first feedback text using a speech recognition algorithm; determining the user's first feedback tone using a feature extraction algorithm based on the first feedback audio; semantically segmenting the first feedback text using a word segmentation module in a pre-trained model to obtain first word segments; determining the feature tag set of the first word segments using an encoder in the pre-trained model; determining the user's first risk value based on the first feedback tone and the feature tag set; and performing first processing on the user based on the first risk value. This method, by utilizing the user's first feedback tone and first feedback text to determine the user's risk value, thereby judging the reasonableness of the user's refund request and improving the reasonableness of processing customer refund requests; furthermore, by determining the credibility of the second risk value, it further ensures the accuracy of the first risk value, further improving the accuracy of judging the reasonableness of the user's refund request and protecting the rights and interests of both customers and merchants.

[0081] Figure 3 This is a flowchart illustrating an order processing method provided in an embodiment of this disclosure. Figure 3 As shown, based on Figure 1 The illustrated embodiment shows that the method includes the following steps.

[0082] Step 301: Obtain the training dataset.

[0083] In some embodiments, the training dataset includes at least user problem description data, customer service feedback data, and user risk probability, wherein the problem description data, customer service feedback data, and user risk probability correspond one-to-one.

[0084] For example, using statement A as user problem description data and statement B as customer service feedback data, with parentheses indicating annotations, the training dataset can be as follows:

[0085] B: Is this the order?

[0086] A: Yes.

[0087] B: Were you injured? Please describe the situation at the time.

[0088] A: I was injured, I even broke a bone (very brief, without describing which part).

[0089] B: We suggest you call 120 immediately... (before he could finish speaking).

[0090] A: I know, I know all that. Let's not talk about that. I just want you to cancel this order for me. I've already claimed insurance (risk probability is 60%, tone is impatient).

[0091] For example, using statement C as user problem description data and statement D as customer service feedback data, with parentheses indicating annotations, the training dataset can be as follows:

[0092] D: Could you please describe the situation at that time?

[0093] C: The headlights weren't working. I fell while riding, but nothing serious happened. There was a small pothole on the road (description contains details).

[0094] D: Are you injured?

[0095] C: I'm definitely injured. I have a 5-6 cm gash from a small slippery surface (specific cause of injury).

[0096] D: What are your requests?

[0097] C: First, cancel the fee for this order. Second, fix the car's headlights (to be more objective). We suggest you go through insurance.

[0098] D: Okay.

[0099] C: Also, I would like to suggest that this vehicle with non-functional lights is a malfunctioning vehicle and should not be deployed further. Also, it is unreasonable for you to charge me (the risk probability is 10%, so I am describing the problem objectively again).

[0100] D: Okay, I suggest you report this order as faulty.

[0101] Step 302: Construct the model to be trained based on the semantic model and the long short-term memory network.

[0102] In some embodiments, the semantic model is, for example, the BERT model, the improved RoBERT model, the ALBERT model, etc., which are not limited in this disclosure.

[0103] In some instances, the output of the last hidden state of the semantic model can be used as the input to a long short-term memory network.

[0104] In some alternative embodiments, a two-layer long short-term memory network (LSTM) can be configured to make the training of the model contextually relevant, but this disclosure does not limit the scope of the LSM network.

[0105] In some alternative embodiments, a dropout layer can be set for the model to be trained to make the hidden units in the long short-term memory network more closely connected and improve the adaptability of the pre-trained model. This disclosure does not limit the dropout rate, which is, for example, 0.3.

[0106] In some alternative embodiments, a fully connected layer can also be set for the model to be trained, so that the output of the fully connected layer is used as the output of the model to be trained.

[0107] In some alternative embodiments, the loss of the model to be trained can still be the classification cross-entropy function.

[0108] Step 303: Determine the problem description data as the input sequence dataset for the model to be trained, and determine the customer service feedback data and user risk probability as the output label dataset for training the model.

[0109] In some embodiments, problem description data can be determined as the input sequence dataset of the model to be trained, and customer service feedback data and user risk probability can be determined as the output label dataset of the model to be trained, so that the model to be trained learns the relationship between problem description data, customer service feedback data and user risk probability.

[0110] Step 304: Based on the input sequence data and the output label dataset, train the model to be trained to obtain a pre-trained model.

[0111] In some embodiments, the model to be trained is trained based on the input sequence data and the output label dataset. When the test results of the model to be trained meet the expectations, or when the iteration of the model to be trained reaches the preset number of iterations, the training is stopped, and the model obtained by training is determined as the pre-trained model.

[0112] In summary, the order processing method disclosed herein includes: obtaining a training dataset; constructing a model to be trained based on a semantic model and a long short-term memory network; determining the problem description data as the input sequence dataset of the model to be trained, and determining the customer service feedback data and user risk probability as the output label dataset of the training model; and training the model to be trained based on the input sequence data and the output label dataset to obtain a pre-trained model. This method, by determining the problem description data as the input sequence dataset of the model to be trained and the customer service feedback data and user risk probability as the output label dataset of the training model, allows the pre-trained model to better understand the relationship between the problem description data, customer service feedback data, and user risk probability, thereby improving the accuracy of the first risk value.

[0113] Figure 4 This is a schematic diagram of the structure of an order processing device 400 provided in an embodiment of this disclosure. Figure 4 As shown, the device includes:

[0114] The acquisition unit 410 is used to acquire the user's first feedback tone and first feedback text in response to the user's refund request;

[0115] The determination unit 420 is used to determine the first risk value corresponding to the user based on the first feedback tone and the first feedback text, using a pre-trained model;

[0116] Processing unit 430 is used to perform first processing on the user based on a first risk value.

[0117] In some embodiments, the acquisition unit 410 is further configured to: acquire the user's first feedback audio based on the refund request; convert the first feedback audio into first feedback text using a speech recognition algorithm based on the first feedback audio; and determine the user's first feedback tone in the first feedback audio using a feature extraction algorithm based on the first feedback audio.

[0118] In some embodiments, the acquisition unit 410 is further configured to extract the fundamental frequency parameter, sound intensity parameter, and prosodic feature parameter from the first feedback audio using a feature extraction algorithm; and determine the first feedback tone based on at least one of the pitch information indicated by the fundamental frequency parameter, the speech loudness information indicated by the sound intensity parameter, and the rhythm and intonation information indicated by the prosodic feature parameter.

[0119] In some embodiments, the determining unit 420 is further configured to: semantically segment the first feedback text based on the word segmentation module in the pre-trained model to obtain the first word segment; determine the feature label set of the first word segment using the encoder in the pre-trained model based on the first word segment, wherein the feature label set includes at least one of the following: similarity features, time features, spatial features, sentiment features, and context features; and determine the user's first risk value based on the tone of the first feedback and the feature label set.

[0120] In some embodiments, the determining unit 420 is further configured to: determine the credibility of the second risk value based on the first feedback tone and the feature tag set; when the credibility is greater than or equal to a preset credibility, determine the second risk value as the first risk value; when the credibility is less than the preset credibility, obtain the user's second feedback audio based on the first feedback tone and the first feedback text to determine the first risk value.

[0121] In some embodiments, the processing unit 430 is further configured to: acquire a training dataset, the training dataset including at least user problem description data, customer service feedback data, and user risk probability; construct a model to be trained based on a semantic model and a long short-term memory network; determine the problem description data as the input sequence dataset of the model to be trained, and determine the customer service feedback data and user risk probability as the output label dataset of the training model; and train the model to be trained based on the input sequence data and the output label dataset to obtain a pre-trained model.

[0122] In summary, the order processing apparatus disclosed herein includes: an acquisition unit, configured to acquire the user's first feedback tone and first feedback text in response to a user's refund request; a determination unit, configured to determine a first risk value corresponding to the user based on the first feedback tone and first feedback text using a pre-trained model; and a processing unit, configured to perform a first processing on the user based on the first risk value. This apparatus, by utilizing the user's first feedback tone and first feedback text to determine the user's risk value, thereby judging whether the user's refund request is reasonable, improves the reasonableness of processing customer refund requests and protects the rights and interests of both customers and merchants.

[0123] The methods and apparatus provided in the embodiments of this application have been described above. To implement the functions of the methods provided in the embodiments of this application, the electronic device may include a hardware structure and software modules, and may implement the above functions in the form of a hardware structure, software modules, or a hardware structure plus software modules. One of the above functions may be executed in the form of a hardware structure, software modules, or a hardware structure plus software modules.

[0124] Figure 5 This is a block diagram illustrating an electronic device 500 for implementing the above-described method according to an exemplary embodiment. For example, the electronic device 500 may be a mobile phone, computer, messaging device, game console, tablet device, medical device, fitness equipment, personal digital assistant, etc.

[0125] Reference Figure 5 The electronic device 500 may include one or more of the following components: processing component 502, memory 504, power supply component 506, multimedia component 508, audio component 510, input / output (I / O) interface 512, sensor component 514, and communication component 516.

[0126] Processing component 502 typically controls the overall operation of electronic device 500, such as operations associated with display, telephone calls, data communication, camera operation, and recording. Processing component 502 may include one or more processors 520 to execute instructions to perform all or part of the steps of the methods described above. Furthermore, processing component 502 may include one or more modules to facilitate interaction between processing component 502 and other components. For example, processing component 502 may include a multimedia module to facilitate interaction between multimedia component 508 and processing component 502.

[0127] Memory 504 is configured to store various types of data to support the operation of electronic device 500. Examples of such data include instructions for any application or method operating on electronic device 500, contact data, phonebook data, messages, pictures, videos, etc. Memory 504 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0128] Power supply component 506 provides power to various components of electronic device 500. Power supply component 506 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 500.

[0129] Multimedia component 508 includes a screen that provides an output interface between electronic device 500 and user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of touch or swipe actions but also the duration and pressure associated with the touch or swipe operation. In some embodiments, multimedia component 508 includes a front-facing camera and / or a rear-facing camera. When electronic device 500 is in an operating mode, such as a shooting mode or video mode, the front-facing camera and / or rear-facing camera may receive external multimedia data. Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

[0130] Audio component 510 is configured to output and / or input audio signals. For example, audio component 510 includes a microphone (MIC) configured to receive external audio signals when electronic device 500 is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 504 or transmitted via communication component 516. In some embodiments, audio component 510 also includes a speaker for outputting audio signals.

[0131] I / O interface 512 provides an interface between processing component 502 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0132] Sensor assembly 514 includes one or more sensors for providing state assessments of various aspects of electronic device 500. For example, sensor assembly 514 may detect the on / off state of electronic device 500, the relative positioning of components such as the display and keypad of electronic device 500, changes in position of electronic device 500 or a component of electronic device 500, the presence or absence of user contact with electronic device 500, orientation or acceleration / deceleration of electronic device 500, and temperature changes of electronic device 500. Sensor assembly 514 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 514 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 514 may also include an accelerometer, gyroscope, magnetometer, pressure sensor, or temperature sensor.

[0133] Communication component 516 is configured to facilitate wired or wireless communication between electronic device 500 and other devices. Electronic device 500 can access wireless networks based on communication standards, such as WiFi, 2G or 3G, 4G LTE, 5G NR (NewRadio), or combinations thereof. In one exemplary embodiment, communication component 516 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 516 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0134] In an exemplary embodiment, the electronic device 500 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the methods described above.

[0135] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 504 including instructions, which can be executed by a processor 520 of an electronic device 500 to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0136] Embodiments of this disclosure also provide a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to perform the methods described in the above embodiments of this disclosure.

[0137] Embodiments of this disclosure also provide a computer program product, including a computer program that is executed by a processor using the methods described in the above embodiments of this disclosure.

[0138] Figure 6 This is a schematic diagram illustrating the structure of a chip 600 for implementing the above method according to an exemplary embodiment. (Refer to...) Figure 6 The chip 600 includes a communication interface 601 and at least one processor 602. The communication interface 601 is used to receive signals input to the chip 600 or signals output from the chip 600. The processor 602 communicates with the communication interface 601 and implements the methods described in the above embodiments of this disclosure through logic circuits or executing code instructions.

[0139] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0140] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in at least one embodiment or example.

[0141] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0142] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processing module, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having at least one wiring (control method), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0143] It should be understood that various parts of the embodiments of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0144] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0145] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc.

[0146] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. An order processing method, characterized by, The method includes: In response to a user's refund request, obtain the user's first feedback tone and first feedback text; Based on the first feedback tone and the first feedback text, a pre-trained model is used to determine the first risk value corresponding to the user. Based on the first risk value, the first process is performed on the user.

2. The method of claim 1, wherein, In response to a user's refund request, obtaining the user's feedback data includes: Based on the refund request, obtain the user's first feedback audio. Based on the feedback audio, a speech recognition algorithm is used to convert the feedback audio into first feedback text; Based on the feedback audio, a feature extraction algorithm is used to determine the user's first feedback tone in the first feedback audio.

3. The method of claim 2, wherein, The step of determining the user's first feedback tone in the feedback audio using a feature extraction algorithm based on the feedback audio includes: Using feature extraction algorithms, the fundamental frequency parameter, sound intensity parameter, and prosodic feature parameter are extracted from the feedback audio. The first feedback tone is determined based on the pitch information indicated by the fundamental frequency parameter, the speech loudness information indicated by the sound intensity parameter, and the rhythm and intonation information indicated by the prosodic feature parameter.

4. The method of claim 1, wherein, The step of determining the first risk value corresponding to the user based on the feedback data and using a pre-trained model includes: Based on the word segmentation module in the pre-trained model, the first feedback text is semantically segmented to obtain the first word; Based on the first word segmentation, the encoder in the pre-trained model is used to determine the feature label set of the first word segmentation, wherein the feature label set includes at least one of the following: similarity features, temporal features, spatial features, sentiment features, and contextual features; Based on the first feedback tone and the feature tag set, the user's first risk value is determined.

5. The method of claim 4, wherein, The process of determining the user's first risk value based on the feedback expectation and the feature label set includes: Based on the first feedback tone and the feature label set, the credibility of the second risk value is determined; When the confidence level is greater than or equal to the preset confidence level, the second risk value is determined as the first risk value; When the credibility is less than the preset credibility, the user's second feedback audio is obtained based on the first feedback tone and the first feedback text to determine the first risk value.

6. The method of claim 1, wherein, The first processing of the user based on the first risk value includes: When the first risk value is less than the preset risk value and the user has no history of refunds, a refund will be issued to the user. When the first risk value is less than the preset risk value, and the user has the historical refund record, obtain the on-site video or image corresponding to the refund request to determine whether to refund the user; When the first risk value is greater than or equal to the preset risk value, the refund to the user is refused.

7. The method of claim 1, wherein, The method further includes: Obtain a training dataset, which includes at least user problem description data, customer service feedback data, and user risk probability; A training model is constructed based on a semantic model and a long short-term memory network. The problem description data is determined as the input sequence dataset of the model to be trained, and the customer service feedback data and the user risk probability are determined as the output label dataset of the training model. The model to be trained is trained based on the input sequence data and the output label dataset to obtain the pre-trained model.

8. An order processing apparatus, the apparatus comprising: The acquisition unit is used to acquire the user's first feedback tone and first feedback text in response to the user's refund request; The determining unit is used to determine the first risk value corresponding to the user based on the first feedback tone and the first feedback text, using a pre-trained model; A processing unit is configured to perform a first processing on the user based on the first risk value.

9. An electronic device, comprising: include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A non-transitory computer-readable storage medium having stored thereon computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-7.

11. A chip, characterized by It includes at least one processor and a communication interface; the communication interface is used to receive signals input to the chip or signals output from the chip, and the processor communicates with the communication interface and implements the method as described in any one of claims 1 to 7 through logic circuits or executing code instructions.

12. A computer program product comprising a computer program, wherein, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images