Incoming call question automatic answering method, device and equipment and storage medium
By converting text and word segmentation to the incoming phone calls received by bank customer service, combining classifiers and FAQ databases, automated answers are achieved, which solves the burden of bank customer service personnel in dealing with repetitive customer complaints and improves user experience.
Patent Information
- Application Number
- CN202411922979.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-25
- Publication Date
- 2025-04-01
AI Technical Summary
Bank customer service personnel need to frequently deal with repetitive customer complaint calls, which leads to excessive burden and affects the user experience.
By converting the received incoming voices and processing the word segmentation, and using preset classifiers and FAQ databases to automatically determine questions and answers, generate voice feedback, and achieve automated answers.
It reduces the burden of answering repetitive questions by bank customer service personnel, improves the response speed and efficiency of user questions, and improves the user experience.
Smart Images
Figure CN119743553A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of digital employee technology, and in particular to a method, device, equipment and storage medium for automatically answering incoming call questions. Background Art
[0002] With the development of telephone banking, the workload of bank agents is also increasing. Bank customer service staff often receive customer complaint calls about bank system failures, such as "a certain function does not work well", "a certain function cannot be opened", "a certain function cannot be found", etc. Bank customer service staff need to answer each customer complaint call one by one.
[0003] However, the questions in the customer complaint calls received by bank customer service staff are often highly repetitive, so bank customer service staff often need to answer single repetitive questions, resulting in a heavy burden on bank customer service staff; and because bank customer service staff need to answer each customer complaint call one by one, users need to queue up to get the corresponding answer, resulting in a poor user experience. Summary of the invention
[0004] In order to reduce the burden on bank customer service personnel and improve user experience, the present application provides a method, device, equipment and storage medium for automatically answering incoming call questions.
[0005] In a first aspect, the present application provides a method for automatically answering incoming call questions, comprising:
[0006] Convert the received incoming call voice into text to obtain the target text;
[0007] Performing word segmentation processing on the target text to obtain a target word segmentation result;
[0008] Determine a target question in the FAQ database corresponding to the incoming call voice based on the target word segmentation result, a preset classifier and a FAQ database;
[0009] An answer voice is obtained based on the target question and a frequently asked question and answer database, and the answer voice is fed back to the user.
[0010] Through the above implementation, the user's incoming call voice can be automatically converted into text and segmented in sequence, and the frequently asked questions corresponding to the incoming call voice can be further determined from the frequently asked questions database, and then the voice of the answer corresponding to the frequently asked questions can be synthesized and automatically fed back to the user. In this way, there is no need for bank customer service personnel to answer each customer complaint call one by one, and the user's incoming call questions can be answered in a short time, which can reduce the burden on bank customer service personnel and improve user experience.
[0011] Preferably, the step of converting the received incoming call voice into text to obtain the target text includes:
[0012] Processing the received incoming call voice based on a preset acoustic model to obtain a digital sound signal;
[0013] The digital sound signal is processed based on a preset speech model to obtain a target text.
[0014] Through the above implementation, the user's incoming call voice can be converted into corresponding text, which makes it easier to determine the answer corresponding to the incoming call voice based on the text later.
[0015] Preferably, the performing word segmentation processing on the target text to obtain a target word segmentation result includes:
[0016] Segmenting the target text based on a preset segmentation library to obtain a primary segmentation result;
[0017] The stop words in the primary word segmentation result are removed to obtain the target word segmentation result.
[0018] Through the above implementation, stop words in the primary word segmentation results are removed, which facilitates improving the accuracy of the subsequent determination of the answer corresponding to the electronic voice.
[0019] Preferably, determining a target question in the FAQ database corresponding to the incoming call voice based on the target word segmentation result, a preset classifier and a FAQ database includes:
[0020] Determine a similarity set based on the target word segmentation result, a preset classifier and a frequently asked question database;
[0021] A target question corresponding to the incoming call voice is determined based on the similarity set and the frequently asked questions and answers database.
[0022] Through the above implementation, the question that best matches the incoming call voice can be determined from the FAQ database according to the similarity, thereby facilitating improving the accuracy and efficiency of the determined question.
[0023] Preferably, the determining of the similarity set based on the target word segmentation result, a preset classifier and a frequently asked questions database includes:
[0024] Vectorizing the target word segmentation result to obtain a result vector;
[0025] Determine the call intent by processing the result vector based on a preset classifier;
[0026] A similarity set is obtained by performing similarity calculation based on the incoming call intention and a preset frequently asked questions database.
[0027] Through the above implementation, by vectorizing the target word segmentation result, it is convenient for the classifier to determine the call intent corresponding to the target word segmentation result, so as to facilitate the subsequent similarity calculation based on the call intent and the frequently asked questions database.
[0028] Preferably, obtaining the answer voice based on the target question and the frequently asked questions and answers database includes:
[0029] Determine a target answer corresponding to the target question from the FAQ database;
[0030] Perform speech synthesis on the target answer to obtain the answer speech.
[0031] Through the above implementation, the obtained target answer is voice synthesized, which facilitates feedback of the answer in voice form to the user, thereby facilitating improving the user experience.
[0032] In a second aspect, the present application provides a device for automatically answering incoming call questions, comprising:
[0033] A text conversion module, used to convert the received incoming call voice into a target text;
[0034] A word segmentation processing module, used to perform word segmentation processing on the target text to obtain a target word segmentation result;
[0035] A question determination module, configured to determine a target question in the FAQ database corresponding to the incoming call voice based on the target word segmentation result, a preset classifier and a FAQ database;
[0036] The voice feedback module is used to obtain an answer voice based on the target question and the frequently asked questions and answers database, and to feed back the answer voice to the user.
[0037] Through the above implementation, the user's incoming call voice can be automatically converted into text and segmented in sequence, and the frequently asked questions corresponding to the incoming call voice can be further determined from the frequently asked questions database, and then the voice of the answer corresponding to the frequently asked questions can be synthesized and automatically fed back to the user. In this way, there is no need for bank customer service personnel to answer each customer complaint call one by one, and the user's incoming call questions can be answered in a short time, which can reduce the burden on bank customer service personnel and improve user experience.
[0038] In a third aspect, the present application provides a computer device, the computer device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the steps in the above method when executing the computer program.
[0039] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which implements the steps in the above method when executed by a processor.
[0040] In a fifth aspect, the present application further provides a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, the steps in any of the above method embodiments are implemented.
[0041] The above-mentioned method, device, equipment and storage medium for automatically answering incoming call questions obtain a target text by converting the received incoming call voice into text; performing word segmentation processing on the target text to obtain a target word segmentation result; determining the target question corresponding to the incoming call voice in the FAQ database based on the target word segmentation result, a preset classifier and a FAQ database; obtaining an answer voice based on the target question and the FAQ database, and feeding back the answer voice to the user. Through the above implementation, the user's incoming call voice can be automatically converted into text and segmented in sequence, and the FAQ corresponding to the incoming call voice can be further determined from the FAQ database, and then the voice corresponding to the answer to the FAQ is synthesized, and the user is automatically fed back. In this way, there is no need for bank customer service personnel to answer each customer complaint call one by one, and the user's incoming call questions can be answered in a short time, which can reduce the burden on bank customer service personnel and improve the user experience.
[0042] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.
[0044] Figure 1 A flow chart of a method for automatically answering incoming call questions provided in an embodiment of the present application;
[0045] Figure 2 A schematic diagram of the structure of a device for automatically answering incoming call questions provided in an embodiment of the present application;
[0046] Figure 3 A schematic diagram of the structure of a computer device provided in an embodiment of the present application;
[0047] Figure 4 This is a diagram of the internal structure of a computer-readable storage medium provided in an embodiment of the present application. DETAILED DESCRIPTION
[0048] In order to make the purpose, technical solution and advantages of the present disclosure more clear, the present disclosure is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure and are not used to limit the present disclosure.
[0049] It should be noted that the terms "first", "second", etc. in the specification and claims of this article and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of this article described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or equipment that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.
[0050] In this article, the term "and / or" is only a description of the association relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the related objects before and after are in an "or" relationship.
[0051] Embodiment 1
[0052] Figure 1 This is a flow chart of a method for automatically answering incoming call questions provided in Example 1 of the present application, refer to Figure 1 The method may be performed by a device for performing the method, and the device may be implemented by software and / or hardware. The method includes:
[0053] S110: Convert the received incoming call voice into text to obtain a target text.
[0054] Among them, if the user encounters a problem when using the banking system, he or she can report the problem to the remote banking center platform by phone. For example, the problems that customers may encounter include: a certain function of the banking system is not easy to use, a certain function of the banking system cannot be opened, and a certain function of the banking system cannot be found; after the user makes a call to the remote banking center platform, the remote banking center platform automatically receives the incoming voice from the user; and uses the built-in voice recognition system to convert the incoming voice into text, which means converting the incoming voice into corresponding text in text form, and recording the text in text form as the target text.
[0055] Specifically, in response to receiving the incoming voice from the user, the remote banking center platform converts the incoming voice into text through a built-in voice recognition system to obtain the target text.
[0056] S120: Perform word segmentation processing on the target text to obtain a target word segmentation result.
[0057] Wherein, the target text is a paragraph, and the remote bank center platform can perform word segmentation processing on the target text, which means dividing the target text into multiple words in a certain way, and the target word segmentation result is the multiple words obtained after the word segmentation processing of the target text. For example, the target text is "the ID card photo recognition function in the bank system is not easy to use", and the target word segmentation results obtained by word segmentation processing of the target text are "bank system", "ID card photo recognition function", and "not easy to use".
[0058] Specifically, the remote bank center platform performs word segmentation processing on the target text to obtain a target word segmentation result.
[0059] S130: Determine a target question in the FAQ database corresponding to the incoming call voice based on the target word segmentation result, a preset classifier, and a FAQ database.
[0060] Among them, the preset classifier is used to process the target word segmentation result input therein to determine the type of call intention corresponding to the target word segmentation result. In this embodiment, the types of call intentions are mainly divided into three types: consultation, processing and complaint. In other embodiments, there is no specific limitation; a plurality of frequently asked questions and answer pairs are stored in the frequently asked questions database, each question and answer pair includes a frequently asked question and the answer to the question corresponding to the frequently asked question; by comparing the type of call intention output by the classifier with the frequently asked questions in the frequently asked questions database, the frequently asked questions that are closest in meaning to the type of call intention can be determined, and the frequently asked questions that are closest in meaning to the type of call intention are recorded as the target questions corresponding to the incoming call voice.
[0061] Specifically, the target word segmentation result is input into a preset classifier for processing to obtain the type of incoming call intention, and then the type of incoming call intention is compared with the common questions in the FAQ database to determine the common questions that are closest in meaning to the type of incoming call intention, and the common questions are used as the target questions corresponding to the incoming call voice.
[0062] S140: Obtain an answer voice based on the target question and a frequently asked question and answer database, and feed back the answer voice to the user.
[0063] Among them, the target question is a common question in the FAQ database. The common question has a corresponding answer in the FAQ database. The answer to the question is a text answer. The answer to the question is processed by voice processing to obtain a voice answer, and the voice answer is recorded as the answer voice; the remote banking center platform generates the answer voice and feeds back the answer voice to the user in the form of a telephone call, so that the user can obtain the answer corresponding to the incoming call voice.
[0064] Specifically, the answer to the question corresponding to the target question is obtained from the FAQ database, the answer to the question is processed by voice processing to obtain the answer voice, and then the answer voice is fed back to the user in the form of a telephone call.
[0065] It should be noted that in this embodiment, the target text is obtained by text conversion of the received incoming call voice; the target text is segmented to obtain the target segmentation result; the target question corresponding to the incoming call voice in the FAQ database is determined based on the target segmentation result, the preset classifier and the FAQ database; the answer voice is obtained based on the target question and the FAQ database, and the answer voice is fed back to the user. Through the above implementation, the user's incoming call voice can be automatically converted to text and segmented in sequence, and the FAQ corresponding to the incoming call voice can be further determined from the FAQ database, and then the voice corresponding to the answer to the FAQ is synthesized and automatically fed back to the user. In this way, there is no need for bank customer service personnel to answer each customer complaint call one by one, and the user's call question can be answered in a short time, which can reduce the burden on bank customer service personnel and improve the user experience.
[0066] Embodiment 2
[0067] A method for automatically answering incoming call questions provided in the second embodiment of the present application is a refinement of the "converting the received incoming call voice into text to obtain a target text" in the first embodiment; it should be noted that for the parts not described in detail in this embodiment, reference can be made to the descriptions of other embodiments, and the method includes:
[0068] S211. Process the received incoming call voice based on a preset acoustic model to obtain a digital sound signal.
[0069] Among them, the voice recognition system is provided with an acoustic model, which is used to process the incoming call voice input into a digital sound signal. Exemplarily, the digital sound signal is a signal in the form of syllables or pinyin expressed in digital form that can be processed by the machine.
[0070] Specifically, the received incoming call voice is input into a preset acoustic model for processing, thereby obtaining a corresponding digital sound signal.
[0071] S212: Process the digital sound signal based on a preset speech model to obtain a target text.
[0072] Among them, the speech recognition system also has a speech model, which is used to process the digital sound signal into a target text, that is, text in text form.
[0073] Specifically, the obtained digital sound signal is input into a preset speech model for processing, thereby obtaining the corresponding target text.
[0074] S210: Perform word segmentation processing on the target text to obtain a target word segmentation result.
[0075] S210: Determine a target question in the FAQ database corresponding to the incoming call voice based on the target word segmentation result, a preset classifier, and a FAQ database.
[0076] S210: Obtain an answer voice based on the target question and a frequently asked question and answer database, and feed back the answer voice to the user.
[0077] Embodiment 3
[0078] A method for automatically answering incoming call questions provided in Embodiment 3 of the present application refines the "performing word segmentation processing on the target text to obtain a target word segmentation result" in Embodiment 1; it should be noted that for the parts not described in detail in this embodiment, reference may be made to the descriptions of other embodiments, and the method includes:
[0079] S310: Convert the received incoming call voice into text to obtain a target text.
[0080] S321. Segment the target text based on a preset segmentation library to obtain a primary segmentation result.
[0081] Among them, the preset word segmentation library in this embodiment specifically adopts the Python third-party open source library jieba, which is not specifically limited in other embodiments. The command "jieba.lcut()" in the word segmentation library can be used to segment the target text in an accurate model. The word segmentation library also combines a custom dictionary to enhance the ambiguity correction capability to segment the target text, thereby improving the accuracy of the target text segmentation; the primary word segmentation result is the result obtained after the preset word segmentation library processes the target text; exemplary. The target text is "the ID card photo recognition function in the bank system is not easy to use", and the primary word segmentation results corresponding to the target text are "bank system", "in", "ID card photo recognition function", "not easy to use" and ".
[0082] Specifically, the target text is segmented by using the command "jieba.lcut()" in the preset third-party open-source library jieba of Python, so as to obtain the primary segmentation result.
[0083] S322. Remove the stop words in the primary segmentation result to obtain the target segmentation result.
[0084] Among them, the primary segmentation result contains some words with low semantic contribution, and the words with low semantic contribution are recorded as stop words. The stop words will affect the accuracy of subsequent processing of the target text. Therefore, it is necessary to remove the stop words in the primary segmentation result. In this embodiment, a statistical-based automatic learning algorithm is specifically used to identify the stop words in the primary segmentation result and delete the stop words. In other embodiments, there is no specific limitation. Exemplarily, common stop words include: auxiliary words or modal particles such as "de", "le", "en", etc. The result obtained after deleting the stop words from the primary segmentation result is also the target segmentation result.
[0085] Specifically, a statistical-based automatic learning algorithm is used to identify the stop words in the primary segmentation result and delete the stop words, so as to obtain the target segmentation result.
[0086] S330. Based on the target segmentation result, a preset classifier, and a common question-and-answer database, determine the target question in the common question-and-answer database corresponding to the incoming call voice.
[0087] S340. Obtain an answer voice based on the target question and the common question-and-answer database, and feedback the answer voice to the user.
[0088] Embodiment 4
[0089] A method for automatically answering incoming call questions provided in Embodiment 4 of the present application refines the "determine the target question in the common question-and-answer database corresponding to the incoming call voice based on the target segmentation result, a preset classifier, and the common question-and-answer database" in Embodiment 1. It should be noted that for parts not detailed in this embodiment, reference can be made to the descriptions of other embodiments. The method includes:
[0090] S410. Perform text conversion on the received incoming call voice to obtain the target text.
[0091] S420. Perform segmentation processing on the target text to obtain the target segmentation result.
[0092] S431. Based on the target segmentation result, a preset classifier, and the common question-and-answer database, determine a similarity set.
[0093] Among them, the corresponding call intention type can be obtained by processing the target word segmentation result through a preset classifier. The call intention types in this embodiment specifically include consultation, processing and complaint. The similarity set can be determined by further processing the call intention types and the frequently asked questions database; the similarity set includes the similarity corresponding to each common question in the frequently asked questions database, and the similarity is used to characterize the similarity between the call intention type and the corresponding common question.
[0094] Specifically, the target word segmentation result is processed by a preset classifier to obtain the corresponding incoming call intention category, and the similarity between the incoming call intention category and each common question in the FAQ database is calculated to obtain a similarity set.
[0095] S432: Determine a target question corresponding to the incoming call voice based on the similarity set and the frequently asked questions database.
[0096] Among them, the similarity with the largest value in the similarity set is recorded as the target similarity, and the target similarity corresponds to a common question in the FAQ database. The target question is also a common question corresponding to the target similarity in the FAQ database.
[0097] Specifically, the similarity with the largest value in the similarity set is determined to obtain the target similarity; then a common question corresponding to the target similarity in the FAQ database is determined as the target question.
[0098] S440: Obtain an answer voice based on the target question and the FAQ database, and feed back the answer voice to the user.
[0099] Embodiment 5
[0100] A method for automatically answering incoming call questions provided in Example 5 of the present application refines the "determining a similarity set based on the target word segmentation result, a preset classifier and a frequently asked question database" in Example 4; it should be noted that for the parts not described in detail in this embodiment, reference may be made to the descriptions of other embodiments, and the method includes:
[0101] S510: Convert the received incoming call voice into text to obtain a target text.
[0102] S520: Perform word segmentation processing on the target text to obtain a target word segmentation result.
[0103] S5311. Vectorize the target word segmentation result to obtain a result vector.
[0104] Among them, the preset classifier specifically adopts the LSTM network (Long-Short Term Memory). It should be noted that a gating mechanism is added to the LSTM network, and a special memory layer is introduced. The LSTM network includes an input gate, a memory layer, a forget gate and an output gate in sequence; the LSTM network is used to vectorize the input target word segmentation result, and vectorization is to convert the target word segmentation result into a feature vector of a fixed length. The feature vector contains the key features of the target word segmentation result, and the converted feature vector is recorded as the result vector.
[0105] Specifically, the target word segmentation result is vectorized through the LSTM network to obtain a result vector.
[0106] S5312: Determine the call intent by processing the result vector based on a preset classifier.
[0107] Among them, the classifier can classify the result vector, so as to determine the type of incoming call intention corresponding to the result vector. The types of incoming call intention in this embodiment specifically include consultation, processing and complaint.
[0108] Specifically, a preset classifier classifies the result vector to determine the incoming call intention corresponding to the result vector.
[0109] S5313: Perform similarity calculation based on the incoming call intention and a preset frequently asked questions database to obtain a similarity set.
[0110] Among them, in this embodiment, there is a corresponding mapping relationship between the call intention and a preset database query statement, such as an SQL statement, and the database query statement corresponding to the call intention can be determined through the mapping relationship; the similarity calculation is to calculate the similarity between the database query statement and each common question in the FAQ database, and the similarity set is the set of similarities corresponding to each common question.
[0111] Specifically, a database query statement corresponding to the caller's intention is determined through a preset mapping relationship; the similarity between the database query statement and each frequently asked question in the frequently asked questions database is calculated to obtain a similarity set.
[0112] S532: Determine a target question corresponding to the incoming call voice based on the similarity set and the frequently asked questions database.
[0113] S540: Obtain an answer voice based on the target question and the FAQ database, and feed back the answer voice to the user.
[0114] Embodiment 6
[0115] A method for automatically answering incoming call questions provided in Example 6 of the present application is a refinement of the "obtaining an answer voice based on the target question and the frequently asked question and answer database" in Example 1; it should be noted that for the parts not described in detail in this embodiment, reference may be made to the descriptions of other embodiments, and the method includes:
[0116] S610: Convert the received incoming call voice into text to obtain a target text.
[0117] S620: Perform word segmentation processing on the target text to obtain a target word segmentation result.
[0118] S630: Determine a target question in the FAQ database corresponding to the incoming call voice based on the target word segmentation result, a preset classifier, and a FAQ database.
[0119] S641: Determine a target answer corresponding to the target question from the FAQ database.
[0120] The frequently asked questions and answers database includes the answer to the target question, and the answer to the target question is recorded as the target answer.
[0121] Specifically, an answer corresponding to the target question is determined from the Q&A database as the target answer.
[0122] S642: Perform speech synthesis on the target answer to obtain an answer speech.
[0123] Among them, the remote banking center platform has a built-in speech synthesis system, which means that the speech synthesis system processes the target answer in text form into the corresponding answer in speech form, and the answer in speech form is also the answer voice.
[0124] Specifically, the remote banking center platform processes the target answer through the built-in speech synthesis system to obtain the answer voice.
[0125] S643: Feedback the answer voice to the user.
[0126] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.
[0127] Embodiment 7
[0128] Based on the same inventive concept, this embodiment also provides an automatic answering device for incoming call questions for implementing the automatic answering method for incoming call questions involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more embodiments of the automatic answering device for incoming call questions provided below can refer to the limitations of the automatic answering method for incoming call questions above, and will not be repeated here.
[0129] In this embodiment, Figure 2 As shown, a device for automatically answering incoming call questions is provided, comprising:
[0130] A text conversion module, used to convert the received incoming call voice into a target text;
[0131] A word segmentation processing module, used to perform word segmentation processing on the target text to obtain a target word segmentation result;
[0132] A question determination module, configured to determine a target question in the FAQ database corresponding to the incoming call voice based on the target word segmentation result, a preset classifier and a FAQ database;
[0133] The voice feedback module is used to obtain an answer voice based on the target question and the frequently asked questions and answers database, and to feed back the answer voice to the user.
[0134] Each module in the above-mentioned automatic answering device for incoming call questions can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.
[0135] It should be noted that this embodiment obtains the target text by converting the received incoming call voice into text; performs word segmentation processing on the target text to obtain the target word segmentation result; determines the target question corresponding to the incoming call voice in the FAQ database based on the target word segmentation result, the preset classifier and the FAQ database; obtains the answer voice based on the target question and the FAQ database, and feeds back the answer voice to the user. Through the above implementation, the user's incoming call voice can be automatically converted into text and word segmented in sequence, and the FAQ corresponding to the incoming call voice can be further determined from the FAQ database, and then the voice corresponding to the answer to the FAQ is synthesized, and the user is automatically fed back. In this way, there is no need for bank customer service personnel to answer each customer complaint call one by one, and the user's incoming call question can be answered in a short time, which can reduce the burden on bank customer service personnel and improve the user experience.
[0136] In one embodiment, in converting the received incoming call voice into a target text, the text conversion module is specifically used to:
[0137] Processing the received incoming call voice based on a preset acoustic model to obtain a digital sound signal;
[0138] The digital sound signal is processed based on a preset speech model to obtain a target text.
[0139] In one embodiment, in performing word segmentation processing on the target text to obtain a target word segmentation result, the word segmentation processing module is specifically used to:
[0140] Segmenting the target text based on a preset segmentation library to obtain a primary segmentation result;
[0141] The stop words in the primary word segmentation result are removed to obtain the target word segmentation result.
[0142] In one embodiment, in determining the target question corresponding to the incoming call voice in the FAQ database based on the target word segmentation result, the preset classifier and the FAQ database, the question determination module is specifically used to:
[0143] Determine a similarity set based on the target word segmentation result, a preset classifier and a frequently asked question database;
[0144] A target question corresponding to the incoming call voice is determined based on the similarity set and the frequently asked questions and answers database.
[0145] In one embodiment, in determining a similarity set based on the target word segmentation result, a preset classifier and a FAQ database, the question determination module is specifically used to:
[0146] Vectorizing the target word segmentation result to obtain a result vector;
[0147] Determine the call intent by processing the result vector based on a preset classifier;
[0148] A similarity set is obtained by performing similarity calculation based on the incoming call intention and a preset frequently asked questions database.
[0149] In one embodiment, in terms of obtaining an answer voice based on the target question and the FAQ database, the voice feedback module is specifically used to:
[0150] Determine a target answer corresponding to the target question from the FAQ database;
[0151] Perform speech synthesis on the target answer to obtain the answer speech.
[0152] Embodiment 8
[0153] In this embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for automatically answering incoming call questions is implemented.
[0154] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the scheme of the present disclosure, and does not constitute a limitation on the computer device to which the scheme of the present disclosure is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0155] Embodiment 9
[0156] In this embodiment, a computer readable storage medium is provided. Figure 4 As shown, a computer program is stored thereon, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.
[0157] Embodiment 10
[0158] In this embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.
[0159] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties.
[0160] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided by the present disclosure can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided by the present disclosure may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited thereto. The processor involved in each embodiment provided by the present disclosure may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited thereto.
[0161] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0162] The above-described embodiments only express several implementation methods of the present disclosure, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present disclosure, and these all belong to the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the attached claims.
Claims
1. A method for automatically answering incoming call questions, characterized in that: include: Convert the received incoming call voice into text to obtain the target text; Performing word segmentation processing on the target text to obtain a target word segmentation result; Determine a target question in the FAQ database corresponding to the incoming call voice based on the target word segmentation result, a preset classifier and a FAQ database; An answer voice is obtained based on the target question and a frequently asked question and answer database, and the answer voice is fed back to the user.
2. The method according to claim 1, characterized in that The step of converting the received incoming call voice into text to obtain a target text includes: Processing the received incoming call voice based on a preset acoustic model to obtain a digital sound signal; The digital sound signal is processed based on a preset speech model to obtain a target text.
3. The method according to claim 1, characterized in that The performing word segmentation processing on the target text to obtain a target word segmentation result includes: Segmenting the target text based on a preset segmentation library to obtain a primary segmentation result; The stop words in the primary word segmentation result are removed to obtain the target word segmentation result.
4. The method according to claim 1, characterized in that: The determining the target question corresponding to the incoming call voice in the FAQ database based on the target word segmentation result, a preset classifier and a FAQ database includes: Determine a similarity set based on the target word segmentation result, a preset classifier and a frequently asked question database; A target question corresponding to the incoming call voice is determined based on the similarity set and the frequently asked questions and answers database.
5. The method according to claim 4, characterized in that The determining of the similarity set based on the target word segmentation result, a preset classifier and a frequently asked question and answer database includes: Vectorizing the target word segmentation result to obtain a result vector; Determine the call intent by processing the result vector based on a preset classifier; A similarity set is obtained by performing similarity calculation based on the incoming call intention and a preset frequently asked questions database.
6. The method according to claim 1, characterized in that The step of obtaining an answer voice based on the target question and the frequently asked questions database includes: Determine a target answer corresponding to the target question from the FAQ database; Perform speech synthesis on the target answer to obtain the answer speech.
7. A device for automatically answering incoming call questions, characterized in that: The device comprises: A text conversion module, used to convert the received incoming call voice into a target text; A word segmentation processing module, used to perform word segmentation processing on the target text to obtain a target word segmentation result; A question determination module, configured to determine a target question in the FAQ database corresponding to the incoming call voice based on the target word segmentation result, a preset classifier and a FAQ database; The voice feedback module is used to obtain an answer voice based on the target question and the frequently asked questions and answers database, and to feed back the answer voice to the user.
8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.