Automatic interactive question system and question method using the same

Abstract

An automatic interactive survey system according to an embodiment of the present invention comprises: a survey input unit that receives content according to the format of a survey; a preprocessing unit connected to the survey input unit that analyzes the input content by item and generates questions sequentially; an output unit connected to the preprocessing unit that outputs voice or text for each item; a response input unit that receives the voice or text of a survey subject; a judgment control unit connected to the preprocessing unit and the response input unit that determines the situation regarding the received response content, classifies it according to a scenario, and responds accordingly; and a storage unit connected to the judgment control unit that stores survey content information.

Description

Technology Field

[0001] The present invention relates to an automated interactive survey system and a survey method using the same, and more particularly to an automated interactive survey system and a survey method using the same that can efficiently conduct surveys targeting people who have difficulty responding to surveys by utilizing artificial intelligence technology. Background Technology

[0002] A survey is a research method that collects and analyzes data on social phenomena through pre-structured questionnaires or interviews. The purpose of such surveys is to obtain information from a large number of respondents presumed to represent a specific population. While printed questionnaires were used in the past, online survey methods have recently gained popularity due to advancements in information and communication technologies, such as the Internet and smartphones.

[0003] The survey results derived from such surveys must be guaranteed to be reliable, but this reliability is compromised by various factors and is particularly heavily dependent on the quality of the data required to derive the results.

[0004] Conventional surveys are conducted by having interviewers read questions directly to subjects and record their responses, which requires a significant amount of time and manpower. In particular, this conventional method can pose even greater difficulties for the elderly, children, and people with limited mobility.

[0005] For example, if a subject responds to a survey insincerely, giving answers unrelated to the survey content, or if response data provided with malicious intent is reflected in the survey results, there is a problem that the results derived from normal response data alone can be distorted. Prior art literature

[0006] Patent Document: Published Patent Application No. 10-2023-0069366 The problem to be solved

[0007] The present invention has been devised to resolve the above-mentioned problems, and the objective of the present invention is to provide an automated conversational survey system that can efficiently conduct surveys targeting people who find it difficult to respond to surveys by utilizing artificial intelligence technology.

[0008] Another objective of the present invention is to provide a survey method using an automated conversational survey system that can efficiently conduct surveys targeting people who have difficulty responding to surveys by utilizing artificial intelligence technology. means of solving the problem

[0009] An automatic interactive survey system according to an embodiment of the present invention comprises: a survey input unit that receives content according to the format of a survey; a preprocessing unit connected to the survey input unit that analyzes the input content by item and generates questions sequentially; an output unit connected to the preprocessing unit that outputs voice or text for each item; a response input unit that receives the voice or text of a survey subject; a judgment control unit connected to the preprocessing unit and the response input unit that determines the situation regarding the received response content, classifies it according to a scenario, and responds accordingly; and a storage unit connected to the judgment control unit that stores survey content information.

[0010] In an automatic interactive survey system according to an embodiment of the present invention, the preprocessing unit is characterized by linking to an external large language model (LLM) module via an application programming interface (API).

[0011] In an automatic interactive survey system according to an embodiment of the present invention, the judgment control unit is characterized by determining the situation regarding the received response content by linking with an external LLM module via an API.

[0012] Additionally, a survey method according to another embodiment of the present invention comprises: (A) a step in which, when survey content is entered into a survey form through a survey input unit, a judgment control unit uploads the survey content along with the survey form; (B) a step in which a preprocessing unit analyzes the survey content by linking with an external LLM (large language model) module via an API (application programming interface) and generates questions in sequence; (C) a step in which the received questions are output to a respondent in the form of voice or text through an output unit according to the control of the judgment control unit; (D) a step in which a response input unit collects the response voice and the text of the response message of the survey subject regarding the outputted questions; and (E) a step in which the judgment control unit analyzes the collected responses by linking with an external LLM module via an API and determines their validity.

[0013] A survey method according to another embodiment of the present invention is characterized by further including the step of (F) outputting an additional question to the respondent through the preprocessing unit and the output unit when the judgment control unit determines that the collected response is an inappropriate response.

[0014] A survey method according to another embodiment of the present invention is characterized by further including the step of (G) storing the collected response in a storage unit as the judgment control unit determines that the collected response is a valid response for each scenario.

[0015] In a survey method according to another embodiment of the present invention, step (B) is characterized in that the preprocessing unit combines similar items among the items of the input survey content in conjunction with the LLM module to generate an implied question.

[0016] The features and advantages of the present invention will become more apparent from the following detailed description based on the accompanying drawings.

[0017] Prior to this, terms and words used in this specification and claims should not be interpreted in their ordinary and dictionary meanings, but should be interpreted in a meaning and concept consistent with the technical spirit of the invention, based on the principle that the inventor can appropriately define the concept of the terms to best describe his invention. Effects of the invention

[0018] A survey method using an automatic interactive survey system according to an embodiment of the present invention utilizes artificial intelligence to conduct a convenient and efficient survey targeting people who have difficulty responding to surveys, and has the effect of saving time and costs.

[0019] The automatic interactive survey method according to an embodiment of the present invention has the effect of improving the reliability of data by increasing the accuracy and consistency of responses. Brief explanation of the drawing

[0020] FIG. 1 is a configuration diagram of an automatic interactive survey system according to an embodiment of the present invention. FIG. 2 is a flowchart for explaining a survey method using an automatic interactive survey system according to an embodiment of the present invention. Specific details for implementing the invention

[0021] The objects, specific advantages, and novel features of the present invention will become more apparent from the following detailed description and preferred embodiments in conjunction with the accompanying drawings. It should be noted that in assigning reference numbers to the components of each drawing in this specification, the same components are assigned the same number whenever possible, even if they are shown in different drawings. Furthermore, terms such as "first," "second," etc., may be used to describe various components, but said components are not to be limited by said terms. Such terms are used solely for the purpose of distinguishing one component from another. Additionally, in describing the present invention, detailed descriptions of related prior art are omitted if it is determined that such detailed descriptions would unnecessarily obscure the essence of the invention.

[0022] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. FIG. 1 is a configuration diagram of an automatic interactive survey system according to an embodiment of the present invention.

[0023] The automatic conversational survey system according to an embodiment of the present invention is an automatic conversational survey system using artificial intelligence to efficiently conduct surveys targeting people who have difficulty responding to surveys, namely the elderly, children, and people with disabilities whose movement is limited.

[0024] Specifically, an automatic interactive survey system according to an embodiment of the present invention includes a survey input unit (110) that receives content according to the format of the survey, a preprocessing unit (120) connected to the survey input unit (110) that analyzes the input content by item and creates questions in sequence, an output unit (130) connected to the preprocessing unit (120) that outputs voice or text by item, a response input unit (140) that receives voice or text of a survey subject, a judgment control unit (150) that judges the situation regarding the received response content and classifies and responds according to a scenario, and a storage unit (160) connected to the judgment control unit (150) that stores survey content information.

[0025] The survey input unit (110) receives and processes the structure of the survey and the list of questions written by the user in a survey form in document format such as Hangul, MS Word, or Excel. The user can input the content of the survey as text or upload an existing survey form, and the content of the survey entered in this way is processed in the survey input unit (110) and transmitted to the preprocessing unit (120).

[0026] The preprocessing unit (120) is a part that connects to an external LLM (large language model) module via an API (application programming interface), analyzes the input survey content item by item to generate questions sequentially, and in this process analyzes the input text by connecting with a fine-tuned LLM module, and the LLM module is appropriately trained for each conversation channel to separate each question independently. The questions thus preprocessed are transmitted to the output unit (130).

[0027] Here, the LLM module is a module containing an LLM composed of an artificial neural network having numerous parameters (usually billions of weights or more), which trains a significant amount of unlabeled text using self-supervised or semi-self-supervised learning.

[0028] These LLM modules are built using deep learning technology to process vast amounts of natural language data and generate surveys indistinguishable from human-generated text; they learn from massive amounts of text data and can derive responses to user requests. Examples of such LLMs include OpenAI's GPT (Generative Pre-trained Transformer) series and Google's BERT (Bidirectional Encoder Representations from Transformers) models. These models are used in various applications, including language translation, content generation, and chatbots.

[0029] In particular, the preprocessing unit (120) can analyze the input survey content by question in conjunction with the LLM module and combine similar questions to generate a condensed question. For example, among the survey questions, questions such as 'region of residence', 'neighborhood', and 'town / village / district where you currently live' can be condensed to generate a summarized question such as 'Where is your current address?'.

[0030] The output unit (130) includes a display device or speaker connected to the preprocessing unit (120), and can output the content of the question transmitted from the preprocessing unit (120) as voice or text according to the control of the judgment control unit (150).

[0031] The response input unit (140) includes a microphone for receiving voice or a display device for receiving text in response to a question output through the output unit (130), and transmits the received voice or text information to the judgment control unit (150).

[0032] The judgment control unit (150) controls the automatic conversational survey system overall, judges the situation regarding the response content received from the response input unit (140), and classifies and responds according to the scenario. This judgment control unit (150) is a part that links with an external LLM module via an API, and by linking with this LLM module, it can generate an artificial intelligence chatbot capable of voice recognition based on a database. The artificial intelligence chatbot can utilize NLP (Natural Language Processing), NLU (Natural Language Understanding), and NLG (Natural Language Generation). At this time, NLU extracts keywords and related words and extracts patterns through linkage with a database, and NLG analyzes sentence structures through linkage with a database to generate and respond to rule-based conversations.

[0033] Specifically, the AI ​​chatbot generated in this way can preprocess, interpret, and output results corresponding to the intent of natural language included in at least one scenario. The main task performed in the Natural Language Process stage of the AI ​​chatbot is morphological analysis of natural language. Morphological analysis (POS-Tagging) refers to the process of splitting a raw corpus into morpheme units and attaching part-of-speech information to each form. This allows for the identification of the structure of various linguistic attributes, such as morphemes, roots, prefixes / suffixes, and parts of speech (POS), thereby enabling the chatbot to respond to the content of the response according to the scenario.

[0034] In addition, during the Natural Language Understanding (NLUA) stage of the AI ​​chatbot, the process of decomposing the respondent's response and determining appropriate syntax and semantic systems for the decomposed entities can be performed. At this time, the NLUA stage of the AI ​​chatbot can perform processing such as general named entity recognition, specific named entity recognition, lexical semantic analysis, syntactic analysis, and intent analysis. When extracting named entities, a Specific-NER can be further utilized for the chatbot to operate in a specific domain or task. Furthermore, for named entity recognition or semantic role determination, not only statistical modeling such as Hidden Markov Models, Support Vector Machines, and Conditional Random Fields, but also deep learning, such as Feed-Forward Neural Networks (FFNN), Convolutional Neural Networks (CNN), and Long Short Term Memory (LSTM), can be utilized.

[0035] The storage unit (160) organizes and stores the received responses according to the survey content under the control of the judgment control unit (150). All response data is safely stored in the storage unit (160) and can be used for analysis and report generation thereafter. The storage unit (160) can also perform regular backups to maintain data integrity.

[0036] The automatic interactive survey system according to the embodiment of the present invention configured as described above can efficiently conduct surveys targeting people who have difficulty responding to surveys, namely the elderly, children, and people with disabilities whose movement is limited, by utilizing artificial intelligence.

[0037] Hereinafter, a survey method using an automatic interactive survey system according to an embodiment of the present invention will be described with reference to FIG. 2. FIG. 2 is a flowchart for explaining a survey method using an automatic interactive survey system according to an embodiment of the present invention.

[0038] In a survey method using an automatic interactive survey system according to an embodiment of the present invention, first, when a user inputs survey content in accordance with the survey form through a survey input unit (110), a judgment control unit (150) uploads the survey content along with the survey form (S210).

[0039] The survey input unit (110) receives and processes the structure of the survey and the list of questions written by the user in a survey form in document format such as Hangul, MS Word, or Excel. The user can input the content of the survey as text or upload an existing survey form, and the content of the survey thus entered is processed in the survey input unit (110) under the control of the judgment control unit (150) and transmitted to the preprocessing unit (120).

[0040] After uploading the survey content, the preprocessing unit (120) analyzes the survey content by connecting to an external LLM (large language model) module via an API (application programming interface) and generates questions in sequence (S220).

[0041] That is, the preprocessing unit (120) analyzes the input text in conjunction with a fine-tuned LLM module, and the LLM module is appropriately trained for each conversation channel to independently separate and analyze each question, and thus the preprocessed questions can be transmitted to the output unit (130).

[0042] In particular, the preprocessing unit (120) can analyze the input survey content by question in conjunction with the LLM module and combine similar questions to generate a condensed question. For example, among the survey questions, questions such as 'region of residence', 'neighborhood', and 'town / village / district where you currently live' can be condensed to generate a summarized question such as 'Where is your current address?'.

[0043] Afterwards, the output unit (130) can output the received preprocessed questions to the respondent in voice or text form according to the control of the judgment control unit (150) (S230).

[0044] The output unit (130) includes a display device or a speaker and can output the question content transmitted from the preprocessing unit (120) to the respondent in the form of voice or text according to the control of the judgment control unit (150).

[0045] For the questions output in this way, the response input unit (140) collects the response voice and text of the response message of the survey subject (S240).

[0046] To this end, the response input unit (140) includes a microphone that receives the voice of a survey subject responding or a display device that receives the text of a response message, and transmits the received voice or text information to the judgment control unit (150).

[0047] The judgment control unit (150) analyzes the collected response by linking with an external LLM module via API and determines its validity (S250).

[0048] Specifically, the judgment control unit (150) can generate an AI chatbot capable of voice recognition based on a database in conjunction with an LLM module, and can preprocess, interpret, and output results corresponding to the intent of the natural language included in the collected response. At this time, the work mainly performed in the Natural Language Process stage involves analyzing morphemes of the natural language, and in the Natural Language Understanding stage, after decomposing the respondent's response content, appropriate syntax and semantic systems are determined for the decomposed entities to perform processing such as general named entity recognition, detailed named entity recognition, lexical semantic analysis, syntactic analysis, and intent analysis, thereby determining validity for each scenario.

[0049] Accordingly, the judgment control unit (150) performs a process to output additional questions to the respondent through the preprocessing unit (120) and the output unit (130) if the collected response is inappropriate (S260).

[0050] On the other hand, the judgment control unit (150) stores the response information in the storage unit (160) when the collected response is valid for each scenario (S270).

[0051] For example, in the first scenario, when a respondent provides a response that fits the survey, the judgment control unit (150) determines it as a valid response and stores the response information in the storage unit (160).

[0052] In the second scenario, if the respondent does not respond to the survey within a set time, the judgment control unit (150) can output a prompting message or voice, such as "Could you please say it again?" to the respondent through the output unit (130).

[0053] If there is no response thereafter, a prompting message or voice output is repeated according to the setting value of the number of waiting times or time, and then the process moves on to the next question. At this time, the response information is recorded and stored as “no response” in the storage unit (160).

[0054] In the third scenario, if the respondent gives an incomplete or inappropriate response, the judgment control unit (150) may output additional questions to the respondent through the output unit (130) in the form of prompting messages or voice, such as "Could you please answer clearly again?" or "The answer is incomplete. Please tell me more about ~." At this time, the judgment control unit (150) specifically explains which part is incomplete and requests it, and if the incomplete response is repeated despite this, the response information is recorded and stored in the storage unit (160) as is.

[0055] In the fourth scenario, if the respondent's response is determined to be in an incomprehensible form, the judgment control unit (150) classifies it as an incomprehensible response and outputs a message or voice such as "I'm sorry. This is an incomprehensible answer. Could you please repeat it?" to the respondent through the output unit (130) to repeat the re-questioning. At this time, the judgment control unit (150) repeats the re-questioning according to the set number of times, and if it is determined to be a repeated incomprehensible response, the response information is recorded and stored in the storage unit (160) as is.

[0056] In the fifth scenario, if the respondent misunderstands the question and the response is judged to be a misunderstanding, the judgment control unit (150) may explain the question in a different way or provide an example, and output to the respondent through the output unit (130) in the form of a message or voice saying, "I will ask the question again in a different way." At this time, the judgment control unit (150) may appropriately reconstruct the explanation of the question according to the respondent's age, gender, and other given information and situation entered in advance, and output it to the respondent through the output unit (130). Nevertheless, if it is judged to be a repeated misunderstanding response, the judgment control unit (150) records and stores the response information as is in the storage unit (160).

[0057] In the sixth scenario, as the respondent expresses their refusal to answer the question, the judgment control unit (150) classifies it as a refusal to answer and outputs to the respondent via the output unit (130) in the form of a message or voice, "I will skip this question and move on to the next question." Subsequently, the judgment control unit (150) classifies the response information as "refusal to answer" and stores it in the storage unit (160), and moves on to the next question.

[0058] A survey method using an automated interactive survey system according to an embodiment of the present invention, which includes such a process, utilizes artificial intelligence to conduct a convenient and efficient survey targeting people who have difficulty responding to surveys, thereby saving time and costs. Furthermore, the automated interactive survey method according to an embodiment of the present invention can improve data reliability by increasing the accuracy and consistency of responses.

[0059] Although the technical concept of the present invention has been specifically described according to the preferred embodiments above, it should be noted that the aforementioned embodiments are for illustrative purposes only and are not intended to be limiting.

[0060] Furthermore, a person skilled in the art will understand that various implementations are possible within the scope of the technical concept of the present invention. Explanation of the symbols

[0061] 110: Survey Input Section 120: Preprocessing Section 130: Output section 140: Response input section 150: Judgment control unit 160: Storage unit

Claims

Claim 1 An automatic conversational survey system comprising: a survey input unit that receives content according to the format of a survey; a preprocessing unit connected to the survey input unit that analyzes the input content by item and generates questions sequentially; an output unit connected to the preprocessing unit that outputs voice or text for each item; a response input unit that receives the voice or text of a survey subject; a judgment control unit connected to the preprocessing unit and the response input unit that determines the situation regarding the received response content, classifies it according to a scenario, and responds accordingly; and a storage unit connected to the judgment control unit that stores survey content information; wherein the judgment control unit determines the situation regarding the received response content by linking with an external LLM module via an API, and the judgment control unit, if it determines that the received response content is incomplete or inappropriate as a result of judgment regarding the received response content, outputs an additional question to the survey subject through the output unit that explains the incomplete part and requests a response regarding it, and if it determines that the response is incorrectly understood or misunderstood by the survey subject, reconstructs the explanation of the question according to the age, gender, and situation previously entered by the survey subject and outputs it to the survey subject through the output unit. Claim 2 An automatic interactive survey system according to claim 1, characterized in that the preprocessing unit links to an external LLM (large language model) module via an API (application programming interface). Claim 3 An automatic interactive survey system according to claim 1, characterized in that the judgment control unit determines the situation regarding the received response content by linking with an external LLM module via an API. Claim 4 (A) a step in which, when survey content is entered into a survey form through a survey input unit, a judgment control unit uploads the survey content along with the survey form; (B) a step in which a preprocessing unit analyzes the survey content by linking with an external LLM (large language model) module via an API (application programming interface) and generates questions sequentially; (C) a step in which, under the control of the judgment control unit, the received questions are output to a survey subject in the form of voice or text via an output unit; (D) a step in which a response input unit collects the voice response and text response message of the survey subject regarding the output questions; and (E) a step in which the judgment control unit analyzes the collected responses by linking with an external LLM module via an API and determines their validity; wherein, if the judgment control unit determines that the collected responses are incomplete or inappropriate, an additional question is output to the survey subject via the output unit explaining the incomplete parts and requesting a response thereto. A survey method further comprising the step of, if the collected response is determined to be a response that the survey subject has misunderstood or misinterpreted, reconstructing the explanation of the question to match the age, gender, and situation previously entered by the survey subject and outputting it to the survey subject through the output unit. Claim 5 delete Claim 6 A survey method according to claim 4, further comprising the step of (G) storing the collected response in a storage unit as the judgment control unit determines that the collected response is a valid response content for each scenario. Claim 7 A survey method according to claim 4, wherein step (B) is characterized in that the preprocessing unit combines similar items among the items of the input survey content in conjunction with the LLM module to generate an implied question.

Images