CN114860742A

CN114860742A - Artificial intelligence-based AI customer service interaction method, device, equipment and medium

Info

Publication number: CN114860742A
Application number: CN202210451154.7A
Authority: CN
Inventors: 卢斯宇; 毛星越; 李婷; 王坚
Original assignee: Shenzhen Pingan Integrated Financial Services Co ltd
Current assignee: Shenzhen Pingan Integrated Financial Services Co ltd
Priority date: 2022-04-26
Filing date: 2022-04-26
Publication date: 2022-08-05

Abstract

The application provides an AI customer service interaction method, device, electronic equipment and storage medium based on artificial intelligence, and the AI customer service interaction method based on artificial intelligence comprises the following steps: calling a voice communication text to construct an initial question-answer data set; evaluating the initial question-answer data set according to a preset question value index to obtain a question value evaluation set; screening the initial question-answer dataset based on the question value evaluation set to obtain a preferred question dataset; taking the voice communication text in the preferred question data set as a question template for interaction between the AI customer service and the customer, and acquiring a customer intention classification set according to the interaction result between the AI customer service and the customer; and updating the question template based on the client intention classification set so as to optimize the interaction process of AI customer service and the client. The method and the device can quantitatively evaluate and screen the problems appearing in the telemarketing, and effectively improve the marketing effect and the communication efficiency of the telemarketing.

Description

Artificial intelligence-based AI customer service interaction method, device, equipment and medium

Technical Field

The application relates to the technical field of artificial intelligence, in particular to an AI customer service interaction method and device based on artificial intelligence, an electronic device and a storage medium.

Background

Along with the increasing popularization of intelligent AI outbound marketing products on the market, the proportion of AI outbound used by financial institutions in the electricity sales scene is increasing day by day, the large-scale knowledge processing technology, the natural language understanding technology, the knowledge management technology, the automatic question and answer system, the reasoning technology and the like of the intelligent AI have industrial universality, the fine-grained knowledge management technology is provided for enterprises, a quick and effective technical means based on natural language is established for communication between the enterprises and mass users, and statistical analysis information required by fine management can be provided for the enterprises.

However, the unified standardized interaction flow is difficult to balance the screening precision and the customer experience in the AI interaction process, and enterprises are difficult to effectively select necessary problems, so that the screening precision is difficult to ensure, and the interaction flow is as concise as possible to improve the retention rate of target customers, thereby reducing the marketing effect of telemarketing.

Disclosure of Invention

In view of the above, there is a need to provide an AI customer service interaction method, device, electronic device and storage medium based on artificial intelligence to solve the technical problem of how to improve the marketing effect of telemarketing.

The application provides an AI customer service interaction method based on artificial intelligence, which comprises the following steps:

calling a voice communication text from a preset intelligent voice management platform database to construct an initial question-answer data set;

evaluating the initial question-answer data set according to a preset question value index to obtain a question value evaluation set;

screening the initial question and answer data set based on the question value evaluation set to obtain a preferred question data set;

taking the voice communication text in the preferred question data set as a question template for interaction between the AI customer service and the customer, and acquiring a customer intention classification set according to the interaction result between the AI customer service and the customer;

and updating the question template based on the client intention classification set so as to optimize the interaction process of AI customer service and the client.

Therefore, the problem occurring in the telemarketing is quantitatively evaluated through the preset problem value index to obtain the problem value evaluation set, and then the corresponding threshold value is calculated according to the self-defined threshold value algorithm to screen the problem occurring in the telemarketing, so that the high-value problem can be reserved for telemarketing, and the marketing effect and the communication efficiency of the telemarketing are effectively improved.

In some embodiments, the retrieving the voice communication text in the preset intelligent voice management platform database to construct the initial question-answer data set includes:

analyzing call voice according to an intelligent voice management platform to obtain a voice communication text, wherein the voice communication text comprises a question text and an answer text;

classifying the answer text based on the question text to obtain an initial classification dataset;

and carrying out dimension reduction on the initial classification data set according to a data dimension reduction algorithm to obtain the initial question-answer data set.

Therefore, the answers of the clients to the same question in the communication process can be classified into one category, so that the subsequent process can conveniently evaluate the corresponding question according to the answer on the basis, meanwhile, the storage space of a large amount of communication data can be effectively reduced through data dimension reduction, and the processing efficiency of the communication data is improved.

In some embodiments, the question value indicators include a customer group proportion indicator and a customer group conversion rate indicator, and the evaluating the initial question-answer data set according to a preset question value indicator to obtain a question value score set includes:

calculating text similarity between different answer texts corresponding to the same question text in the initial question-answer data set according to a text similarity algorithm to obtain answer semantic degrees;

calculating the variance of the answer semantic degree to be used as a passenger group proportion index;

performing secondary classification on different answer texts corresponding to the same question text in the initial question-answer data set to obtain a conversion success answer set and a conversion failure answer set of the question text;

calculating the data quantity ratio of the conversion success answer set and the conversion failure answer set to be used as a guest group conversion rate index;

evaluating the initial question-answer dataset based on the customer proportion index and the customer conversion rate index to obtain the set of question value evaluations.

Therefore, the problem occurring in the telemarketing is quantitatively evaluated according to the preset problem value index, so that the problem value evaluation set is obtained, the subsequent process is facilitated to screen out more effective problems according to the problem value evaluation set, and the marketing effect is improved.

In some embodiments, the set of question value scores satisfies the relationship:

wherein, P _i Scoring a value of an ith question text, m, in the set of question value scores _i The passenger group occupation ratio index, n, corresponding to the ith question text ⁺ _i The data volume, n, in the conversion success answer set corresponding to the ith question text ^- _i The data volume, n, in the conversion failure answer set corresponding to the ith question text _i The customer group conversion rate index, n, corresponding to the ith question text _i ＝max(n ⁺ _i ,n ^- _i )/min((n ⁺ _i ,n ^- _i )。

Therefore, each problem occurring in the telemarketing can be quantified according to the customer group occupation ratio index and the customer group conversion rate index, and the value score corresponding to each problem can be obtained more accurately.

In some embodiments, said screening said initial set of question and answer data based on said set of question value ratings to obtain a preferred set of question data comprises:

calculating the problem value scoring set according to a user-defined threshold value selection algorithm to obtain a problem value scoring threshold value;

screening the initial question-answer dataset based on the question value score threshold to obtain the preferred question dataset.

Therefore, more accurate and appropriate threshold values can be obtained through a user-defined threshold value selection algorithm to screen the initial question-answer data set, and therefore the accuracy of the obtained optimal question data set is improved.

In some embodiments, the taking the voice communication text in the preferred question data set as a question template for the interaction between the AI customer service and the customer, and obtaining the customer intention classification set according to the interaction result between the AI customer service and the customer includes:

sorting the preferred problem data sets according to the sequence of the problem value scores from large to small to obtain a preferred problem sorting set;

taking the voice communication texts in the preferred question sequencing set as question templates for interaction between AI customer service and clients in sequence to obtain real-time communication voice;

analyzing the real-time communication voice based on the intelligent voice management platform to obtain a communication text vector set;

and inputting the communication text vector set into a preset classification model to obtain the client intention classification set.

Therefore, the intention information of the user can be acquired in real time in the process of communication between the customer service and the client, and then the client with high intention strength is correctly screened, so that efficient and accurate marketing is realized.

In some embodiments, the set of client intent classifications includes a set of high value clients and a set of low value clients, and updating the questioning template based on the set of client intent classifications to optimize the AI customer service interaction process with the client includes:

selecting a subsequent communication mode based on the client intention classification set to acquire a subsequent communication data set;

and updating the question template based on the subsequent communication data set so as to optimize the interaction process of AI customer service and the customer.

Therefore, the initial question-answer data set can be updated in real time according to the communication data between the customer service and the client every time, the questions to be asked by the customer service are dynamically evaluated, the questions are guaranteed to have high value, and the telemarketing effect is continuously improved.

The embodiment of the present application further provides an AI customer service interaction device based on artificial intelligence, the device includes:

the voice communication system comprises a construction unit, a voice communication unit and a voice communication unit, wherein the construction unit is used for calling a voice communication text in a preset intelligent voice management platform database to construct an initial question-answer data set;

the evaluation unit is used for evaluating the initial question-answer data set according to a preset question value index so as to acquire a question value evaluation set;

a screening unit for screening the initial question-answer dataset based on the question value evaluation set to obtain an optimal question dataset;

the acquisition unit is used for taking the voice communication text in the preferred question data set as a questioning template for interaction between the AI customer service and the customer and acquiring a customer intention classification set according to the interaction result between the AI customer service and the customer;

and the updating unit is used for updating the questioning template based on the client intention classification set so as to optimize the interaction process of AI customer service and the client.

An embodiment of the present application further provides an electronic device, where the electronic device includes:

a memory storing at least one instruction;

and the processor executes the instructions stored in the memory to realize the artificial intelligence based AI customer service interaction method.

An embodiment of the present application further provides a computer-readable storage medium, where at least one instruction is stored in the computer-readable storage medium, and the at least one instruction is executed by a processor in an electronic device to implement the artificial intelligence based AI customer service interaction method.

Drawings

FIG. 1 is a flow chart of a preferred embodiment of an artificial intelligence based AI customer service interaction method to which the present application is directed.

FIG. 2 is a flow chart of the preferred embodiment of the present application for retrieving voice communication text in a pre-defined intelligent voice management platform database to construct an initial question-answer dataset.

FIG. 3 is a flow chart of a preferred embodiment of the present application for evaluating the initial question-answer data set according to a predetermined question value index to obtain a set of question value evaluations.

Fig. 4 is a flowchart of a preferred embodiment of the present application, in which the voice communication text in the preferred question data set is used as a question template for the interaction between the AI customer service and the customer, and a customer intention classification set is obtained according to the interaction result between the AI customer service and the customer.

Fig. 5 is a functional block diagram of a preferred embodiment of an artificial intelligence based AI customer service interaction device to which the present application is directed.

Fig. 6 is a schematic structural diagram of an electronic device according to a preferred embodiment of the artificial intelligence based AI customer service interaction method according to the present application.

Detailed Description

For a clearer understanding of the objects, features and advantages of the present application, reference is made to the following detailed description of the present application along with the accompanying drawings and specific examples. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict. In the following description, numerous specific details are set forth to provide a thorough understanding of the present application, and the described embodiments are merely a subset of the embodiments of the present application and are not intended to be a complete embodiment.

Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", may explicitly or implicitly include one or more of the described features. In the description of the present application, "a plurality" means two or more unless specifically limited otherwise.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the description of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

The embodiment of the Application provides an AI customer service interaction method based on artificial intelligence, which can be applied to one or more electronic devices, wherein the electronic devices are devices capable of automatically performing numerical calculation and/or information processing according to preset or stored instructions, and hardware of the electronic devices includes but is not limited to a microprocessor, an Application Specific Integrated Circuit (ASIC), a Programmable Gate Array (FPGA), a Digital Signal Processor (DSP), an embedded device, and the like.

The electronic device may be any electronic product capable of performing human-computer interaction with a client, for example, a Personal computer, a tablet computer, a smart phone, a Personal Digital Assistant (PDA), a game machine, an Internet Protocol Television (IPTV), an intelligent wearable device, and the like.

The electronic device may also include a network device and/or a client device. The network device includes, but is not limited to, a single network server, a server group consisting of a plurality of network servers, or a Cloud Computing (Cloud Computing) based Cloud consisting of a large number of hosts or network servers.

The Network where the electronic device is located includes, but is not limited to, the internet, a wide area Network, a metropolitan area Network, a local area Network, a Virtual Private Network (VPN), and the like.

Fig. 1 is a flowchart of a preferred embodiment of the AI customer service interaction method based on artificial intelligence according to the present application. The order of the steps in the flow chart may be changed and some steps may be omitted according to different needs.

And S10, calling the voice communication text in a preset intelligent voice management platform database to construct an initial question-answer data set.

Referring to fig. 2, in an alternative embodiment, the retrieving the voice communication text from the preset intelligent voice management platform database to construct the initial question-answer data set includes:

s101, analyzing call voice according to an intelligent voice management platform to obtain a voice communication text, wherein the voice communication text comprises a question text and an answer text.

In an optional embodiment, the intelligent voice management platform may use an AI intelligent voice call center system of the Enjoy Talk, the voice calls an asr (automatic Speech recognition) voice recognition engine to perform parsing through an IP-PBX by using a network RTP transmission protocol, and after the parsing is completed to obtain a recognition text, the recognition text is returned to the AI intelligent voice management platform by using a TCP protocol to perform classified storage so as to serve as the voice communication text.

In the alternative embodiment, the IP-PBX is an IP-based corporate telephone system that enables all communications to be unimpeded, establishing a unified voice and data network that can connect office locations and customer services distributed around the world; the ASR telephone voice recognition model searches available voice recognition service from public network resources through network connection to analyze call voice into voice communication texts, and the voice communication texts are transmitted back to the intelligent voice management platform for classified storage; by means of borrowing external available resources, a voice recognition system of the intelligent voice management platform does not need to be established, and construction cost of the intelligent voice management platform can be saved.

S102, classifying the answer texts based on the question texts to obtain an initial classification data set.

In this optional embodiment, because each question asked by the AI customer service corresponds to a different answer of a different customer during the communication with a large number of customers, the specific process of returning the voice communication text to the intelligent voice management platform for classified storage includes: classifying the answer texts based on the question texts, namely classifying each question text and different answer data corresponding to the question text into a category, and taking data contained in all the classified categories as the initial classification data set.

S103, reducing the dimension of the initial classification data set according to a data dimension reduction algorithm to obtain the initial question-answer data set.

In this optional embodiment, since the AI customer service correspondingly generates a large number of voice communication texts in the process of communicating with a large number of customers, the obtained initial classification data set is subjected to dimension reduction processing to reduce the number of answer texts corresponding to each question text. The dimensionality reduction processing is to map data points in an original high-dimensional space to a low-dimensional space by adopting a certain mapping method, so that redundant information and noise information contained in original data are reduced, the accuracy of the original data is improved, and the data volume is reduced.

In this alternative embodiment, the data dimension reduction algorithm may use pca (principal component analysis) principal component analysis. The principal idea of PCA is to map r-dimensional features onto k-dimensions, which are completely new orthogonal features, also called principal components, and k-dimensional features reconstructed on the basis of the original r-dimensional features.

In this alternative embodiment, the PCA works by finding a set of mutually orthogonal axes sequentially from the original space, and the selection of a new axis is closely related to the data itself in the initial classification data set. The first new coordinate axis is selected to be the direction with the largest variance of all data in the initial classification data set, the second new coordinate axis is selected to be the plane which is orthogonal to the first coordinate axis to enable the variance to be the largest, and the third axis is the plane which is orthogonal to the 1 st axis and the 2 nd axis and has the largest variance. By analogy, r such coordinate axes can be obtained. With the new axes obtained in this way, we have found that most of the variances are contained in the preceding k axes, and the variance contained in the following axes is almost 0. Thus, we can ignore the remaining axes and only keep the first k axes containing the most variance. In fact, this is equivalent to only retaining the dimension feature containing most of the variance, and neglecting the feature dimension containing the variance of almost 0, so as to implement the dimension reduction processing on the data feature.

For example, although the question text a corresponds to 10000 different answer text data, there are many identical or similar parts in the 10000 answer text data, and some answer text data themselves are meaningless noise or have no statistical significance, so that through PCA dimension reduction, the top 10 answer text data categories with the largest category are selected according to the variance, and the rest are discarded, and finally, there are 9600 answer text data included in the top 10 answer text data categories.

And S11, evaluating the initial question-answer data set according to a preset question value index to obtain a question value evaluation set.

In this optional embodiment, the question value index includes a customer group proportion index and a customer group conversion rate index, the customer group proportion index is used to represent the difference degree of customer groups corresponding to different answer texts of the same question text, and the customer group conversion rate index is used to represent the difference degree of customer groups corresponding to different answer texts of the same question text, which are finally converted into effective customers.

Referring to fig. 3, in an alternative embodiment, the evaluating the initial question-answer data set according to a preset question value index to obtain a question value score set includes:

and S111, calculating text similarity between different answer texts corresponding to the same question text in the initial question-answer data set according to a text similarity algorithm to obtain answer semantic degrees.

In this alternative embodiment, the text similarity algorithm may use a Vector Space Model (VSM) algorithm, where the VSM forms the answer text into one point in space, and provides the point in space in a vector form, so as to simplify the processing of the answer text into the operation of vectors in a vector space, thereby reducing the complexity of performing similarity calculation on the answer text.

In this optional embodiment, for different answer texts of the same question text, the similarity of the corresponding vector between each answer text and each of the rest answer texts is respectively calculated according to the VSM algorithm as the text similarity, the text similarity obtained for each answer text is subjected to mean value calculation, and the average value of the text similarities obtained for each text answer is used as the answer semantic degree.

And S112, calculating the variance of the answer semantic degree as a passenger group occupation ratio index.

In this optional embodiment, the smaller the difference between different answer texts corresponding to the same question text is, the better the difference is, it is indicated that the corresponding question text can more stably obtain the corresponding answer text, so in this scheme, variance calculation is performed on the answer semantic degrees of all answer texts corresponding to the same question text, and the obtained variance result is used as the passenger group occupation index m of the question text.

S113, performing two classifications on different answer texts corresponding to the same question text in the initial question-answer data set to obtain a conversion success answer set and a conversion failure answer set of the question text.

In this optional embodiment, the result after the customer service and the customer communicate each time in the historical data is counted, if the customer successfully purchases the product sold by the customer service through the telephone after communication, the customer conversion is successful, and if the customer does not purchase the product sold by the customer service through the telephone after communication, the customer is unsuccessful.

In this optional embodiment, two classifications are performed on different answer texts corresponding to the same question text in the initial question-answer data set according to a result after the customer service communicates with the customer every time in the historical data, wherein if the customer conversion is successful, the answer text corresponding to the customer is assigned to the conversion successful answer set, and if the customer conversion is not successful, the answer text corresponding to the customer is assigned to the conversion failed answer set, wherein each question text corresponds to one conversion successful answer set and one conversion failed answer set.

And S114, calculating a data quantity ratio of the conversion success answer set and the conversion failure answer set to serve as a customer group conversion rate index.

In this alternative embodiment, let n ⁺ For the data volume in the conversion success answer set, n ^- Calculating n for the data amount in the conversion failure answer set ⁺ And n ^- The ratio of the medium-large value to the small value is used as the index n of conversion rate of guest group, i.e. n is max (n) ⁺ ,n ^- )/min((n ⁺ ,n ^- ) Wherein n is>The larger 1, n indicates the larger difference in data amount between the conversion successful answer set and the conversion failed answer set.

S115, evaluating the initial question and answer data set based on the passenger group occupation ratio index and the passenger group conversion rate index to obtain the question value evaluation set.

In this alternative embodiment, the question value evaluation set satisfies the relationship:

In this alternative embodiment, the value score of each question text in the initial question-and-answer data set is calculated, and the value scores of all question texts are used as the set of question value scores.

S12, screening the initial question and answer data set based on the question value evaluation set to obtain a preferred question data set.

In an alternative embodiment, said screening said initial question and answer data set based on said set of question value scores to obtain a preferred question data set comprises:

and S121, calculating the problem value scoring set according to a user-defined threshold selection algorithm to obtain a problem value scoring threshold.

In this optional embodiment, the process of calculating the question value scoring set by the user-defined threshold selection algorithm to obtain the question value scoring threshold is as follows:

firstly, calculating an average value M of value scores of all question texts in the question value score set, and then optionally selecting a value score t in the question value score set, wherein the value score t can divide the question value score set into two parts, namely A and B;

respectively calculating the average value of the value scores corresponding to the A part and the B part, and recording the average value as MA and MB, wherein the ratio of the value score number of the question texts included in the A part to the total number of the value scores in the question value score set is recorded as PA, and the ratio of the value score number of the question texts included in the B part to the total number of the value scores in the question value score set is recorded as PB;

illustratively, there are 100 question value scores in total, corresponding to an average value M of 60, optionally one value score t of 55, where there are 30 question value scores greater than 55 in the 100 question value scores, 70 question value scores not greater than 55 as part a, 70 question value scores as part B, PA of 30/100 of 3/10, and PB of 70/100 of 7/10.

Then according to the relation ICV ═ PA ═ MA-M ² +PB*(MB-M) ² And traversing the value t, and calculating the value of the inter-class variance ICV at which the value t is taken to be the maximum value, wherein the corresponding value t at the moment is the problem value scoring threshold.

S122, screening the initial question-answer data set based on the question value scoring threshold value to obtain the preferred question data set.

In an optional embodiment, if the value score corresponding to the question text in the initial question-and-answer data set is greater than the question value score threshold, the question text is taken as a preferred question, and finally all preferred questions are taken as the preferred question data set.

And S13, taking the voice communication text in the preferred question data set as a questioning template for the interaction between the AI customer service and the customer, and acquiring a customer intention classification set according to the interaction result between the AI customer service and the customer.

Referring to fig. 4, in an alternative embodiment, the taking the voice communication text in the preferred question data set as a question template for the interaction between the AI customer service and the customer, and obtaining the customer intention classification set according to the interaction result between the AI customer service and the customer includes:

s131, sorting the preferred problem data sets according to the sequence of the problem value scores from large to small to obtain a preferred problem sorting set.

And S132, sequentially using the voice communication texts in the preferred question sequencing set as question templates for interaction between the AI customer service and the customer to obtain real-time communication voice.

S133, analyzing the real-time communication voice based on the intelligent voice management platform to obtain a communication text vector set.

In this optional embodiment, the ASR speech recognition engine analyzes the real-time communication speech to obtain corresponding word vectors and word vectors, and concatenates the word vectors and word vectors according to the occurrence time sequence of the real-time communication speech to form word combination vectors, and uses all the word combination vectors as the communication text vector set.

S134, inputting the communication text vector set into a preset classification model to obtain the client intention classification set.

In an optional embodiment, the communication text vector set is input into a preset classification model as input data, and abnormal influence on intention recognition caused by word segmentation errors is reduced on the basis of not losing word features. The classification model in the scheme can be a long-short term memory (LSTM) model, and information of the whole sentence is fused in each hidden layer state in the LSTM model, so that the last hidden layer state is used as a model output vector to be classified, the output of each hidden layer state can be comprehensively utilized to classify each word to obtain an intention classification result of each word, and finally the intention classification of the whole sentence is determined through voting, so that the intention classification of a client in each communication is obtained.

And S14, updating the question template based on the client intention classification set to optimize the interaction process of AI customer service and the client.

In an optional embodiment, the updating the question template based on the client intention classification set to optimize the interaction process of the AI customer service and the client comprises:

s141, selecting a subsequent communication mode based on the client intention classification set to acquire a subsequent communication data set.

In this alternative embodiment, different subsequent communication modes are selected for different client intention categories in the client intention classification set for communication. For the customers with high value in the customer set, the intelligent voice management platform can automatically send marketing short messages to inform the customers, and manual customer service can carry out secondary efficient and accurate telemarketing follow-up; for the customers in the low-value customer set, the call is made again according to a preset period to carry out the telephone sales work, and the preset period can be one month.

In this alternative embodiment, the speech content generated according to the different subsequent communication modes is parsed into text data by the ASR speech recognition engine to construct the subsequent communication data set.

And S142, updating the question template based on the subsequent communication data set so as to optimize the interaction process of the AI customer service and the customer.

In an optional embodiment, the interactive template is supplemented according to different answer texts corresponding to the question texts in the subsequent communication data set, so that dynamic update of the initial question data set can be realized, accuracy of subsequent scoring screening is guaranteed, accurate marketing is performed by the AI customer service according to the updated question and answer data set, and marketing effect is further improved.

Therefore, the initial question-answer data set can be updated in real time according to communication data between customer service and clients each time, the questions to be asked by the customer service are dynamically evaluated, and the questions are guaranteed to have high value, so that the telemarketing effect is continuously improved.

Referring to fig. 5, fig. 5 is a functional block diagram of a preferred embodiment of the AI customer service interaction device based on artificial intelligence according to the present application. The AI customer service interaction device 11 based on artificial intelligence comprises a construction unit 110, an evaluation unit 111, a screening unit 112, an acquisition unit 113, and an update unit 114. A module/unit as referred to herein is a series of computer readable instruction segments capable of being executed by the processor 13 and performing a fixed function, and is stored in the memory 12. In the present embodiment, the functions of the modules/units will be described in detail in the following embodiments.

In an alternative embodiment, the constructing unit 110 is configured to retrieve the voice communication text from a preset intelligent voice management platform database to construct the initial question-answer data set.

In an optional embodiment, the retrieving the voice communication text in the preset intelligent voice management platform database to construct the initial question-answer data set includes:

In an alternative embodiment, the evaluation unit 111 is configured to evaluate the initial question and answer data set according to a preset question value index to obtain a set of question value evaluations.

In an optional embodiment, the question value index includes a customer group proportion index and a customer group conversion rate index, and the evaluating the initial question-answer data set according to a preset question value index to obtain a question value scoring set includes:

classifying different answer texts corresponding to the same question text in the initial question-answer data set for two times to obtain a conversion success answer set and a conversion failure answer set of the question text;

In this alternative embodiment, let n ⁺ For the data volume in the conversion success answer set, n ^- Calculating n for the data amount in the conversion failure answer set ⁺ And n ^- The ratio of the medium-large value to the small value is used as the index n of conversion rate of guest group, i.e. n is max (n) ⁺ ,n ^- )/min((n ⁺ ,n ^- )。

And collecting to obtain the question value evaluation set.

In this alternative embodiment, the set of question value scores satisfies the relationship:

In an alternative embodiment, the screening unit 112 is configured to screen the initial question and answer data sets based on the sets of question value scores to obtain preferred question data sets.

calculating the problem value scoring set according to a user-defined threshold selection algorithm to obtain a problem value scoring threshold;

Then the relationship ICV ═ PA (MA-M) can be followed ² +PB*(MB-M) ² And traversing the value t, and calculating the value of the inter-class variance ICV at which the value t is taken to be the maximum value, wherein the corresponding value t at the moment is the problem value scoring threshold.

In an optional embodiment, if the value score corresponding to the question text in the initial question-and-answer data set is greater than the question value score threshold, the question text is used as a preferred question, and finally all preferred questions are used as the preferred question data set.

In an optional embodiment, the obtaining unit 113 is configured to use the voice communication text in the preferred question data set as a question template for the AI customer service to interact with the customer, and obtain the customer intention classification set according to an interaction result between the AI customer service and the customer.

In an optional embodiment, the taking the voice communication text in the preferred question data set as a question template for the interaction between the AI customer service and the customer, and obtaining the customer intention classification set according to the interaction result between the AI customer service and the customer includes:

In an alternative embodiment, the updating unit 114 is configured to update the question template based on the client intention classification set to optimize the interaction process of the AI customer service and the client.

In an alternative embodiment, the set of client intent classifications includes a set of high value clients and a set of low value clients, and the updating the questioning template based on the set of client intent classifications to optimize the interaction process of AI customer service with the clients includes:

and updating the questioning template based on the subsequent communication data set so as to optimize the interaction process of the AI customer service and the customer.

According to the technical scheme, the problems appearing in the telemarketing can be quantitatively evaluated through the preset problem value indexes to obtain the problem value evaluation sets, and then the corresponding threshold value is calculated according to the user-defined threshold value algorithm to screen the problems appearing in the telemarketing, so that the high-value problems can be reserved for telemarketing, and the marketing effect and the communication efficiency of the telemarketing are effectively improved.

Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application. The electronic device 1 comprises a memory 12 and a processor 13. The memory 12 is used for storing computer readable instructions, and the processor 13 is used for executing the computer readable instructions stored in the memory to implement the artificial intelligence based AI customer service interaction method according to any one of the above embodiments.

In an alternative embodiment, the electronic device 1 further comprises a bus, a computer program stored in said memory 12 and executable on said processor 13, such as an artificial intelligence based AI customer service interaction program.

Fig. 6 only shows the electronic device 1 with the memory 12 and the processor 13, and it will be understood by those skilled in the art that the structure shown in fig. 6 does not constitute a limitation of the electronic device 1, and may comprise fewer or more components than shown, or a combination of certain components, or a different arrangement of components.

In conjunction with fig. 1, the memory 12 in the electronic device 1 stores a plurality of computer-readable instructions to implement an artificial intelligence based AI customer service interaction method, and the processor 13 can execute the plurality of instructions to implement:

screening the initial question-answer dataset based on the question value evaluation set to obtain a preferred question dataset;

Specifically, the processor 13 may refer to the description of the relevant steps in the embodiment corresponding to fig. 1 for a specific implementation method of the instruction, which is not described herein again.

It will be understood by those skilled in the art that the schematic diagram is only an example of the electronic device 1, and does not constitute a limitation to the electronic device 1, the electronic device 1 may have a bus-type structure or a star-shaped structure, the electronic device 1 may further include more or less hardware or software than those shown in the figures, or different component arrangements, for example, the electronic device 1 may further include an input and output device, a network access device, etc.

It should be noted that the electronic device 1 is only an example, and other existing or future electronic products, such as those that may be adapted to the present application, should also be included in the scope of protection of the present application, and are included by reference.

Memory 12 includes at least one type of readable storage medium, which may be non-volatile or volatile. The readable storage medium includes flash memory, removable hard disks, multimedia cards, card type memory (e.g., SD or DX memory, etc.), magnetic memory, magnetic disks, optical disks, etc. The memory 12 may in some embodiments be an internal storage unit of the electronic device 1, for example a removable hard disk of the electronic device 1. The memory 12 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like provided on the electronic device 1. The memory 12 may be used not only for storing application software installed in the electronic device 1 and various types of data, such as codes of an artificial intelligence based AI customer service interaction program, etc., but also for temporarily storing data that has been output or is to be output.

The processor 13 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or may be composed of a plurality of integrated circuits packaged with the same or different functions, including one or more Central Processing Units (CPUs), microprocessors, digital Processing chips, graphics processors, and combinations of various control chips. The processor 13 is a Control Unit (Control Unit) of the electronic device 1, connects various components of the electronic device 1 by various interfaces and lines, and executes various functions and processes data of the electronic device 1 by running or executing programs or modules (e.g., executing an Artificial Intelligence (AI) customer service interaction program, etc.) stored in the memory 12 and calling data stored in the memory 12.

The processor 13 executes an operating system of the electronic device 1 and various installed application programs. The processor 13 executes the application program to implement the steps in each of the artificial intelligence based AI customer service interaction method embodiments described above, such as the steps shown in fig. 1-4.

Illustratively, the computer program may be partitioned into one or more modules/units, which are stored in the memory 12 and executed by the processor 13 to accomplish the present application. The one or more modules/units may be a series of computer-readable instruction segments capable of performing certain functions, which are used to describe the execution of the computer program in the electronic device 1. For example, the computer program may be divided into a building unit 110, an evaluation unit 111, a screening unit 112, an acquisition unit 113, an update unit 114.

The integrated unit implemented in the form of a software functional module may be stored in a computer-readable storage medium. The software functional module is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a computer device, or a network device) or a processor (processor) to execute parts of the artificial intelligence based AI customer service interaction method according to the embodiments of the present application.

The integrated modules/units of the electronic device 1 may be stored in a computer-readable storage medium if they are implemented in the form of software functional units and sold or used as separate products. Based on such understanding, all or part of the processes in the methods of the embodiments described above may be implemented by a computer program, which may be stored in a computer-readable storage medium and executed by a processor, to implement the steps of the embodiments of the methods described above.

Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U-disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), random-access Memory and other Memory, etc.

Further, the computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.

The block chain referred by the application is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

The bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one arrow is shown in FIG. 6, but this does not indicate only one bus or one type of bus. The bus is arranged to enable connection communication between the memory 12 and at least one processor 13 or the like.

An embodiment of the present application further provides a computer-readable storage medium (not shown), where the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions are executed by a processor in an electronic device to implement the artificial intelligence based AI customer service interaction method according to any of the foregoing embodiments.

In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and other divisions may be realized in practice.

The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.

In addition, functional modules in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional module.

Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the specification may also be implemented by one unit or means through software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.

Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present application and not for limiting, and although the present application is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions can be made on the technical solutions of the present application without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. An AI customer service interaction method based on artificial intelligence, which is characterized by comprising the following steps:

2. The AI customer service interaction method based on artificial intelligence of claim 1, wherein the retrieving of the voice communication text in the pre-defined intelligent voice management platform database to construct the initial question-answer dataset comprises:

3. The artificial intelligence based AI customer service interaction method of claim 1, wherein the question value indicators comprise a customer group proportion indicator and a customer group conversion rate indicator, and the evaluating the initial question-answer data set according to a preset question value indicator to obtain a question value score set comprises:

4. The artificial intelligence based AI customer service interaction method of claim 3 wherein the question value scoring set satisfies the relationship:

wherein, P _i For text of ith question in the set of question value scoresValue score, m _i The passenger group occupation ratio index, n, corresponding to the ith question text ⁺ _i The data volume, n, in the conversion success answer set corresponding to the ith question text ^- _i The data volume, n, in the conversion failure answer set corresponding to the ith question text _i The customer group conversion rate index, n, corresponding to the ith question text _i ＝max(n ⁺ _i ,n ^- _i )/min((n ⁺ _i ,n ^- _i )。

5. The artificial intelligence based AI customer service interaction method of claim 1 wherein screening the initial question-answer dataset based on the set of question value scores to obtain a preferred question dataset comprises:

6. The artificial intelligence based AI customer service interaction method of claim 1, wherein the using the voice communication text in the preferred question data set as a questioning template for the interaction of the AI customer service with the customer, and obtaining the set of the intention classifications of the customer according to the interaction result of the AI customer service with the customer comprises:

taking the voice communication texts in the preferred question sequencing set as question templates for interaction between AI customer service and customers in sequence to obtain real-time communication voice;

7. The artificial intelligence based AI customer service interaction method of claim 1, wherein the set of customer intent classifications includes a set of high value customers and a set of low value customers, and wherein updating the questioning template based on the set of customer intent classifications to optimize an AI customer service and customer interaction process comprises:

8. An artificial intelligence based AI customer service interaction device, the device comprising:

and the updating unit is used for updating the question template based on the client intention classification set so as to optimize the interaction process of the AI customer service and the client.

9. An electronic device, characterized in that the electronic device comprises:

a memory storing computer readable instructions; and

a processor executing computer readable instructions stored in the memory to implement the artificial intelligence based AI customer service interaction method of any of claims 1 to 7.

10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon computer-readable instructions which, when executed by a processor, implement the artificial intelligence based AI customer service interaction method according to any of claims 1 to 7.