Method, apparatus and program for providing hr support service through ai interview

Abstract

The present disclosure relates to a method for providing a recruitment support service through an artificial intelligence interview. A method performed by a computing device comprises the steps of: obtaining a user’s recruitment information and an interviewee’s application information; generating an interview question based on at least some of the recruitment information and the application information using a generative artificial intelligence model (Generative AI); providing the generated interview question to the interviewee; obtaining the interviewee’s answer; and evaluating the interviewee based on the obtained interviewee’s answer.

Description

Technology Field

[0001] The present disclosure relates to a method, apparatus, and program for providing recruitment support services through artificial intelligence interviews. Background Technology

[0002] With the recent rapid advancement of Artificial Intelligence (AI) technology, AI-based automation systems are being introduced across various industries, and AI-powered recruitment and interview systems are gaining particular attention in the field of corporate Human Resources (HR).

[0003] The traditional interview process involved interviewers reviewing applicants' resumes and cover letters, generating questions, and evaluating them through face-to-face interviews. This conventional method suffered from drawbacks, such as a heavy reliance on the interviewers' subjective judgments and the difficulty in maintaining consistency in evaluation criteria. Furthermore, interviewing a large number of applicants individually during the recruitment process required significant time and cost, and there was a risk that fairness could be compromised due to interviewer fatigue.

[0004] Accordingly, research is underway to systematically analyze applicants' backgrounds and conduct more objective evaluations during the recruitment process by utilizing data analysis and Natural Language Processing (NLP) technologies. In particular, technologies are being developed to extract key information from applicants' submitted documents and perform more precise analyses, with attempts being made to enhance the efficiency of the recruitment procedure through these efforts.

[0005] Traditional interview questions were primarily structured in a standardized manner, posing a problem in that they made it difficult to adequately reflect the individual characteristics of applicants. Consequently, there is a need for technology to construct interview questions in a more flexible and systematic manner, and research is being conducted on customized evaluation methods that consider applicants' careers and experiences.

[0006] The aforementioned background technology is one that the inventor possessed or acquired in the process of deriving the content of the present disclosure, and it cannot be considered as prior art disclosed to the general public prior to the filing of this application. The problem to be solved

[0007] The problem that the present disclosure aims to solve is to develop a technology that effectively analyzes an interviewee's self-introduction and resume data using a Large Language Model (LLM) and automatically generates interview questions based on this analysis. Additionally, by utilizing Retrieval-Augmented Generation (RAG) technology, the technology can retrieve additional information related to the applicant's answers and provide a more in-depth evaluation based on this information.

[0008] The problems that this disclosure aims to solve are not limited to those mentioned above, and other unmentioned problems will be clearly understood by a person skilled in the art from the description below. means of solving the problem

[0009] A method for providing a recruitment support service through an artificial intelligence interview according to an embodiment of the present disclosure for solving the aforementioned problem is disclosed. In one embodiment, the method performed by a computing device may include the steps of: obtaining recruitment information of a user and application information of an interviewee; generating an interview question based on at least some of the recruitment information and the application information using a generative artificial intelligence model (Generative AI); providing the generated interview question to the interviewee; obtaining an answer from the interviewee; and evaluating the interviewee based on the obtained answer from the interviewee.

[0010] In various embodiments, the step of generating the interview question may include generating input data requesting the generation of the interview question based on at least some of the recruitment information and the support information, inputting the generated input data into the generative artificial intelligence model, and obtaining the output of the generative artificial intelligence model.

[0011] In various embodiments, the step of generating the interview question may include the step of obtaining the interviewee's previous interview answer and the step of generating additional interview questions based on the obtained previous interview answer.

[0012] In various embodiments, the step of generating the interview question may include generating a prompt including conditions for generating the interview question and a command requesting the generation of the interview question, and generating input data including the recruitment information, the support information, and the generated prompt.

[0013] In various embodiments, the step of generating the interview question may include the step of obtaining a target question from an interview question database and the step of generating the interview question based on the target question, the recruitment information, and the support information using the generative artificial intelligence model.

[0014] In various embodiments, the step of evaluating the interviewer may include the step of obtaining the interviewer's previous interview answer, the step of calculating a similarity with the obtained previous interview answer based on the obtained interviewer's answer, and the step of determining the reliability of the interviewer's answer based on the calculated similarity.

[0015] In various embodiments, the step of evaluating the interviewee may include the step of preprocessing the acquired interview answer data and the step of evaluating the interviewee using the preprocessed interview answer data.

[0016] In various embodiments, the preprocessing step may include extracting information related to the user's employment information as necessary information among a plurality of pieces of information included in the acquired interview answer data, or filtering information not related to the user's employment information as unnecessary information.

[0017] In various embodiments, the method may include the steps of organizing necessary information extracted from the preprocessed interview answer data to create a database that enables searching, obtaining search keywords from the user, searching for information corresponding to the search keywords, and providing one or more interview answers related to the searched information and information about the interviewer who responded with the one or more interview answers.

[0018] In various embodiments, the step of evaluating the interviewer may include the step of obtaining unlisted information among the interviewer's answers that is not included in the interviewer's application information, and the step of correcting the interviewer's application information based on the obtained unlisted information.

[0019] In various embodiments, the apparatus comprises a memory for storing one or more instructions and a processor for executing the one or more instructions stored in the memory, wherein the processor performs a method comprising the steps of: obtaining user recruitment information and interviewee application information by executing the one or more instructions; generating interview questions based on at least some of the recruitment information and the application information using a generative artificial intelligence model (Generative AI); providing the generated interview questions to the interviewee; obtaining the interviewee's answer; and evaluating the interviewee based on the obtained interviewee's answer.

[0020] A computer program stored on a computer-readable recording medium to enable a method comprising, in various embodiments, combining with a computer which is hardware, obtaining user recruitment information and interviewee application information; generating interview questions based on at least some of said recruitment information and said application information using a generative artificial intelligence model (Generative AI); providing said interview questions to said interviewee; obtaining said interviewee's answers; and evaluating said interviewee based on said obtained answers.

[0021] Other specific details of the present disclosure are included in the detailed description and drawings. Effects of the invention

[0022] According to the first disclosure, the present invention can automatically generate customized interview questions by analyzing an interviewee's self-introduction data using AI. This allows for a departure from conventional standardized interview questioning methods and has the effect of more precisely evaluating the applicant's individual experiences and capabilities.

[0023] By applying an AI-based evaluation model that can analyze and evaluate the interviewee's answers in real time upon initiation, the objectivity of the interview process can be enhanced and the consistency of evaluation criteria can be maintained.

[0024] Upon initiation, RAG technology can be utilized to search for additional information related to the applicant's responses and conduct a more in-depth evaluation based on this. This reduces the intervention of subjective judgment by evaluators that can occur in traditional interview methods and enables the fair assessment of the applicant's capabilities based on quantitative analysis.

[0025] In addition, after analyzing the interviewee's responses, follow-up questions can be automatically generated to supplement areas of weakness, enabling a more systematic and in-depth interview process.

[0026] The effects of the present disclosure are not limited to those mentioned above, and other unmentioned effects will be clearly understood by a person skilled in the art from the description below. Brief explanation of the drawing

[0027] The following drawings attached to this specification illustrate preferred embodiments of the present disclosure and serve to further enhance understanding of the technical concept of the present disclosure together with the detailed description of the invention; therefore, the present disclosure should not be interpreted as being limited only to the matters described in such drawings. FIG. 1 is a diagram illustrating a system for providing recruitment support services through artificial intelligence interviews according to one embodiment of the present disclosure. FIG. 2 is a diagram illustrating the hardware configuration of a computing device that performs a method for providing a recruitment support service through an artificial intelligence interview according to another embodiment of the present disclosure. FIG. 3 is a flowchart illustrating a method for providing recruitment support services through artificial intelligence interviews in various embodiments. FIG. 4 is a flowchart illustrating a method for generating interview questions in various embodiments. FIG. 5 is a flowchart illustrating a method for evaluating an interviewee in various embodiments. FIG. 6 is a flowchart illustrating a method for determining whether each interviewee passes an interview based on the answers of multiple interviewees obtained in various embodiments. FIG. 7 is a flowchart illustrating a method for evaluating an interviewee in various embodiments. FIGS. 8 to 10 are drawings illustrating a user interface (UI) that provides a recruitment support service through artificial intelligence interviews in various embodiments. Specific details for implementing the invention

[0028] The advantages and features of the present disclosure and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure is complete and to fully inform those skilled in the art of the scope of the present disclosure, and the present disclosure is defined only by the scope of the claims.

[0029] The terms used herein are for describing the embodiments and are not intended to limit the disclosure. In this specification, the singular form includes the plural form unless specifically stated otherwise in the text. As used herein, "comprises" and / or "comprising" do not exclude the presence or addition of one or more other components in addition to the components mentioned.

[0030] Throughout this specification, the same reference numerals refer to the same components, and "and / or" includes each of the mentioned components and all combinations of one or more thereof. Although terms such as "first," "second," etc., are used to describe various components, they are not limited by these terms. These terms are used merely to distinguish one component from another. Accordingly, the first component mentioned below may be the second component within the technical scope of this disclosure.

[0031] As used herein, the terms “part” or “module” refer to hardware components such as software, FPGAs, or ASICs, and the “part” or “module” perform certain roles. However, the meaning of “part” or “module” is not limited to software or hardware. The “part” or “module” may be configured to reside in an addressable storage medium or may be configured to run on one or more processors. Thus, by example, the “part” or “module” includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. The functions provided within the components and “parts” or “modules” may be combined into a smaller number of components and “parts” or “modules,” or further separated into additional components and “parts” or “modules.”

[0032] Expressions such as "first," "second," or "first," "second" as used in this specification are used to distinguish one object from another when referring to a plurality of objects of the same kind, unless otherwise indicated by the context, and do not limit the order or importance of said objects.

[0033] Expressions used herein such as “A, B, and C,” “A, B, or C,” “A, B, and / or C,” or “at least one of A, B, and C,” “at least one of A, B, or C,” “at least one of A, B, and / or C,” “at least one selected from A, B, and C,” “at least one selected from A, B, or C,” “at least one selected from A, B, and / or C,” etc., may mean each of the listed items or all possible combinations of the listed items. For example, “at least one selected from A and B” may refer to (1) A, (2) at least one of A, (3) B, (4) at least one of B, (5) at least one of A and at least one of B, (6) at least one of A and B, (7) at least one of B and A, and (8) all of A and B.

[0034] As used herein, the expression “based on” is used to describe one or more factors affecting an act or action of a decision or judgment described in the phrase or sentence containing such expression, and such expression does not exclude additional factors affecting said act or action of a decision or judgment.

[0035] As used in this specification, the expression that a certain component (e.g., a first component) is "connected" or "connected" to another component (e.g., a second component) may mean that the said certain component is not only directly connected or connected to the said other component, but is also connected or connected through a new other component (e.g., a third component).

[0036] As used herein, the expression "configured to" may have meanings such as "set to," "capable of," "modified to," "made to," or "capable of." Such expression is not limited to the meaning of "specifically designed in hardware," and, for example, a processor configured to perform a specific operation may mean a generic-purpose processor capable of performing that specific operation by executing software.

[0037] Unless otherwise defined, all terms used herein (including technical and scientific terms) may be used in a meaning commonly understood by those skilled in the art to which this disclosure pertains. Additionally, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise.

[0038] In this specification, the term "computer" refers to any type of hardware device comprising at least one processor, and may be understood to include software configurations operating on said hardware device according to the embodiments. For example, the term "computer" may be understood to include smartphones, tablet PCs, desktops, laptops, and user clients and applications running on each of these devices, but is not limited thereto.

[0039] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the attached drawings.

[0040] Each step described in this specification is described as being performed by a computer, but the subject of each step is not limited thereto, and depending on the embodiment, at least some of each step may be performed on different devices.

[0041] FIG. 1 is a diagram illustrating a method for providing a recruitment support service through an artificial intelligence interview according to one embodiment of the present disclosure.

[0042] Referring to FIG. 1, a method for providing a recruitment support service through an artificial intelligence interview according to one embodiment of the present disclosure may include a computing device (100), a terminal (200), an external server (300), and a network (400).

[0043] Here, the method of providing a recruitment support service through an artificial intelligence interview illustrated in FIG. 1 is according to one embodiment, and the components thereof are not limited to the embodiment illustrated in FIG. 1 and may be added, changed, or deleted as needed.

[0044] In one embodiment, the computing device (100) can provide a method for providing a recruitment support service through an artificial intelligence interview.

[0045] In various embodiments, the computing device (100) can obtain user recruitment information and interviewee application information, and can generate interview questions based on at least some of the recruitment information and application information using a generative artificial intelligence model. Additionally, the generated interview questions can be provided to the interviewee to obtain the interviewee's answers, and the interviewee can be evaluated based on the obtained answers from the interviewee.

[0046] Here, an artificial intelligence model (e.g., a neural network) is composed of one or more network functions, and one or more network functions may consist of a set of interconnected computational units that can generally be referred to as 'nodes'. These 'nodes' may also be referred to as 'neurons'. One or more network functions are composed of at least one node. The nodes (or neurons) constituting one or more network functions may be interconnected by one or more 'links'.

[0047] In an artificial intelligence model, one or more nodes connected via links can form a relative relationship between an input node and an output node. The concepts of input and output nodes are relative; any node in an output node relationship with respect to one node may be in an input node relationship with respect to another node, and vice versa. As previously mentioned, the input node versus output node relationship can be generated based on links. One or more output nodes may be connected to a single input node via links, and vice versa.

[0048] In a relationship between input and output nodes connected via a single link, the value of the output node can be determined based on data input into the input node. Here, the nodes interconnecting the input and output nodes may have weights. These weights can be variable and may be varied by a user or an algorithm to enable the artificial intelligence model to perform desired functions. For example, if one or more input nodes are interconnected to a single output node via respective links, the output node value can be determined based on the values ​​input into the input nodes connected to the output node and the weights set on the links corresponding to each input node.

[0049] As described above, an artificial intelligence model consists of one or more nodes interconnected through one or more links, forming input and output node relationships within the model. The characteristics of an artificial intelligence model can be determined by the number of nodes and links within the model, the relationships between the nodes and links, and the weight values ​​assigned to each link. For example, if two artificial intelligence models exist with the same number of nodes and links but different weight values ​​between the links, the two models may be perceived as different from each other.

[0050] Some of the nodes constituting an artificial intelligence model may form a layer based on their distances from the initial input node. For example, a set of nodes with a distance of n from the initial input node may form n layers. The distance from the initial input node can be defined by the minimum number of links that must be traversed to reach that node from the initial input node. However, this definition of a layer is arbitrary for illustrative purposes, and the degree of a layer within the artificial intelligence model may be defined in a way different from that described above. For example, the layer of nodes may be defined by their distance from the final output node.

[0051] The initial input node may refer to one or more nodes within the artificial intelligence model to which data is directly input without passing through links in relation to other nodes. Alternatively, within the artificial intelligence model network, in terms of relationships between nodes based on links, it may refer to nodes that do not have other input nodes connected by links. Similarly, the final output node may refer to one or more nodes within the artificial intelligence model that do not have output nodes in relation to other nodes. Furthermore, the hidden node may refer to nodes constituting the artificial intelligence model that are neither the initial input node nor the final output node. An artificial intelligence model according to one embodiment of the present invention may have more nodes in the input layer than nodes in the hidden layer that are close to the output layer, and may be an artificial intelligence model in which the number of nodes decreases as it progresses from the input layer to the hidden layer.

[0052] An artificial intelligence model may include one or more hidden layers. The hidden nodes of a hidden layer can take the output of the previous layer and the output of neighboring hidden nodes as input. The number of hidden nodes for each hidden layer may be the same or different. The number of nodes in the input layer may be determined based on the number of data fields in the input data and may be the same or different from the number of hidden nodes. The input data fed into the input layer can be processed by the hidden nodes of the hidden layer and output by the fully connected layer (FCL), which is the output layer.

[0053] In various embodiments, the artificial intelligence model may be a deep learning model.

[0054] A deep learning model (e.g., a deep neural network (DNN)) can refer to an artificial intelligence model that includes multiple hidden layers in addition to input and output layers. Using a deep neural network, one can identify the latent structures of data. That is, one can identify the latent structures of photos, text, videos, voice, and music (e.g., what objects are in a photo, what the content and emotions of the text are, what the content and emotions of the voice are, etc.).

[0055] Deep neural networks may include, but are not limited to, convolutional neural networks (CNN), recurrent neural networks (RNN), autoencoders, Generative Adversarial Networks (GAN), restricted Boltzmann machines (RBM), deep belief networks (DBN), Q networks, U networks, and Siamese networks.

[0056] In various embodiments, the network function may include an autoencoder. Here, the autoencoder may be a type of artificial neural network for outputting output data similar to the input data.

[0057] An autoencoder may include at least one hidden layer, and an odd number of hidden layers may be placed between the input and output layers. The number of nodes in each layer may be reduced from the number of nodes in the input layer to an intermediate layer called the bottleneck layer (encoding), and then expanded symmetrically from the bottleneck layer to the output layer (symmetrical to the input layer). The nodes of the dimensionality reduction layer and the dimensionality restoration layer may or may not be symmetrical. Additionally, the autoencoder can perform non-linear dimensionality reduction. The number of input and output layers may correspond to the number of sensors remaining after the preprocessing of the input data. In the autoencoder structure, the number of nodes in the hidden layers included in the encoder may have a structure where it decreases as it moves away from the input layer. Since the number of nodes in the bottleneck layer (the layer with the fewest nodes located between the encoder and the decoder) may not transmit a sufficient amount of information if it is too small, it may be maintained at a certain number or higher (e.g., more than half the number of the input layer).

[0058] In various embodiments, the artificial intelligence model may be a generative artificial intelligence model (Generative AI).

[0059] Generative artificial intelligence models can refer to natural language models that have been pre-trained on large-scale data using Transformers. Transformer models are constructed based on an attention mechanism and are characterized by an encoder-decoder structure that improves upon the shortcomings of existing sequence-to-sequence models and designs an encoder-decoder that does not use recurrent neural networks.

[0060] Attention was previously used to calibrate recurrent neural networks in sequence-to-sequence models, but the Transformer designed an encoder-decoder using only attention without recurrent neural networks. The attention technique used in the Transformer is called Multi-Head Attention or Self-Attention.

[0061] The Transformer model is characterized by being trained to increase the probability value of the correct answer (the next word in the output sequence) given the encoder input (input sequence) and the decoder input (words constituting the output sequence).

[0062] A pre-trained language model can refer to a language model that has learned the process of performing a specific task using large-scale language data (corpus). For example, GPT-family models are trained on the task of predicting the next word given previous words, and undergo a training process to update the model to correctly predict the next word. As another example, BERT-family models are trained on the task of predicting the empty word in the middle from given words on both sides, and undergo a training process to update the model to correctly predict the empty word.

[0063] The characteristic of these tasks is that the computer can perform self-supervised learning by automatically covering and guessing words one by one, even when a large amount of language data, which is relatively easy to collect, is input without any separate preliminary work. Once a model pre-trained on a large amount of language data in this way is completed, it can be utilized by fine-tuning it to perform various detailed tasks.

[0064] Fine-tuning is a type of transfer learning method that retrains a model trained on a specific task to perform a different task; it is a method that leverages the experience of a pre-trained model to enable it to learn new tasks relatively easily.

[0065] For example, a pre-trained language model is trained to fill in the blanks, but by fine-tuning it, you can obtain a model that generates responses to questions or classifies documents.

[0066] By significantly increasing the scale of these pre-trained language models and pre-training them with an overwhelmingly large amount of data, results were obtained that greatly surpassed the performance of existing artificial intelligence models. Consequently, the AI ​​models developed in this way are called Hyperscale AI or Large Language Models (LLM).

[0067] In various embodiments, the computing device (100) can analyze the interviewee's answers to identify additional information not included in the resume and automatically update it. For example, if the interviewee mentioned specific skills or project experience during the interview but it was not listed in the application, the computing device (100) can extract such information to supplement the interviewee's application information. The interviewee can more clearly reflect the strengths and skills revealed during the interview process, and the hiring manager can evaluate the candidate based on more accurate information. Additionally, even if the interviewee does not manually edit their resume, the AI ​​can automatically analyze the interview answers and supplement the missing parts of the application.

[0068] In various embodiments, the computing device (100) may generate interview questions and provide them to the user. The computing device (100) automatically generates interview questions based on the interviewee's application information and recruitment information, and the user may review the questions generated by the computing device (100) and select or modify some of them. Additionally, the user may directly add questions, or additional questions may be generated immediately based on the interviewee's response.

[0069] Additionally, the computing device (100) can generate the next interview question based on the interviewee's interview answer data or the generated interview question through an artificial intelligence model. More specifically, the computing device (100) can identify parts of the interviewee's answer that are unclear or require further explanation, and based on this, can generate the next interview question by deepening or expanding the interview question.

[0070] In various embodiments, the computing device (100) may provide the interview service as an independent service or in combination with a recruitment platform. In the case of an independent interview service, the user sends an interview notification email to candidates who have passed the document screening, and the email may include a link to conduct the interview. The candidate's responses provided during the interview process may be recorded, summarized, evaluated, and organized, and then transmitted to the computing device (100).

[0071] Additionally, the interview service may be provided in combination with a recruitment platform. In this case, the user can manage the entire process from candidate management to conducting interviews in an integrated manner on the recruitment platform. For example, when the user selects a candidate who has passed the document screening, an interview is automatically requested via a computing device (100) within the platform, and after the interview is completed, the results can be organized and provided.

[0072] In various embodiments, the computing device (100) may provide a mock interview service to enable an interviewee to improve their interview skills before an actual interview. The mock interview is conducted using the computing device (100), and customized questions may be generated based on the interviewee's resume, career, skill stack, and job position. The interviewee may practice answers in an environment similar to the actual interview through a virtual interview with the computing device (100), and the responses provided during the interview process may be automatically analyzed to provide feedback.

[0073] In various embodiments, the computing device (100) may provide feedback on the interviewee's interview response by analyzing the response provided during the interview process. More specifically, the computing device (100) may analyze the interviewee's text, voice, or video interview response data to identify the logical structure, job relevance, intonation, eye contact, posture, facial expressions, etc., of the interview response and provide overall feedback on the interview based thereon. At this time, the computing device (100) may provide feedback on the interview content and interview response by analyzing the interview response data and using a large-scale language model (LLM).

[0074] In various embodiments, the computing device (100) may be connected to a user terminal (200) via a network (400) and may provide a UI that provides a recruitment support service through an artificial intelligence interview to the user terminal (200).

[0075] Here, the terminal (200) may refer to any form of entity(s) in a system having a mechanism for communicating with a computing device (100). For example, such a terminal (200) may include a PC (personal computer), a notebook, a mobile terminal, a smartphone, a tablet PC, and a wearable device, and may include any type of terminal capable of connecting to a wired or wireless network. Additionally, the terminal (200) may include any computing device implemented by at least one of an agent, an API (Application Programming Interface), and a plug-in. Additionally, the terminal (200) may include an application source and / or a client application.

[0076] Additionally, the network (400) may refer to a connection structure capable of exchanging information between each node, such as multiple terminals and servers. For example, the network (400) may include a Local Area Network (LAN), a Wide Area Network (WAN), the World Wide Web (WWW), a wired / wireless data network, a telephone network, a wired / wireless television network, a Controller Area Network (CAN), and Ethernet.

[0077] Wireless data communication networks may include, but are not limited to, 3G, 4G, 5G, 3GPP (3rd Generation Partnership Project), 5GPP (5th Generation Partnership Project), LTE (Long Term Evolution), WIMAX (World Interoperability for Microwave Access), Wi-Fi, Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth network, NFC (Near-Field Communication) network, satellite broadcasting network, analog broadcasting network, DMB (Digital Multimedia Broadcasting) network, etc.

[0078] In one embodiment, an external server (300) may be connected to a computing device (100) via a network (400) and may store and manage various information and data necessary for the computing device (100) to perform a method of providing a recruitment support service through an artificial intelligence interview, or the computing device (100) may collect, store, and manage various information and data derived as it performs the method of providing a recruitment support service through an artificial intelligence interview. For example, the external server (300) may be a storage server separately provided outside the computing device (100), but is not limited thereto. Hereinafter, with reference to FIG. 2, the hardware configuration of the computing device (100) that performs the method of providing a recruitment support service through an artificial intelligence interview will be described.

[0079] FIG. 2 is a diagram illustrating the hardware configuration of a computing device that performs a method for providing a recruitment support service through an artificial intelligence interview according to another embodiment of the present disclosure.

[0080] Referring to FIG. 2, a computing device (100) according to another embodiment of the present disclosure may include one or more processors (110), a memory (120) for loading a computer program (151) executed by the processor (110), a bus (130), a communication interface (140), and a storage (150) for storing the computer program (151). Here, FIG. 2 illustrates only the components related to the embodiments of the present disclosure. Accordingly, a person skilled in the art to which the present disclosure pertains will understand that other general-purpose components may be included in addition to the components illustrated in FIG. 2.

[0081] The processor (110) controls the overall operation of each component of the computing device (100). The processor (110) may be configured to include a CPU (Central Processing Unit), an MPU (Micro Processor Unit), an MCU (Micro Controller Unit), a GPU (Graphic Processing Unit), or any form of processor well known in the art of the present disclosure.

[0082] Additionally, the processor (110) may perform operations for at least one application or program for executing the method according to the embodiments of the present disclosure, and the computing device (100) may have one or more processors.

[0083] In various embodiments, the processor (110) may further include Random Access Memory (RAM) (not shown) and Read-Only Memory (ROM) (not shown) for temporarily and / or permanently storing signals (or data) processed within the processor (110). Additionally, the processor (110) may be implemented in the form of a System on Chip (SoC) comprising at least one of a graphics processing unit, RAM, and ROM.

[0084] Memory (120) stores various data, instructions and / or information. Memory (120) may load a computer program (151) from storage (150) to execute a method / operation according to various embodiments of the present disclosure. When the computer program (151) is loaded into memory (120), the processor (110) may perform the method / operation by executing one or more instructions constituting the computer program (151). Memory (120) may be implemented as a volatile memory such as RAM, but the technical scope of the present disclosure is not limited thereto.

[0085] The bus (130) provides communication functions between components of the computing device (100). The bus (130) can be implemented as various types of buses, such as an address bus, a data bus, and a control bus.

[0086] The communication interface (140) supports wired and wireless internet communication of the computing device (100). Additionally, the communication interface (140) may support various communication methods other than internet communication. To this end, the communication interface (140) may be configured to include a communication module well known in the art of the present disclosure. In some embodiments, the communication interface (140) may be omitted.

[0087] Storage (150) can store a computer program (151) non-temporarily. When performing a recruitment support service provision process through an artificial intelligence interview via a computing device (100), storage (150) can store various information necessary to provide the recruitment support service provision process through an artificial intelligence interview.

[0088] The storage (150) may be configured to include non-volatile memory such as ROM (Read Only Memory), EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), flash memory, a hard disk, a removable disk, or any form of computer-readable recording medium well known in the art to which the present disclosure belongs.

[0089] A computer program (151) may include one or more instructions that cause a processor (110) to perform a method / operation according to various embodiments of the present disclosure when loaded into memory (120). That is, the processor (110) may perform the method / operation according to various embodiments of the present disclosure by executing the one or more instructions.

[0090] In one embodiment, the computer program (151) may include one or more instructions for performing a method of providing a recruitment support service through a generative AI interview, the step of obtaining a user's recruitment information and an interviewee's application information, and the step of generating an interview question based on at least some of the obtained recruitment information and application information using a pre-trained generative AI model.

[0091] The steps of the method or algorithm described in connection with the embodiments of the present disclosure may be implemented directly in hardware, implemented as a software module executed by hardware, or implemented by a combination thereof. The software module may reside in RAM (Random Access Memory), ROM (Read Only Memory), EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), Flash Memory, a hard disk, a removable disk, a CD-ROM, or any form of computer-readable recording medium well known in the art to which the present disclosure belongs.

[0092] The components of the present disclosure may be implemented as a program (or application) and stored on a medium to be executed in combination with a computer, which is hardware. The components of the present disclosure may be executed as software programming or software elements, and similarly, embodiments may be implemented in programming or scripting languages ​​such as C, C++, Java, assembler, etc., including various algorithms implemented as a combination of data structures, processes, routines, or other programming configurations. Functional aspects may be implemented as algorithms executed on one or more processors. Hereinafter, with reference to FIGS. 3 through 10, a method for providing a recruitment support service through an artificial intelligence interview performed by a computing device (100) will be described.

[0094] FIG. 3 is a flowchart of a method for providing a recruitment support service through an artificial intelligence interview according to another embodiment of the present invention.

[0095] Referring to FIG. 3, in step S110, the computing device (100) can obtain the user's recruitment information and the interviewee's application information.

[0096] Here, the user's recruitment information refers to information utilized by a company during the recruitment process, and may include, but is not limited to, job descriptions, recruitment categories, employment types (full-time, contract, intern, etc.), number of hires, job descriptions, and mandatory and preferred requirements for selecting the talent desired by the company.

[0097] In addition, applicant information refers to the data submitted when an applicant applies for a company's job posting, and can be used to generate customized questions during the interview process. Applicant information may include, but is not limited to, the applicant's basic profile (name, age, etc.), educational background, work experience, technical skills and certifications, personal statement and motivation, and portfolio.

[0098] In various embodiments, the computing device (100) can acquire recruitment information and applicant application information data in text form. For example, the computing device (100) can receive and analyze detailed information included in the job posting and job description via a user terminal (200), and can provide a UI for submitting a resume to collect data such as the applicant's application, resume, self-introduction, technical stack, and certifications held, and use this to generate customized interview questions. Additionally, the data can be processed to evaluate specific job competencies based on the applicant's submitted portfolio and project experience, and if a request for additional information occurs during the real-time interview process, immediate feedback can be exchanged through the user terminal (200) to acquire recruitment information and applicant application information data.

[0099] In step S120, the computing device (100) can generate interview questions based on at least a portion of the recruitment information and interviewee application information data obtained through step S110.

[0100] In various embodiments, the computing device (100) may generate input data requesting a generative artificial intelligence model to generate interview questions based on at least some of the recruitment information and application information. For example, the computing device (100) may generate input data including a prompt requesting a generative artificial intelligence model to generate interview questions based on recruitment information and the applicant's application information data, and may obtain result data from the generative artificial intelligence model that is output as the generated input data is input into the generative artificial intelligence model. The result data output from the generative artificial intelligence model may include one or more interview questions.

[0101] In step S130, the computing device (100) can provide the interview question generated in step S120 to the interviewer through the user terminal (200).

[0102] More specifically, interview questions may be provided as text, voice, or video through the UI of the user terminal (200). Additionally, the tone of the interview questions may be adjusted in various ways depending on the purpose and nature of the interview, and the language in which the interview is conducted may not be limited to a specific language but may be conducted in multiple languages. For example, input data requesting the generation or modification of interview questions to a generative AI model may include information specifying the tone or language of the interview questions. The generative AI model may generate interview questions composed of various tones, nuances, and languages ​​according to the request.

[0103] In step S140, the computing device (100) can obtain the interviewee's answer obtained through the user terminal (200) as interview answer data. The interviewee's answer can be collected in various forms, such as text, voice, or video.

[0104] In step S150, the computing device (100) can evaluate the interviewee based on the acquired interviewee's answers. The interviewee evaluation process may be performed automatically according to predefined evaluation criteria, or may be performed using interview and recruitment data learned by artificial intelligence.

[0105] In various embodiments, the computing device (100) may perform the step of generating interview questions based on at least some of the recruitment information and the applicant's application information. To do this, the computing device (100) may first generate input data requesting the generation of interview questions.

[0106] For example, input data can be configured based on the job requirements in the job posting and the interviewee's career and skills, and the computing device (100) can input the generated input data into a generative artificial intelligence model to automatically generate interview questions related to the input data.

[0107] The computing device (100) can select the most appropriate question for the applicant's application information and recruitment information by reviewing the difficulty, question type, and suitability with evaluation criteria of the generated interview questions based on the generated interview questions. For example, if there is project experience highlighted in the applicant's resume, specific questions regarding that experience may be added.

[0108] Through this process, the computing device (100) can generate customized interview questions that reflect the applicant's application information and recruitment information.

[0109] Below, various embodiments of a method for a computing device (100) to generate interview questions are described in detail.

[0110] In various embodiments, the computing device (100) may generate a prompt including conditions for generating an interview question and information requesting the generation of an interview question.

[0111] More specifically, the computing device (100) may generate a prompt including conditions for generating interview questions and a command requesting the generation of interview questions. For example, the conditions for generating interview questions may include the format, level, type, difficulty, evaluation criteria, evaluation items, etc. of the interview.

[0112] In various embodiments, the computing device (100) can generate input data including recruitment information, support information, and generated prompts.

[0113] More specifically, at the stage where a generative AI model generates interview questions, interview questions can be constructed based on recruitment information, application information, and generated prompts. For example, input data may include, but is not limited to, the format of interview questions, the interconnectedness between questions, job relevance, whether the interviewee's career and skill level are reflected, evaluation methods, interview methods (remote, in-person), job postings, applicants' resumes, cover letters, technical and career information, major projects, competencies possessed, analysis data of the interviewee's personality, industry-specific competencies, fit with corporate culture, analysis of the applicant's strengths and weaknesses, essential competencies for the relevant role, response prediction patterns, evaluation criteria for the interviewee's answer style and logic, question difficulty adjustment factors, and conditions for generating additional questions.

[0114] In various embodiments, the computing device (100) can construct personality questions based on input data.

[0115] More specifically, personality questions may be constructed based on personality test result data or personality analysis data of the interviewer prepared in advance by the interviewer, supporting information, and generated prompts. For example, a computing device (100) may generate personality interview questions based on the content of a self-introduction letter that matches keywords associated with personality test result data or personality analysis data of the interviewer, and personality interview questions may be generated using various metaphors.

[0116] In various embodiments, the computing device (100) can obtain target questions from an interview question database.

[0117] More specifically, the interview question database may be a database containing pre-generated questions necessary for hiring good employees. The computing device (100) can utilize the database to obtain the necessary target questions and modify the obtained target questions to suit the target to be hired.

[0118] In addition, the computing device (100) can generate not only basic common questions but also customized questions tailored to the individual characteristics of the interviewee by analyzing recruitment information and the applicant's application information, and by searching for and obtaining appropriate target questions from the interview question database.

[0119] In various embodiments, the computing device (100) can generate interview questions by utilizing an interview question database and RAG technology. RAG technology is a technology that searches for relevant information in real time from an external database and generates natural text based on it. RAG technology can increase the accuracy and efficiency of information generation by ensuring that the text generation process and the information search process operate complementarily.

[0120] For example, the computing device (100) can search an interview question database to search for target questions suitable for the interviewee in real time. The computing device (100) can select and analyze the searched target questions according to the characteristics of the interviewee, and can generate interview questions by modifying and improving them according to the characteristics of the interviewee. In addition, the computing device (100) can modify the expression method of the interview questions according to the user's requirements, and according to the embodiment, it can also generate similar interview questions anew with different expressions (e.g., use of synonyms, use of metaphors, or change of sentence structure).

[0121] In various embodiments, the computing device (100) can determine the reliability of an interviewee's answer by comparing the interviewee's answer to an interview question with the interviewee's supporting information, and if the reliability is below a preset standard, it can regenerate the interview question and provide it to the interviewee. Here, the method for determining reliability can be applied in the same way as the method described below with reference to FIG. 5. For example, if there is content where the interviewee's answer and supporting information do not match, the computing device (100) can regenerate the interview question by changing the expression method of the interview question associated with the interviewee's answer and then provide it to the interviewee.

[0122] In various embodiments, the computing device (100) may obtain the interviewer's previous interview answers and generate additional interview questions based on the obtained previous interview answers. This will be explained in more detail below with reference to FIG. 4.

[0123] FIG. 4 is a flowchart illustrating a method for generating additional interview questions through a generative artificial intelligence model in various embodiments.

[0124] Referring to FIG. 4, in step S210, the computing device (100) can obtain the answers provided by the interviewer in a previous interview. The previous interview may refer to the interviewer's responses to interview questions previously conducted within the interview, a preceding interview stage in a multi-stage interview, or a past interview experience.

[0125] In step S220, the computing device (100) can generate additional interview questions based on the previous interview answers obtained in step S210.

[0126] More specifically, the computing device (100) can formulate follow-up questions based on the experience or technical competencies emphasized by the interviewee in the previous interview. Through this, it is possible to evaluate whether the interviewee's answers are limited to a simple conceptual understanding or if there is actual experience in applying them in practice, and to verify whether areas lacking in the previous interview have been improved.

[0127] More specifically, when a previous interview answer is obtained, the computing device (100) can analyze the obtained previous interview answer through natural language processing (NLP) technology and extract key keywords from the interviewer's response.

[0128] For example, if an interviewee answers, "I have experience developing AI-based recommendation systems," "AI recommendation system development" can be extracted as a key keyword, and this information can be used to generate additional interview questions.

[0129] As another example, if an interviewee has previously applied to the same company, interview questions and answers used in the previous interview can be obtained, and keywords such as technical competencies or project experience emphasized in the previous interview can be extracted; this information can then be used to generate additional interview questions.

[0130] In another embodiment, the computing device (100) may input the acquired previous interview answers into a generative artificial intelligence model and enable the generative artificial intelligence model to generate additional interview questions based thereon.

[0131] In various embodiments, the computing device (100) obtains the interviewer's previous interview answer, calculates a similarity with the obtained previous interview answer based on the content of the obtained previous interview answer, and determines the reliability of the interviewer's answer based on the calculated similarity. Hereinafter, with reference to FIG. 5, it will be explained in more detail.

[0132] FIG. 5 is a flowchart illustrating a method for evaluating an interviewer based on the interviewer's answers obtained in various embodiments.

[0133] Referring to FIG. 5, at step S310, the computing device (100) can obtain the answers provided by the interviewer in a previous interview. The previous interview may refer to the interviewer's responses to interview questions previously conducted within the interview, a preceding interview stage in a multi-stage interview, or a past interview experience.

[0134] More specifically, when a previous interview answer is obtained, the computing device (100) can analyze the obtained previous interview answer to determine whether the previous interview answer maintains logical consistency or contains conflicting information.

[0135] In step S320, the computing device (100) can calculate the similarity between the obtained previous interview answer and the current interview answer based on the previous interview answer content obtained through step S310.

[0136] In various embodiments, natural language processing (NLP) technologies and various similarity measurement techniques may be utilized. For example, two answers may be vectorized and cosine similarity may be utilized, edit distance using the difference between two answers may be used, and contextual similarity between two answers may be compared, but is not limited thereto.

[0137] For example, using cosine similarity, similarity can be measured by vectorizing the previous interview answer and the current interview answer and comparing the angles between the vectors. A cosine similarity value closer to 1 indicates that the two answers are very similar, while a value closer to 0 indicates that they are not similar.

[0138] As another example, similarity can be measured using Edit Distance, which is the difference between a previous interview answer and a current interview answer. Edit Distance refers to the minimum number of character or word insertions, deletions, or replacements required to transform one answer into another, and a smaller value indicates that the two answers are more similar.

[0139] In step S330, the computing device can determine the reliability of the interviewer's answer based on the similarity between the previous interview answer obtained through step S320 and the current interview answer.

[0140] More specifically, the computing device (100) can evaluate whether the current interview answer is consistent with the answer provided in the previous interview, or whether it contains conflicting information.

[0141] At this time, a certain threshold (e.g., 0.7) can be set to determine the reliability of the interviewer's answer. For example, if the similarity value between the previous interview answer and the current interview answer is greater than or equal to the set threshold (e.g., 0.7), it can be judged that the interviewer's response maintains consistency and evaluated as having high reliability. On the other hand, if the similarity value is lower than or equal to the threshold (e.g., 0.7), it can be judged that the interviewer provided information different from the previous interview or that a logical contradiction occurred, and evaluated as having low reliability.

[0142] In various embodiments, the computing device (100) provides an existing interview answer and a new interview answer as input to a generative artificial intelligence model to obtain a summary of the meaning and a comparative analysis of the core content of the existing answer and the new answer. The generative artificial intelligence model can comprehensively analyze the main content, logical progression, and contextual information of the two input answers to produce a natural language evaluation index regarding response consistency and similarity.

[0143] The calculated evaluation indicators can be compared with preset criteria to determine whether the reliability of the interview response is high or low. Additionally, the computing device (100) can provide a more intuitive and dynamic method for evaluating response reliability by implementing a feedback function that provides reliability information based on the analysis results or automatically presents additional questions or supplementary explanations during the interview.

[0144] In various embodiments, the computing device (100) may apply different reliability judgment criteria depending on the type of answer.

[0145] In one embodiment, there may be an important answer in which a specific keyword is consistently maintained. For example, there may be a school attended, a previous workplace, tools or programming languages ​​that can be handled, or the name of a project participated in. The computing device (100) analyzes the type of interview question and answer, and if it determines that the question and answer require verification regarding a specific keyword, it can extract a key keyword from the answer and perform a comparison with the keyword of the previous answer.

[0146] In other embodiments, there may be parts where numerical values, such as periods, must match. For example, periods of employment, periods of participation in projects, work experience, etc. The computing device (100) analyzes the types of interview questions and answers, and if it determines that the questions and answers require verification of numerical values, it can extract numerical information from the answers and perform a comparison with the numerical information of the previous answers.

[0147] In one embodiment, the computing device (100) can analyze the type of interview question and answer and determine whether the answer requires a variable numerical element. For example, the tenure at the current company may change over time. On the other hand, the tenure at a project that has already ended or at a company from which one has left does not change over time.

[0148] The computing device (100) can determine whether a value is variable if there is a discrepancy between the previous answer and the current answer. For example, it can be determined that a value is variable if the item corresponding to the value on the resume is currently in an ongoing state rather than a currently terminated state (e.g., regarding work experience), but is not limited thereto. As another example, it can be determined whether the value of the item is variable based on a pre-set judgment criterion or table. As yet another example, a generative artificial intelligence model can determine whether the value is variable.

[0149] When determining the reliability of a variable value, the computing device (100) can analyze the time difference between the time when the previous answer was obtained and the current time, and determine the reliability of the variable value based on this. For example, if the answer from 3 months ago stated that the period of employment was 1 year and 6 months, but the current answer states that the period of employment is 1 year and 9 months, the reliability can be determined to be high. As another example, if the answered value increased more significantly than the time interval between the previous answer and the current answer, the reliability can be determined to be low.

[0150] In another embodiment, the computing device (100) can determine the reliability of the qualitative content of the answers. For example, in the case of a question about the interviewer's disposition, even if specific keywords do not match, the overall context or content may be consistent. In this case, the computing device (100) can determine similarity through an embedding technique instead of whether the keywords included in the answers match. As another example, the computing device (100) may summarize the content of the answers or extract the context and determine the reliability between the answers based thereon.

[0151] In one embodiment, the computing device (100) can determine the reason for the change in content if there is a change between the interviewer's previous interview answer and the current interview answer. The computing device (100) can derive what the changed content is and determine the attributes of the changed content. For example, if the changed content is related to specific keywords such as the school attended or previous workplace, the computing device (100) can determine that the reason for the change is not valid and evaluate the reliability of the answer as low.

[0152] As another example, if the change concerns the presence or absence of specific work experience—that is, if the previous answer stated that there was no specific work experience but the current answer states that there is—an additional process may be performed to determine the reliability of the answer. For example, the computing device (100) checks the elapsed time between the previous answer and the current answer, and evaluates the reliability of the answer relatively highly by assessing the possibility that the interviewer has newly experienced the relevant work as the elapsed time is longer. As another example, the computing device (100) may evaluate the reliability of the answer by determining whether the relevant work experience is listed in the user's application information during the period between the previous answer and the current answer, or by determining whether the company or department where the user worked during that period is a place where the relevant work experience can be acquired.

[0153] As another example, if the changed content is about a period, the computing device (100) can determine the reliability of the answer based on the time elapsed between the previous answer and the current answer as described above.

[0154] In various embodiments, when a change between a previous answer and a current answer is detected, the computing device (100) may generate additional questions regarding the change and provide them to the interviewer. Through this, the computing device (100) may obtain additional answers regarding the change from the interviewer, analyze them to determine the reliability of the answers, or provide additional information to the user.

[0155] In various embodiments, the computing device (100) can preprocess the acquired interview answer data and evaluate the interviewee using the preprocessed interview answer data.

[0156] More specifically, the computing device (100) may perform a step of preprocessing the acquired interview response data. The preprocessing process is a step of refining the interview response data and converting it so that it can be analyzed, and may include a process of processing data in various forms such as text, voice, and video.

[0157] For example, when the computing device (100) acquires interview response data in the form of voice, it may acquire voice data by recording voice from the user terminal (200) from the time the interviewer starts the interview (e.g., start of interview) to the time the interviewer ends the interview (e.g., end of interview). However, it is not limited to this, and the computing device (100) may acquire interview response data in various forms in various ways. The computing device (100) may convert the acquired voice interview response data into text by applying Speech-to-Text (STT) technology, and may perform sentence-level analysis on the converted text data. In addition, the computing device (100) may evaluate the interviewer's tone of voice, emotion, pronunciation accuracy, speed, etc., by performing sentiment analysis on the voice data.

[0158] As another example, when the computing device (100) acquires interview response data in text form, it can refine the data using natural language processing (NLP) technology. It can examine logical consistency by analyzing the structure of the sentences through the removal of unnecessary symbols, deletion of duplicate words, spelling correction, and sentence segmentation.

[0159] As another example, when the computing device (100) acquires video-type interview response data, it can analyze the interviewee's facial expressions, eye movements, gestures, etc. by utilizing video analysis technology. Through such analysis, the computing device (100) can evaluate the interviewee's confidence, trustworthiness, concentration, etc.

[0160] FIG. 6 is a flowchart illustrating a method for determining whether each interviewee passes an interview based on the answers of multiple interviewees obtained in various embodiments.

[0161] Referring to FIG. 6, in step S410, the computing device (100) can analyze the interview response data of multiple interviewees and provide a report on the interview results or interview content.

[0162] The computing device (100) can provide an interview result report by analyzing the acquired answer data. In preparing the interview result report, the computing device (100) can provide a report containing a comprehensive interview evaluation, such as overall response trends, logical development methods, job relevance, and attitude, by analyzing the interviewee's answer data and the evaluation results of the interview answers.

[0163] Additionally, the computing device (100) can provide information regarding feedback to the interviewee by analyzing the answer data provided by the interviewee through an interview content report. The computing device (100) can analyze key keywords, logicality of expression, emotional stability, technical specificity, etc. derived from the answer data, and based on the analysis results, can provide an interview content report containing feedback on points that the interviewee needs to improve in the future or points that need to be additionally emphasized.

[0164] In various embodiments, the computing device (100) may perform the step of evaluating an interviewee using preprocessed interview response data.

[0165] More specifically, the computing device (100) can perform an evaluation by analyzing the interviewee's answers and assigning scores according to evaluation criteria set in advance by the interviewer.

[0166] For example, the interviewer may pre-set evaluation criteria by considering the job to be filled and the competencies required for that job. The evaluation criteria may include items such as job suitability, problem-solving ability, creativity, and communication ability, and each item may be converted into a quantitative or qualitative score. These evaluation criteria and score indicators may be pre-entered into a computing device (100) and used to automatically calculate a score by analyzing how well the interviewee's answers match the evaluation criteria. Once the score for each response item of the interviewee is calculated, the computing device (100) may generate a comprehensive evaluation result and provide it to the interviewer, which may help the interviewer make a final hiring decision by referring to it.

[0167] Additionally, when analyzing the interviewer's answers, the computing device (100) may analyze and provide answers that do not conform to the overall response trend of multiple interviewers or answers containing keywords that have not been used more than a preset frequency in the answer data of multiple interviewers, along with an evaluation of each of the interviewers' answers (e.g., good or bad, positive or negative, suitability, score). The computing device (100) may highlight and display in a different color the main keywords of answers evaluated as good or bad or positive or negative in the comprehensive evaluation results, the main keywords of answers that do not conform to the overall response trend, and keywords that have not been used more than a preset frequency.

[0168] In step S420, the computing device (100) can determine whether each interviewee has passed the interview.

[0169] More specifically, the computing device (100) can extract a comprehensive interview score by analyzing answer data for multiple interviewees. The computing device (100) can preprocess the answer data for multiple interviewees and, using the preprocessed answer data, evaluate logic, job suitability, expressiveness, technical competence, etc. through natural language processing (NLP) and a pre-trained artificial intelligence model, and determine the total score of the interview results for each of the multiple interviewees by assigning weights to each item. In addition, the computing device (100) can assign weights to specific keywords related to the job and reflect the evaluation of answers containing specific keywords (i.e., good or bad, or positive and negative) in the total score of the interview results.

[0170] The computing device (100) can determine whether an interviewee passes based on a preset threshold using the total score of the interview results. The computing device (100) can apply different criteria for each job, and the weight or preset threshold may be adjusted according to the different criteria. Accordingly, the computing device (100) can determine whether each interviewee passes based on different criteria for each job.

[0171] In various embodiments, the computing device (100) may input information for evaluating an interviewee, such as model answers or qualitative judgment criteria, into a generative artificial intelligence model and allow the generative artificial intelligence model to judge the interviewee's interview answers.

[0172] In various embodiments, the computing device (100) may provide information on various interview questions and model answers thereto to a generative artificial intelligence model as a database. The generative artificial intelligence model may obtain the interviewee's interview answers, search for model answers using RAG technology, and then output an evaluation result for the interviewee.

[0173] In various embodiments, the computing device (100) may extract information related to the user's employment information among a plurality of pieces of information included in the acquired interview answer data as necessary information, or filter information not related to the user's employment information as unnecessary information.

[0174] More specifically, the computing device (100) can use natural language processing (NLP) technology to extract information related to recruitment information from the interviewee's response as necessary information and filter out information not related to recruitment information as unnecessary information.

[0175] In various embodiments, the computing device (100) can filter unnecessary information based on the judgment of a generative artificial intelligence model. For example, the computing device (100) may input interview response content into the generative artificial intelligence model and request it to classify content related to the topic of the interview and content unrelated to it.

[0176] For example, the computing device (100) can extract information as necessary information when experience or skills related to the essential competencies listed in the job description are mentioned in the interview answers, and filter out information as unnecessary information when personal beliefs, past experiences unrelated to the job, or answers unrelated to a specific company or industry are mentioned.

[0177] In various embodiments, the computing device (100) can organize necessary information extracted from preprocessed interview answer data and database it so that it can be searched.

[0178] More specifically, the computing device (100) can database the interviewee's interview answers by preprocessing the interview answer data to extract necessary information, converting the necessary information into a structured form, and applying predefined classification criteria such as by job, competency, skill, and experience to the preprocessed interview answer data to classify and store it in a database.

[0179] In various embodiments, the computing device (100) may obtain a search keyword from a user and search for information corresponding to the search keyword. Additionally, the computing device (100) may provide one or more interview answers related to the searched information and information about the interviewer who responded to the one or more interview answers.

[0180] More specifically, the computing device (100) can obtain search keywords entered by the user through the user terminal (200). In one embodiment, the search keywords may include natural language forms or filtering conditions based on predefined categories and tags. Additionally, the computing device (100) can search for information corresponding to the search keywords entered by the user. The search process may be performed on interview answer data and interviewer information stored in a database, and various search algorithms (e.g., matching-based search, vector embedding, TF-IDF, etc.) may be applied to increase search accuracy.

[0181] Additionally, the searched interview answers may be sorted according to similarity scores or relevance rankings with respect to search keywords, and the search results may be provided in a summarized form to make them easier for users to understand. Similarity scores or relevance rankings may be calculated by vectorizing each keyword and utilizing cosine similarity, or by using edit distances based on the differences between each keyword, but are not limited to these.

[0182] Additionally, when a specific interview question is entered as a search keyword through the user terminal (200), the computing device (100) can search for interview answers from multiple interviewees regarding the specific interview question and provide statistical information classifying the interview answers by type, or evaluation results by type of answer, for example, the ratio of good and bad or positive and negative evaluations by answer type, or the pass or fail result or final score of the interviewee who gave the answer.

[0183] In addition, a function to view detailed information regarding interview answers found in the disclosed embodiment may be provided. For example, when a specific interview answer is selected from the search results, the entire interview record related to that answer may be viewed, and a data visualization function may be applied to analyze the interviewee's answer style, answer patterns, and reactions to follow-up questions. Through this, the interviewer can identify not only the found interview answers but also the interviewee's overall response trends, and can precisely review candidates suitable for a specific job.

[0184] FIG. 7 is a flowchart illustrating a method for evaluating an interviewer based on the interviewer's answers obtained in various embodiments.

[0185] Referring to Fig. 7, in step S510, information not included in the interviewer's application information can be obtained from the interviewer's answer content.

[0186] More specifically, the computing device (100) may utilize natural language processing (NLP) technology to identify unlisted information that is not included in the support information among the interviewer's answers, but is not limited thereto.

[0187] For example, the computing device (100) can convert the applicant's application information into text data and also convert the applicant's answer into text data to extract the main keywords and concepts used, and compare this with existing data within the application information to determine whether new information is included. When comparing the application information and the applicant's answer, it may be important to evaluate semantic similarity at the sentence level rather than simply determining whether they match at the word level. To this end, the computing device (100) can vectorize the application information and the applicant's answer using a sentence embedding technique, and then measure the similarity between the two sentences by applying mathematical methods such as cosine similarity. If the similarity is low, it can be determined that the answer is likely to contain new information, and such information can be identified as unlisted information, but the method is not limited to this.

[0188] In step S520, the computing device (100) can correct the interviewer's support information based on the acquired unlisted information.

[0189] More specifically, the computing device (100) can compare the acquired unlisted information with existing support information and determine whether there is a need to correct it. In this process, the decision on whether to correct and the method of correction can be determined by considering the similarity, importance, and job relevance with the existing support information.

[0190] For example, if unlisted information is identified in step S510, the computing device (100) may directly add the unlisted information to the support information, or expand, supplement, and remove existing content by reflecting additional details provided in the interviewer's answer.

[0191] As another example, if there is a conflict between the support information and the interviewer's answer, the computing device (100) can supplement the support information after evaluating the reliability of the interviewer's answer. The supplemented support information may be applied immediately or through the interviewer's final approval. To this end, the computing device (100) may apply an automatic recommendation-based update function or allow the interviewer to review it directly and decide whether to finally reflect it. Additionally, the computing device (100) may provide a function that allows the interviewer to directly check the supplemented information and request a modification.

[0192] Through this process, the disclosed embodiment does not simply modify existing information based solely on the interviewer's response, but evaluates the consistency between the existing information and the interviewer's response, and if necessary, allows the interviewer's application content to be supplemented through the interviewer's review.

[0193] FIGS. 8 to 10 are drawings illustrating a user interface (UI) that provides a recruitment support service through artificial intelligence interviews in various embodiments.

[0194] Referring to FIG. 8, in various embodiments, the computing device (100) can manage multiple interviewees on a project basis. The computing device (100) can classify basic personal information and information on interview results for each interviewee and provide them in a list format, thereby enabling an administrator to compare the results of multiple interviewees at a glance.

[0195] Referring to FIG. 9, in various embodiments, the computing device (100) may provide a generated interview question to the interviewee. The computing device (100) may record a series of processes in which the interview is conducted, and may provide the next question or provide a step in the interview process when the interview answer is completed. Additionally, the computing device (100) may output the video currently being recorded along with a shooting guide for filming the interviewee's interview video, as shown in the lower left corner of FIG. 9. The interviewee may adjust the camera angle or face position according to the shooting guide so that the interview video is filmed consistently, and the computing device (100) may track the interviewee's face position movement and gaze movement on the screen during the interview video filming and provide an alarm to the interviewee if movement outside a preset range occurs.

[0196] Referring to FIG. 10, in various embodiments, the computing device (100) may provide results of evaluating the interviewee based on the acquired interviewee's answers. The computing device (100) may classify the content regarding the interviewee and the job applied for, and analyze the answer text and interview video regarding the interview content to provide summary information. In addition, the computing device (100) may provide an overall score for the interview results by synthesizing the results for each evaluation item, such as the logicality of the analyzed answers, job suitability, and communication ability.

[0198] The method for providing a recruitment support service through an AI interview described above has been explained with reference to the flowchart illustrated in the drawings. For the sake of simplicity, the method for providing a recruitment support service through an AI interview has been illustrated and described using a series of blocks; however, the present disclosure is not limited to the order of the blocks, and some blocks may be performed in a different order or simultaneously than those illustrated and described in this specification. Furthermore, new blocks not described in this specification and drawings may be added, or some blocks may be deleted or modified.

[0199] Although embodiments of the present disclosure have been described above with reference to the attached drawings, those skilled in the art will understand that the present disclosure may be implemented in other specific forms without altering its technical concept or essential features. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. Explanation of the symbols

[0200] 100: Computing device 200 : Terminal 300 : External server 400 : Network

Claims

Claim 1 A method performed by a computing device comprises: a step of obtaining user recruitment information and interviewee application information; a step of generating interview questions based on at least some of the recruitment information and application information using a generative artificial intelligence model (Generative AI); a step of providing the generated interview questions to the interviewee; a step of obtaining the interviewee's answer; and a step of evaluating the interviewee based on the obtained interviewee's answer, wherein the step of evaluating the interviewee includes: a step of obtaining the interviewee's previous interview answer; a step of calculating a similarity with the obtained previous interview answer based on the obtained interviewee's answer; a step of determining the reliability of the interviewee's answer based on the calculated similarity and the type of the obtained interviewee's answer; and a step of obtaining unlisted information among the content of the interviewee's answer that is not included in the interviewee's application information. The method includes a step of correcting the applicant's application information based on the aforementioned unlisted information obtained above, and the step of determining the reliability of the applicant's answer comprises: a step of analyzing the type of the interview question and answer; a step of extracting key keywords from the applicant's answer and comparing them with keywords from the previous interview answer if verification of specific keywords is required according to the type of the interview question and answer, if such verification is required; a step of determining whether the keywords are changeable information if there is information that has changed between the applicant's answer and the previous interview answer based on the comparison of keywords, and if they are changeable information, a step of analyzing the difference between the time when the previous answer was obtained and the current time to determine the reliability of the changed information; and a step of extracting numerical information from the applicant's answer and comparing it with numerical information from the previous interview answer if verification of numerical values ​​is required according to the type of the interview question and answer.A method for providing a recruitment support service through an artificial intelligence interview, comprising: a step of determining whether the numerical information is variable when there is a discrepancy between the numerical information of the interviewer's answer and the numerical information of the previous interview answer, and if it is variable, analyzing the difference between the time when the previous answer was obtained and the current time to determine the reliability of the variable numerical information. Claim 2 A method for providing a recruitment support service through an artificial intelligence interview, wherein, in claim 1, the step of generating the interview question comprises: generating input data requesting the generation of the interview question based on at least some of the recruitment information and the support information; inputting the generated input data into the generative artificial intelligence model; and obtaining the output of the generative artificial intelligence model. Claim 3 A method for providing a recruitment support service through an artificial intelligence interview, wherein, in paragraph 2, the step of generating the input data comprises: a step of generating a prompt including conditions for generating the interview question and a command requesting the generation of the interview question; and a step of generating input data including the recruitment information, the support information, and the generated prompt. Claim 4 A method for providing a recruitment support service through an artificial intelligence interview, wherein, in claim 1, the step of generating the interview question comprises: a step of obtaining the interviewee's previous interview answer; and a step of generating additional interview questions based on the content of the obtained previous interview answer. Claim 5 A method for providing a recruitment support service through an artificial intelligence interview, wherein, in claim 1, the step of generating the interview question comprises: a step of obtaining a target question from an interview question database; and a step of generating the interview question based on the target question, the recruitment information, and the support information using the generative artificial intelligence model. Claim 6 delete Claim 7 A method for providing a recruitment support service through an artificial intelligence interview, wherein the step of evaluating the interviewee in claim 1 comprises: a step of preprocessing interview answer data obtained based on the interviewee's answers; and a step of evaluating the interviewee using the preprocessed interview answer data. Claim 8 A method for providing a recruitment support service through an artificial intelligence interview, wherein the preprocessing step comprises, among a plurality of pieces of information included in the acquired interview answer data, extracting information related to the user's recruitment information as necessary information, or filtering information not related to the user's recruitment information as unnecessary information. Claim 9 A method for providing a recruitment support service through an artificial intelligence interview, further comprising: a step of organizing necessary information extracted from the preprocessed interview answer data to make it searchable and creating a database; a step of obtaining a search keyword from the user; and a step of searching for information corresponding to the search keyword and providing information about one or more interview answers related to the searched information and the interviewer who responded with the one or more interview answers. Claim 10 delete Claim 11 A device comprising: a memory for storing one or more instructions; and a processor for executing the one or more instructions stored in the memory, wherein the processor performs the method of claim 1 by executing the one or more instructions. Claim 12 A computer program stored on a computer-readable recording medium that is combined with a computer, which is hardware, to perform the method of claim 1.

Images