KR102264173B1 - Method to generate consultation analysis data - Google Patents

Abstract

A computer program stored in a computer-readable storage medium is disclosed according to an embodiment of the present disclosure. The computer program, when executed on one or more processors of a computing device, causes the computer program to perform operations for generating consultation analysis data, the operations comprising: receiving customer consultation data; generating analysis data on the customer consultation data by using the analysis data generation method; receiving feedback on the analysis data; and updating the analysis data generating method or storing the analysis data based on the feedback.

Description

How to generate consultation analysis data {METHOD TO GENERATE CONSULTATION ANALYSIS DATA}

The present invention relates to a method for generating consultation analysis data, and more particularly, to a method for generating analysis data based on customer consultation data.

Due to the growth of the Internet and mobile industries, a big data environment, which means a large amount of structured or unstructured data beyond the scale that can be collected, managed, stored, and analyzed with existing database management tools, is coming. Such big data is a large-scale data in various forms generated in a digital environment, and is regarded as a key resource that can influence future competitiveness of companies, etc.

In addition, extracting economically necessary values from big data may be referred to as big data technology. Utilizing big data can create various types of value, such as establishing marketing strategies, optimizing processes, improving productivity, and preventing fraud. Therefore, big data technology to effectively utilize such big data can be said to be a means of securing corporate competitiveness that should be considered first in the era of unlimited competition regardless of the type of company.

The insurance industry can also use big data to secure the competitiveness of the company concerned. For example, an insurance company could use big data to serve customers in a call center. In the case of telephone consultations conducted in a call center, the agent must directly summarize the contents of the consultation with the customer after responding to the customer or classify the type of the consultation contents. It takes a lot of time for the agent to summarize the consultation contents or classify the types of consultation contents manually after responding to customers, and it may be impossible to respond to other customers while performing the following tasks. Therefore, there is a need in the art for a convenient method for summarizing or classifying types of consultations with customers in order to improve the work efficiency of counselors.

Korean Patent Registration No. 10-1949427 discloses an automatic evaluation system for consultation contents.

The present disclosure has been devised in response to the background art described above, and aims to generate counseling analysis data.

According to an embodiment of the present disclosure for realizing the above-described problems, a computer program stored in a computer-readable storage medium, the computer program being executed in one or more processors of a computing device for generating counseling analysis material perform operations, the operations comprising: receiving customer consultation data; generating analysis data on the customer consultation data by using the analysis data generation method; receiving feedback on the analysis data; and updating the analysis data generating method or storing the analysis data based on the feedback.

In an alternative embodiment of the computer program operations to perform the following operations for generating the consultation analysis material, the analysis material may include summary information summarizing the customer consultation data or the customer consultation data of at least one predetermined type. At least one of type code information classified as a code may be included.

In an alternative embodiment of the computer program operations for performing the following operations for generating the consultation analysis data, the operation of generating the analysis data on the customer consultation data using the analysis data generating method includes: the customer consultation data determining at least one of a main word or a main sentence about and generating at least one of summary information and type code information about the customer consultation data based on at least one of the main word or the main sentence.

In an alternative embodiment of the computer program operations for performing the following operations for generating the consultation analysis data, the operation of generating the analysis data on the customer consultation data using the analysis data generating method includes: the customer consultation data determining the frequency of use of words included in the ; and determining a key word related to the customer consultation data based on the frequency of use of the words.

In an alternative embodiment of the computer program operations for performing the following operations for generating the consultation analysis data, the operation of generating the analysis data on the customer consultation data using the analysis data generating method includes: the customer consultation data determining a co-occurrence relationship of the words when the distance between the words included in the word is less than or equal to a predetermined value; and determining a key word related to the customer consultation data based on the degree of simultaneous appearance of the words.

In an alternative embodiment of the computer program operations to perform the following operations for generating consultation analysis material, determining a key word about the customer consultation data based on a degree of simultaneous occurrence of the words includes: generating two or more nodes corresponding to each of two or more words included in data; generating an edge connecting nodes of words having the co-occurrence relationship; calculating a weight for each of the edges; calculating the importance of each of the two or more nodes based on the weight of the edge; and determining a key word based on the importance of each of the two or more nodes.

In an alternative embodiment of the computer program operations for performing the following operations for generating the consultation analysis data, the operation of generating the analysis data on the customer consultation data using the analysis data generating method includes: the customer consultation data determining, as a representative sentence, one sentence among two or more sentences having a high degree of similarity included in ; and determining a main sentence regarding the customer consultation data based on the representative sentence.

In an alternative embodiment of computer program operations for performing the following operations for generating consultation analysis data, the operation of determining one of two or more sentences with a high degree of similarity included in the customer consultation data as a representative sentence includes: , generating two or more nodes corresponding to each of the two or more sentences included in the customer consultation data; generating an edge connecting nodes of sentences having a similar relationship; calculating a weight for each of the edges; calculating the importance of each of the two or more nodes based on the weight of the edge; and determining a main sentence based on the importance of each of the two or more nodes.

In an alternative embodiment of the computer program operations for performing the following operations for generating the consultation analysis data, the operation of generating the analysis data on the customer consultation data using the analysis data generating method includes: the customer consultation data calculating by using an analysis data generation model including one or more pre-trained network functions; and outputting at least one of a main word or a main sentence by using the analysis data generation model, and the analysis data generation model receives past customer consultation data as an input, and the past customer consultation It may be a model trained with learning data using at least one of a main word or a main sentence that is matched and stored with respect to the data as a label.

In an alternative embodiment of the computer program operations to perform the following operations for generating the consultation analysis material, the analysis material generation model may include a first method for identifying a relationship between two or more words included in the customer consultation data. and a second sub-model for outputting at least one of a main word or a main sentence based on the sub-model and the relationship between the two or more words.

In an alternative embodiment of the computer program operations for performing the following operations for generating the consultation analysis data, the operation of generating the analysis data on the customer consultation data using the analysis data generating method includes: the customer consultation data determining at least one of a main word or a main sentence about and generating type code information based on a type code having a high matching rate by comparing at least one of the words or sentences stored by matching each of the two or more predetermined type codes with at least one of the main words or the main sentences. may include

In an alternative embodiment of the computer program operations for performing the following operations for generating the consultation analysis data, the operation of generating the analysis data on the customer consultation data using the analysis data generating method includes: the customer consultation data , calculating at least one of a main word or a main sentence related to the customer consultation data using a type code determination model including one or more pre-trained network functions; and outputting type code information using the type code determination model, wherein the type code determination model includes at least one of past customer consultation data, key words or main sentences related to past customer consultation data. It may be a model trained with learning data using one as an input, and type code information that is matched and stored with respect to the past customer consultation data as a label.

In an alternative embodiment of the computer program operations to perform the following operations for generating the counseling analysis data, updating the analysis data generating method based on the feedback or storing the analysis data may include: updating the method for generating the analysis data based on the error information when error information on the analysis data is included; Alternatively, when the feedback includes normal information on the analysis data, the method may include at least one operation of matching and storing the customer consultation data and the analysis data.

In an alternative embodiment of the computer program operations to perform the following operations for generating the counseling analysis data, when the feedback includes error information for the analysis data, the method for generating the analysis data based on the error information The updating may include: generating re-learning data based on the customer consultation data and the error information; and updating the analysis data generation method using the re-learning data.

In an alternative embodiment of the computer program operations to perform the following operations for generating the consultation analysis material, the operation may include: receiving a consultation request from a customer terminal; identifying customer information corresponding to the customer terminal; and transmitting at least one of past customer consultation data stored by matching the customer information or analysis data corresponding to the past customer consultation data to the counselor terminal.

According to an embodiment of the present disclosure for realizing the above-described problem, there is provided a method for generating consultation analysis data, the method comprising: receiving customer consultation data; generating analysis data on the customer consultation data by using the analysis data generation method; receiving feedback on the analysis data; and updating the analysis data generating method or storing the analysis data based on the feedback.

A server for generating counseling analysis data according to an embodiment of the present disclosure for realizing the above-described problem, comprising: a processor including one or more cores; and a memory, wherein the processor receives customer consultation data, generates analysis data on the customer consultation data using an analysis data generation method, receives feedback on the analysis data, and responds to the feedback. Based on the method, the analysis data generation method may be updated or the analysis data may be stored.

The present disclosure may provide a method for generating counseling analysis data.

1 is a diagram illustrating a block diagram of a computing device that performs an operation for generating counseling analysis data according to an embodiment of the present disclosure.
2 is a diagram exemplarily illustrating a method for generating counseling analysis data according to an embodiment of the present disclosure.
3 is a diagram exemplarily illustrating a method for generating counseling analysis data according to an embodiment of the present disclosure.
4 is a diagram exemplarily illustrating a method for generating counseling analysis data according to an embodiment of the present disclosure.
5 is a flowchart of a method for generating counseling analysis data according to an embodiment of the present disclosure.
6 is a block diagram of a computing device according to an embodiment of the present disclosure.

Various embodiments are now described with reference to the drawings. In this specification, various descriptions are presented to provide an understanding of the present disclosure. However, it is apparent that these embodiments may be practiced without these specific descriptions.

The terms “component,” “module,” “system,” and the like, as used herein, refer to a computer-related entity, hardware, firmware, software, a combination of software and hardware, or execution of software. For example, a component can be, but is not limited to being, a process running on a processor, a processor, an object, a thread of execution, a program, and/or a computer. For example, both an application running on a computing device and the computing device may be a component. One or more components may reside within a processor and/or thread of execution. A component may be localized within one computer. A component may be distributed between two or more computers. In addition, these components can execute from various computer readable media having various data structures stored therein. Components may communicate via a network such as the Internet with another system, for example via a signal having one or more data packets (eg, data and/or signals from one component interacting with another component in a local system, distributed system, etc.) may communicate via local and/or remote processes depending on the data being transmitted).

In addition, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or.” That is, unless otherwise specified or clear from context, "X employs A or B" is intended to mean one of the natural implicit substitutions. That is, X employs A; X employs B; or when X employs both A and B, "X employs A or B" may apply to either of these cases. It should also be understood that the term “and/or” as used herein refers to and includes all possible combinations of one or more of the listed related items.

Also, the terms "comprises" and/or "comprising" should be understood to mean that the feature and/or element in question is present. However, it should be understood that the terms "comprises" and/or "comprising" do not exclude the presence or addition of one or more other features, elements and/or groups thereof. Also, unless otherwise specified or unless the context is clear as to designating a singular form, the singular in the specification and claims should generally be construed to mean "one or more."

Those skilled in the art will further appreciate that the various illustrative logical blocks, configurations, modules, circuits, means, logics, and algorithm steps described in connection with the embodiments disclosed herein may be implemented in electronic hardware, computer software, or combinations of both. It should be recognized that they can be implemented with To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, configurations, means, logics, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application. However, such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

Descriptions of the presented embodiments are provided to enable those skilled in the art to use or practice the present invention. Various modifications to these embodiments will be apparent to those skilled in the art of the present disclosure. The generic principles defined herein may be applied to other embodiments without departing from the scope of the present disclosure. Thus, the present invention is not limited to the embodiments presented herein. This invention is to be construed in its widest scope consistent with the principles and novel features presented herein.

In an embodiment of the present disclosure, the server may include other configurations for performing a server environment of the server. The server may include any type of device. The server is a digital device, and may be a digital device equipped with a processor, such as a laptop computer, a notebook computer, a desktop computer, a web pad, and a mobile phone, and having a computing capability with a memory. The server may be a web server that processes the service. The above-described types of servers are merely examples, and the present disclosure is not limited thereto.

1 is a diagram illustrating a block diagram of a computing device that performs an operation for generating counseling analysis data according to an embodiment of the present disclosure.

The computing device 100 performing an operation for generating counseling analysis data according to an embodiment of the present disclosure may include a network unit 110 , a processor 120 , and a memory 130 .

The network unit 110 may transmit/receive data for generating consultation analysis data according to an embodiment of the present disclosure to other computing devices, servers, and the like. The network unit 110 may transmit/receive customer consultation data and the like to other computing devices, servers, and the like in order to generate consultation analysis data. In addition, the network unit 110 may enable communication between a plurality of computing devices so that learning of a network function is performed in a distributed manner in each of the plurality of computing devices. The network unit 110 may enable communication between a plurality of computing devices so that generation of analysis data using a network function may be distributed and processed.

The network unit 110 according to an embodiment of the present disclosure includes a Public Switched Telephone Network (PSTN), x Digital Subscriber Line (xDSL), Rate Adaptive DSL (RADSL), Multi Rate DSL (MDSL), VDSL ( A variety of wired communication systems such as Very High Speed DSL), Universal Asymmetric DSL (UADSL), High Bit Rate DSL (HDSL), and Local Area Network (LAN) can be used.

In addition, the network unit 110 presented herein is CDMA (Code Division Multi Access), TDMA (Time Division Multi Access), FDMA (Frequency Division Multi Access), OFDMA (Orthogonal Frequency Division Multi Access), SC-FDMA ( A variety of wireless communication systems may be used, such as Single Carrier-FDMA) and other systems.

In the present disclosure, the network unit 110 may be configured regardless of its communication mode, such as wired and wireless, and may be configured with various communication networks such as a personal area network (PAN) and a wide area network (WAN). can In addition, the network may be a well-known World Wide Web (WWW), and may use a wireless transmission technology used for short-range communication, such as infrared (IrDA) or Bluetooth (Bluetooth).

The techniques described herein may be used in the networks mentioned above, as well as in other networks.

The processor 120 may include one or more cores, and may include a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), and a tensor processing unit (TPU) of a computing device. unit), and may include a processor for generating counseling analysis data, such as deep learning. The processor 120 may read the computer program stored in the memory 130 to generate counseling analysis data according to an embodiment of the present disclosure. According to an embodiment of the present disclosure, the processor 120 may perform an operation for learning the neural network. The processor 120 is for learning of the neural network, such as processing input data for learning in deep learning (DN), extracting features from the input data, calculating an error, and updating the weight of the neural network using backpropagation. calculations can be performed. At least one of a CPU, a GPGPU, and a TPU of the processor 120 may process learning of a network function. For example, the CPU and the GPGPU can work together to process the learning of the network function and the generation of counseling analysis data using the network function. In addition, in an embodiment of the present disclosure, by using the processors of a plurality of computing devices together, learning of a network function and generation of counseling analysis data using the network function may be processed. In addition, the computer program executed in the computing device according to an embodiment of the present disclosure may be a CPU, GPGPU, or TPU executable program.

According to an embodiment of the present disclosure, the memory 130 may store any type of information generated or determined by the processor 120 and any type of information received by the network unit 110 .

According to an embodiment of the present disclosure, the memory 130 is a flash memory type, a hard disk type, a multimedia card micro type, and a card type memory (eg, SD or XD memory, etc.), Random Access Memory (RAM), Static Random Access Memory (SRAM), Read-Only Memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Programmable Read (PROM) -Only Memory), a magnetic memory, a magnetic disk, and an optical disk may include at least one type of storage medium. The computing device 100 may operate in relation to a web storage that performs a storage function of the memory 130 on the Internet. The description of the above-described memory is only an example, and the present disclosure is not limited thereto.

The processor 120 may receive customer consultation data through the network unit 110 .

The customer consultation data may be data including details of consultation between the agent and the customer. The customer consultation data may refer to data collected or generated from a customer consultation start time to a customer consultation end time. For example, the customer consultation data may refer to data collected or generated during a time period between the time when the customer requests the consultation and the time when the customer consultation ends when the customer requests the consultation. For example, customer consultation data includes data recorded on calls between agents and customers in a call center, message data exchanged between agents and customers through chatting, customer questions and answers uploaded to VOC (Voice of Customer) bulletin boards. data and the like. The detailed description of the above-described customer consultation data is only an example, and the present disclosure is not limited thereto.

The customer consultation data may include at least one of voice data and text data. For example, customer consultation data may include voice data when a customer and an agent conduct consultations over the phone, and customer consultation data may include text data when a customer and an agent conduct consultations through chatting. have. The detailed description of the above-described customer consultation data is only an example, and the present disclosure is not limited thereto.

When the customer consultation data includes voice data, the processor 120 may convert the voice data into text data. For example, the processor 120 may convert voice data collected from a call center server into text data through a speech to text (STT) engine. For example, the STT engine may be an artificial neural network model trained to convert speech data into text data. The detailed description of the above-described customer consultation data is only an example, and the present disclosure is not limited thereto.

The processor 120 may store the received customer consultation data or data converted from the customer consultation data in the memory 130 .

The processor 120 may generate analysis data on customer consultation data by using the analysis data generation method. A method for generating analysis data will be described later in detail.

The analysis data may be data obtained by analyzing customer consultation data. The analysis data may include at least one of summary information for summarizing customer consultation data or type code information for classifying customer consultation data into at least one predetermined type code.

The analysis data may include summary information summarizing customer consultation data. The summary information may be information that summarizes customer consultation data including a large amount of data to include only at least some important data. The summary information according to an embodiment of the present disclosure may include a combination of at least some words included in the customer consultation data. For example, customer consultation data says, "Yesterday, I was injured and was hospitalized, and the bone on my left arm was broken. But I saw that Hee-yeon's mother received insurance premiums when she broke a bone in the past. I also signed up for that insurance at that time, but I also had insurance premiums. When including the sentence "Can I get it?", the summary information may be "insurance insurance in case of bone fracture" or "insurance insurance in case of hospitalization", which is a combination of at least some words. The summary information according to an embodiment of the present disclosure may include a combination of at least some sentences included in the customer consultation data. In the above example, the summary information may be a combination of at least some sentences included in the customer service data, "I broke my left arm bone. I also signed up for that insurance at that time, can I get insurance?" Specific description of the above-described analysis data is merely an example, and the present disclosure is not limited thereto.

The analysis data may include type code information in which customer consultation data is classified into at least one predetermined type code. The type code information may be information for classifying the intention of the consultation included in the customer consultation data. The memory 130 may store a type code corresponding to each of the consultation intentions included in the customer consultation data. The customer consultation data may be classified into at least one type code stored in the memory 130 . For example, customer consultation data may be classified into one or more type codes based on three depths. The three depths may include a large classification, a medium classification, and a small classification for customer consultation data. For example, a major category may be insurance, annuity, branch finding, contract inquiry, and insurance claim. A major category included in the major insurance category may be health insurance, retirement preparation, and raising a large sum of money, and a subcategory included in the health insurance category may be cancer insurance. , actual expenses insurance, and the like. The type code may include, for example, keywords, numbers, letters, and the like. For example, customer consultation data says, "Yesterday, I was injured and was hospitalized, and the bone on my left arm was broken. But I saw that Hee-yeon's mother received insurance premiums when she broke a bone in the past. I also signed up for that insurance at that time, but I also had insurance premiums. When including the sentence "Can I get it?", the type code information may include a type code for "Claim claim - insurance type - coverage". Specific description of the above-described analysis data is merely an example, and the present disclosure is not limited thereto.

The processor 120 may determine at least one of a key word or a key sentence related to the customer consultation data. The processor 120 may generate at least one of summary information and type code information regarding customer consultation data based on at least one of a main word or a main sentence. A method of determining a main word or main sentence regarding customer consultation data and generating summary information or a type code based on the determined main word or main sentence will be described later in detail.

The key word may be a word having high importance for classifying the customer's consultation intention among two or more words included in the customer consultation data. For example, customer consultation data says, "Yesterday, I was injured and was hospitalized, and the bone on my left arm was broken. But I saw that Hee-yeon's mother received insurance premiums when she broke a bone in the past. I also signed up for that insurance at that time, but I also had insurance premiums. When including the sentence "Can I get it?", the key words may be "bone", "broken", "premium", "insurance", "subscription". Specific descriptions of the above key words are only examples, and the present disclosure is not limited thereto.

The main sentence may be a sentence having high importance for classifying the customer's counseling intention among two or more sentences included in the customer counseling data. For example, when there are sentences with high similarity among two or more sentences included in the customer consultation data, the processor 120 may determine only one representative sentence as the main sentence. For example, the processor 120 may determine a sentence including a plurality of words matching the type code among two or more sentences included in the customer consultation data as the main sentence. In the above example, the main sentence may be "I signed up for that insurance at the same time, can I get insurance?" Specific description of the above-mentioned main sentences is only an example, and the present disclosure is not limited thereto.

Hereinafter, a method for generating analysis data will be described.

Hereinafter, a method for the processor 120 to pre-process customer consultation data according to an embodiment of the present disclosure will be described.

The processor 120 may perform pre-processing on the customer consultation data. The preprocessing may include an operation of removing syllables from which a customer intention cannot be derived from among syllables included in the customer consultation data in order to determine a key word related to the customer consultation data. The processor 120 may pre-process the customer consultation data using natural language processing. Specifically, the processor 120 may perform pre-processing through morphological analysis on customer consultation data. The processor 120 may remove a syllable from which a customer intention cannot be derived from the stemmed analysis of customer consultation data. The syllable from which the customer's intention cannot be derived may be, for example, a syllable corresponding to a proposition, a suffix, or the like. The processor 120 may perform pre-processing to include a syllable capable of deriving a customer consultation intention with respect to the customer consultation data. The syllable from which the customer consultation intention can be derived may be, for example, a syllable corresponding to a noun, a verb, an adjective, or the like. For example, if the customer service data is "I signed up for the same insurance at that time, can I also receive the premium?", the processor 120 may display "Me", "Insurance", "Together", "Subscription", "Me too" , "insurance premium", "receive" can be preprocessed to include only. The detailed description of the above-described pretreatment is merely an example, and the present disclosure is not limited thereto.

Hereinafter, a method for the processor 120 to determine a key word based on a frequency of use of a word included in customer consultation data according to an embodiment of the present disclosure will be described.

The processor 120 may determine the frequency of use of words included in the customer consultation data. The processor 120 may determine the frequency of use of words included in the preprocessed customer consultation data.

The processor 120 may determine the frequency of use based on the number of times the same word is included in the customer consultation data. When the number of times the same word is included in the customer consultation data is equal to or greater than a predetermined threshold, the processor 120 may determine that the frequency of use is high, or corresponds to the highest number of times among the usage times of the words included in the customer consultation data. It may be determined that the words used are the most frequently used. For example, the processor 120 may classify the frequency of use as high, low, etc., or may determine the frequency of use as a value corresponding to the number of uses. 4, the word "insurance" was used 80 times, the word "treatment cost" was used 24 times, the word "traffic accident" was used 13 times, "Injury" may have been used 18 times. According to an embodiment of the present disclosure, if it is determined that the frequency of use is high for words that are used more than 20 times, the processor 120 may determine that the frequency of use for the words “insurance” and “treatment cost” is high. have. According to another embodiment of the present disclosure, when it is determined that the words belonging to the top three frequencies of use are high in frequency of use, the processor 120 provides information about the words "insurance", "treatment cost" and "injury". The frequency of use may be determined to be high. The specific description regarding the frequency of use of the above-mentioned words is merely an example, and the present disclosure is not limited thereto.

Even when the same word is included in the customer consultation data in a modified form, the processor 120 may determine the frequency of use by determining that the same word is mentioned. The modified word may be, for example, a word having a different spacing, a word having a different ending, or the like. For example, the processor 120 may determine the frequency of use by determining "hospitalization fee" and "hospitalization fee" as the same word. The specific description regarding the frequency of use of the above-mentioned words is merely an example, and the present disclosure is not limited thereto.

The processor 120 may determine a key word for the customer consultation data based on the frequency of use of the words. The processor 120 may determine a key word based on words that are frequently used. The processor 120 may determine words having a higher frequency of use as main words. For example, if the number of times of use for each of the words included in the customer consultation data is 24, 25, 1, 6, 14, 5, and 4, the processor 120 determines the number of times of use of words corresponding to 24 and 25. It may be determined that the frequency of use is high, and words corresponding to the number of times 24 and 25 may be determined as the main word. The detailed description of the above-described key word determination is merely an example, and the present disclosure is not limited thereto.

Hereinafter, a method for the processor 120 to determine a key word based on the degree of simultaneous appearance of words included in customer consultation data according to an embodiment of the present disclosure will be described.

The processor 120 may perform pre-processing on the customer consultation data as described above. The processor 120 may determine a key word based on the degree of simultaneous appearance of words included in the preprocessed customer consultation data.

When the distance between words included in the customer consultation data is equal to or less than a predetermined value, the processor 120 may determine a relationship of simultaneous appearance of words. When the distance between words included in the pre-processed customer consultation data is equal to or less than a predetermined value, the processor 120 may determine the co-occurrence relationship of words. The co-occurrence relationship may be information regarding whether a distance between words included in the customer consultation data is equal to or less than a predetermined value. For example, if the customer consultation data is "I was hospitalized because of an injury yesterday, and my left arm bone was broken", and the predetermined value is 3, the distance between the words "injured" and "hospitalized" is 2, so the processor ( 120 may determine that “injured” and “hospitalized” are co-occurring, and since the distance between the words “injured” and “bone” is 5, the processor 120 determines that “injured” and “bone” are co-occurring. You can decide not to be in a relationship. The detailed description of the above-mentioned co-appearance relationship is only an example, and the present disclosure is not limited thereto.

The processor 120 may perform post-processing on the words based on the degree of simultaneous appearance of the words. Post-processing may refer to an operation of transforming words according to the degree of simultaneous appearance of the words. The post-processing may include grouping into one word when the degree of simultaneous appearance of words is high. When the distance between words is 1, the processor 120 may determine an adjacent word relationship. The processor 120 may re-define the words as one word by grouping the words in the case of the words determined as the adjacent word relationship more than a predetermined threshold number of times. For example, when a case in which the distance between the words "actual cost" and "insurance" is 1 is found 5 times in the customer consultation data, the processor 120 may calculate that it is determined 5 times as an adjacent word relationship. The processor 120 may determine that the words "actual expenses" and "insurance" are determined as adjacent words more than a predetermined threshold number of times, group them into "actual expenses insurance" and redefine them as one word. The detailed description of the post-processing operation described above is merely an example, and the present disclosure is not limited thereto.

The processor 120 may determine a key word regarding the customer consultation data based on the degree of co-occurrence of the words. Hereinafter, a method of determining a key word related to customer consultation data based on the degree of simultaneous appearance of the words will be described with reference to FIGS. 3 and 4 . 3 is a diagram exemplarily illustrating a method for generating counseling analysis data according to an embodiment of the present disclosure. 4 is a diagram exemplarily illustrating a method for generating counseling analysis data according to an embodiment of the present disclosure.

The processor 120 may generate two or more nodes corresponding to each of two or more words included in the customer consultation data. The processor 120 may generate two or more nodes corresponding to each of two or more words in consideration of a post-processing operation for customer consultation data. The processor 120 may generate a word redefined as one word as one node based on the post-processing operation. A circle corresponding to each of A, B, C, and D in FIG. 3A may be a node. Each of A, B, C, and D may correspond to each of the words included in the customer consultation data. For example, if the customer consultation data is "Yesterday I was injured and hospitalized, but my left arm bone was broken.", the processor 120 determines "Yesterday", "Injured", "Hospital", "Hospitalized", " You can create nodes corresponding to each of "left arm", "bone", and "broken". The detailed description of the above-mentioned node is only an example, and the present disclosure is not limited thereto.

The processor 120 may generate an edge connecting nodes of words having a co-occurrence relationship. According to an embodiment of the present disclosure, the processor 120 may generate an edge connecting nodes corresponding to each of words having a co-occurrence relationship. According to another embodiment of the present disclosure, the processor 120 may generate an edge only for words having a co-occurrence relationship more than a predetermined number of times. A line connecting between nodes A and B in FIG. 3A may be an edge. In this example, the words for each of A-B, A-D, B-D, and C-D may be words having a co-occurrence relationship, or words having co-occurrence more than a predetermined number of times. The detailed description of the above-described edge is only an example, and the present disclosure is not limited thereto.

The processor 120 may equally distribute a weight of 1 to each of the two or more nodes. In FIG. 3A , the processor 120 equally distributes a weight of 1 to each of nodes A, B, C, and D, so that each of the nodes may have received a weight of 0.25. The detailed description of the above-described weights is only an example, and the present disclosure is not limited thereto.

The processor 120 may calculate a weight for each edge. The processor 120 may calculate to distribute the weights of each node equally to the edges connected to the node. In FIG. 3B , the processor 120 may calculate to equally distribute the weight 0.25 for the node A to each of the two edges connected to the node B and the node D. The processor 120 may calculate to distribute the weight 0.25 of node A to each edge by 0.125. The processor 120 may calculate to evenly distribute a weight of 0.25 for node D to each of the three edges connected to each of nodes A, B, and C. The processor 120 may calculate to distribute the weight 0.25 of the node D to each edge by 0.083. In FIG. 3(b) , the edge weight calculation with respect to the B node and the C node is omitted. The detailed description of the above-described edge weight calculation is merely an example, and the present disclosure is not limited thereto.

The processor 120 may calculate the importance of each of the two or more nodes based on the weight of the edge. The processor 120 may calculate the importance of the node by multiplying the weight of the node by a predetermined first probability and adding each value obtained by multiplying the weight of the edge connected to the node by the predetermined second probability. A sum of the first predetermined probability and the second predetermined probability may be 100 percent. The initial importance value for node B is calculated by multiplying the weight of node B (0.25) by a first predetermined probability (0.15), and the edge weight between node B and node A (0.125) and the edge weight between node B and node D (0.125). It may be a sum ((0.125+0.083)*0.85+(0.25)*0.15=0.214) of each value obtained by multiplying a value obtained by adding (0.083) to a predetermined second probability (0.85). In FIG. 3(c) , the initial importance values obtained by performing the calculation for each node in the same manner may be 0.214 for the A node, 0.214 for the B node, 0.108 for the C node, and 0.463 for the D node. The processor 120 may calculate the weight of the edge based on the initial importance value of each node and repeat the process of calculating the importance again. If the importance value is repeatedly calculated several times, the importance value may converge to one value for each node. When the importance value converges to one value for each node, the processor 120 may finally determine the corresponding value as the importance value for the node. The specific description regarding the above-mentioned importance value is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the processor 120 may determine a key word based on the importance of each of two or more nodes. According to an embodiment of the present disclosure, the processor 120 may determine, as a key word, a word corresponding to nodes having an importance greater than or equal to a predetermined threshold value among nodes. According to another embodiment of the present disclosure, the processor 120 may determine a word corresponding to nodes having a higher importance value among the nodes as the main word. Referring to the importance 410 of FIG. 4 , the importance value of the node corresponding to the word “insurance” is 0.089849, the importance value of the node corresponding to the word “treatment cost” is 0.026183, and the node corresponding to the word “traffic accident” The importance value of may be 0.017231. According to an embodiment of the present disclosure, when a word having an importance value equal to or greater than a predetermined threshold value of 0.02 is determined as a key word, the processor 120 may determine the words “insurance” and “treatment cost” as key words. According to another embodiment of the present disclosure, when determining words having the top three importance values as key words, the processor 120 may determine the words “insurance”, “treatment cost” and “traffic accident” as key words. . Specific descriptions of the above key words are only examples, and the present disclosure is not limited thereto.

Hereinafter, a method for the processor 120 to determine a main sentence based on the similarity of sentences included in the customer consultation data according to an embodiment of the present disclosure will be described.

According to an embodiment of the present disclosure, the processor 120 may determine one sentence among two or more sentences having a high degree of similarity included in the customer consultation data as the representative sentence. The processor 120 may determine a main sentence regarding the customer consultation data based on the representative sentence. The representative sentence may mean one sentence that can represent two or more sentences with high similarity. For example, the processor 120 may say, "I called to inquire about whether I can receive premiums with the insurance I have purchased." Since the matching rate of the words in the sentence and the sentence "I signed up for that insurance at the same time, can I also get insurance?" sentences is high, it can be determined that the similarity between the two sentences is high. The processor 120 responds to the two sentences, "I called to inquire about whether I can receive premiums with the insurance I subscribed to." A sentence can be determined as a representative sentence. The determination of similarity between sentences will be described later in detail. The processor 120 may determine a representative sentence of at least some of the representative sentences as the main sentence. Specific description of the above-mentioned main sentences is only an example, and the present disclosure is not limited thereto.

The processor 120 may calculate a degree of similarity between two or more sentences included in the customer consultation data. According to an embodiment of the present disclosure, the processor 120 may calculate the similarity by performing analysis based on the meaning and arrangement of one or more words included in each of the sentences, and the relationship between each word. The processor 120 may calculate the similarity based on the matching rate of the meaning and arrangement of one or more words included in each of the sentences. The processor 120 may determine that the higher the matching rate, the higher the similarity. According to another embodiment of the present disclosure, the processor 120 may calculate the similarity between two sentences using natural language processing. The processor 120 may calculate the similarity between the two sentences by calculating the similarity between the sentence vectors by calculating the sentence vector for each of the two sentences. For example, customer service data says, "I called to ask if I could receive insurance premiums with the insurance I signed up for. Yesterday, I was injured and was hospitalized, and the bone on my left arm was broken. But I saw that Heeyeon's mother had a broken bone before. In the case of including the sentences, "I signed up for the insurance at the same time, can I also receive the insurance?", the processor 120 says, "I called to inquire about whether I can receive premiums with the insurance I subscribed to." Since the matching rate of the words in the sentence and the sentence "I signed up for that insurance at the same time, can I also get insurance?" sentences is high, it can be determined that the similarity between the two sentences is high. The detailed description of the above-described sentence similarity is only an example, and the present disclosure is not limited thereto.

The processor 120 may generate two or more nodes corresponding to each of two or more sentences included in the customer consultation data. Hereinafter, it will be described with reference to FIG. 3 . A circle corresponding to each of A, B, C, and D in FIG. 3A may be a node. Each of A, B, C, and D may correspond to each of the sentences included in the customer consultation data. The detailed description of the above-mentioned node is only an example, and the present disclosure is not limited thereto.

The processor 120 may generate an edge connecting nodes of sentences having a similar relationship. The processor 120 may generate an edge only for nodes corresponding to each of the sentences having a similarity greater than or equal to a predetermined threshold value. A line connecting between nodes A and B in FIG. 3A may be an edge. In this example, sentences for each of A-B, A-D, B-D, and C-D may be sentences having a similar relationship (eg, two sentences having a similarity greater than or equal to a predetermined threshold). The detailed description of the above-described edge is only an example, and the present disclosure is not limited thereto.

The processor 120 may equally distribute a weight of 1 to each of the two or more nodes. The processor 120 may calculate a weight for each edge. The processor 120 may calculate the importance of each of the two or more nodes based on the weight of the edge. A method of distributing a weight, calculating a weight for an edge, and calculating an importance for each node is the same as described above.

The processor 120 may determine the main sentence based on the importance for each of the two or more nodes. According to an embodiment of the present disclosure, the processor 120 may determine, as the main sentence, a sentence corresponding to nodes having an importance greater than or equal to a predetermined threshold value among the nodes. According to another embodiment of the present disclosure, the processor 120 may determine a sentence corresponding to nodes having a higher importance value among the nodes as the main sentence. For example, the importance value of the node corresponding to the sentence "I called to inquire about whether I can receive premiums with the insurance I subscribed to" is 0.24, and the importance value of the node corresponding to the sentence "I was hospitalized yesterday because I was injured" is If it is 0.01 and the importance value of the node corresponding to the sentence “I broke my left arm bone” is 0.18, the sentence corresponding to the node with high importance value is “I called to inquire about whether I can receive insurance with my insurance” can be determined as the main sentence. Specific description of the above-mentioned main sentences is only an example, and the present disclosure is not limited thereto.

Hereinafter, a method for the processor 120 to determine a key word or a key sentence using an analysis data generation model including one or more pre-trained network functions according to an embodiment of the present disclosure will be described.

The analysis data generation model includes a first sub-model for identifying a relationship between two or more words included in the customer consultation data and a second for outputting at least one of a main word or a main sentence based on the relationship between the two or more words It can contain sub-models.

The first sub-model may identify a relationship between two or more words included in the customer consultation data. The output value output through the function activated in the node of the hidden layer included in the first sub-model is input to the node included in the next layer of the hidden layer, and also the next operation performed at the node of the hidden layer It can be re-entered by input of . A node included in the hidden layer may perform a recursive operation using, as its input, an output value output from a node included in the hidden layer according to an operation at a previous point in time. The first sub-model may recursively perform operations on phonemes, syllables, or words included in the customer consultation data to determine a relationship between two or more words. The first sub-model may include, for example, a recurrent neural network (RNN). The detailed description of the above-described first sub-model is merely an example, and the present disclosure is not limited thereto.

The second sub-model may output at least one of a main word or a main sentence based on a relationship between two or more words. The second sub-model may include, for example, a convolutional neural network (CNN). A convolutional neural network may include one or more convolutional layers and subsampling layers. The convolutional neural network may extract features from input data by repeatedly performing a convolutional process and a subsampling process (eg, the aforementioned max pooling, etc.). Through an iterative convolutional process and subsampling process, a neural network can extract global features of input data. An output of the convolutional layer or the subsampling layer may be input to a fully connected layer. A fully connected layer is a layer in which all neurons in one layer and all neurons in neighboring layers are connected. The fully connected layer may mean a structure in which all nodes of each layer are connected to all nodes of other layers in a neural network. A description of a specific configuration for a convolutional neural network is discussed more specifically in US Patent No. US9870768B2, which is incorporated herein by reference in its entirety. The detailed description of the above-described second sub-model is merely an example, and the present disclosure is not limited thereto.

Hereinafter, a method for the processor 120 to train an analysis data generation model according to an embodiment of the present disclosure will be described.

The learning data for learning the analysis data generation model may be generated based on the history stored in the memory 130 while the counselor conducts the consultation with the customer in the past. The learning data may include, as an input, past customer consultation data in which an agent and a customer have performed a consultation in the past. The learning data may include, as a label, at least one of a key word or a key sentence that the agent writes and stores after consulting with the customer.

The analysis data generation model according to an embodiment of the present disclosure may be a teacher-trained model so that at least one of a main word or a main sentence is output by inputting customer consultation data. The analysis data generation model may be a model trained with learning data in which past customer consultation data is input, and at least one of a main word or a main sentence stored by matching with respect to the past customer consultation data is used as a label.

The processor 120 inputs past customer consultation data, which is an input included in the learning data, to the analysis data generation model, and at least one of the main words or main sentences that are the output of the analysis data generation model and the label (ie, correct answer) of the learning data. ), the error can be calculated by comparing at least one of the main words or main sentences. The processor 120 may backpropagate from the output layer in the reverse direction to the input layer direction in the network function included in the analysis data generation model based on the calculated error. The processor 120 may cause the connection weight of each node of each layer to be updated according to the backpropagation backpropagated from the output layer to the input layer.

The processor 120 may calculate the customer consultation data using an analysis data generation model including one or more pre-trained network functions. The processor 120 may output at least one of a main word or a main sentence using the analysis data generation model.

Hereinafter, a method for the processor 120 to generate summary information based on key words or key sentences related to customer consultation data according to an embodiment of the present disclosure will be described.

According to an embodiment of the present disclosure, the processor 120 may determine a main word or a main sentence related to customer consultation data as summary information. According to another embodiment of the present disclosure, when there are two or more main words or two or more main sentences related to customer consultation data, the processor 120 combines two or more main words or two or more main sentences to generate summary information. can create The processor 120 may generate summary information by arranging the main words or main sentences based on the order of the main words or main sentences in the customer consultation data. For example, customer service data says, "I called to inquire if I could receive premiums with the insurance I signed up for. Yesterday, I was injured and was hospitalized, and the bone on my left arm was broken. But it turns out that Hee-yeon's mother must have broken a bone before. I received the premium at the time. I signed up for the insurance at the same time, but can I get the premium too?" sentences, "I called to ask if I can receive premiums with the insurance I bought" and "I broke my left arm bone." "The sentences can each be the main sentence. In the customer consultation data, the sentence “I called to inquire about whether I can receive premiums with the insurance I signed up for” is a sentence that comes before the sentence “I broke my left arm bone”, so “I want to ask if I can receive premiums with the insurance I have purchased. You can create summary information in the following order: "I called, but the bone on my left arm was broken." The detailed description of the above summary information is only an example, and the present disclosure is not limited thereto.

Hereinafter, a method for the processor 120 to generate type code information based on a main word or a main sentence related to customer consultation data according to an embodiment of the present disclosure will be described.

The processor 120 may determine at least one of a key word or a key sentence related to the customer consultation data. The processor 120 compares at least one of the words or sentences stored by matching each of the two or more predetermined type codes with at least one of the main words or main sentences to generate type code information based on the type code having a high matching rate. can

Two or more type codes may be stored in the memory 130 . One or more words may be matched and stored in the memory 130 for each of two or more type codes. For example, the words "claim", "guarantee", "compensation amount", and "payment amount" may be matched and stored with respect to the type code of "claim insurance - insurance type - coverage coverage" in the memory 130 . One or more sentences for each of the two or more type codes may be matched and stored in the memory 130 . For example, a plurality of sentences including the sentence “I want to know the amount of compensation when claiming insurance” may be stored in the memory 130 with respect to the “insurance claim-insurance type-insurance coverage” type code. Specific description of the above type code is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the processor 120 may calculate a matching rate of words stored by matching key words and type codes related to customer consultation data. The processor 120 may calculate a matching rate based on whether or not the words stored by matching the type code with the main words related to the customer consultation data match or whether the words are similar. The processor 120 may determine, as the type code related to the customer consultation data, a type code in which words matching the main word exist more than a predetermined threshold value or words similar to the main word exist more than a predetermined threshold value.

According to an embodiment of the present disclosure, the processor 120 may calculate a matching rate of a sentence stored by matching key words and type codes related to customer consultation data. The processor 120 may calculate a matching rate based on whether a word included in a sentence stored by matching a type code with a key word related to customer consultation data matches or is similar to a word. The processor 120 determines, as the type code related to the customer consultation data, a type code corresponding to a sentence in which words matching the main word exist more than a predetermined threshold value, or words similar to the main word exist more than a predetermined threshold value. can

According to an embodiment of the present disclosure, the processor 120 may calculate a matching rate of words stored by matching the main sentences and type codes related to customer consultation data. The processor 120 may calculate a matching rate based on whether a word included in a main sentence relating to customer consultation data matches a word stored by matching the type code or whether the word is similar. The processor 120 determines, as a type code regarding customer consultation data, a type code in which words matching words included in the main sentence exist more than a predetermined threshold value, or words similar to the main word exist more than a predetermined threshold value can

In an embodiment of the present disclosure, the processor 120 may calculate a matching rate of a sentence stored by matching the main sentence and the type code related to the customer consultation data. The processor 120 may calculate a matching rate based on whether a word included in a main sentence related to customer consultation data matches a type code and matches a word included in a stored sentence or whether the word is similar. The processor 120 generates a type code corresponding to a sentence in which words matching words included in the main sentence exist above a predetermined threshold value, or in which words similar to the main word exist above a predetermined threshold value for customer consultation data. It can be determined by the type code.

Hereinafter, a method for the processor 120 to determine type code information using a type code determination model including one or more pre-trained network functions according to an embodiment of the present disclosure will be described.

Hereinafter, a method for the processor 120 to train the type code determination model according to an embodiment of the present disclosure will be described.

The training data for learning the type code determination model may be generated based on a history stored in the memory 130 while an agent consults with a customer in the past. After the agent consults with the customer, he or she may write key words or main sentences to summarize the contents of the consultation with the customer, and write a type code to classify the contents of the consultation with the customer. The history stored in the memory 130 may include information about main words or main sentences and type codes written by past counselors according to counseling performance. The learning data may include, as an input, at least one of main words or main sentences written and stored after an agent and a customer have consulted in the past. The learning data may include type code information created and stored after an agent and a customer have consulted in the past as a label.

The type code determination model according to an embodiment of the present disclosure may be a teacher-learned model to output type code information by inputting at least one of a main word or a main sentence. The type code determination model is a learning that takes as an input at least one of a key word or a main sentence regarding past customer consultation data, past customer consultation data, and matches the stored type code information with respect to the past customer consultation data as a label It may be a model trained on data.

The processor 120 inputs at least one of a main word or a main sentence, which is an input included in the training data, to the type code determination model, and the type code information that is an output of the type code determination model and a label (ie, correct answer) of the training data. The error can be calculated by comparing the type code information. The processor 120 may backpropagate from the output layer in the reverse direction to the input layer direction in the network function included in the analysis data generation model based on the calculated error. The processor 120 may cause the connection weight of each node of each layer to be updated according to the backpropagation backpropagated from the output layer to the input layer.

The processor 120 may calculate at least one of the customer consultation data, key words or main sentences related to the customer consultation data using a type code determination model including one or more pre-trained network functions. The processor 120 may output type code information using the type code determination model.

When an agent conducts a consultation with a customer in a call center, the agent directly selects one of hundreds of type codes and has to write a summary of the consultation content. There is a deviation in the summary of the consultation contents depending on the competency of the agent, and there are many cases where it is not possible to write an accurate summary because the consultation with the next customer is continuously coming in. Also, there have been many cases in which an incorrect type code is specified by an agent selecting one type code from among hundreds of type codes. When the analysis data generating method according to the present invention is used, analysis data including summary information or type code is generated while the counselor and the customer have a consultation, so that the counselor's work efficiency and work accuracy can be improved.

Hereinafter, a method for the processor 120 to store analysis data or update a method for generating analysis data based on feedback according to an embodiment of the present disclosure will be described.

The processor 120 may transmit information on the analysis data to the counselor terminal through the network unit 110 . The processor 120 may receive feedback on the analysis data from the counselor terminal through the network unit 110 .

The feedback may be evaluation information on analysis data derived based on customer consultation data. Feedback may include error information or normal information about the analysis data.

The error information may be negative evaluation information for the analyzed data. The error information may include correction information for information that is not normal when the analysis data includes information that is not normal (ie, incorrect answer). For example, in summary information that includes “insurance”, “claim”, and “annuity”, if “annuity” is an incorrect summary and “hospital expense” is a correct summary, the error information is “annuity” as “hospital expense” Corrected information may be included. The detailed description of the above-described error information is merely an example, and the present disclosure is not limited thereto.

The normal information may be positive evaluation information for the analyzed data. The normal information may include information that the analysis data is correct when all of the information included in the analysis data is normal information (ie, correct answers).

The agent terminal may communicate with the computing device 100 or the user terminal. Also, the agent terminal may perform operations performed by the above-described computing device 100 . For example, the agent terminal converts the voice of a customer on a call into a character string to generate customer consultation data, identifies words in the customer consultation data, determines a key word among the identified words, and types based on the key word. An operation such as determining code information may be performed. Also, the agent terminal may receive analysis data from the computing device 100 , generate feedback on the analysis data, and transmit the feedback to the computing device 100 . The detailed description of the agent terminal described above is only an example, and the present disclosure is not limited thereto.

The processor 120 may store the analysis data in the memory 130 based on the feedback. The processor 120 may match the customer consultation data, the analysis data, and the feedback and store it in the memory 130 .

When the feedback includes normal information on the analysis data, the processor 120 may match the customer consultation data with the analysis data and store them in the memory 130 . When the feedback includes normal information, since the analysis data includes accurate information, the analysis data itself may be stored in the memory 130 .

According to an embodiment of the present disclosure, when the feedback includes error information on the analysis data, the processor 120 may match the customer consultation data with the feedback and store the match in the memory 130 . When the feedback includes error information, since the analysis data may include inaccurate information, the analysis data may not be stored, and only the feedback, which is correction information on the analysis data, may be stored in the memory 130 .

According to an embodiment of the present disclosure, when the feedback includes error information on the analysis data, the processor 120 may match the customer consultation data with the analysis data and the feedback and store the matching data in the memory 130 . The error information may include only correction information for information that is not normal in some of the data included in the analysis data. Therefore, some normal information among the data included in the analysis data may not be included in the error information. The processor 120 may match the correction information for some abnormal information included in the analysis data included in the customer consultation data and the feedback and some other normal information included in the analysis data and store it in the memory 130 . have.

According to another embodiment of the present disclosure, when the feedback includes error information on the analysis data, the processor 120 may match the customer consultation data with the analysis data corrected based on the feedback and store it in the memory 130 . . The analysis data corrected based on the feedback may be analysis data in which some information that is not normal among the information included in the analysis data is changed based on the correction information.

Hereinafter, a method for the processor 120 to update an analysis data generation method according to an embodiment of the present disclosure will be described.

The processor 120 may update the analysis data generating method based on the feedback. Since the feedback can be interpreted as an accurate label of the training data for training the model, the performance of the model can be improved when the model is retrained or updated based on the feedback.

According to an embodiment of the present disclosure, the processor 120 may generate re-learning data based on error information. The processor 120 may update the analysis data generation method based on the error information. The processor 120 may generate the re-learning data using the customer consultation data as an input of the re-learning data, and using the changed analysis data as a label based on the error information.

In order to re-learn the analysis data generation model according to an embodiment of the present disclosure, the processor 120 uses customer consultation data as an input of the re-learning data, and at least one of the main words or main sentences changed based on the error information You can create re-learning data with labels.

In order to retrain the type code determination model according to an embodiment of the present disclosure, the processor 120 uses the customer consultation data, key words or main sentences related to the customer consultation data as input for the re-learning data, and based on the error information Thus, it is possible to generate re-learning data using the changed type code information as a label.

According to another embodiment of the present disclosure, the processor 120 may generate re-learning data based on normal information and error information. The processor 120 may update the analysis data generation method based on the normal information and the error information.

The processor 120 may retrain at least one of the previously learned analysis data generation model and the type code determination model. The processor 120 may set an initial weight of the re-learning model to share at least a portion of the weight of the previously trained model, and train the re-learning model using the re-learning data.

Hereinafter, a method of using customer consultation data when receiving a consultation request from a customer will be described.

The processor 120 may receive information from the agent terminal indicating that the agent terminal has received a consultation request from the customer terminal.

The processor 120 may identify customer information corresponding to the customer terminal. The customer information may include information for distinguishing and identifying the customer from other customers. For example, in order to distinguish between customer A and customer B, personal information such as each customer's name, date of birth, and gender may be included. Such customer information may be encrypted and stored in accordance with relevant laws and regulations.

The processor 120 may transmit at least one of past customer consultation data stored in matching with customer information in the memory 130 or analysis data corresponding to the past customer consultation data to the counselor terminal.

By providing the agent with analysis data on the customer's past customer consultation data who have requested the current consultation, when the customer calls the agent again for consultation, the agent can use the analysis data ( In other words, it is possible to quickly respond to customers by viewing only summary information that summarizes customer consultation data or type code information that classifies customer consultation data into at least one predetermined type code, thereby increasing the agent's work efficiency. have.

2 is a diagram exemplarily illustrating a method for generating counseling analysis data according to an embodiment of the present disclosure.

The analysis data generating processor 202 may correspond to the processor 120 included in the computing device 100 . The call center agent terminal 204 may correspond to the computing device 100 or may be a separate computing device. The database 206 may correspond to the memory 130 included in the computing device 100 .

The call center agent terminal 204 may receive a consultation request from the customer terminal and perform consultation with the customer ( 212 ). Customer consultation data may be generated according to the consultation of the call center agent terminal 204 .

The analysis data generation processor 202 may analyze a key word or key sentence based on the customer consultation data, and may generate 214 summary information based on the key word or key sentence.

The analysis data generation processor 202 may generate 216 type code information based on at least one of a key word, a key sentence, or summary information.

The call center agent terminal 204 may receive 218 the generated summary information and type code information. The call center agent terminal 204 may generate feedback for summary information and type code information. The call center agent terminal 204 may generate positive feedback when the summary information and the type code information are normal, and generate negative feedback when the summary information and the type code information include at least some of the data that is not normal. can

When the feedback generated by the call center agent terminal 204 is positive feedback, the consultation data, summary information, and type code information may be stored in the database 206 (220).

When the summary information and the type code information include at least a part of the data that is not normal, the call center agent terminal 204 may generate error information including correction information for the summary information or the type code information ( 222 ).

When the feedback generated by the call center agent terminal 204 is negative feedback, consultation data and error information may be stored in the database 206 (224).

The analysis data generating processor 202 may update 226 the analysis data generating method based on the error information including the summary information or correction information for the type code information.

5 is a flowchart of a method for generating counseling analysis data according to an embodiment of the present disclosure.

The computing device 100 may receive 510 customer consultation data.

The computing device 100 may generate 520 analysis data on the customer consultation data by using the analysis data generation method.

The analysis data may include at least one of summary information summarizing the customer consultation data or type code information in which the customer consultation data is classified into at least one predetermined type code. The summary information may be information that summarizes customer consultation data including a large amount of data to include only at least some important data. The type code information may be information for classifying the intention of the consultation included in the customer consultation data.

The computing device 100 may determine at least one of a main word or a main sentence regarding the customer consultation data. The computing device 100 may generate at least one of summary information and type code information regarding the customer consultation data based on at least one of a main word or a main sentence. The key word may be a word having high importance for classifying the customer's consultation intention among two or more words included in the customer consultation data. The main sentence may be a sentence having high importance for classifying the customer's counseling intention among two or more sentences included in the customer counseling data.

According to an embodiment of the present disclosure, the computing device 100 may determine the frequency of use of words included in the customer consultation data. The computing device 100 may determine a key word related to the customer consultation data based on the frequency of use of the words.

According to an embodiment of the present disclosure, when the distance between words included in the customer consultation data is equal to or less than a predetermined value, the computing device 100 may determine the relationship of simultaneous appearance of the words. The computing device 100 may generate two or more nodes corresponding to each of the two or more words included in the customer consultation data. The computing device 100 may generate an edge connecting nodes of words having the co-occurrence relationship. The computing device 100 may calculate a weight for each of the edges. The computing device 100 may calculate the importance of each of the two or more nodes based on the weight of the edge. The computing device 100 may determine a key word based on the importance of each of the two or more nodes.

According to an embodiment of the present disclosure, the computing device 100 may determine one sentence among two or more sentences having a high similarity included in the customer consultation data as a representative sentence. The computing device 100 may determine a main sentence regarding the customer consultation data based on the representative sentence.

According to an embodiment of the present disclosure, the computing device 100 may generate two or more nodes corresponding to each of two or more sentences included in the customer consultation data. The computing device 100 may generate an edge connecting nodes of sentences having a similar relationship. The computing device 100 may calculate a weight for each of the edges. The computing device 100 may calculate the importance of each of the two or more nodes based on the weight of the edge. The computing device 100 may determine the main sentence based on the importance of each of the two or more nodes.

According to an embodiment of the present disclosure, the computing device 100 may calculate the customer consultation data using an analysis data generation model including one or more pre-trained network functions. The computing device 100 may output at least one of a main word or a main sentence by using the analysis data generation model. The analysis data generation model may be a model trained with learning data using past customer consultation data as an input and using as a label at least one of key words or main sentences that are matched and stored with respect to the past customer consultation data. The analysis data generation model outputs at least one of a main word or a main sentence based on a first sub-model for identifying a relationship between two or more words included in the customer consultation data and a relationship between the two or more words It may include a second sub-model for

According to an embodiment of the present disclosure, the computing device 100 may determine at least one of a main word or a main sentence regarding the customer consultation data. The computing device 100 compares at least one of the words or sentences stored by matching each of the two or more predetermined type codes with at least one of the main words or the main sentences to obtain type code information based on the type code having a high matching rate. can create

According to an embodiment of the present disclosure, the computing device 100 uses a type code determination model including one or more network functions pre-trained for at least one of the customer consultation data, the main word or the main sentence related to the customer consultation data. can be calculated by The computing device 100 may output type code information using the type code determination model. The type code determination model inputs at least one of a key word or a main sentence related to past customer consultation data, past customer consultation data, and matches stored type code information with respect to the past customer consultation data as a label It may be a model trained with training data.

The computing device 100 may receive 530 feedback regarding the analysis data. The feedback may be evaluation information on analysis data derived based on customer consultation data. Feedback may include error information or normal information about the analysis data.

The computing device 100 may update the method of generating the analysis data based on the feedback or store 540 the analysis data.

When the feedback includes error information on the analysis data, the computing device 100 may update the method of generating the analysis data based on the error information. The computing device 100 may generate re-learning data based on the customer consultation data and the error information. The error information may be negative evaluation information for the analyzed data. The error information may include correction information for information that is not normal when the analysis data includes information that is not normal (ie, incorrect answer). The computing device 100 may update the analysis data generation method using the re-learning data.

When the feedback includes normal information on the analysis data, the computing device 100 may match and store the customer consultation data and the analysis data. The normal information may be positive evaluation information for the analyzed data. The normal information may include information that the analysis data is correct when all of the information included in the analysis data is normal information (ie, correct answers).

The computing device 100 may receive information from the agent terminal indicating that the agent terminal has received a consultation request from the customer terminal. The computing device 100 may identify customer information corresponding to the customer terminal. The computing device 100 may transmit at least one of past customer consultation data stored by matching the customer information or analysis data corresponding to the past customer consultation data to the counselor terminal.

6 is a block diagram of a computing device according to an embodiment of the present disclosure.

6 is a simplified, general schematic diagram of an exemplary computing environment in which embodiments of the present disclosure may be implemented.

Although the present disclosure has been described above generally in the context of computer-executable instructions that may be executed on one or more computers, those skilled in the art will appreciate that the present disclosure may be implemented in combination with other program modules and/or as a combination of hardware and software. will be.

Generally, program modules include routines, programs, components, data structures, etc. that perform particular tasks or implement particular abstract data types. In addition, those skilled in the art will appreciate that the methods of the present disclosure are suitable for single-processor or multiprocessor computer systems, minicomputers, mainframe computers as well as personal computers, hand held computing devices, microprocessor-based or programmable consumer electronics, etc. (each of which is It will be appreciated that other computer system configurations may be implemented, including those that may operate in connection with one or more associated devices.

The described embodiments of the present disclosure may also be practiced in distributed computing environments where certain tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

Computers typically include a variety of computer-readable media. Any medium accessible by a computer may be a computer-readable medium. Computer-readable media includes volatile and non-volatile media, transitory and non-transitory media, removable and non-removable media. By way of example, and not limitation, computer-readable media may include computer-readable storage media and computer-readable transmission media. Computer-readable storage media includes volatile and non-volatile media, transitory and non-transitory media, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. includes media. Computer storage media may include RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital video disk (DVD) or other optical disk storage device, magnetic cassette, magnetic tape, magnetic disk storage device or other magnetic storage device; or any other medium that can be accessed by a computer and used to store the desired information.

Computer-readable transmission media typically embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as other transport mechanisms, and includes all information delivery media. do. The term modulated data signal means a signal in which one or more of the characteristics of the signal is set or changed so as to encode information in the signal. By way of example, and not limitation, computer-readable transmission media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above are also intended to be included within the scope of computer-readable transmission media.

An example environment 1100 implementing various aspects of the disclosure is shown including a computer 1102 , the computer 1102 including a processing unit 1104 , a system memory 1106 , and a system bus 1108 . do. The system bus 1108 couples system components, including but not limited to system memory 1106 , to the processing device 1104 . The processing device 1104 may be any of a variety of commercially available processors. Dual processor and other multiprocessor architectures may also be used as processing unit 1104 .

The system bus 1108 may be any of several types of bus structures that may further be interconnected to a memory bus, a peripheral bus, and a local bus using any of a variety of commercial bus architectures. System memory 1106 includes read only memory (ROM) 1110 and random access memory (RAM) 1112 . A basic input/output system (BIOS) is stored in non-volatile memory 1110, such as ROM, EPROM, EEPROM, etc., which is the basic input/output system (BIOS) that helps transfer information between components within computer 1102, such as during startup contains routines. RAM 1112 may also include high-speed RAM, such as static RAM, for caching data.

Computer 1102 may also be configured for external use within an internal hard disk drive (HDD) 1114 (eg, EIDE, SATA) - this internal hard disk drive 1114 may also be configured for external use within a suitable chassis (not shown). Yes, magnetic floppy disk drive (FDD) 1116 (eg, for reading from or writing to removable diskette 1118), and optical disk drive 1120 (eg, CD-ROM) for reading from, or writing to, disk 1122, or other high capacity optical media such as DVD. The hard disk drive 1114 , the magnetic disk drive 1116 , and the optical disk drive 1120 are connected to the system bus 1108 by the hard disk drive interface 1124 , the magnetic disk drive interface 1126 , and the optical drive interface 1128 , respectively. ) can be connected to The interface 1124 for implementing an external drive includes at least one or both of Universal Serial Bus (USB) and IEEE 1394 interface technologies.

These drives and their associated computer readable media provide non-volatile storage of data, data structures, computer executable instructions, and the like. In the case of computer 1102, drives and media correspond to storing any data in a suitable digital format. Although the description of computer readable media above refers to HDDs, removable magnetic disks, and removable optical media such as CDs or DVDs, those skilled in the art will use zip drives, magnetic cassettes, flash memory cards, cartridges, etc. It will be appreciated that other tangible computer-readable media such as etc. may also be used in the exemplary operating environment and any such media may include computer-executable instructions for performing the methods of the present disclosure.

A number of program modules may be stored in the drive and RAM 1112 , including an operating system 1130 , one or more application programs 1132 , other program modules 1134 , and program data 1136 . All or portions of the operating system, applications, modules, and/or data may also be cached in RAM 1112 . It will be appreciated that the present disclosure may be implemented in various commercially available operating systems or combinations of operating systems.

A user may enter commands and information into the computer 1102 via one or more wired/wireless input devices, for example, a pointing device such as a keyboard 1138 and a mouse 1140 . Other input devices (not shown) may include a microphone, IR remote control, joystick, game pad, stylus pen, touch screen, and the like. Although these and other input devices are often connected to the processing unit 1104 through an input device interface 1142 that is connected to the system bus 1108, parallel ports, IEEE 1394 serial ports, game ports, USB ports, IR interfaces, It may be connected by other interfaces, etc.

A monitor 1144 or other type of display device is also coupled to the system bus 1108 via an interface, such as a video adapter 1146 . In addition to the monitor 1144, the computer typically includes other peripheral output devices (not shown), such as speakers, printers, and the like.

Computer 1102 may operate in a networked environment using logical connections to one or more remote computers, such as remote computer(s) 1148 via wired and/or wireless communications. Remote computer(s) 1148 may be workstations, computing device computers, routers, personal computers, portable computers, microprocessor-based entertainment devices, peer devices, or other common network nodes, and are typically connected to computer 1102 . Although it includes many or all of the components described for it, only memory storage device 1150 is shown for simplicity. The logical connections shown include wired/wireless connections to a local area network (LAN) 1152 and/or a larger network, eg, a wide area network (WAN) 1154 . Such LAN and WAN networking environments are common in offices and companies, and facilitate enterprise-wide computer networks, such as intranets, all of which can be connected to a worldwide computer network, for example, the Internet.

When used in a LAN networking environment, the computer 1102 is connected to the local network 1152 through a wired and/or wireless communication network interface or adapter 1156 . Adapter 1156 may facilitate wired or wireless communication to LAN 1152 , which also includes a wireless access point installed therein for communicating with wireless adapter 1156 . When used in a WAN networking environment, the computer 1102 may include a modem 1158, be connected to a communication computing device on the WAN 1154, or establish communications over the WAN 1154, such as over the Internet. have other means. A modem 1158 , which may be internal or external and a wired or wireless device, is coupled to the system bus 1108 via a serial port interface 1142 . In a networked environment, program modules described for computer 1102 or portions thereof may be stored in remote memory/storage device 1150 . It will be appreciated that the network connections shown are exemplary and other means of establishing a communication link between the computers may be used.

Computer 1102 may be associated with any wireless device or object that is deployed and operates in wireless communication, for example, printers, scanners, desktop and/or portable computers, portable data assistants (PDAs), communications satellites, wireless detectable tags. It operates to communicate with any device or place, and phone. This includes at least Wi-Fi and Bluetooth wireless technologies. Accordingly, the communication may be a predefined structure as in a conventional network or may simply be an ad hoc communication between at least two devices.

Wi-Fi (Wireless Fidelity) makes it possible to connect to the Internet, etc. without a wire. Wi-Fi is a wireless technology such as cell phones that allows these devices, eg, computers, to transmit and receive data indoors and outdoors, ie anywhere within range of a base station. Wi-Fi networks use a radio technology called IEEE 802.11 (a, b, g, etc.) to provide secure, reliable, and high-speed wireless connections. Wi-Fi can be used to connect computers to each other, to the Internet, and to wired networks (using IEEE 802.3 or Ethernet). Wi-Fi networks may operate in unlicensed 2.4 and 5 GHz radio bands, for example, at 11 Mbps (802.11a) or 54 Mbps (802.11b) data rates, or in products that include both bands (dual band). have.

One of ordinary skill in the art of this disclosure will understand that information and signals may be represented using any of a variety of different technologies and techniques. For example, data, instructions, instructions, information, signals, bits, symbols and chips that may be referenced in the above description are voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields particles or particles, or any combination thereof.

A person of ordinary skill in the art of the present disclosure will recognize that the various illustrative logical blocks, modules, processors, means, circuits and algorithm steps described in connection with the embodiments disclosed herein include electronic hardware, (convenience , it will be understood that it may be implemented by various forms of program or design code (referred to herein as "software") or a combination of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. A person skilled in the art of the present disclosure may implement the described functionality in various ways for each specific application, but such implementation decisions should not be interpreted as a departure from the scope of the present disclosure.

The various embodiments presented herein may be implemented as methods, apparatus, or articles of manufacture using standard programming and/or engineering techniques. The term "article of manufacture" includes a computer program or media accessible from any computer-readable device. For example, computer-readable media include magnetic storage devices (eg, hard disks, floppy disks, magnetic strips, etc.), optical disks (eg, CDs, DVDs, etc.), smart cards, and flash memory. devices (eg, EEPROMs, cards, sticks, key drives, etc.). Also, various storage media presented herein include one or more devices and/or other machine-readable media for storing information.

It is understood that the specific order or hierarchy of steps in the presented processes is an example of exemplary approaches. Based on design priorities, it is understood that the specific order or hierarchy of steps in the processes may be rearranged within the scope of the present disclosure. The appended method claims present elements of the various steps in a sample order, but are not meant to be limited to the specific order or hierarchy presented.

The description of the presented embodiments is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art of the present disclosure. The generic principles defined herein may be applied to other embodiments without departing from the scope of the present disclosure. Thus, the present disclosure is not intended to be limited to the embodiments presented herein, but is to be construed in the widest scope consistent with the principles and novel features presented herein.

Claims (17)

A computer program stored on a computer-readable storage medium, the computer program, when executed on one or more processors of a computing device, to perform operations for generating counseling analysis material, the operations comprising:
receiving customer service data;
generating analysis data on the customer consultation data by using the analysis data generation method;
receiving feedback on the analysis data; and
updating the analysis data generating method or storing the analysis data based on the feedback;
including,
The operation of generating analysis data on the customer consultation data using the analysis data generation method includes:
generating two or more nodes and at least one edge connecting the nodes;
calculating a weight of each node by equally distributing weights to the two or more nodes;
calculating the weight of each edge by equally distributing the weight of the node to edges connected to each node; and
calculating an importance for each of the two or more nodes based on a first predetermined probability for the weight of the node and a second predetermined probability for the weight of the edge;
comprising,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The analysis data includes at least one of summary information summarizing the customer consultation data or type code information in which the customer consultation data is classified into at least one predetermined type code,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The operation of generating analysis data on the customer consultation data using the analysis data generation method includes:
determining at least one of a main word or a main sentence related to the customer consultation data; and
generating at least one of summary information or type code information about the customer consultation data based on at least one of the main word or the main sentence;
comprising,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The operation of generating analysis data on the customer consultation data using the analysis data generation method includes:
determining the frequency of use of words included in the customer consultation data; and
determining a key word related to the customer consultation data based on the frequency of use of the words;
comprising,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The operation of generating analysis data on the customer consultation data using the analysis data generation method includes:
determining a co-occurrence relationship of the words when the distance between the words included in the customer consultation data is less than or equal to a predetermined value; and
determining a key word related to the customer consultation data based on the degree of simultaneous appearance of the words;
comprising,
A computer program stored on a computer-readable storage medium.
6. The method of claim 5,
The operation of determining a key word related to the customer consultation data based on the degree of simultaneous appearance of the words includes:
determining a key word based on the importance for each of the two or more nodes;
including,
The two or more nodes correspond to each of the two or more words included in the customer consultation data,
The edge is an edge connecting nodes of words having a co-occurrence relationship,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The operation of generating analysis data on the customer consultation data using the analysis data generation method includes:
determining one sentence as a representative sentence based on the similarity between the two or more sentences calculated based on one or more words included in each of the two or more sentences included in the customer consultation data; and
determining a main sentence regarding the customer consultation data based on the representative sentence;
comprising,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The operation of generating analysis data on the customer consultation data using the analysis data generation method includes:
determining a main sentence based on the importance for each of the two or more nodes;
including,
The two or more nodes correspond to each of the two or more sentences included in the customer consultation data,
The edge is an edge connecting nodes of sentences having a similar relationship,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The operation of generating analysis data on the customer consultation data using the analysis data generation method includes:
calculating the customer consultation data using an analysis data generation model based on a deep learned neural network; and
outputting at least one of a main word or a main sentence using the analysis data generation model;
including, and
The analysis data generation model is a model trained with learning data using past customer consultation data as an input, and matching at least one of the stored key words or main sentences with respect to the past customer consultation data as a label,
A computer program stored on a computer-readable storage medium.
10. The method of claim 9,
The analysis data generation model outputs at least one of a main word or a main sentence based on a first sub-model for identifying a relationship between two or more words included in the customer consultation data and a relationship between the two or more words including a second sub-model for
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The operation of generating analysis data on the customer consultation data using the analysis data generation method includes:
determining at least one of a main word or a main sentence related to the customer consultation data; and
Type code information based on a type code in which a matching rate calculated by comparing at least one of the stored words or sentences matching each of the two or more predetermined type codes with at least one of the main word or the main sentence is equal to or greater than a predetermined threshold value creating an action;
comprising,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The operation of generating analysis data on the customer consultation data using the analysis data generation method includes:
calculating at least one of a main word or a main sentence related to the customer consultation data using a type code determination model based on a deep learned neural network; and
outputting type code information using the type code determination model;
including, and
The type code determination model receives at least one of a main word or a main sentence regarding past customer consultation data as an input, and is trained as learning data using the stored type code information as a label to match the past customer consultation data. model,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
The operation of updating the analysis data generating method or storing the analysis data based on the feedback includes:
updating the analysis data generating method based on the error information when the feedback includes error information on the analysis data; and
matching and storing the customer consultation data and the analysis data when the feedback includes normal information on the analysis data;
comprising at least one operation of
A computer program stored on a computer-readable storage medium.
14. The method of claim 13,
If the feedback includes error information on the analysis data, updating the analysis data generating method based on the error information includes:
generating re-learning data based on the customer consultation data and the error information; and
updating the analysis data generation method using the re-learning data;
comprising,
A computer program stored on a computer-readable storage medium.
The method of claim 1,
receiving, from the agent terminal, information that the agent terminal has received a consultation request from the customer terminal;
identifying customer information corresponding to the customer terminal; and
transmitting at least one of past customer consultation data stored by matching the customer information and analysis data corresponding to the past customer consultation data to the agent terminal;
further comprising,
A computer program stored on a computer-readable storage medium.
As a method for generating counseling analysis data using a computing device,
receiving, by the processor included in the computing device, customer consultation data;
generating, by the processor, analysis data on the customer consultation data using the analysis data generation method;
receiving, by the processor, feedback on the analysis data; and
updating, by the processor, the method of generating the analysis data or storing the analysis data based on the feedback;
including,
The step of generating analysis data on the customer consultation data by using the analysis data generation method includes:
generating, by the processor, two or more nodes and at least one edge connecting the nodes;
calculating, by the processor, a weight of each node by equally distributing weights to the two or more nodes;
calculating, by the processor, the weight of each edge by equally distributing the weight of the node to edges connected to each node; and
calculating, by the processor, a degree of importance for each of the two or more nodes based on a first predetermined probability for the weight of the node and a second predetermined probability for the weight of the edge;
containing,
A method for generating counseling analysis data using a computing device.
As a server for generating consultation analysis data,
a processor including one or more cores; and
Memory;
including,
The processor is
receive customer service data;
generating analysis data about the customer consultation data using the analysis data generation method;
receive feedback regarding the analysis; and
Update the analysis data generation method based on the feedback or store the analysis data,
The method for generating the analysis data includes the following steps:
generating two or more nodes and at least one edge connecting the nodes;
calculating a weight of each node by equally distributing weights to the two or more nodes;
calculating a weight of each edge by equally distributing the weight of the node to edges connected to each node; and
calculating an importance for each of the two or more nodes based on a first predetermined probability for the weight of the node and a second predetermined probability for the weight of the edge;
characterized in that it comprises,
Server for generating consultation analysis data.

Images