Information sending method, device, equipment, computer readable medium and program product

Abstract

The embodiment of the invention discloses an information sending method, device and equipment, a computer readable medium and a program product. A specific embodiment of the method comprises the steps of obtaining a reservation result for a target reservation object; obtaining communication data for the target reservation object in response to determining that the reservation result represents that the target reservation object is communicated with each other; performing data index analysis on the communication data to generate a first communication index information set; determining reservation object emotion prediction information corresponding to the communication data; and sending the first communication index information set and the reservation object emotion prediction information to a communication anomaly detection end. The implementation mode is related to artificial intelligence, abnormity check can be accurately and efficiently carried out on the reservation result corresponding to the reservation object, and the reservation success rate is guaranteed.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technologies, and more particularly, to an information sending method, apparatus, device, computer-readable medium, and program product. Background Art

[0002] Currently, for placing an order for an item, the relevant supplier needs to make a proper delivery to the order recipient to schedule the item delivery event. The current event scheduling method is usually: driven by artificial intelligence technology, using speech recognition technology to achieve event scheduling.

[0003] However, the inventors found that when using the above method to schedule an event, the following technical problems often exist:

[0004] The speech recognition technology is not accurate enough in recognizing the user's conversation intention, and there are often situations such as answering off-topic and recognition errors in the conversation, resulting in inaccurate scheduling and low scheduling efficiency.

[0005] The above information disclosed in this background art section is only used to enhance the understanding of the background of the inventive concept, and thus, it may include information that does not form the prior art known to those of ordinary skill in the art in this country. Summary of the Invention

[0006] This summary of the disclosure is used to introduce concepts in a brief form, which will be described in detail in the following detailed implementation section. This summary of the disclosure is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to be used to limit the scope of the claimed technical solution.

[0007] Some embodiments of the present disclosure provide an information sending method, apparatus, electronic device, computer-readable medium, and program product to solve the technical problems mentioned in the above background art section.

[0008] In a first aspect, some embodiments of the present disclosure provide an information sending method, including: obtaining a scheduling result for a target scheduling object; in response to determining that the scheduling result indicates that communication has been established with the target scheduling object, obtaining communication data for the target scheduling object; performing data index analysis on the communication data to generate a first communication index information set; determining prediction information on the emotion of the scheduling object corresponding to the communication data; and sending the first communication index information set and the prediction information on the emotion of the scheduling object to a communication anomaly detection end.

[0009] Optionally, the above method further includes: in response to determining that the above appointment result indicates that there is no communication with the above target appointment object, determining a second communication metric information set for the above appointment result, where the above second communication metric information set includes: appointment connection rate information, object location connection rate information, appointment timeout rate information; generating at least one prediction information on the reason for non-communication for the above second communication metric information set; and sending the above at least one prediction information on the reason for non-communication and the above second communication metric information set to the above communication anomaly detection end.

[0010] Optionally, the above communication data is voice communication data; and the above data metric analysis of the above communication data to generate a first communication metric information set includes: performing data voice metric analysis on the above communication data to generate at least one first communication metric.

[0011] Optionally, the above at least one first communication metric information includes: abnormal speech metric information and negative keyword metric information; and the above data voice metric analysis of the above communication data to generate at least one first communication metric information includes: performing text conversion on the above communication data to obtain a communication text; generating the above abnormal speech metric information for a preset communication abnormal speech set according to the above communication text; generating the above negative keyword metric information for a preset negative keyword set according to the above communication text.

[0012] Optionally, the above appointment object emotion prediction information includes: text emotion prediction information; and the above determination of the appointment object emotion prediction information corresponding to the above communication data includes: performing text preprocessing on the above communication text to obtain a text preprocessing result; inputting the above text preprocessing result into a pre-trained text emotion generation model to generate the above text emotion prediction information.

[0013] Optionally, the above appointment object emotion prediction information includes: voice emotion prediction information; and the above determination of the appointment object emotion prediction information corresponding to the above communication data includes: performing audio preprocessing on the above communication data to obtain processed voice data; inputting the above processed voice data into a pre-trained voice emotion generation model to generate the above voice emotion prediction information.

[0014] Optionally, the above-mentioned speech emotion generation model includes: an audio feature extraction network, an attention mechanism network, a multi-layer convolutional neural network, a bidirectional long short-term memory network, and a fully connected network; and the above-mentioned process of inputting the processed speech data into a pre-trained speech emotion generation model to generate the above-mentioned speech emotion prediction information includes: inputting the processed speech data into the above-mentioned audio feature extraction network to generate audio feature information; inputting the above-mentioned audio feature information into the above-mentioned attention mechanism network to generate attention feature information; inputting the above-mentioned attention feature information into the above-mentioned multi-layer convolutional neural network to generate convolutional feature information; performing feature processing on the above-mentioned convolutional feature information to obtain feature processing information; inputting the above-mentioned feature processing information into the above-mentioned bidirectional long short-term memory network to output long short-term memory feature information; inputting the above-mentioned long short-term memory feature information into the above-mentioned fully connected network to output the above-mentioned speech emotion prediction information.

[0015] Optionally, the above-mentioned first communication metric information set includes: call duration metric information and communication round metric information; and the above-mentioned process of performing data metric analysis on the above-mentioned communication data to generate the first communication metric information set includes: performing data voice metric analysis on the above-mentioned communication data to generate the above-mentioned call duration metric information and the above-mentioned communication round metric information.

[0016] Optionally, the above-mentioned reservation result is an intelligent automatic reservation result.

[0017] In a second aspect, some embodiments of the present disclosure provide an information sending device, including: a first obtaining unit configured to obtain a reservation result for a target reservation object; a second obtaining unit configured to obtain communication data for the target reservation object in response to determining that the reservation result indicates that communication has occurred with the target reservation object; an analysis unit configured to perform data metric analysis on the communication data to generate a first communication metric information set; a determination unit configured to determine reservation object emotion prediction information corresponding to the communication data; and a sending unit configured to send the first communication metric information set and the reservation object emotion prediction information to a communication anomaly detection end.

[0018] Optionally, the device further includes: in response to determining that the reservation result indicates that communication has not occurred with the target reservation object, determining a second communication metric information set for the reservation result, where the second communication metric information set includes: reservation connection rate information, object location connection rate information, reservation timeout rate information; generating at least one uncommunicated reason prediction information for the second communication metric information set; and sending the at least one uncommunicated reason prediction information and the second communication metric information set to the communication anomaly detection end.

[0019] Optionally, the above communication data is voice communication data; and the analysis unit can be configured to: analyze the voice metrics of the above communication data to generate at least one first communication metric information.

[0020] Optionally, the above at least one first communication metric information includes: abnormal speech metric information and negative keyword metric information; and the analysis unit can be configured to: perform text conversion on the above communication data to obtain communication text; generate the above abnormal speech metric information for a preset set of abnormal communication speech based on the above communication text; generate the above negative keyword metric information for a preset set of negative keywords based on the above communication text.

[0021] Optionally, the above appointment object emotion prediction information includes: text emotion prediction information; and the analysis unit can be configured to: perform text preprocessing on the above communication text to obtain a text preprocessing result; input the above text preprocessing result into a pre-trained text emotion generation model to generate the above text emotion prediction information.

[0022] Optionally, the above appointment object emotion prediction information includes: voice emotion prediction information; and the analysis unit can be configured to: perform audio preprocessing on the above communication data to obtain processed voice data; input the above processed voice data into a pre-trained voice emotion generation model to generate the above voice emotion prediction information.

[0023] Optionally, the above voice emotion generation model includes: an audio feature extraction network, an attention mechanism network, a multi-layer convolutional neural network, a bidirectional long short-term memory network, and a fully connected network; and the analysis unit can be configured to: input the above processed voice data into the above audio feature extraction network to generate audio feature information; input the above audio feature information into the above attention mechanism network to generate attention feature information; input the above attention feature information into the above multi-layer convolutional neural network to generate convolutional feature information; perform feature processing on the above convolutional feature information to obtain feature processing information; input the above feature processing information into the above bidirectional long short-term memory network to output long short-term memory feature information; input the above long short-term memory feature information into the above fully connected network to output the above voice emotion prediction information.

[0024] Optionally, the above first communication metric information set includes: call duration metric information and communication round metric information; and the analysis unit can be configured to: analyze the voice metrics of the above communication data to generate the above call duration metric information and the above communication round metric information.

[0025] Optionally, the above appointment result is an intelligent automatic appointment result.

[0026] In a third aspect, some embodiments of the present disclosure provide an electronic device, including: one or more processors; a storage device storing one or more programs thereon, and when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the method described in any implementation manner of the first aspect.

[0027] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium storing a computer program thereon, wherein when the program is executed by a processor, the method described in any implementation manner of the first aspect is implemented.

[0028] In a fifth aspect, some embodiments of the present disclosure provide a computer program product including a computer program, and when the computer program is executed by a processor, the method described in any implementation manner of the first aspect above is implemented.

[0029] The above various embodiments of the present disclosure have the following beneficial effects: Through the information sending method of some embodiments of the present disclosure, the abnormal check of the reservation result corresponding to the reservation object can be accurately and efficiently performed, ensuring the reservation success rate. Specifically, the reasons for the inaccuracy and inefficiency of the relevant reservations are as follows: The speech recognition technology is not accurate enough in recognizing the user's conversation intention, and there are often situations such as answering questions irrelevantly and misidentifying in the conversation, resulting in inaccurate reservations and low reservation efficiency. Based on this, the information sending method of some embodiments of the present disclosure first obtains the reservation result for the target reservation object for subsequent communication anomaly detection. Then, in response to determining that the above reservation result indicates that communication has been made with the above target reservation object, communication data for the above target reservation object is obtained as data support for subsequent determination of communication metric information and reservation object emotion prediction information. Next, data metric analysis is performed on the above communication data to generate a first set of communication metric information. Here, by generating communication metric information for subsequent communication anomaly detection by the communication anomaly detection end, it not only facilitates the subsequent detection of communication anomalies but also ensures the accuracy of communication anomaly detection. Furthermore, the reservation object emotion prediction information corresponding to the above communication data is determined. Here, by determining the reservation object emotion prediction information, the accuracy of subsequent communication anomaly judgment can be further improved, and the specific reservation situation for the target reservation object can be understood indirectly. Finally, the above first set of communication metric information and the reservation object emotion prediction information are sent to the communication anomaly detection end, so that the communication anomaly detection end can perform precise regulation of the reservation based on the received first set of communication metric information and the above reservation object emotion prediction information, ensuring the efficiency and accuracy of subsequent reservations. In summary, by generating the first set of communication metric information and the reservation object emotion prediction information, the reservation situation corresponding to the reservation result can be effectively monitored, ensuring the effective control of the accuracy of subsequent event reservations. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In combination with the accompanying drawings and with reference to the following specific embodiments, the above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic and the elements and elements are not necessarily drawn to scale.

[0031] Figure 1 is a schematic diagram of an application scenario of an information sending method according to some embodiments of the present disclosure;

[0032] Figure 2 is a flowchart of some embodiments of the information sending method according to the present disclosure;

[0033] Figure 3 is a flowchart of some other embodiments of the information sending method according to the present disclosure;

[0034] Figure 4 is a schematic diagram of a monitoring strategy in some embodiments of the information sending method according to the present disclosure on the premise that the reservation result representation and the above-mentioned target reservation object do not communicate with each other;

[0035] Figure 5 is a schematic structural diagram of some embodiments of the information sending device according to the present disclosure;

[0036] Figure 6 is a schematic structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure. Specific Embodiments

[0037] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not used to limit the protection scope of the present disclosure.

[0038] In addition, it should be noted that for the sake of convenience of description, only parts related to the relevant invention are shown in the drawings. Without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other.

[0039] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence relationship of the functions performed by these devices, modules or units.

[0040] It should be noted that the modifiers "one" and "multiple" mentioned in this disclosure are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise clearly specified in the context, it should be understood as "one or more".

[0041] The names of the messages or information exchanged between multiple devices in the embodiments of this disclosure are only for illustrative purposes and are not used to limit the scope of these messages or information.

[0042] Regarding operations such as the collection, storage, and use of data (such as communication data) involved in this disclosure, before performing the corresponding operations, relevant organizations or individuals shall fulfill obligations including conducting a data security impact assessment, fulfilling the obligation of informing the data subject, and obtaining the prior authorization and consent of the data subject.

[0043] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0044] Figure 1 It is a schematic diagram of an application scenario of an information sending method according to some embodiments of the present disclosure.

[0045] In Figure 1 In the application scenario, first, the electronic device 101 can obtain a reservation result 103 for a target reservation object. In this application scenario, the target reservation object can be the user 102. The reservation result 103 can be "have communicated with each other with the user 102". Then, in response to determining that the above reservation result 103 indicates that communication has been made with the above target reservation object, the electronic device 101 can obtain communication data 104 for the above target reservation object. Next, the electronic device 101 can perform data index analysis on the above communication data 104 to generate a first communication index information set 105. In this application scenario, the first communication index information set 105 includes: first communication index information 1051 and first communication index information 1052. The first communication index information 1051 can be "actual duration of communication voice: 3 minutes". The first communication index information 1052 can be "tone of voice: calm". Furthermore, the electronic device 101 can determine reservation object emotion prediction information 106 corresponding to the above communication data 104. In this application scenario, the reservation object emotion prediction information 106 can be "calm". Finally, the electronic device 101 can send the above first communication index information set 105 and the above reservation object emotion prediction information 106 to the communication anomaly detection end 107.

[0046] It should be noted that the above electronic device 101 can be hardware or software. When the electronic device is hardware, it can be implemented as a distributed cluster composed of multiple servers or terminal devices, or as a single server or a single terminal device. When the electronic device is embodied as software, it can be installed in the above-listed hardware devices. It can be implemented as, for example, multiple software or software modules for providing distributed services, or as a single software or software module. No specific limitation is made here.

[0047] It should be understood that Figure 1 the number of electronic devices in

[0048] Continuing to refer to Figure 2 , a flow 200 of some embodiments of the information sending method according to the present disclosure is shown. The information sending method includes the following steps:

[0049] Step 201, obtaining a reservation result for a target reservation object.

[0050] In some embodiments, the execution subject of the above information sending method (such as Figure 1 the electronic device 101 shown) can obtain the reservation result for the target reservation object through a wired connection method or a wireless connection method. Among them, the target reservation object can be an object for which an event reservation is to be made. In practice, the above target reservation object can be a reservation user for a large item order confirmation event. The above reservation result can be a result indicating whether a reservation communication has been made with the target reservation object. For example, for a reservation result of "01", the reservation result can indicate that a reservation communication has been made with the target reservation object. For a reservation result of "00", the reservation result can indicate that no reservation communication has been made with the target reservation object.

[0051] In some alternative implementation manners of some embodiments, the above reservation result is an intelligent automatic reservation result. That is, the reservation result can be a result of intelligent automatic reservation based on an artificial intelligence voice reservation terminal. The artificial intelligence voice reservation terminal can include, but is not limited to, at least one of the following: a voice recognition module, a voice conversion module, an automatic reservation module, and a telephone system module.

[0052] Step 202, in response to determining that the above reservation result indicates that mutual communication has been made with the above target reservation object, obtaining communication data for the above target reservation object.

[0053] In some embodiments, in response to determining that the above reservation result indicates that the above target reservation object has communicated with each other, the above execution entity may obtain communication data for the above target reservation object through a wired connection method or a wireless connection method. Among them, the communication data may be the reservation communication content with the target reservation object. The above communication data may be various forms of data. For example, the communication data may be chat record data with the target reservation object.

[0054] Step 203, perform data index analysis on the above communication data to generate a first set of communication index information.

[0055] In some embodiments, the above execution entity may perform data index analysis on the above communication data to generate a first set of communication index information. Among them, the first communication index information may be the index value of the first communication index. The above first communication index may represent the communication performance with respect to the target reservation object. The above first communication index may reflect to a certain extent whether there is abnormal communication during the reservation process with the target reservation object. In practice, the above first communication index may be but is not limited to one of the following: communication frequency, communication feedback duration, communication feedback timeout duration. The above communication frequency may represent the number of communications with the target reservation object. The communication feedback duration may be the total duration consumed by the target reservation object to feedback the reservation satisfaction. The communication feedback timeout duration may be the duration when the time for the target reservation satisfaction exceeds the predetermined time.

[0056] As an example, first, the above execution entity may perform data preprocessing on the communication data to obtain preprocessed data. Then, input the preprocessed data into a multi-communication index information generation model to output a first set of communication index information. In practice, the above multi-communication index information generation model may be a multi-head classification neural network model based on a temporal neural network.

[0057] In some optional implementation manners of some embodiments, the communication data is voice communication data. The voice communication data may be the voice content of the voice communication with the target reservation object. In practice, the above voice communication data may be telephone communication content.

[0058] Optionally, the above execution entity may perform data voice index analysis on the above communication data to generate at least one piece of first communication index information. Among them, the first communication index information in the at least one piece of first communication index information may be communication voice index information. In practice, the at least one piece of first communication index information may include: actual communication voice duration, voice intonation. The above actual communication voice duration may be the duration of the voice after removing the segments without the main body's voice in the communication voice. The voice intonation may be the intonation of the target reservation object's speech. For example, the voice intonation may be "peaceful".

[0059] As an example, first, the above-mentioned execution entity can perform data preprocessing on the communication data to obtain preprocessed data. Then, the preprocessed data is input into the voice metric information generation model to generate at least one first communication metric information. The above-mentioned voice metric information generation model can be a model for generating voice metric information. In practice, the above-mentioned voice metric information generation model can be a Hidden Markov Model (HMM)-Gaussian Mixed Model (GMM).

[0060] Optionally, the at least one first communication metric information includes: abnormal speech pattern metric information and negative keyword metric information. Among them, the abnormal speech pattern metric information can be the metric information of the included abnormal speech. The above-mentioned abnormal speech can be the speech indicating an abnormal reservation. The negative keyword metric information can be the metric information of the included negative keywords. The above-mentioned negative keywords can be words that indirectly indicate an abnormal reservation. For example, the negative keyword can be "disagree", or it can also be "no time". Among them, the abnormal speech pattern metric information can be the abnormal sentences in the communication data associated with the abnormal speech pattern. For example, the abnormal speech pattern can be "I want to make a complaint". The negative keyword metric information can be the abnormal sentences in the communication data associated with the negative keyword.

[0061] Optionally, the above-mentioned analysis of the data voice metrics of the communication data to generate at least one first communication metric information may include the following steps:

[0062] In the first step, the above-mentioned execution entity can perform text conversion on the above-mentioned communication data to obtain a communication text.

[0063] As an example, the above-mentioned execution entity can use a text conversion model to convert the communication data in audio form into text form to obtain a communication text. The text conversion model can be a model that converts audio into corresponding text.

[0064] In the second step, according to the above-mentioned communication text, the above-mentioned abnormal speech pattern metric information for the preset communication abnormal speech pattern set is generated. Among them, the preset communication abnormal speech can be the pre-set speech indicating an abnormal reservation. In practice, the abnormal speech pattern can be "I want to make a complaint".

[0065] As an example, first, the above-mentioned execution entity can perform sentence segmentation on the communication text to obtain a set of sentences. Then, for each sentence in the set of sentences, it is determined whether there is a speech pattern in the preset communication abnormal speech pattern set with the same sentence content as the above-mentioned sentence. Finally, according to the obtained determination information, the above-mentioned abnormal speech pattern metric information for the preset communication abnormal speech pattern set is generated.

[0066] Step 3: Generate the above-mentioned negative keyword index information for the preset negative keyword set according to the above communication text.

[0067] Among them, the preset negative keywords can be pre-set words that characterize appointment anomalies.

[0068] As an example, first, the above-mentioned execution entity can perform word segmentation on the communication text to obtain a word set. Then, match the word set with the preset negative keyword set to generate a matching result. Finally, generate negative keyword index information according to the matching result.

[0069] Optionally, the above-mentioned first communication index information set includes: call duration index information and communication round index information. Among them, the call duration index information can be the total audio duration corresponding to the communication data. That is, the duration consumed from the start to the end of the audio. The communication round index information can be the number of rounds of mutual responses with the target appointment object.

[0070] Optionally, the above-mentioned execution entity can perform data voice index analysis on the above-mentioned communication data to generate the above-mentioned call duration index information and the above-mentioned communication round index information.

[0071] As an example, the above-mentioned execution entity can perform data duration statistics and round statistics on the communication data to generate the above-mentioned call duration index information and the above-mentioned communication round index information.

[0072] Step 204: Determine the appointment object emotion prediction information corresponding to the above-mentioned communication data.

[0073] In some embodiments, the above-mentioned execution entity can determine the appointment object emotion prediction information corresponding to the above-mentioned communication data. Among them, the appointment object emotion prediction information can be the emotion prediction information of the target appointment object. In practice, the above-mentioned appointment object emotion prediction information can be one of the following: happy emotion, angry emotion, disappointed emotion, neutral emotion.

[0074] As an example, first, the above-mentioned execution entity can convert the communication data into text form to generate a communication text. Then, select at least one emotion keyword from the communication text. Finally, determine the appointment object emotion prediction information corresponding to the above-mentioned communication data according to the at least one emotion keyword.

[0075] In some optional implementation manners of some embodiments, the above-mentioned appointment object emotion prediction information includes: text emotion prediction information. Among them, the text emotion prediction information can be the emotion prediction information of the target appointment object reflected in the text.

[0076] Optionally, the above-mentioned determination of the appointment object emotion prediction information corresponding to the above-mentioned communication data may include the following steps:

[0077] In the first step, perform text preprocessing on the above communication text to obtain the text preprocessing result.

[0078] Among them, the above text preprocessing result can be in vector form.

[0079] As an example, first, the above execution entity can perform text tokenization on the communication text to obtain the tokenization result. Then, the above execution entity can perform word embedding on the tokenization result to obtain the text preprocessing result.

[0080] In the second step, input the above text preprocessing result into a pre-trained text emotion generation model to generate the above text emotion prediction information. Among them, the text emotion generation model can be a model for generating text emotion information. In practice, the text emotion generation model can be a bidirectional recurrent neural network (RNN) model.

[0081] In some optional implementation manners of some embodiments, the above appointment object emotion prediction information includes: voice emotion prediction information. Among them, the voice emotion prediction information can be the emotion prediction information that can be reflected in the voice and characterizes the target appointment object.

[0082] Optionally, determining the appointment object emotion prediction information corresponding to the above communication data may include the following steps:

[0083] In the first step, perform audio preprocessing on the above communication data to obtain the processed voice data.

[0084] As an example, first, the above execution entity can perform audio preprocessing on the communication data to convert it into a spectrogram as the processed voice data.

[0085] In the second step, input the above processed voice data into a pre-trained voice emotion generation model to generate the above voice emotion prediction information. Among them, the voice emotion generation model can be a model for generating voice emotion. In practice, the above voice emotion generation model can be a multi-layer convolutional neural network model.

[0086] Optionally, the above-mentioned speech emotion generation model includes: an audio feature extraction network, an attention mechanism network, a multi-layer convolutional neural network, a bidirectional long short-term memory network (Bi-LSTM, Bi-directional Long Short-Term Memory), and a fully connected network. Among them, the attention mechanism network can be a network based on spectrogram spatio-temporal attention + convolutional channel attention. The attention mechanism network can filter out the influence of low-value features on model training, determine which parts of the features on the spectrogram are key features, and determine which parts of the features are key features. The bidirectional long short-term memory network can utilize the context of the context to extract temporal feature information. In practice, the multi-layer convolutional neural network can be a multi-layer serially connected convolutional neural network. The fully connected network can be a classification network. The fully connected network can include: at least one serially connected fully connected layer. In practice, the above-mentioned fully connected network can be a 4-classification network. The output classification results of the fully connected network can include "happy, disappointed, angry, neutral".

[0087] Optionally, inputting the processed speech data into a pre-trained speech emotion generation model to generate the speech emotion prediction information may include the following steps:

[0088] First step, input the processed speech data into the audio feature extraction network to generate audio feature information. Among them, the audio feature extraction network can be a network model for extracting audio features. In practice, the above-mentioned audio feature extraction model can be a recurrent neural network model.

[0089] Second step, input the audio feature information into the attention mechanism network to generate attention feature information.

[0090] Third step, input the attention feature information into the multi-layer convolutional neural network to generate convolutional feature information.

[0091] Fourth step, perform feature processing on the convolutional feature information to obtain feature processing information.

[0092] Fifth step, input the feature processing information into the bidirectional long short-term memory network to output long short-term memory feature information.

[0093] Sixth step, input the long short-term memory feature information into the fully connected network to output the speech emotion prediction information.

[0094] Step 205, send the first communication metric information set and the appointment object emotion prediction information to the communication anomaly detection end.

[0095] In some embodiments, the above-mentioned execution entity may send the above-mentioned first communication metric information set and the above-mentioned reserved object emotion prediction information to the communication anomaly detection end. Among them, the communication anomaly detection end may be a terminal that detects whether a reservation is abnormal based on the first communication metric information set and the above-mentioned reserved object emotion prediction information.

[0096] The above-mentioned various embodiments of the present disclosure have the following beneficial effects: Through the information sending method of some embodiments of the present disclosure, it is possible to accurately and efficiently check for anomalies in the reservation results corresponding to the reserved objects, ensuring the reservation success rate. Specifically, the reasons for the inaccuracy and inefficiency of relevant reservations are as follows: The speech recognition technology is not precise enough in recognizing the conversation intentions of users, and there are often situations where the conversation answers are off-topic or misrecognized, resulting in inaccurate reservations and low reservation efficiency. Based on this, in some embodiments of the information sending method of the present disclosure, first, obtain the reservation result for the target reserved object for subsequent communication anomaly detection. Then, in response to determining that the above-mentioned reservation result indicates that communication has occurred with the above-mentioned target reserved object, obtain the communication data for the above-mentioned target reserved object as data support for subsequent determination of communication metric information and reserved object emotion prediction information. Next, perform data metric analysis on the above-mentioned communication data to generate a first communication metric information set. Here, by generating communication metric information for subsequent communication anomaly detection by the communication anomaly detection end, it not only facilitates subsequent communication anomaly detection but also ensures the accuracy of communication anomaly detection. Furthermore, determine the reserved object emotion prediction information corresponding to the above-mentioned communication data. Here, by determining the reserved object emotion prediction information, the accuracy of subsequent communication anomaly judgment can be further improved, and the specific reservation situation for the target reserved object can be understood indirectly. Finally, send the above-mentioned first communication metric information set and the above-mentioned reserved object emotion prediction information to the communication anomaly detection end so that the communication anomaly detection end can perform precise regulation of the reservation based on the received first communication metric information set and the above-mentioned reserved object emotion prediction information, ensuring the efficiency and accuracy of subsequent reservations. In summary, by generating the first communication metric information set and the reserved object emotion prediction information, it is possible to effectively monitor the reservation situation corresponding to the reservation result and ensure the effective control of the accuracy of subsequent event reservations.

[0097] Further referring to Figure 3 , flowchart 300 of some other embodiments of the information sending method according to the present disclosure is shown. The information sending method includes the following steps:

[0098] Step 301, obtain the reservation result for the target reserved object.

[0099] Step 302, in response to determining that the above-mentioned reservation result indicates that communication has occurred with the above-mentioned target reserved object, obtain the communication data for the above-mentioned target reserved object.

[0100] Step 303: Analyze the communication data to generate a first set of communication metric information.

[0101] Step 304: Determine the emotion prediction information of the appointment object corresponding to the above communication data.

[0102] Step 305: Send the above first set of communication metric information and the emotion prediction information of the appointment object to the communication anomaly detection end.

[0103] In some embodiments, for the specific implementation of steps 301 - 305 and the technical effects brought thereby, reference can be made to Figure 2 Steps 201 - 205 in the corresponding embodiments, which will not be elaborated here.

[0104] Step 306: In response to determining that the above appointment result representation and the target appointment object have not communicated with each other, determine a second set of communication metric information for the above appointment result.

[0105] In some embodiments, in response to determining that the above appointment result representation and the target appointment object have not communicated with each other, the execution subject (such as Figure 1 the electronic device 101 shown) can determine a second set of communication metric information for the above appointment result. Among them, the above second set of communication metric information includes: appointment connection rate information, connection rate information of the object's place of origin, appointment timeout rate information. The appointment connection rate information can be the appointment connection rate within a predetermined time period. The connection rate information of the object's place of origin can be the appointment connection rate of the place of origin of the target appointment object within a predetermined time period. The appointment timeout rate information can be the appointment timeout rate of the place of origin of the target appointment object within a predetermined time period. The appointment timeout rate can characterize the timeout situation of the appointment object's connection.

[0106] As an example, first, the above execution subject can determine the appointment time for the target appointment object. Then, determine the predetermined time period to which the above appointment time belongs. For example, the appointment time is 10 o'clock. The predetermined time period can be one of the following: 8 am - 12 pm, 12 pm - 6 pm, 6 pm - 11 pm, 11 pm - 8 am. Then, the predetermined time period to which the appointment time belongs is 8 am - 12 pm. Finally, determine the appointment connection rate information, connection rate information of the object's place of origin, and appointment timeout rate information within the predetermined time period to which the appointment time belongs.

[0107] Step 307: Generate at least one prediction information on the reason for non - communication for the above second set of communication metric information.

[0108] In some embodiments, the above-mentioned execution entity may generate at least one prediction information on the reason for non-communication for the above-mentioned second communication metric information set. Among them, the prediction information on the reason for non-communication may be the predicted reason for the target appointment object not to communicate. In practice, the above-mentioned prediction information on the reason for non-communication may be one of the following: blocked by local operator, user blacklisted, system anomaly.

[0109] As an example, the above-mentioned execution entity may generate at least one prediction information on the reason for non-communication by analyzing the metrics of the second communication metric information set.

[0110] As an example, referring to Figure 4 , a monitoring strategy is shown under the premise that the appointment result representation and the above-mentioned target appointment object do not communicate with each other. The specific monitoring strategy can be specifically monitored from two aspects. They are "monitoring granularity" and "monitoring metrics" (that is, the second communication metric information set). The above-mentioned "monitoring granularity" may include: "region" granularity, "time period" granularity, "operator" granularity. The "monitoring metrics" may include: "connection rate", "proportion of overdue orders", and "orders that cannot be connected due to appointment failure".

[0111] Step 308, send the above-mentioned at least one prediction information on the reason for non-communication and the above-mentioned second communication metric information set to the above-mentioned communication anomaly detection end.

[0112] In some embodiments, the above-mentioned execution entity may send the above-mentioned at least one prediction information on the reason for non-communication and the above-mentioned second communication metric information set to the above-mentioned communication anomaly detection end.

[0113] From Figure 3 it can be seen that compared with the description of some embodiments corresponding to Figure 2 , the process 300 of the information sending method in some embodiments corresponding to Figure 3 performs metric analysis and reason analysis through the situation where the appointment result representation and the above-mentioned target appointment object do not communicate with each other, which is convenient for subsequent effective monitoring of object appointment based on at least one prediction information on the reason for non-communication and the above-mentioned second communication metric information set, ensuring the appointment success rate and improving the appointment efficiency.

[0114] Further referring to Figure 5 , as an implementation of the methods shown in the above figures, some embodiments of an information sending device are provided in the present disclosure. These device embodiments correspond to those method embodiments shown in Figure 2 , and the information sending device can be specifically applied to various electronic devices.

[0115] As Figure 5As shown, an information sending device 500 includes: a first acquisition unit 501, a second acquisition unit 502, an analysis unit 503, a determination unit 504, and a sending unit 505. Among them, the first acquisition unit 501 is configured to acquire a reservation result for a target reservation object; the second acquisition unit 502 is configured to, in response to determining that the above reservation result indicates that communication has been established with the above target reservation object, acquire communication data for the above target reservation object; the analysis unit 503 is configured to perform data index analysis on the above communication data to generate a first communication index information set; the determination unit 504 is configured to determine reservation object emotion prediction information corresponding to the above communication data; the sending unit 505 is configured to send the above first communication index information set and the above reservation object emotion prediction information to a communication anomaly detection end.

[0116] In some alternative implementation manners of some embodiments, the above device 500 further includes: an information determination unit, an information generation unit, and an information sending unit (not shown in the figure). Among them, the information determination unit may be configured to: in response to determining that the above reservation result indicates that communication has not been established with the above target reservation object, determine a second communication index information set for the above reservation result, where the above second communication index information set includes: reservation connection rate information, object location connection rate information, reservation timeout rate information; the information generation unit may be configured to: generate at least one prediction information on the reason for non-communication for the above second communication index information set; the information sending unit may be configured to: send the above at least one prediction information on the reason for non-communication and the above second communication index information set to the above communication anomaly detection end.

[0117] In some alternative implementation manners of some embodiments, the above communication data is voice communication data; and the analysis unit 503 may be configured to: perform data voice index analysis on the above communication data to generate at least one first communication index information.

[0118] In some alternative implementation manners of some embodiments, the above at least one first communication index information includes: abnormal speech index information and negative keyword index information; and the analysis unit 503 may be configured to: perform text conversion on the above communication data to obtain communication text; generate the above abnormal speech index information for a preset communication abnormal speech set according to the above communication text; generate the above negative keyword index information for a preset negative keyword set according to the above communication text.

[0119] In some alternative implementations of some embodiments, the above-mentioned reserved object emotion prediction information includes: text emotion prediction information; and the analysis unit 503 can be configured to: perform text preprocessing on the above-mentioned communication text to obtain a text preprocessing result; input the above-mentioned text preprocessing result into a pre-trained text emotion generation model to generate the above-mentioned text emotion prediction information.

[0120] In some alternative implementations of some embodiments, the above-mentioned reserved object emotion prediction information includes: voice emotion prediction information; and the analysis unit 503 can be configured to: perform audio preprocessing on the above-mentioned communication data to obtain processed voice data; input the above-mentioned processed voice data into a pre-trained voice emotion generation model to generate the above-mentioned voice emotion prediction information.

[0121] In some alternative implementations of some embodiments, the above-mentioned voice emotion generation model includes: an audio feature extraction network, an attention mechanism network, a multi-layer convolutional neural network, a bidirectional long short-term memory network, and a fully connected network; and the analysis unit 503 can be configured to: input the above-mentioned processed voice data into the above-mentioned audio feature extraction network to generate audio feature information; input the above-mentioned audio feature information into the above-mentioned attention mechanism network to generate attention feature information; input the above-mentioned attention feature information into the above-mentioned multi-layer convolutional neural network to generate convolutional feature information; perform feature processing on the above-mentioned convolutional feature information to obtain feature processing information; input the above-mentioned feature processing information into the above-mentioned bidirectional long short-term memory network to output long short-term memory feature information; input the above-mentioned long short-term memory feature information into the above-mentioned fully connected network to output the above-mentioned voice emotion prediction information.

[0122] In some alternative implementations of some embodiments, the above-mentioned first communication metric information set includes: call duration metric information and communication round metric information; and the analysis unit 503 can be configured to: perform data voice metric analysis on the above-mentioned communication data to generate the above-mentioned call duration metric information and the above-mentioned communication round metric information.

[0123] In some alternative implementations of some embodiments, the above-mentioned reservation result is an intelligent automatic reservation result.

[0124] It can be understood that the various units described in the information sending device 500 correspond to the respective steps in the method described with reference to Figure 2 Therefore, the operations, features, and beneficial effects described above for the method also apply to the information sending device 500 and the units included therein, and will not be repeated here.

[0125] Next, with reference to Figure 6 which shows an electronic device suitable for implementing some embodiments of the present disclosure (for exampleFigure 1 Schematic diagram of the structure of the electronic device 101)600 in Figure 6 The electronic device shown is merely an example and should not impose any limitations on the functions and scope of use of the embodiments of the present disclosure.

[0126] As Figure 6 shown, the electronic device 600 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 601, which may perform various appropriate actions and processes according to the program stored in the read-only memory 602 or the program loaded from the storage device 608 into the random access memory 603. In the random access memory 603, various programs and data required for the operation of the electronic device 600 are also stored. The processing device 601, the read-only memory 602, and the random access memory 603 are connected to each other through a bus 604. The input / output interface 605 is also connected to the bus 604.

[0127] Generally, the following devices may be connected to the input / output interface 605: an input device 606 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 6 the electronic device 600 with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. Instead, more or fewer devices may be implemented or had. Figure 6 Each block shown in

[0128] Specifically, according to some embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes program codes for performing the methods shown in the flowcharts. In such some embodiments, the computer program may be downloaded and installed from the network through the communication device 609, or installed from the storage device 608, or installed from the read-only memory 602. When the computer program is executed by the processing device 601, the above functions defined in the methods of some embodiments of the present disclosure are executed.

[0129] It should be noted that, in some embodiments of the present disclosure, the above-mentioned computer-readable medium may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, 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 with 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 or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. 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, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0130] In some embodiments, the client and the server can communicate using any currently known or future-developed network protocol such as HTTP (HyperText Transfer Protocol), and can be interconnected with digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include local area networks ("LANs"), wide area networks ("WANs"), the Internet (e.g., the Internet), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0131] The above computer-readable medium may be included in the above electronic device; or may exist separately without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device is caused to: obtain a reservation result for a target reservation object; in response to determining that the reservation result indicates that communication has occurred with the target reservation object, obtain communication data for the target reservation object; perform data index analysis on the communication data to generate a first communication index information set; determine reservation object emotion prediction information corresponding to the communication data; and send the first communication index information set and the reservation object emotion prediction information to a communication anomaly detection end.

[0132] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may execute entirely on the user's computer, partially on the user's computer, execute as a stand-alone software package, execute partially on the user's computer and partially on a remote computer, or execute entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., by connecting through the Internet using an Internet service provider).

[0133] 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 disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than 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 and / or flowchart, and combinations of blocks in the block diagram and / 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.

[0134] The units described in some embodiments of the present disclosure can be implemented in software or in hardware. The described units can also be provided in a processor. For example, it can be described as: a processor includes a first acquisition unit, a second acquisition unit, an analysis unit, a determination unit, and a sending unit. Among them, the names of these units do not constitute a limitation on the unit itself in some cases. For example, the first acquisition unit can also be described as "the unit for acquiring the reservation result for the target reservation object".

[0135] The functions described above herein can be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that can be used include: Field Programmable Gate Array (FPGA), Application Specific Integrated Circuit (ASIC), Application Specific Standard Product (ASSP), System on Chip (SOC), Complex Programmable Logic Device (CPLD), and so on.

[0136] Some embodiments of the present disclosure also provide a computer program product, including a computer program, which implements any of the above information sending methods when executed by a processor.

[0137] The above description is only some preferred embodiments of the present disclosure and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by mutually replacing the above features with the (but not limited to) technical features having similar functions disclosed in the embodiments of the present disclosure.

Claims

1. An information sending method, comprising: Obtain the reservation result for the target reservation object; In response to determining that the reservation result indicates that there has been mutual communication with the target reservation object, obtain the communication data for the target reservation object; Perform data index analysis on the communication data to generate a first set of communication index information; Determine the reservation object emotion prediction information corresponding to the communication data; Send the first set of communication index information and the reservation object emotion prediction information to the communication anomaly detection end.

2. The method according to claim 1, wherein, The method further includes: In response to determining that the reservation result indicates that there has been no mutual communication with the target reservation object, determine a second set of communication index information for the reservation result, where the second set of communication index information includes: reservation connection rate information, object location connection rate information, reservation timeout rate information; Generate at least one prediction information on the reason for no communication for the second set of communication index information; Send the at least one prediction information on the reason for no communication and the second set of communication index information to the communication anomaly detection end.

3. The method according to claim 1, wherein, The communication data is voice communication data; And The performing data index analysis on the communication data to generate a first set of communication index information includes: Perform data voice index analysis on the communication data to generate at least one first communication index information.

4. The method according to claim 3, wherein, The at least one first communication index information includes: abnormal speech pattern index information and negative keyword index information; and The performing data voice index analysis on the communication data to generate at least one first communication index information includes: Perform text conversion on the communication data to obtain communication text; Generate the abnormal speech pattern index information for a preset set of communication abnormal speech patterns according to the communication text; Generate the negative keyword index information for a preset set of negative keywords according to the communication text.

5. The method according to claim 4, wherein, The reservation object emotion prediction information includes: text emotion prediction information; and The determining the reservation object emotion prediction information corresponding to the communication data includes: Perform text preprocessing on the communication text to obtain a text preprocessing result; Input the text preprocessing result into a pre-trained text emotion generation model to generate the text emotion prediction information.

6. The method according to claim 3, wherein, The reservation object emotion prediction information includes: voice emotion prediction information; and The determining the reservation object emotion prediction information corresponding to the communication data includes: Perform audio preprocessing on the communication data to obtain processed voice data; Input the processed voice data into a pre-trained voice emotion generation model to generate the voice emotion prediction information.

7. The method according to claim 6, wherein, The voice emotion generation model includes: an audio feature extraction network, an attention mechanism network, a multi-layer convolutional neural network, a bidirectional long short-term memory network, and a fully connected network; and The inputting the processed voice data into a pre-trained voice emotion generation model to generate the voice emotion prediction information includes: Input the processed voice data into the audio feature extraction network to generate audio feature information; Input the audio feature information into the attention mechanism network to generate attention feature information; Input the attention feature information into the multi-layer convolutional neural network to generate convolutional feature information; Perform feature processing on the convolutional feature information to obtain feature processing information; Input the feature processing information into the bidirectional long short-term memory network to output long short-term memory feature information; Input the long short-term memory feature information into the fully connected network to output the speech emotion prediction information.

8. The method according to claim 3, wherein The first communication metric information set includes: call duration metric information and communication round metric information; and The data metric analysis of the communication data to generate the first communication metric information set includes: Perform data voice metric analysis on the communication data to generate the call duration metric information and the communication round metric information.

9. The method according to claim 1, wherein The reservation result is an intelligent automatic reservation result.

10. An information sending device, comprising: A first acquisition unit configured to acquire a reservation result for a target reservation object; A second acquisition unit configured to, in response to determining that the reservation result indicates that mutual communication has occurred with the target reservation object, acquire communication data for the target reservation object; An analysis unit configured to perform data metric analysis on the communication data to generate a first communication metric information set; A determination unit configured to determine reservation object emotion prediction information corresponding to the communication data; A sending unit configured to send the first communication metric information set and the reservation object emotion prediction information to a communication anomaly detection end.

11. An electronic device, comprising: One or more processors; A storage device having stored thereon one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1-9.

12. A computer-readable medium having a computer program stored thereon, wherein The computer program, when executed by a processor, implements the method according to any one of claims 1-9.

13. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1-9.

Images