KR20230062711A - Chemical substance estimation method using symptom and chemical structure, terminal and server performing the same - Google Patents

Abstract

In accordance with one embodiment of the present invention, provided are a chemical substance estimation method using a symptom and a chemical structure, and a terminal and a server performing the same. The chemical substance estimation method includes: an information collection step in which a terminal senses human body information generated by sensing the body of a user, and environment information generated by sensing an environment in which the user is located, to a server; a symptom recognition step in which a server recognizes a symptom occurring in the body of the user based on the human body information and the environment information; a chemical substance estimation step in which the server estimates a chemical substance exposed to the body of the user based on the recognized symptom; and a structure checking step in which the server checks whether the estimated chemical substance is a chemical substance exposed to the body of the user by comparing the symptom occurring in the body of the user to a symptom expected based on the chemical structure of the chemical substance. Therefore, the present invention is capable of more quickly and more accurately estimating the chemical substance exposed to the user.

Description

Chemical substance estimation method using symptom and chemical structure, terminal and server performing the same}

The present invention relates to a method for estimating chemical substances using symptoms and chemical structures, and a terminal and server for performing the same.

When random chemical substances are leaked from various places such as industrial sites, homes, laboratories, and multi-use facilities and come into contact with the human body or when contact is suspected, it is necessary to quickly find and deal with the contacted chemicals. The human body exhibits various types of symptoms when exposed to chemicals. Chemical substances have different levels of risk to the human body depending on their types, and specific methods for responding to chemical substances are also different. Therefore, it is necessary to estimate what chemicals the human body has come into contact with and respond differently depending on the chemicals.

KR 10-1957382 B1

An object according to an embodiment of the present invention is to receive symptoms appearing in the human body, and to estimate what chemical substances are exposed to the human body, using the symptoms appearing in the human body and the chemical structure of the chemical substance related to the symptom. It is to provide an estimation method, a terminal and a server that performs the estimation.

A method for estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention provides body information generated by a terminal sensing a user's body and environmental information generated by sensing an environment in which a user is located to a server. information collection step of doing, a symptom recognition step in which the server recognizes symptoms appearing on the user's body based on the body information and environment information, and the server estimates chemical substances exposed to the user's body based on the recognized symptoms A chemical estimation step may be included.

In addition, the information collection step is a body sensing step in which the terminal worn on the user's body senses the user's body to generate body information and provide it to the server, the terminal worn on the user's body or the space where the user is located. An environment sensing step in which the deployed terminal senses the environment where the user is located to generate environment information and provide it to the server, and a voice conversation step in which symptoms appearing on the user's body are received through voice conversation between the terminal and the user. can

In addition, the voice conversation step can be performed through questions and answers between the terminal and the user, and the voice conversation starts while speaking the symptoms of the body recognized by the user to the terminal, or the body sensed by the terminal. The server may analyze the information or environment information and start by asking a user a question through the terminal by voice.

In addition, in the symptom recognizing step, the server derives a list of symptoms likely to appear to the user based on the body information and environment information, analyzes voice conversations, recognizes words expressing symptoms, and recognizes symptoms included in the symptom list. It is possible to recognize the symptoms appearing on the user's body by matching with.

In addition, the chemical estimating step is to determine the chemical substance causing the symptoms appearing in the user's body by using a symptom database including a knowledge graph generated by learning the chemical substance and symptoms appearing when the chemical substance is exposed to the body. It may include a symptom-based estimation step of searching, a determination step of determining one or more chemical substances with a high probability as the final exposed chemical substance when the number of substances estimated to be chemical substances causing symptoms in the user's body is smaller than the reference number. .

In addition, the chemical substance estimating step compares a symptom expected based on the chemical structure of the chemical substance estimated in the symptom-based estimating step with a symptom appearing on the user's body, and the estimated chemical substance is exposed to the user's body. It may further include a structure confirmation step to confirm that the chemical substance is correct.

In addition, the chemical substance estimating step may further include a question generation step of generating a question for confirming whether a symptom that is not expected to appear or a symptom that is expected to appear exists using the symptom database and the structure database. there is.

In addition, the method for estimating chemicals using symptoms and chemical structures according to an embodiment of the present invention further includes a notification step of providing information on the chemicals estimated in the chemical substance estimation step to the user through the terminal. can do.

In addition, a terminal performing the method of estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention includes a voice input unit for receiving information spoken by a user, and a voice output unit for outputting questions to the user in voice. , and a control unit for controlling the voice input unit and the audio output unit to perform questions and answers with the user by voice and provide the information to the server.

In addition, the question may include content in which the server analyzes the user's body information or environment information and inquires whether or not there are symptoms that may occur in the user's body.

In addition, the answer may include an answer to the question or content for confirming whether or not a symptom exists in the user's body as the user's judgment.

In addition, a terminal performing the method of estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention includes a body sensor unit that senses the user's body to generate body information and senses the environment in which the user is located. It may further include an environment sensor unit that generates environment information by doing so.

Also, the control unit may control the body sensor unit and environment sensor unit to measure the body information and environment information at predetermined intervals, and to measure the body information and environment information at intervals shorter than the predetermined period while the question and answer are in progress.

In addition, the server performing the chemical estimation method using symptoms and chemical structures according to an embodiment of the present invention includes the user's body information received from the terminal, the environment information of the environment in which the user is located, and the relationship between the user and the terminal. A symptom recognition unit that recognizes symptoms by analyzing voice conversations, a symptom database including a knowledge graph generated by learning chemicals and symptoms that appear when the chemical is exposed to the body, and symptoms recognized by the symptom recognition unit A substance estimating unit may be included to search and estimate the chemical substance to be generated by searching the symptom database and, when the number of the estimated chemical substance is smaller than the reference number, determine the finally exposed chemical substance in order of high probability.

In addition, the server performing the method of estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention includes a structure database including a knowledge graph generated by learning the chemical structure of a chemical substance and symptoms indicated by the chemical structure. , And a structural analysis unit for extracting the chemical structure constituting the chemical substance, extracting the connection relationship of a plurality of chemical structures, and updating the symptom database by relating symptoms caused by the toxicity of the chemical structure acting on the body. can

In addition, the substance estimating unit searches the symptom database and compares a symptom expected based on the chemical structure of the estimated chemical substance with a symptom appearing on the user's body so that the estimated chemical substance is exposed to the chemical substance on the user's body. You can check if the material is right.

In addition, the server performing the chemical estimation method using symptoms and chemical structures according to an embodiment of the present invention merges the knowledge graph of the symptom database and the knowledge graph of the structure database, so that the relationship between symptoms and chemicals , a complex database including a knowledge graph formed by the relationship between chemical substances and structures, and between structures and symptoms may be further included.

According to one embodiment of the present invention, in estimating the chemical substance exposed to the human body, it is estimated what chemical substance causes the input symptom, and the symptoms caused by the specific chemical structure of the estimated chemical substance and the human body By comparing the symptoms you see, you can more accurately estimate the chemical.

1 is a diagram showing an environment in which a chemical substance estimation method using symptoms and chemical structures according to an embodiment of the present invention is performed.
2 is a diagram showing a terminal and a server performing a method of estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention.
3 is a flowchart showing each step of a method for estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention.
4 is a diagram illustrating functions performed by a terminal and a server that perform a chemical estimation method using symptoms and chemical structures according to an embodiment of the present invention.
5 is a flowchart showing in detail a method for estimating chemicals using symptoms and chemical structures according to an embodiment of the present invention.
6 is a diagram showing a knowledge graph of a symptom database used in a method for estimating chemicals using symptoms and chemical structures according to an embodiment of the present invention.
7 is a diagram showing the chemical structure of chemicals used in the method for estimating chemicals using symptoms and chemical structures according to an embodiment of the present invention.
8 is a diagram showing a knowledge graph of a structure database used in a method for estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention.
9 is a diagram showing a knowledge graph in which a symptom knowledge graph and a structure knowledge graph used in a method for estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention are integrated.

Objects, specific advantages and novel features of an embodiment of the present invention will become more apparent from the following detailed description and preferred embodiments taken in conjunction with the accompanying drawings. In adding reference numerals to components of each drawing in this specification, it should be noted that the same components have the same numbers as much as possible, even if they are displayed on different drawings. In addition, terms such as "one side", "the other side", "first", and "second" are used to distinguish one component from another, and the components are not limited by the above terms. no. Hereinafter, in describing an embodiment of the present invention, a detailed description of related known technologies that may unnecessarily obscure the subject matter of an embodiment of the present invention will be omitted.

Hereinafter, with reference to the accompanying drawings, an embodiment of the present invention will be described in detail.

1 is a diagram showing an environment in which a chemical substance estimation method using symptoms and chemical structures according to an embodiment of the present invention is performed. 2 is a diagram showing a terminal 100 and a server 200 performing a method of estimating a chemical substance using symptoms and chemical structures according to an embodiment of the present invention. 3 is a flowchart showing each step of a method for estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention.

In this specification, the terminal 100 performing the chemical estimation method using symptoms and chemical structures can be simply described as 'terminal 100', and the server 200 performing the chemical estimation method using symptoms and chemical structures ) can be simply described as 'server 200'. In this specification, a chemical substance may include substances that cause symptoms when exposed to the body, substances that may harm the body, and various other chemical substances.

See Figures 1, 2, and 3. When certain chemicals and the user 10 are present in one place, there may be a risk that the user 10 is exposed to the chemicals. For example, chemical substance A, chemical substance B, chemical substance C and the user 10 may exist in place A. When the user 10 is exposed to chemical substance A, symptoms caused by chemical substance A may occur in the body of the user 10 . Symptoms include reactions in the body of the user 10, and may include various symptoms such as odor, fever, stinging eyes, shortness of breath, increased heart rate, flames, smoke, visually confirmed container damage or leakage. It may also include the impact on the surrounding environment.

The user 10 may perform a voice conversation with the terminal 100 disposed in place A and input symptoms displayed in the body to the terminal 100 . The terminal 100 may transmit a voice conversation to the server 200 through the network 20 . The terminal 100 may sense the body of the user 10 to generate body information and provide it to the server 200 . Body information may include body temperature, blood pressure, pulse, electromyography, and the like. The terminal 100 may generate environment information by sensing the environment of the place where the user 10 is located and provide the environment information to the server 200 . Environmental information may include temperature, humidity, air pressure, gas leakage, smoke generation, vibration, and inclination. The server 200 may estimate what chemicals the user 10 has been exposed to by analyzing voice conversations, body information, and environmental information.

The server 200 may provide information on the estimated chemical substance to the terminal 100 and the responding agency 30, and the terminal 100 may provide the user 10 with a voice message to the chemical substance the user 10 is exposed to. The type of substance, response method, etc. can be informed. The response agency 30 may include a fire station, a hospital, a police station, a ward office, a city hall, other administrative agencies, site managers, and the like necessary to cope with a chemical leakage accident.

In general, since each chemical causes the same or different symptoms, and it is impossible for the user 10 to know the symptoms for all chemicals, it is very difficult for the user 10 to immediately determine which chemical he or she has been exposed to. difficult. According to one embodiment of the present invention, even if the user 10 does not know what chemical substance he has been exposed to, by performing a voice conversation with the terminal 100, the user 10 will be guided to the type of chemical substance to which he is exposed, how to respond, and the like. can

A plurality of terminals 100 may be disposed in one place. For example, terminal 100 A and terminal 100 B may be disposed in place A. The place may include various places such as industrial facilities, construction sites, laboratories, and multi-use facilities. The server 200 may analyze voice conversations, body information, and environmental information provided from a plurality of terminals 100 respectively disposed in a plurality of places, recognize symptoms, and estimate chemical substances. The terminal 100 and the server 200 may be directly connected by wire or wireless, or connected through a wired or wireless network 20 .

See Figure 2. The terminal 100 performing the method of estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention includes a voice input unit 110 that receives information spoken by the user 10 by voice, and the user 10 A voice output unit 120 that outputs questions by voice, and controls the voice input unit 110 and the voice output unit 120 to perform questions and answers with the user 10 and provide them to the server 200. A controller 130 may be included. Further, the terminal 100 includes a body sensor unit 150 that senses the body of the user 10 to generate body information, and an environment sensor unit 140 that senses the environment in which the user 10 is located to generate environment information. can include The terminal 100 is connected to the server 200 through the network 20, the terminal communication unit 160 transmits and receives data, and the program code necessary for the terminal 100 to perform the chemical substance estimation method using symptoms and chemical structures and a terminal storage unit 170 for storing other data. The terminal 100 may be implemented as an AI speaker having an AI assistant function, an electronic product such as an air purifier, and the like. The terminal 100 may be implemented as a vest, helmet, wristwatch, belt, etc. that can be worn on the body of the user 10 .

The voice input unit 110 may include a microphone. The voice input unit 110 may convert a sentence input as voice by the user 10 into an electrical signal. The control unit 130 may provide the voice input through the voice input unit 110 to the server 200 through the terminal communication unit 160 . The audio output unit 120 may include a speaker. The control unit 130 may convert the question or notification received from the server 200 into voice through the audio output unit 120 and provide it to the user 10 .

Here, the question may include content inquiring whether the body information or environmental information of the user 10 is analyzed by the server 200 and whether symptoms that may occur in the body of the user 10 exist. The question is a symptom database 250 including a knowledge graph generated by learning a chemical substance and symptoms that appear when the chemical substance is exposed to the body by the server 200, the chemical structure of the chemical substance, and the symptoms represented by the chemical structure. It can be created to reduce the range of chemicals exposed to the body of the user 10 by using the result of searching the structure database 260 including the knowledge graph created by learning.

Here, the answer is an answer to a question or a judgment of the user 10, and may include content confirming whether symptoms exist in the body of the user 10. In response to the question of the terminal 100, the user 10 may answer whether symptoms exist in the body. The user 10 may tell the terminal 100 that a symptom exists in the body according to the user 10's own judgment even if the terminal 100 does not ask a question. The terminal 100 provides the server 200 with a conversation that the symptom exists from the user 10, and the server 200 may start a process of estimating a chemical substance that causes the symptom.

The environment sensor unit 140 may include one or more sensors. Sensors included in the environment sensor unit 140 may include various types of sensors such as a gas sensor, a smoke sensor, a camera, a microphone, a temperature sensor, an optical sensor, a pressure sensor, and a vibration sensor. The environment sensor unit 140 may monitor a place where the terminal 100 is located. The environment sensor unit 140 provides the measured value of the sensor to the control unit 130 . The control unit 130 may detect changes in the environment of the place through measurement values received from the environment sensor unit 140 .

The control unit 130 may provide information about the change to the server 200 when the change recognized by the environmental sensor unit 140 is out of a predetermined range. For example, when the measured value of the gas sensor included in the environmental sensor unit 140 is higher than a predetermined value, it is recognized that a change has occurred and the type of sensor (gas sensor) and measured value information can be provided to the server 200. there is.

The body sensor unit 150 may include one or more sensors. Sensors included in the body sensor unit 150 may include a temperature sensor, a blood pressure sensor, an EMG sensor, a pulse sensor, a pressure sensor, and the like. The body sensor unit 150 may monitor the state of the user 10 . The environment sensor unit 140 provides the measured value of the sensor to the control unit 130 . The control unit 130 may detect a change in the state of the body through a measurement value received from the body sensor unit 150 .

The controller 130 may provide information about the change to the server 200 when the change recognized by the body sensor unit 150 is out of a predetermined range. For example, if the body temperature of the user 10 measured by the temperature sensor included in the body sensor unit 150 is higher than a predetermined value, it is recognized that a change has occurred, and the type of sensor (temperature sensor) and the measurement value are sent to the server 200. Value information can be provided.

The controller 130 may control the body sensor unit 150 and the environment sensor unit 140 to measure body information and environment information at predetermined cycles, and during questions and answers, at intervals shorter than the predetermined cycle. . The controller 130 may control the terminal 100 to generate body information and environmental information at predetermined intervals and provide the generated body information and environment information to the server 200 . If the terminal 100 is performing a voice conversation, the controller 130 may generate body information and environment information at a shorter period than a predetermined period and provide the generated body information and environment information to the server 200 . The ongoing voice conversation means that a change has occurred in the environment or body. Accordingly, the control unit 130 frequently measures body information and environment information and provides the information to the server 200, thereby helping to recognize symptoms or estimating chemicals, and increasing the accuracy of estimating chemicals.

The terminal 100 includes an interactive terminal 100 including a voice input unit 110, a voice output unit 120, and a control unit 130, an installation type terminal 100 having an environment sensor unit 140, and a control unit 130. , the body sensor unit 150 and the wearable terminal 100 having the control unit 130, respectively. Alternatively, one or more of the interactive terminal 100 , the installable terminal 100 , and the wearable terminal 100 may be included in one physical terminal 100 .

The server 200, which performs the method of estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention, includes information about the body of the user 10 received from the terminal 100 and the environment in which the user 10 is located. The symptom recognition unit 210 for recognizing symptoms by analyzing environmental information and voice conversations between the user 10 and the terminal 100, and a knowledge graph created by learning chemicals and symptoms that appear when the body is exposed to chemicals The symptom database 250 including the symptom database 250 and chemicals causing symptoms recognized by the symptom recognition unit 210 are estimated by searching the symptom database 250, and if the number of estimated chemicals is smaller than the reference number, the probability is It may include a substance estimator 230 that determines the last exposed chemical substance in the order of high.

In addition, the server 200 according to an embodiment of the present invention includes a structure database 260 including a knowledge graph generated by learning the chemical structure of a chemical substance and symptoms represented by the chemical structure, and Further comprising a structure analysis unit 220 that extracts a chemical structure, extracts a connection relationship between a plurality of chemical structures, and updates the structure database 260 by relating symptoms to the body based on the toxicity of the chemical structure. can do. At this time, the substance estimator 230 searches the symptom database 250 and compares the symptoms expected based on the chemical structure of the chemical substance estimated by the user 10 with the symptom appearing on the body of the user 10, so that the estimated chemical substance is the user ( 10) can confirm whether the chemical substance exposed to the body is correct.

In addition, the server 200 according to an embodiment of the present invention merges the knowledge graph of the symptom database 250 and the knowledge graph of the structure database 260 to determine the relationship between symptoms and chemicals, chemical structures and symptoms. It may include a composite database including a knowledge graph formed by relationships. The composite database may replace symptom database 250 and structure database 260 .

In addition, the server 200 according to an embodiment of the present invention includes a notification providing unit 240 that delivers information on the chemical substance determined by the substance estimation unit 230 to the terminal 100 or the corresponding institution 30, A server communication unit 270 connected to the terminal 100 through the network 20 to transmit and receive data, and the server 200 stores program codes and other data necessary to perform a chemical estimation method using symptoms and chemical structures It may further include a server storage unit 280 that does.

See FIG. 3 . A method for estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention includes body information generated by sensing the body of a user 10 by a terminal 100 and sensing an environment in which the user 10 is located. information collection step of providing the generated environment information to the server 200 (S10), and symptom recognition step in which the server 200 recognizes symptoms appearing on the body of the user 10 based on body information and environment information (S20) ), and a chemical substance estimation step (S30) of estimating the chemical substances exposed to the body of the user 10 based on the symptoms recognized by the server 200. And, in the method for estimating chemicals using symptoms and chemical structures according to an embodiment of the present invention, information on chemicals estimated in the chemical substance estimation step (S30) is provided to the user 10 through the terminal 100. A providing notification step (S30) may be further included.

4 is a diagram illustrating functions performed by a terminal 100 and a server 200 performing a chemical estimation method using symptoms and chemical structures according to an embodiment of the present invention.

See Figures 2, 3, and 4. The information collection step (S10) is a voice conversation step (S11) of receiving symptoms appearing on the body of the user 10 through voice conversation between the terminal 100 and the user 10, A body sensing step (S12) in which the terminal 100 senses the body of the user 10 to generate body information and provides it to the server 200, and the terminal 100 worn on the body of the user 10 or the user ( 10) may include an environment sensing step ( S13 ) in which the terminal 100 disposed in the space where the user 10 is located senses the environment where the user 10 is located to generate environment information and provide it to the server 200 .

The body sensing step ( S12 ) may be performed through the wearable terminal 100 worn by the user 10 . In the body sensing step (S12), the terminal 100 worn on the body of the user 10 senses the body of the user 10 using the body sensor unit 150 included in the terminal 100, and the measured data Body information including may be generated and the body information may be provided to the server 200 . For example, in the body sensing step (S12), the terminal 100 may measure the body temperature of the user 10, include it in body information, and provide it to the server 200.

The environment sensing step S13 may be performed through the wearable terminal 100 worn by the user 10 or the installable terminal 100 installed in the environment where the user 10 is located. In the body sensing step (S12), the installable terminal 100 senses the environment of the place where the user 10 is located using the environment sensor unit 140 included in the terminal 100, and environmental information including the measured data. and may provide environment information to the server 200. For example, in the environment sensing step ( S13 ), the terminal 100 may measure the gas leaked in the place where the user 10 is located, include it in environment information, and provide the information to the server 200 .

The voice conversation step (S11) includes questions and answers exchanged by voice between the user 10 and the terminal 100. The terminal 100 may receive information on bodily symptoms from the user 10 as voice input, and may output a question asking the user 10 whether or not there are bodily symptoms.

The voice conversation step (S11) may be performed through questions and answers between the terminal 100 and the user 10, and the user 10 speaks voice to the terminal 100 about the symptoms of the body recognized by the user 10 and voice conversation. starts, or the server 200 analyzes the body information or environmental information sensed by the terminal 100 and may start by asking a question to the user 10 through the terminal 100 by voice.

The user 10 may start a conversation by first telling the terminal 100 about his symptoms. When the user 10 is exposed to a chemical substance, the user 10 may be aware of the symptoms appearing in the body, and the user 10 inputs the symptoms he/she is aware of into the terminal 100 by voice to the terminal ( 100) to start a conversation. For example, the user 10 may tell the terminal 100 the symptoms of itchy eyes and a headache through voice.

When the user 10 inputs a symptom into the terminal 100 as a voice, the voice input unit 110 of the terminal 100 may receive the user's 10 voice and provide it to the control unit 130 . The controller 130 may provide voice to the server 200 through the terminal communication unit 160 . The server 200 may perform a symptom recognition step of recognizing a symptom from the voice received from the terminal 100 (S20) and a chemical substance estimation step of estimating a substance based on the recognized symptom (S30).

The terminal 100 may start a conversation by asking the user 10 whether symptoms exist. The body sensor unit 150 or the environment sensor unit 140 of the terminal 100 measures the body and environment of the user 10, and when a value exceeding a predetermined range is measured, the control unit 130 indicates that a state change has occurred. The server 200 may be notified. The server 200 may create and provide a question to the terminal 100 asking whether there is a symptom related to the state change, and the terminal 100 may output the question to the user 10 as a voice.

The sensor included in the terminal 100 may detect a leaked chemical substance at a low concentration before it causes symptoms in the human body, or it may first detect the symptoms appearing in the human body. Although the user 10 has been exposed to chemicals, there is a possibility that he or she may not be aware of symptoms of the body. In this case, when the terminal 100 first asks the user 10 about physical symptoms, the user 10 may know the possibility of chemical leakage and have an opportunity to pay attention to whether there are physical symptoms. Therefore, when the terminal 100 makes a conversation request while asking the user 10 via voice first whether symptoms exist, this itself has a useful effect.

When the terminal 100 starts a voice conversation first, the user 10 replies to the terminal 100, and the terminal 100 provides an answer to the server 200, the server 200 performs a symptom recognition step (S20). ) and chemical substance estimation step (S30) can be performed.

Regardless of whether the terminal 100 starts the voice conversation first or the user 10 starts the voice conversation first, the server 200 performs the symptom recognition step (S20) and the chemical substance estimation step (S30), If further confirmation of symptoms is required, the voice conversation can be continued by creating a new question. The voice conversation may be repeatedly performed until the substance estimation unit 230 of the server 200 determines the chemical substance.

5 is a flowchart showing in detail a method for estimating chemicals using symptoms and chemical structures according to an embodiment of the present invention.

See FIGS. 2 , 3 , and 5 . In the symptom recognition step (S20), the server 200 derives a list of symptoms likely to appear to the user 10 based on body information and environment information, analyzes voice conversations, recognizes words expressing symptoms, and lists the symptoms. It is possible to recognize the symptoms appearing on the body of the user 10 by matching with the symptoms included in .

In the symptom recognition step (S20), the symptom recognition unit 210 of the server 200 analyzes the body information, environment information, and voice conversation received from the terminal 100 to recognize symptoms appearing in the body of the user 10. It is a process of

In the symptom recognition step ( S20 ), a list of symptoms that may appear to the user 10 may be derived by analyzing body information and environment information. In body information, when body temperature rises, fever, fainting, exhaustion, etc. can be expected, and in environmental information, breathing difficulties, increased heart rate, rash, stinging eyes, etc. can be expected in case of gas leakage. . Symptoms related to changes in body information or environmental information are stored in the server storage unit 280 .

In the symptom recognition step (S20), the symptom recognition unit 210 may convert a sentence spoken by the user 10 into text using voice recognition technology. The symptom recognizing unit 210 may extract words related to symptoms included in text sentences. The symptom recognizing unit 210 may recognize expressions of various users 10 as expressing specific symptoms. When the user 10 expresses a symptom appearing in the body, each user 10 may use a different word. In this case, the symptom recognizing unit 210 may recognize a specific symptom even if the user 10 describes the symptom displayed on the body in various expressions. The symptom recognizing unit 210 may recognize one or more symptoms described by the user 10 , respectively.

The symptom recognizing unit 210 may determine whether a symptom list derived from the body information and environment information matches a symptom recognized in the voice conversation, and may determine that the corresponding symptom is recognized if they match. If they do not match, it may be determined that the symptom input by the user 10 through voice conversation has been recognized.

The chemical substance estimating step (S30) may be performed by the substance estimating unit 230. In the chemical substance estimation step (S30), symptoms appearing on the body of the user 10 are determined by using the symptom database 250 including a knowledge graph generated by learning chemicals and symptoms that appear when the body is exposed to the chemicals. A symptom-based estimation step of searching for a chemical substance causing symptoms (S31), and when the number of substances estimated to be chemicals causing symptoms in the body of the user 10 is smaller than the reference number, one or more chemicals with high probability are selected as final exposure chemicals. It may include a determination step (S33) of determining the material.

The symptom-based estimating step (S31) is a process of searching for a chemical substance capable of causing a corresponding symptom based on a symptom recognized as appearing on the body of the user 10.

6 is a diagram showing a knowledge graph of a symptom database 250 used in a method for estimating chemicals using symptoms and chemical structures according to an embodiment of the present invention.

See FIGS. 2 and 6 . The symptom database 250 of the server 200 may include a knowledge graph generated by learning chemicals and symptoms. Symptoms include changes in the body that occur when the body is exposed to a chemical. Chemicals and their associated symptoms can be formed by learning previously published databases. The data that the symptom database 250 learns is, for example, material safety data sheet (MSDS) provided by the Korea Occupational Safety and Health Agency, Hazardous Substance Data (HSDB) included in the US NIH's WISER (Wireless Information System for Emergency Responders) Bank), databases open to the public by the US National Institute for Occupational Safety and Health (NIOSH).

The symptom database 250 of the server 200 is, for example, as shown in part in FIG. 6, chemicals (chemicals A, B, C, D) and related symptoms (symptoms A, B, . .. I) may have a form in which each relational value (relational value AA, AB, ... DI) is connected. Since one chemical substance can show multiple symptoms, one chemical substance can be connected with multiple symptoms through relational values. Since one symptom can be manifested by various chemicals, one symptom can be connected to multiple chemicals through relational values. In FIG. 6, the relationship value is shown as a line connecting the chemical substance and the symptom, and various methods can be used to display the relationship value. In this specification, the relationship value indicated as relationship value AB is symptom B caused by chemical substance A. It is exemplarily defined and explained to mean.

If there is a relationship value between a chemical substance and a symptom, it can be judged that there is a symptom caused by a chemical substance in the body. The relational value may include various information such as the amount of a chemical substance to be exposed to cause symptoms, the fatality of symptoms occurring when exposed to a chemical substance, and the order in which symptoms occur when exposed to a chemical substance.

When a symptom is input from the symptom recognizing unit 210, the substance estimator 230 may perform a symptom-based estimating step S31 of searching for the corresponding symptom in the symptom database 250. The symptom-based estimation step (S31) is a process of searching for a chemical substance that causes a symptom appearing in the body of the user 10. First, chemical substances connected through the entered symptoms and relationship values are searched. For example, when symptom A is input from the symptom recognizing unit 210, chemicals (chemical substances A, B, C, and D) that can cause symptom A are searched for.

Next, the substance estimator 230 may perform a judgment step (S33) of determining whether it is appropriate to finally output the retrieved chemical substances as chemical substances exposed to the user 10. In the determination step (S33), the substance estimation unit 230 may determine based on whether the number of searched chemicals is greater than the reference number. There may be a variety of chemicals that produce any symptom, and the required response may be different. Therefore, outputting too many chemicals can make it difficult to respond. Therefore, the reference number can be set to a number small enough for the user 10 to respond to the chemical substance. For example, if the standard number is 2 and the searched chemicals are 4 of A, B, C, and D, in the determination step (S33), the substance estimator 230 determines that the searched chemicals are chemicals exposed to the user 10. It can be determined that it is difficult to estimate (N).

When it is determined (N) that it is difficult to estimate that the chemical substance retrieved in the determination step (S33) is a chemical substance exposed to the user 10, the question generation step (S34) may be performed. The question generated in the question generating step (S34) is transmitted to the terminal 100, and the terminal 100 outputs the question to the user 10 by voice to perform a voice conversation. The question generation step (S34) generates a question for estimating the chemical substance exposed to the body of the user 10 when the chemical substance exposed to the body of the user 10 is not determined in the determination step (S33), and the terminal (100) is provided. The question generating step ( S34 ) may be performed by the substance estimation unit 230 .

The substance estimator 230 may search for other symptoms indicated by the searched chemicals and generate a question asking the user 10 whether one or more of the other symptoms exist. For example, symptoms caused by chemicals A, B, C, and D in the body can be retrieved as having symptoms A, B, ... I. Among the searched symptoms, a question asking whether symptom E exists may be generated. At this time, if the user 10 answers that symptom E exists, chemicals C and D may be excluded, and the user 10 If the answer is that symptom E does not exist, chemicals A and B can be excluded. In this way, the substance estimator 230 may generate a question inquiring about symptoms that may occur to the user 10 using the question generating step (S34). The question is provided to the terminal 100 through the server communication unit 270, and the terminal 100 provides the user 10 with the question as a voice so that the voice conversation step (S11) can be performed again. When an answer is received from the user 10 by performing the voice conversation step (S11), the symptom recognition step (S20), the symptom-based estimation step (S31), the judgment step (S33), the question generation step (S34), or substance determination Step S35 may be performed.

The symptom recognition step (S20) is performed, the symptom-based estimation step (S31) of the chemical substance estimation step (S30), the judgment step (S33), the question generation step (S34) are performed, and the voice conversation step (S11) is performed again. If this is repeated, it is possible to repeatedly determine the presence or absence of symptoms appearing in the body of the user 10, and it is possible to continue narrowing down the range of chemicals. If this process is repeatedly performed, the number of chemicals found in the determination step (S33) can be filtered to be smaller than the reference number.

In the determination step (S33), if the number of substances determined to be chemical substances exposed to the user 10 is smaller than the reference number, it is determined that the searched chemical substance is a chemical substance exposed to the user 10 (Y), and the substance is determined. Step S35 may be performed. In the substance determination step S35 , one or more chemicals may be determined in order of a high probability of being exposed to the user 10 from among the searched chemicals. The probability of exposure to the user 10 may be determined by factors such as whether the chemical substance is used in the environment where the user 10 is located, the frequency of use of the chemical substance, and the work of the user 10 . The determined chemical substance may be provided to the notification provider 240 . The notification provider 240 may provide the terminal 100 or the responding institution 30 with the determined type of chemical substance, response method, and other necessary information.

By performing this process, the user 10 inputs symptoms appearing in his/her body through voice to the terminal 100, and the server 200 determines the user 10's body based on the symptoms input by the user 10. It is possible to determine which substances are presumed to have been exposed to. Since the user 10 communicates with the terminal 100 using voice, he or she can conveniently input symptoms and receive a response method.

7 is a diagram showing the chemical structure of chemicals used in the method for estimating chemicals using symptoms and chemical structures according to an embodiment of the present invention.

In the method for estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention, the chemical substance estimation step (S30) is based on the chemical structure of the chemical substance estimated in the symptom-based estimation step (S31). A structure confirmation step (S32) of comparing the expected symptom with the symptom appearing on the body of the user 10 and confirming whether the estimated chemical substance is the chemical substance exposed to the body of the user 10 (S32) may be further included.

Because there are many types of chemicals, searching for all chemicals and all symptoms during the search process is likely to reduce the accuracy of estimating exposure to the body and can be time consuming. Therefore, a structure confirmation step (S32) of analyzing the chemical structures of the chemical substances retrieved as a result of performing the symptom-based estimation step (S31) and confirming whether they match the symptoms displayed by the user 10 may be additionally performed.

를 가질 수 있다. As shown in Figure 7, chemicals can have a variety of structures. For example, chemical substance A may have a structure in which structures 1 and 3 are connected. Chemical substance B is a structure where structure 1, structure 3, structure 4, and structure 7 are connected.

can have

를 갖고 있는지 판단할 수 있다. 구조 분석부(220)는 SMILES, InChI 등의 방식으로 표현된 화학물질을 분석하여 구조를 확인할 수 있다. 구조 분석부(220)는 증상기반 추정 단계(S31)에서 검색된 화학물질들 각각의 구조를 분석할 수 있다. 구조 분석부(220)는 화학물질 C를 분석하여, 화학물질 C는 구조 1, 구조 2가 연결되고, 구조 2에 구조 1과 다른 방향으로 구조 5가 연결되는 구조를 갖는 것을 확인할 수 있다. The structure analysis unit 220 of the server 200 analyzes the structure of a chemical substance to

You can determine if you have The structure analyzer 220 may analyze a chemical substance expressed in a method such as SMILES or InChI to confirm the structure. The structure analysis unit 220 may analyze the structure of each of the chemical substances retrieved in the symptom-based estimation step (S31). The structure analyzer 220 analyzes chemical substance C and confirms that chemical substance C has a structure in which structures 1 and 2 are connected, and structure 5 is connected to structure 2 in a direction different from structure 1.

When a structure of a new chemical substance is input, the structure analyzer 220 of the server 200 may analyze the chemical substance structure and display it as a connection of a plurality of detailed structures. For example, by analyzing chemical substance D, the structure analysis unit 220 can confirm that chemical substance D has a structure in which structures 1, 7, and 3 are connected.

를 변성시키거나, 체액과 결합하여 산성도를 변화시키는 등의 작용을 할 수 있다. 이와 같이 화학물질의 구조 변화는 인체에서 정해진 기능을 방해할 수 있다. 정리하면, 화학물질의 특정한 구조들은 인체에 노출되는 경우 인체의 생명활동을 방해하는 독성을 가질 수 있다. 화학물질의 특정한 구조가 갖는 독성은 인체에 특정한 증상 또는 다양한 증상을 발생시킬 수 있다. 따라서 화학물질의 구조와 증상은 서로 관련될 수 있다. 화학물질의 특정 구조는 특정 증상을 발생시킬 수 있고, 특정 구조가 다른 특정 구조와 함께 연결되면 특정 증상을 발생시키지 않을 수도 있다. The structures constituting the chemical substance (structure 1, structure 2, ...) may include chemical functional groups. Functional groups can chemically combine with other chemicals to change the structure of the chemical. The structures that make up chemicals are called proteins.

It can denature or combine with bodily fluids to change acidity. As such, changes in the structure of chemicals can interfere with their intended functions in the human body. In summary, certain structures of chemical substances can have toxicity that interferes with the vital activity of the human body when exposed to the human body. The toxicity of a specific structure of a chemical substance can cause specific symptoms or various symptoms in the human body. Thus, the structure and symptoms of a chemical can be related. A particular structure of a chemical can cause a particular symptom, and a particular structure may not cause a particular symptom when linked together with another particular structure.

A relationship between a chemical substance and a symptom stored in the knowledge graph of the symptom database 250 and a relationship between a chemical structure and a symptom stored in the knowledge graph of the structure database 260 do not match. Therefore, if the symptom is additionally confirmed based on the structure using the structure database 260, the chemical substance can be more accurately estimated.

The structure analysis unit 220 of the server 200 may predict symptoms of C, a new chemical substance, based on symptoms caused by each of the structures stored in the structure database 260. The structure analyzer 220 may update the symptom database 250 by associating a new chemical substance C with a symptom caused by the chemical substance C.

8 is a diagram showing a knowledge graph of a structure database 260 used in a method for estimating chemicals using symptoms and chemical structures according to an embodiment of the present invention.

The substance estimator 230 may perform a structure confirmation step (S32) on a plurality of chemical substances resulting from the symptom-based estimation step (S31). In the structure confirmation step (S32), the structure analysis unit 220 analyzes the chemical substance, which is the result of the search in the symptom-based estimation step, to obtain the structure and connection of the chemical substance. The chemical structure and connection are searched in the structure database 260 to obtain related symptoms. The symptoms obtained by searching the structure database 260 are compared with the symptoms displayed on the body of the user 10 . If the symptoms obtained by searching the structure database 260 include the symptoms appearing on the body of the user 10, the chemical substance is maintained as a search result, and if not included, it is excluded from the search results.

For example, when chemical substance B in FIG. 7 is searched as a result of performing the symptom-based estimation step (S31), the structure analysis unit 220 analyzes the structure of chemical substance B, and chemical substance B is structure 1, structure 3, structure 4, and structure 7 can be analyzed.

In the knowledge graph shown in FIG. 8, when a symptom and a structure are connected by a solid line, a specific structure causes a specific symptom, and when a symptom and a structure are connected by a dotted line, a relationship that does not cause a specific symptom when a plurality of structures are connected. In other words, when symptoms and structures are connected by a dashed-dotted line, it is a relationship in which a specific symptom occurs only when multiple structures are connected. When a specific structure is connected to another structure, another chemical reaction can cause toxicity to disappear or toxicity to be created that did not exist. Such information is open to the public, and a knowledge graph of a structure database can be created by learning these relationships.

The substance estimator 230 uses the knowledge graph of the structure database 260 of FIG. Symptom H related to structure 4, symptom F related to structure 7, and symptom E related to structure 7 can be acquired. Symptom D cannot be caused when structure 2, structure 3, and structure 6 are all connected, so symptom D is not a symptom that can be caused by chemical substance B. Symptom H is not a probable symptom because structure 3 and structure 4 are connected. Thus, Symptom A, Symptom B, Symptom E, Symptom G, and Symptom F can be analyzed as possible symptoms caused by chemical substance B.

The substance estimator 230 may exclude chemical substance B from the estimated substances when the user 10 has symptom D. Alternatively, the substance estimation unit 230 does not exclude chemical substance B when symptoms A, symptoms B, and symptoms E are recognized by the user 10, but symptoms G and symptoms F are not recognized.

Referring again to FIG. 5 . The chemical substance estimation step (S30) is a question generation step of generating a question confirming whether a symptom that is not expected to appear or a symptom that is expected to appear exists using the symptom database 250 and the structure database 260 ( S34) may be further included.

In the question generation step (S34), the symptom-based estimation step (S31) or the structure confirmation step (S32) is performed, and as a result of the determination step (S33), when the number of searched chemicals is greater than the reference number (N) can be performed That is, if too many chemicals are searched, additional queries may be performed if needed to improve accuracy.

The question generating step ( S34 ) may be performed by the substance estimation unit 230 . As a result of performing the symptom-based estimation step (S31) or the structure confirmation step (S32), the substance estimation unit 230 determines a plurality of chemicals, a plurality of symptoms that are likely to occur, and a plurality of symptoms that are unlikely to occur. symptoms can be obtained. In the question generating step (S34), a question asking whether a specific symptom exists may be generated. The symptom to be confirmed is the number and possibility of occurrence of a specific symptom among chemicals that are search results, excluding symptoms recognized as having already appeared to the user 10 or symptoms recognized as not appearing. Select the symptom with the smallest difference in the number of chemicals without For example, there are 8 chemicals that can cause symptom A, 2 chemicals that cannot cause symptom A, 6 chemicals that can cause symptom B, and 2 chemicals that cannot cause symptom B. If there are 4 substances, choose Symptom B.

9 is a diagram showing a knowledge graph in which a symptom knowledge graph and a structure knowledge graph used in a method for estimating chemical substances using symptoms and chemical structures according to an embodiment of the present invention are integrated.

In the integrated database, the symptom database 250 and the structure database 260 are integrated. The knowledge graph of the integrated database consists of three layers. The first layer contains chemicals, the second layer contains symptoms, and the third layer contains chemical structures. Chemicals A, B, C, and D are located in the first layer, and symptoms A, B, C, D, E, F, G, etc. associated with chemicals A, B, C, and D are located in the second layer, Structures 1, 2, 3, 4, 5, etc. are located in the third layer. Chemicals are related to symptoms, and symptoms are related to structures.

The substance estimator 230 may perform a search from the second layer where the symptom of the integrated database is located to the first layer where the chemical substance is located in the symptom-based estimation step (S31). The searched chemical substance may be analyzed by the structure analysis unit 220 and recognized as a connection of one or more structures. The substance estimator 230 may perform a search from the third layer where the structure of the integrated database is located to the second layer where the symptom is located in the structure confirmation step (S32). At this time, the symptom of the second layer, which is the starting point of the symptom-based estimation step (S31), and the symptom of the second layer, which is the destination point of the structure confirmation step (S32), may or may not match. If the chemical substance associated with the symptom that matches the structure associated with the symptom that matches is included, it may first be asked whether a symptom related to the corresponding chemical substance exists in the question generation step S34.

Although the present invention has been described in detail through specific examples, this is for explaining the present invention in detail, the present invention is not limited thereto, and within the technical spirit of the present invention, by those skilled in the art It will be clear that the modification or improvement is possible.

All simple modifications or changes of the present invention fall within the scope of the present invention, and the specific protection scope of the present invention will be clarified by the appended claims.

10: User
20: network
30: Response Agency
100: terminal
110: voice input unit
120: audio output unit
130: control unit
140: environmental sensor unit
150: body sensor unit
160: terminal communication unit
170: terminal storage unit
200: server
210: symptom recognition unit
220: structural analysis unit
230: material estimation unit
240: notification providing unit
250: update unit
260: symptom database
270 structure database
280: server communication unit
290: server storage unit

Claims (13)

An information collection step of providing, by a terminal, body information generated by sensing a user's body and environment information generated by sensing an environment in which the user is located to a server;
a symptom recognition step in which a server recognizes a symptom displayed on a user's body based on the body information and environment information; and
A method of estimating chemicals using symptoms and chemical structures, comprising a chemical estimation step of estimating, by a server, chemicals exposed to a user's body based on the recognized symptoms. The method of claim 1,
The information gathering step
a voice conversation step of receiving symptoms appearing on the user's body through voice conversation between the terminal and the user;
a body sensing step in which the terminal worn on the user's body senses the user's body, generates body information, and provides the body information to a server; and
Chemical substances using symptoms and chemical structures, including an environment sensing step in which a terminal worn on the user's body or a terminal disposed in the space where the user is located senses the environment where the user is located to generate environment information and provide it to a server estimation method. The method of claim 2,
The voice conversation step is
It can be performed through questions and answers between the terminal and the user, and a voice conversation is started while speaking the symptoms of the body perceived by the user to the terminal by voice, or body information or environmental information sensed by the terminal is transmitted to the server. A method for estimating chemicals using symptoms and chemical structures, which is started by analyzing and asking a user a question by voice through the terminal. The method of claim 1,
The symptom recognition stage is
The server derives a list of symptoms that may appear to the user based on the body information and environment information, analyzes voice conversations to recognize words expressing symptoms, and matches the symptoms included in the symptom list to the user's body. A method for estimating chemical substances using symptoms and chemical structures, recognizing symptoms shown in . The method of claim 1,
The chemical estimation step is
A symptom-based estimating step of searching for a chemical substance causing a symptom appearing in the user's body by using a symptom database including a chemical substance and a knowledge graph generated by learning symptoms appearing when the chemical substance is exposed to the body;
Chemical substances using symptoms and chemical structures, including a determination step of determining one or more chemical substances with a high probability as the final exposure chemical substance when the substance estimated to be a chemical substance causing symptoms in the user's body is smaller than the reference number estimation method. The method of claim 5,
The chemical estimation step is
A structure for confirming whether the estimated chemical substance is the chemical substance exposed to the user's body by comparing the symptom expected based on the chemical structure of the chemical substance estimated in the symptom-based estimation step with the symptom appearing on the user's body A chemical estimation method using symptoms and chemical structures, further comprising an identification step. The method of claim 6,
The chemical estimation step is
Chemical substances using symptoms and chemical structures, further comprising a question generating step of generating a question confirming whether a symptom that is not expected to appear or a symptom that is expected to appear exists using the symptom database and the structure database estimation method. The method of claim 1,
A chemical substance estimation method using symptoms and chemical structures, further comprising a notification step of providing information on the chemical substance estimated in the chemical substance estimation step to the user through the terminal. a voice input unit that receives information spoken by a user through voice;
a voice output unit outputting a question to the user by voice; and
A control unit controlling the voice input unit and the audio output unit to perform questions and answers with the user by voice and provide the information to the server;
The above question
The server analyzes the user's body information or environmental information and inquires whether or not there are symptoms that may occur in the user's body,
The above answer is
A terminal that performs a method of estimating a chemical substance using a symptom and a chemical structure, including information for confirming whether a symptom exists in the user's body, as an answer to the question or as a judgment of the user. The method of claim 9,
a body sensor unit configured to sense the user's body and generate body information; and
Further comprising an environment sensor unit for generating environment information by sensing an environment in which the user is located;
The control unit
Chemical substance estimation using symptoms and chemical structures that controls the body sensor unit and the environment sensor unit to measure the body information and environment information at predetermined cycles, and to measure the body information and environment information at intervals shorter than the predetermined cycle while the question and answer are in progress. A terminal that performs the method. a symptom recognition unit for recognizing a symptom by analyzing body information of a user received from a terminal, environment information of an environment in which the user is located, and voice conversation between the user and the terminal;
a symptom database including a chemical substance and a knowledge graph generated by learning symptoms that appear when the chemical substance is exposed to the body; and
Chemical substances that cause symptoms recognized by the symptom recognizing unit are estimated by searching the symptom database, and if the number of estimated chemicals is smaller than the reference number, substance estimation that determines the finally exposed chemical substances in order of high probability. A server that performs a method for estimating chemical substances using symptoms and chemical structures, including parts. The method of claim 12,
A structure database including a knowledge graph generated by learning the chemical structure of a chemical substance and the symptom indicated by the chemical structure; and
Further comprising a structure analysis unit that extracts chemical structures constituting the chemical substance, extracts a connection relationship between a plurality of chemical structures, and updates a symptom database by relating symptoms caused by the toxicity of the chemical structure acting on the body,
The substance estimation unit
Checking whether the estimated chemical substance is a chemical substance exposed to the user's body by comparing the symptom expected based on the chemical structure of the chemical substance estimated by searching the symptom database and the symptom appearing in the user's body, A server that performs chemical estimation methods using symptoms and chemical structures. The method of claim 12,
Symptoms and chemical structures further comprising a composite database including a knowledge graph formed by merging the knowledge graph of the symptom database and the knowledge graph of the structure database to form a relationship between a symptom and a chemical substance and a relationship between a chemical structure and a symptom. A server that performs the chemical estimation method used.

Images