KR20230010802A - Method, apparatus and computer program for providing synthetic material development process using artificial intelligence model - Google Patents

Abstract

A synthetic material development process providing method using an artificial intelligence model and a device and computer program thereof are provided. According to various embodiments of the present invention, the synthetic material development process providing method using an artificial intelligence model performed by a computing device comprises the following steps of: obtaining one or more conditions; and extracting information about a synthetic material that satisfies the one or more obtained conditions and extracting guide information for synthesizing the synthetic material based on the extracted information about the synthetic material.

Description

Method, device and computer program for providing synthetic material development process using artificial intelligence model

Various embodiments of the present invention relate to a method, apparatus, and computer program for providing a synthetic material development process using an artificial intelligence model.

In general, bio and nano research is conducted through the steps of synthesizing substances and materials, confirming structures and configurations, measuring various properties to confirm application possibilities, and commercializing applications.

Here, if the structure and composition of an optimized material are not derived in the step of confirming the structure and composition, or if the desired properties are not specified in the step of measuring various properties for confirming application possibility, the first step, the structure and composition of the material Since the synthesizing step must be repeated, and the step of synthesizing the structure and composition of the material must be repeatedly performed to find the optimized material, it takes a minimum of several months to a maximum of several years to find the optimized material. There is a problem that a lot of cost is spent on synthesis and inspection.

On the other hand, in order to solve these conventional problems, a method of using an artificial intelligence model learned from various information and data related to bio and nano research has been proposed.

Here, in order to use the artificial intelligence model, it is essential to create research record data as data for learning the artificial intelligence model. Since it is written, there are various problems such as causing hassle such as repetition of similar records, omission of important conditions for experiments, necessity of reprocessing when using data, and difficulties in analysis and management due to the generation of large amounts of data.

The problem to be solved by the present invention is to process unstructured research record information to automatically generate standardized research record data, thereby increasing the convenience of records, significantly improving the accuracy of records, analyzing and sharing research processes and results and management, as well as deriving a synthetic material that satisfies specific conditions and guide information for synthesizing the synthetic material through an artificial intelligence model learned using the research record data standardized according to the above method as learning data To provide a synthetic material development process using an artificial intelligence model that can assist in developing synthetic materials more conveniently and quickly by providing a synthetic material development process, a device, and a computer program.

The problems to be solved by the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the description below.

A method for providing a synthetic material development process using an artificial intelligence model according to an embodiment of the present invention for solving the above problems is a method performed by a computing device, comprising the steps of acquiring one or more conditions and the obtained one or more conditions. The method may include extracting information on a synthetic material that satisfies a condition and extracting guide information for synthesizing the synthetic material based on the extracted information on the synthetic material.

In various embodiments, the extracting of the guide information may include generating a first artificial intelligence model, attribute information including structure and characteristic information of each of a plurality of synthetic materials, and synthesizing each of the plurality of synthetic materials. The first artificial intelligence model generated by using information - the information for synthesizing each of the plurality of synthetic materials includes information on materials, equipment, and synthesis methods for synthesizing each of the plurality of synthetic materials - as learning data and predicting a synthetic material having a structure and characteristics that satisfy the acquired one or more conditions by using the obtained one or more conditions as input values of the learned first artificial intelligence model, and the predicted synthetic material It may include a step of extracting result values including materials, equipment, and synthesis methods for synthesizing.

In various embodiments, the step of training the generated first artificial intelligence model may include, in response to providing the extracted result value to a user, feedback information from the user - the feedback information includes the predicted synthetic material. Including materials for synthesis, equipment, and information on synthetic materials generated according to the synthesis method - Step of receiving an input and re-learning the learned first artificial intelligence model using the input feedback information as learning data can include

In various embodiments, the step of extracting the guide information is predicted to be synthesized according to the specific material, equipment, and synthesis method by using the specific material, equipment, and synthesis method as input values of the learned first artificial intelligence model A step of extracting information about the synthetic material as a result value may be further included.

In various embodiments, the step of learning the generated first artificial intelligence model includes learning the learned first artificial intelligence model using a yield according to a plurality of synthesis methods as learning data, and the result value The step of extracting is, when scale-up for the predicted synthetic material is to be performed, the prediction through the first artificial intelligence model learned using the yield according to the plurality of synthesis methods as learning data It may include re-extracting the synthesis method for synthesizing the synthesized material.

In various embodiments, the extracting of the guide information may include generating a second artificial intelligence model, information on a plurality of synthetic materials and information on a test - the information on the test may include information on the plurality of synthetic materials. Learning the generated second artificial intelligence model using conditions set for conducting a preclinical test or clinical trial, including test results according to the set conditions as learning data, and learning information about the synthetic material and setting a condition for testing the synthetic material using the input value of the second artificial intelligence model, wherein the step of learning the generated second artificial intelligence model is performed on the synthetic material according to the set condition. Re-learning the learned second artificial intelligence model using a result of performing a test for the learning data as learning data.

In various embodiments, in the step of extracting the guide information, a test result prediction value of the synthetic material is extracted as a result value by using a specific condition for the test of the synthetic material as an input value of the learned second artificial intelligence model It may further include steps to do.

In various embodiments, the step of extracting the guide information may include generating a third artificial intelligence model, information on a plurality of test subjects, information on a plurality of synthetic materials developed as new materials, and a plurality of materials developed as new materials. learning the generated third artificial intelligence model using the test results of each of the synthetic materials as learning data, and using the information on the synthetic material as an input value of the third artificial intelligence model New material including the synthetic material A step of extracting, as a result value, information about a target suitable for using the material including the synthetic material and the amount used during development may be included.

An apparatus for providing a synthetic material development process using an artificial intelligence model according to another embodiment of the present invention for solving the above problems is a processor, a network interface, a memory, and a computer loaded into the memory and executed by the processor. Including a program, wherein the computer program extracts instructions for acquiring one or more conditions and information about synthetic materials that satisfy the obtained one or more conditions, and based on the information about the extracted synthetic materials It may include instructions for extracting guide information for synthesizing the synthetic material.

A computer program recorded on a computer-readable recording medium according to another embodiment of the present invention for solving the above problems is combined with a computing device, obtaining one or more conditions, and the obtained one or more conditions. A method for providing a synthetic material development process using an artificial intelligence model comprising the step of extracting information about a synthetic material that satisfies and extracting guide information for synthesizing the synthetic material based on the extracted synthetic material information. It may be stored in a computer-readable recording medium in order to be executed.

Other specific details of the invention are included in the detailed description and drawings.

According to various embodiments of the present invention, by processing non-standard research record information and automatically generating standardized research record data, the convenience of record is increased, the accuracy of record is greatly improved, the research process and result analysis, In addition to making it easy to share and manage, synthetic materials that satisfy specific conditions and guide information for synthesizing the synthetic materials through an artificial intelligence model learned using research record data standardized according to the above method as learning data By providing a synthetic material development process that derives, there is an advantage in being able to assist in developing synthetic materials more conveniently and quickly.

The effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description below.

1 is a diagram illustrating a synthetic material development process providing system using an artificial intelligence model according to an embodiment of the present invention.
2 is a hardware configuration diagram of a synthetic material development process providing apparatus using an artificial intelligence model according to another embodiment of the present invention.
3 is a flowchart illustrating a method of automatically generating standardized research record data for learning of an artificial intelligence model, in various embodiments.
4 is a flowchart illustrating a method of automatically generating standardized research record data using the research record information in the form of a line in various embodiments.
5 to 7 are diagrams illustrating a process of automatically generating standardized research record data using the research record information in the form of a line in various embodiments.
8 is a flowchart illustrating a method of automatically generating standardized research record data through keyword input, in various embodiments.
9 is a diagram illustrating a process of automatically generating standardized research record data through keyword input in various embodiments.
10 is a diagram illustrating a process of automatically generating standardized research record data using voice-type research record information in various embodiments.
11 is a flowchart illustrating a method of generating research record information in a line form using standardized research record data, in various embodiments.
12 to 14 are diagrams illustrating a process of generating research record information in a line form using standardized research record data, in various embodiments.
15 is a flowchart illustrating a method of performing a simulation based on standardized study record data, in various embodiments.
16 is a diagram for explaining a method of extracting a synthetic material and guide information for synthesizing the synthetic material through a first artificial intelligence model, in various embodiments.
17 is a diagram for explaining a method of setting test conditions for a synthetic material through a second artificial intelligence model, in various embodiments.
18 is a view for explaining a method of deriving a patient and dosage suitable for a new drug when developing a new drug through a third artificial intelligence model, in various embodiments.
19 is a diagram illustrating a process of recommending and providing unstructured data based on standardized research record data, in various embodiments.
20 is a diagram illustrating a process of recommending and matching a researcher based on standardized research record data, in various embodiments.
21 to 24 are diagrams illustrating forms of user interfaces applicable to various embodiments by way of example.

Advantages and features of the present invention, and methods of achieving them, will become clear with reference to the detailed description of the following embodiments taken in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, only these embodiments are intended to complete the disclosure of the present invention, and are common in the art to which the present invention belongs. It is provided to fully inform the person skilled in the art of the scope of the invention, and the invention is only defined by the scope of the claims.

Terminology used herein is for describing the embodiments and is not intended to limit the present invention. In this specification, singular forms also include plural forms unless specifically stated otherwise in a phrase. As used herein, "comprises" and/or "comprising" does not exclude the presence or addition of one or more other elements other than the recited elements. Like reference numerals throughout the specification refer to like elements, and “and/or” includes each and every combination of one or more of the recited elements. Although "first", "second", etc. are used to describe various components, these components are not limited by these terms, of course. These terms are only used to distinguish one component from another. Accordingly, it goes without saying that the first element mentioned below may also be the second element within the technical spirit of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used in this specification may be used with meanings commonly understood by those skilled in the art to which the present invention belongs. In addition, terms defined in commonly used dictionaries are not interpreted ideally or excessively unless explicitly specifically defined.

The term "unit" or "module" used in the specification means a hardware component such as software, FPGA or ASIC, and "unit" or "module" performs certain roles. However, "unit" or "module" is not meant to be limited to software or hardware. A “unit” or “module” may be configured to reside in an addressable storage medium and may be configured to reproduce one or more processors. Thus, as an example, a “unit” or “module” may refer to components such as software components, object-oriented software components, class components, and task components, processes, functions, properties, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. Functions provided within components and "units" or "modules" may be combined into smaller numbers of components and "units" or "modules" or may be combined into additional components and "units" or "modules". can be further separated.

The spatially relative terms "below", "beneath", "lower", "above", "upper", etc. It can be used to easily describe a component's correlation with other components. Spatially relative terms should be understood as including different orientations of elements in use or operation in addition to the orientations shown in the drawings. For example, if you flip a component that is shown in a drawing, a component described as "below" or "beneath" another component will be placed "above" the other component. can Thus, the exemplary term “below” may include directions of both below and above. Components may also be oriented in other orientations, and thus spatially relative terms may be interpreted according to orientation.

In this specification, a computer means any kind of hardware device including at least one processor, and may be understood as encompassing a software configuration operating in a corresponding hardware device according to an embodiment. For example, a computer may be understood as including a smartphone, a tablet PC, a desktop computer, a laptop computer, and user clients and applications running on each device, but is not limited thereto.

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

Although each step described in this specification is described as being performed by a computer, the subject of each step is not limited thereto, and at least a part of each step may be performed in different devices according to embodiments.

1 is a diagram illustrating a synthetic material development process providing system using an artificial intelligence model according to an embodiment of the present invention.

1, a synthetic material development process providing system using an artificial intelligence model according to an embodiment of the present invention includes a synthetic material development process providing apparatus 100, a user terminal 200, an external server 300, and a network ( 400) may be included.

Here, the synthetic material development process providing system using the artificial intelligence model shown in FIG. 1 is according to an embodiment, and its components are not limited to the embodiment shown in FIG. may be deleted. For example, a synthetic material development process providing system using an artificial intelligence model does not separately include an external server 300 for storing and managing various information and data, and a storage space provided in the synthetic material development process providing apparatus 100. You can store and manage various types of information and data using .

In one embodiment, the synthetic material development process providing apparatus 100 may receive research record information about an experiment from a user, and automatically process and record the input research record information according to a preset format, thereby forming an artificial intelligence model. It is possible to create standardized research record data for learning.

In addition, the synthetic material development process providing apparatus 100 may record and store standardized research record data, and computes various information included in the stored standardized research record data to virtually simulate a specific experiment. Coded research record data can be coded to create coded study record data. In this case, the standardized research record data and the coded research record data may be correlated and stored.

In various embodiments, the synthetic material development process providing apparatus 100 predicts a synthetic material that satisfies a specific condition through a pre-learned artificial intelligence model using research record data standardized according to the above method as learning data, It is possible to provide a synthetic material development process providing method using an artificial intelligence model that extracts and provides guide information for synthesizing a predicted synthetic material.

In various embodiments, the synthetic material development process providing apparatus 100 may be connected to the user terminal 200 through the network 400, and the user terminal 200 automatically records standardized research data for learning of an artificial intelligence model. To provide a user interface (UI) (e.g., a graphical user interface (GUI), 10 of FIGS. 20 to 24) for providing a method of providing a synthetic material development process using a generation method and an artificial intelligence model. can

Here, the network 400 may refer to a connection structure capable of exchanging information between nodes such as a plurality of terminals and servers. For example, the network 400 includes a local area network (LAN), a wide area network (WAN), a world wide web (WWW), a wired and wireless data communication network, a telephone network, a wired and wireless television communication network, and the like. can do.

In addition, here, the wireless data communication networks are 3G, 4G, 5G, 3GPP (3rd Generation Partnership Project), 5GPP (5th Generation Partnership Project), LTE (Long Term Evolution), WIMAX (World Interoperability for Microwave Access), Wi-Fi (Wi-Fi) Fi), Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth network, A near-field communication (NFC) network, a satellite broadcasting network, an analog broadcasting network, a digital multimedia broadcasting (DMB) network, and the like may be included. However, it is not limited thereto, and other general-purpose networks applicable to the same or similar technical fields may be further included.

In various embodiments, the synthetic material development process providing apparatus 100 may generate standardized research record data based on research record information input from a specific user, and share the standardized research record data with other users. A standardized research record data sharing service can be provided.

For example, the synthetic material development process providing apparatus 100 may store and manage a plurality of standardized research record data generated based on research record information input from a plurality of users, and provide each of the plurality of users with It is possible to provide a search and reproduction service for stored standardized research record data, and through the search and reproduction service, users can search and confirm the standardized research record data they want (e.g., check and simulate information recorded in data). can be implemented so that

In addition, the synthetic material development process providing apparatus 100 provides an import function for importing and storing standardized research record data from the outside and an export function for exporting previously stored standardized research record data to the outside, It may be implemented so that the user can download a plurality of pre-stored standardized research record data or upload standardized research record data generated externally from the user.

At this time, the computing device 100 supports various extensions, and wide compatibility, such as converting the extension of standardized research record data exported to the outside according to the user's request or uploading standardized research record data having various extensions from the outside. It can be implemented to have

In various embodiments, the synthetic material development process providing apparatus 100 provides a method for automatically generating standardized research record data for learning of a web or application-based artificial intelligence model and an artificial intelligence model according to a user's request. It is possible to provide a method for providing a synthetic material development process using, but is not limited thereto.

In one embodiment, the user terminal 200 may be connected to the synthetic material development process providing apparatus 100 through the network 400, and input research record information through the UI provided from the synthetic material development process providing apparatus 100. and, in response to input of the research record information, standardized research record data generated in response to the research record information may be provided.

In addition, the user terminal 200 may input a specific condition through the UI provided from the synthetic material development process providing apparatus 100, and has a structure and characteristics satisfying the specific condition in response to inputting the specific condition. A synthetic material and guide information for synthesizing the synthetic material may be provided.

In various embodiments, the user terminal 200 downloads, installs, and executes an application provided from the apparatus for providing a synthetic material development process 100, thereby providing a standardized system for learning an artificial intelligence model from the apparatus for providing a synthetic material development process 100. You can be provided with a method for automatically generating research record data and a method for providing a synthetic material development process using an artificial intelligence model.

To this end, the user terminal 200 may be a smartphone including an operating system capable of driving applications and including a display for UI output on at least a portion of the user terminal 200 . However, it is not limited thereto, and the user terminal 200 is a wireless communication device that ensures portability and mobility, and includes navigation, PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet ) terminal, smart pad (Smartpad), tablet PC (Tablet PC), such as all kinds of handheld (Handheld) based wireless communication device can be included.

In one embodiment, the external server 300 may be connected to the synthetic material development process providing apparatus 100 through the network 400, and the synthetic material development process providing apparatus 100 provides a standardized system for learning artificial intelligence models. Various information/data (e.g., data related to experimental materials, data related to research and experiment process, and result data form, etc.) Various information/data (e.g. standardized (or coded) research record data) can be stored and managed.

Here, the external server 300 may be a storage server provided separately outside the synthetic material development process providing apparatus 100, but is not limited thereto. Hereinafter, the hardware configuration of the synthetic material development process providing apparatus 100 will be described with reference to FIG. 2 .

2 is a hardware configuration diagram of a synthetic material development process providing apparatus using an artificial intelligence model according to another embodiment of the present invention.

Referring to FIG. 2 , in various embodiments, an apparatus 100 for providing a synthetic material development process (hereinafter referred to as “computing apparatus 100”) includes one or more processors 110 and a computer program (executed by the processor 110). It may include a memory 120 for loading 151, a bus 130, a communication interface 140, and a storage 150 for storing the computer program 151. Here, in FIG. 2, only components related to the embodiment of the present invention are shown. Therefore, those skilled in the art to which the present invention pertains can know that other general-purpose components may be further included in addition to the components shown in FIG. 2 .

The processor 110 controls the overall operation of each component of the computing device 100 . The processor 110 includes a Central Processing Unit (CPU), a Micro Processor Unit (MPU), a Micro Controller Unit (MCU), a Graphic Processing Unit (GPU), or any type of processor well known in the art of the present invention. It can be.

Also, the processor 110 may perform an operation for at least one application or program for executing a method according to embodiments of the present invention, and the computing device 100 may include one or more processors.

In various embodiments, the processor 110 may temporarily and/or permanently store signals (or data) processed in the processor 110 (RAM: Random Access Memory, not shown) and ROM (ROM: Read -Only Memory, not shown) may be further included. In addition, the processor 110 may be implemented in the form of a system on chip (SoC) including at least one of a graphics processing unit, RAM, and ROM.

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

The bus 130 provides a communication function between components of the computing device 100 . The bus 130 may be implemented in various types of buses such as an address bus, a data bus, and a control bus.

The communication interface 140 supports wired and wireless Internet communication of the computing device 100 . Also, the communication interface 140 may support various communication methods other than Internet communication. To this end, the communication interface 140 may include a communication module well known in the art. In some embodiments, communication interface 140 may be omitted.

The storage 150 may non-temporarily store the computer program 151 . When the method of automatically generating standardized research record data for learning an artificial intelligence model and the method of providing a synthetic material development process using an artificial intelligence model are performed through the computing device 100, the storage 150 stores the artificial intelligence model for learning. It is possible to store various types of information necessary to provide a method for automatically generating standardized research record data and a method for providing a synthetic material development process using an artificial intelligence model.

The storage 150 may be a non-volatile memory such as read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, or the like, a hard disk, a removable disk, or a device well known in the art. It may be configured to include any known type of computer-readable recording medium.

Computer program 151 may include one or more instructions that when loaded into memory 120 cause processor 110 to perform methods/operations in accordance with various embodiments of the invention. That is, the processor 110 may perform the method/operation according to various embodiments of the present disclosure by executing the one or more instructions.

In one embodiment, the computer program 151 may perform the steps of acquiring research record information about an experiment, processing the obtained research record information based on previously stored data related to an experiment, and standardizing the research record information using the processed research record information. It may include one or more instructions for performing a method of automatically generating standardized research record data for learning of an artificial intelligence model, including generating research record data.

In addition, the computer program 151 includes a guide for obtaining one or more conditions, extracting information about a synthetic material that satisfies the obtained one or more conditions, and synthesizing the synthetic material based on the extracted information about the synthetic material. It may include one or more instructions for performing a synthetic material development process providing method using an artificial intelligence model including extracting information.

Steps of a method or algorithm described in connection with an embodiment of the present invention may be implemented directly in hardware, implemented in a software module executed by hardware, or implemented by a combination thereof. A software module may include random access memory (RAM), read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, removable disk, CD-ROM, or It may reside in any form of computer readable recording medium well known in the art to which the present invention pertains.

Components of the present invention may be implemented as a program (or application) to be executed in combination with a computer, which is hardware, and stored in a medium. Components of the present invention may be implemented as software programming or software elements, and similarly, embodiments may include various algorithms implemented as data structures, processes, routines, or combinations of other programming constructs, such as C, C++ , Java (Java), can be implemented in a programming or scripting language such as assembler (assembler). Functional aspects may be implemented in an algorithm running on one or more processors. Hereinafter, a method for automatically generating standardized research record data for learning of an artificial intelligence model performed by the computing device 100 will be described with reference to FIGS. 3 to 15 .

3 is a flowchart illustrating a method of automatically generating standardized research record data for learning of an artificial intelligence model, in various embodiments.

Referring to FIG. 3 , in step S110 , the computing device 100 may obtain research record information about the experiment.

Here, the research record information may refer to information generated as a user conducts an experiment or research. For example, the research record information may include, but is not limited to, information about experimental materials, information about experimental equipment, information about experimental and research processes, and information about experimental and research results.

In various embodiments, the computing device 100 provides research record information (eg, information about experiment materials, information about experiment equipment, information about experiments and research processes) for experiments generated as a user conducts experiments or researches. , information on experiment and research results) can be directly input from the user.

In various embodiments, the computing device 100 may collect research information (eg, material, equipment, property, research process, experiment method, etc.) on experiments performed by major domestic and foreign bio-nano manufacturers by web crawling. However, it is not limited thereto, and any method for obtaining study record information can be applied.

In step S120, the computing device 100 may process the research record information obtained through step S110. For example, the computing device 100 extracts only key keywords (eg, keywords indicating materials, equipment, synthesis methods, etc.) from the research record information obtained through step S110 based on data related to pre-stored experiments. , It is possible to process a plurality of pieces of information included in the research record information into a form for generating standardized research record data.

In various embodiments, the computing device 100 may store pre-stored experiments and related data in conjunction with processed research record information. For example, the computing device 100 may store experimental material information generated by processing keywords related to experimental materials included in research record information and data related to experimental materials included in pre-stored experiment-related data in conjunction with each other, In addition, research and experiment process information generated by processing keywords related to research and experiment process included in research record information and data related to experiment materials included in pre-stored experiment related data can be linked and stored. Through this, it can be implemented to enable accurate sample information confirmation.

In addition, the computing device 100 may visualize (eg, image, graph, etc.) keywords related to the research result input from the user based on a preset result data format.

In step S130, the computing device 100 may generate standardized research record data (Labnote) using the research record information processed through step S120. For example, as shown in FIG. 23, the computing device 100 templates processed research record information (eg, experiment material information, research and experiment process information, and visualized research results), thereby standardized research records. data can be generated. Hereinafter, with reference to FIGS. 4 to 12, a method of processing research record information according to the form of the study record information will be described.

4 is a flowchart for explaining a method of automatically generating standardized research record data using research record information in a shortened form in various embodiments, and FIGS. It is a diagram showing the process of automatically generating standardized research record data.

Referring to FIGS. 4 to 7 , in step S210 , the computing device 100 may acquire research record information in the form of lines including one or more text sentences (eg, 21 in FIG. 5 ).

In various embodiments, the computing device 100 may directly receive research record information recorded in a line or handwritten form including one or more sentences from a user.

In various embodiments, the computing device 100 may collect research record information in the form of lines including one or more sentences from the outside through web crawling or the like. However, it is not limited thereto.

In step S220, the computing device 100 tokenizes each of a plurality of words included in the research record information in the form of a line based on previously stored experiment-related data (eg, DB of experimental research-related words extracted from papers and patents). (Tokenization) (e.g., 22 in FIG. 5). For example, the computing device 100 maximizes the likelihood of a corpus with a transformation algorithm of BPE (Byte Pair Encoding, an information compression algorithm widely used in natural language processing models) through a Wordpiece model (WPM). Each of a plurality of words can be tokenized in such a way.

Here, the meaning of a word to be tokenized may be one word, but is not limited thereto, and may be a phrase, clause, or sentence including two or more words in some cases. For example, in the case of "aqueous solution", it is composed of two words "aqueous" and "solution", but since the two words are combined to express one meaning, a plurality of words are classified / You can group them and tokenize each one.

In various embodiments, the computing device 100 may extract keywords related to experimental materials, keywords related to research and experimental processes, and keywords related to research results by analyzing research record information in the form of lines, and correspond to the extracted keywords. You can selectively tokenize only words that are

In various embodiments, the computing device 100 may extract a plurality of keywords from the research record information in the form of lines by analyzing (eg, OCR analysis, etc.) the research record information in the form of lines, and the extracted keywords and database By matching pre-stored data (e.g., keyword data by experimental material, keyword data by research and experiment process, and keyword data by research result), a plurality of extracted keywords are converted into keywords related to experimental materials and keywords related to research and experimental processes. and keywords related to research results.

That is, the computing device 100 determines that each of a plurality of keywords extracted from the research record information in the form of a line is based on pre-stored keyword data for each experimental material, keyword data for each research and experiment process, and keyword data for each research result. It is determined whether the keywords are pointing to, research and experimentation processes, or research results, and keywords may be classified into categories according to the determination result.

Here, the computing device 100 uses pre-stored keyword data for each experimental material, keyword data for each research and experiment process, and keyword data for each research result as learning data, and performs string analysis and keyword extraction operations using a pre-learned artificial intelligence model. It may be performed, but is not limited thereto.

In various embodiments, the computing device 100 extracts keywords related to experimental materials, keywords related to research and experimental processes, and keywords related to research results by analyzing research record information in the form of lines, and standardized research among the extracted keywords. If a keyword corresponding to information essential for generating record data is not extracted, the research record information in the form of lines may be repeatedly reanalyzed until the keyword is extracted.

At this time, the computing device 100 guides information to input information corresponding to the keyword when the corresponding keyword is not extracted even though the research record information in the form of a line is reanalyzed for a preset number of times (N times). can provide. Through this, in generating standardized research record data, it is possible to assist in inputting all important conditions and information that must be input without omission.

In step S230, the computing device 100 labels attribute information for each of a plurality of tokenized words through step S220 to generate processed research record information (eg, 23 in FIG. 5), thereby generating a plurality of Research record information in which attribute information is labeled for each word, that is, research record information that has been processed, may be generated. For example, the computing device 100 may determine properties of each of a plurality of words based on pre-stored experiment-related data (eg, DB of words related to experimental research extracted from papers and patents), and determine the properties according to the determined properties. According to the attribute information, each of a plurality of words may be labeled.

In various embodiments, labeling of tokenized words may be performed through named entity recognition.

In step S240, the computing device 100 may generate a research record graph using the processed research record information generated through step S230. In addition, the computing device 100 may generate standardized research record data using the research record graph.

In various embodiments, the computing device 100 may generate a research record graph using a pre-learned first model (Text2Labnote model).

Here, the pre-learned first model is a domain language learning model based on a transformer structure including a plurality of encoders as shown in FIG. 6 (eg, a BERT (Bidirectional Encoder Representations from Transformers) model) ) can be.

In addition, the pre-learned first model may be a model learned by using as learning data a plurality of research record information in the form of a plurality of lines and a plurality of research record graphs corresponding to each of the plurality of lines of research record information, but is not limited thereto don't

First, the computing device 100 may generate an adjacency matrix by using processed research record information including a plurality of words labeled with attribute information as input data of a pre-learned first model.

)을 이용하여 노드 i와 노드 j 간의 연결 관계를 판단할 수 있고, 복수의 노드 각각의 연결 관계에 따라 복수의 노드 사이를 연결하지 않거나 또는 연결함으로써 연구 기록 그래프를 생성할 수 있다.Then, the computing device 100 may generate a study record graph using the adjacency matrix. In this case, the research record graph may include a node corresponding to each of a plurality of words and an edge connecting each of the plurality of words, but is not limited thereto. For example, as shown in FIG. 7 , the computing device 100 calculates the value of a cell in row i of column j of an adjacency matrix (

) can be used to determine the connection relationship between node i and node j, and a research record graph can be created by not connecting or connecting a plurality of nodes according to the connection relationship of each of a plurality of nodes.

Thereafter, the computing device 100 may generate standardized research record data using the generated research record graph (eg, step S130). For example, the computing device 100 arranges a card 16 corresponding to each of a plurality of words on the UI 10 as shown in FIG. 23, but corresponds to each of a plurality of words according to the research record graph. The arrangement form and arrangement order of the cards 16 may be determined, but are not limited thereto.

In various embodiments, the computing device 100 may convert text-type research record information created by an electronic device as well as research record information manually written by a user into standardized research record data through the first model. . To this end, the computing device 100 provides a first model for learning text-type research record information and corresponding standardized data created by an electronic device, research record information in handwritten form, and standardized research record data corresponding thereto. The learned second model may be separately constructed.

8 is a flowchart illustrating a method of automatically generating standardized research record data through keyword input in various embodiments, and FIG. 9 illustrates a process of automatically generating standardized research record data through keyword input in various embodiments. it is a drawing

Referring to FIGS. 8 and 9, referring to FIG. 9, in step S310, the computing device 100 may provide a UI (eg, 10 of FIGS. 21 to 24) for inputting research record information, and the UI Through this, a keyword including at least one of keywords related to experimental materials, keywords related to research and experimental processes, and keywords related to research results may be input as a search word.

Here, since the research record information includes information on test materials, information on test equipment, and information on experiments and research processes, the UI 10 has a function of receiving information on test materials and information on test equipment. It may include a function of receiving input and a function of receiving information about experiments and research processes.

More specifically, first, as shown in FIGS. 21 and 22 , the computing device 100 displays the UI 14 for inputting information on experimental materials in response to the user selecting the first button 11 . can provide At this time, the computing device 100 may provide the UI 14 for inputting information on the experiment material in a pop-up form as the user selects the first button 11, but is not limited thereto.

Thereafter, the computing device 100 may receive a keyword related to the experimental material through the UI 14 for inputting information on the experimental material. For example, the computing device 100 provides keywords for nicknames, keywords for chemical elements, and product numbers through the UI 14 for inputting information on experimental materials. Keywords, keywords related to state, keywords related to solvent, keywords related to volume, keywords related to volume, and keywords related to concentration (additionally, a separate comment on the corresponding experimental material) may be input. .

Although not shown in the drawings, similarly to the above, the computing device 100 provides a UI for inputting information on the experiment equipment in response to the user selecting the second button 12, thereby providing information on the experiment equipment. It is possible to receive input of keywords related to experimental equipment through the UI for input, and in response to the user selecting the third button 13, by providing a UI for inputting information on experiments and research processes, experiments and Keywords related to the research process may be entered.

In step S320, the computing device 100 may select experiment-related data corresponding to one or more keywords input from the user based on pre-stored experiment-related data. For example, as shown in FIG. 9 , when the user inputs “acetic acid” as a keyword, the computing device 100 selects data including information on acetic acid from among pre-stored experiment-related data. may, but is not limited thereto.

In various embodiments, when there is a plurality of experiment-related data corresponding to one or more keywords input by the user, the computing device 100 provides data related to one or more experiments among the data related to the plurality of experiments based on the user's search history. may be selected and provided to the user as recommendation data, and any one of data related to one or more experiments provided as recommendation data may be selected from the user. For example, if there is a plurality of pieces of data including information on acetic acid, the computing device 100 selects the last searched (or selected) data or the most searched (or selected) data based on the user's search history among the plurality of data. ) may be provided as recommendation data.

In various embodiments, the computing device 100 performs a plurality of experiments based on search histories of a plurality of users researching the same research field as the user when there is a plurality of data related to experiments corresponding to one or more keywords input by the user. Data related to at least one experiment among data related to may be selected and provided to the user as recommended data.

In step S330, the computing device 100 may automatically complete research record information corresponding to one or more keywords input by the user using data related to the experiment selected through step S320. However, it is not limited thereto. For example, the computing device 100 may automatically arrange information included in user-selected data (experiment data corresponding to a keyword) on a template for generating standardized research record data.

Thereafter, the computing device 100 may generate standardized research record data using the automatically completed research record information. Here, the method of generating standardized research record data may be implemented in the same form as step S130 of FIG. 3, but is not limited thereto.

10 is a diagram illustrating a process of automatically generating standardized research record data using voice-type research record information in various embodiments.

Referring to FIG. 10 , in step S410 , research record information in the form of audio may be obtained from the user through the UI. For example, the computing device 100 may be connected to the user terminal 200 and the network 400, and may be connected to a microphone (eg, a smartphone, a laptop computer) or a user terminal 200 separately provided in the user terminal 200. It is possible to collect research record information in the form of voice input through the connected external microphone. However, it is not limited thereto.

In step S420, the computing device 100 may convert the research record information in a text form by natural language processing (NLP) on the voice type research record information acquired through step S410.

Here, various techniques are known for converting voice-type information into text-type information by natural language processing, and since these known techniques can be selectively applied, specific methods are not mentioned.

In addition, the computing device 100 may extract one or more keywords by analyzing research record information in text form. For example, the computing device 100 uses pre-stored keyword data for each experimental material, keyword data for each research and experiment process, and keyword data for each research result as learning data, and analyzes strings and extracts keywords using a pre-learned artificial intelligence model. An operation may be performed, but is not limited thereto.

In various embodiments, the computing device 100 operates the voice input device when the research record information in the form of voice input in step S410 is input in real time through the voice input device provided in the user terminal 200. One or more keywords may be extracted by analyzing (eg, natural language processing) a voice input device input in real time from the point of view.

At this time, the computing device 100 may perform verification of one or more keywords extracted in real time by reanalyzing all research record information in the form of voice input through the voice input device when the driving of the voice input device is terminated. For example, the computing device 100 reanalyzes the entire research record information in the form of audio to determine the validity of the extracted keyword in real time (eg, whether the extracted keyword is correctly extracted or whether the wrong keyword is extracted), It is possible to check whether there are keywords that could not be extracted.

In step S430, the computing device 100 may select experiment-related data corresponding to one or more keywords extracted through step S420 based on pre-stored experiment-related data. Here, the process of selecting experiment data corresponding to the keyword may be implemented in the same or similar form to step S320 of FIG. 8 , but is not limited thereto.

In step S440, the computing device 100 may automatically complete research record information corresponding to one or more keywords input by the user using data related to the experiment selected through step S430. Here, the process of automatically completing the research record information may be implemented in the same or similar form to step S330 of FIG. 8, but is not limited thereto.

In step S450, the computing device 100 may obtain equipment usage information from sensors provided in each of a plurality of equipment used in research and experiments, and materials from sensors provided in each of a plurality of materials used in research and experiments. It is possible to obtain usage information, and by recognizing the use of equipment and materials through the obtained equipment usage information and the acquired material usage information, it is possible to generate processed research record information.

For example, the computing device 100 may be connected to a sensor (eg, a position sensor, a motion sensor, etc.) provided in each equipment used for research and experimentation, and equipment use information based on sensor data collected from the corresponding sensor. (eg, whether or not the equipment is used and the type of equipment used) may be recognized (eg, when the location of a specific equipment moves more than a predetermined distance or when motion of a specific equipment is detected).

In addition, the computing device 100 may recognize and track material usage information (eg, whether or not the material is used, the type and amount of the material used, etc.) by utilizing RFID attached to a test material container or a reagent bottle. However, it is not limited thereto. Hereinafter, with reference to FIGS. 12 to 14, a process of generating research record information in the form of lines using standardized research record data will be described.

Thereafter, the computing device 100 may generate standardized research record data using the automatically completed research record information. Here, the method of generating standardized research record data may be implemented in the same form as step S130 of FIG. 3, but is not limited thereto. Hereinafter, with reference to FIGS. 12 to 14, a process of generating research record information in the form of lines using standardized research record data will be described.

11 is a flowchart illustrating a method of generating research record information in a line form using standardized research record data in various embodiments, and FIGS. 12 to 14 are, in various embodiments, using standardized research record data. It is a diagram showing the process of generating the research record information in the form of a line by writing.

Referring to FIGS. 11 to 14 , in step S510 , the computing device 100 may obtain a request for generating research record information in a line form including standardized research record data from the user.

In step S520, the computing device 100 may generate a research record graph by using the standardized research record data included in the research record information generation request in the form of lines acquired through step S510. Here, since the standardized research record data is generated based on the research record graph, the computing device 100 may generate a research record graph corresponding to the standardized research record data by performing inverse transformation. However, it is not limited thereto.

In step S530, the computing device 100 may generate research record information in the form of a line including one or more text sentences using the research record graph generated through step S520.

In various embodiments, the computing device 100 may generate research record information in the form of lines including one or more text sentences using the research record graph as input data of the pre-learned second model (Labnote2Text model).

Here, the pre-learned second model is a natural language processing model (eg, sequence to sequence natural language processing) based on a transformer structure including an encoder and a decoder, as shown in FIGS. 12 and 13 . model), but is not limited thereto.

In addition, the pre-learned second model is a graph data learning model using an attention mechanism, and the standardized research record data converted into a graph of research records can be used as a base model for the second model. As shown in FIG. 14, a GAP (Graph Attention Network) may be produced/constructed, but is not limited thereto.

In addition, the pre-learned second model may be a model learned by using as learning data a plurality of research record information in the form of a plurality of lines and a plurality of research record graphs corresponding to each of the plurality of lines of research record information, but is not limited thereto don't

In various embodiments, the computing device 100 may learn the first model and the second model through input/output data by interworking the first model and the second model. For example, the first model is a model that outputs a research record graph by taking the research record information in the form of a line, and the second model is a model that outputs the research record information in the form of a line by taking the research record graph as an input. , The output of the first model is used as the input of the second model, the output of the second model is used as the input of the first model, and each of the first model and the second model uses the input and output data as training data to generate the first model and the second model. 2 models can be trained.

Thereafter, the computing device 100 may provide the user with the research record information in the form of a line generated according to the above method.

In other words, the user can more easily manage and share research records by generating standardized research record data as he inputs research record information, and writes research records in the form of lines to write documents such as theses and patents. In the case of writing, there is an advantage in that it is possible to write more quickly and conveniently by converting the standardized research record data back into the research record information in the form of a line.

In various embodiments, when receiving a text conversion request for the first standardized research record data from the first user, the computing device 100 generates a first research record graph by graphing the first standardized research record data; , Research record information in the first line form is generated through the generated first research record graph and provided to the first user, but when at least a portion of the research record information in the first line form is modified by the first user, at least a portion of the research record information in the first line form is modified. The pre-learned second model may be re-learned by using the modified research record information in the form of the first line and the first research record graph as learning data. Through this, it is possible to improve text conversion performance of standardized research record data.

In various embodiments, computing device 100 may simulate experiments and studies using structured study record data. Hereinafter, it will be described with reference to FIG. 15 .

15 is a flowchart illustrating a method of performing a simulation based on standardized study record data, in various embodiments.

Referring to FIG. 15 , in step S610 , the computing device 100 may simulate an experiment according to experimental materials, experimental equipment, and experimental procedures included in the first standardized research record data.

First, the computing device 100 may build a cloud lab (unmanned automatic laboratory, virtual lab) to perform experimental simulation based on standardized research record data.

Here, the cloud lab is a platform-based research and experiment space utilizing a data sharing system such as the cloud, and specific research and experimentation based on external data (eg, standardized research record data provided from the computing device 100). It may mean a space where experiments can be virtually simulated or contents and results related to research and experiments can be shared with multiple users without time and space restrictions. Here, the method of building and using the cloud lab is a previously known technology, and even if a specific method of building and using the cloud lab is not described in this specification, those skilled in the art will be able to easily recognize it.

Thereafter, the computing device 100 generates coded research record data by converting the standardized research record data into a computer-readable code form in order to compute virtual research and experiments based on the standardized research record data. The generated coded research record data can be provided to the cloud lab, and the experiment based on the coded research record data can be simulated by operating the cloud lab. For example, the computing device 100 generates (codes) a first research record graph using the first standardized research record data, and uses the first research record graph to generate (encode) the first standardized research record data included in the first standardized research record data. Experiments can be simulated according to experimental materials, experimental equipment, and experimental procedures. However, it is not limited thereto.

Here, the coded research record data simulated through the cloud lab is automatically performed for all standardized research record data whenever standardized research record data is generated based on the research record information entered by the user, or If a simulation is requested for specific standardized research record data from the company, simulation may be performed only for the standardized research record data.

In step S620, the computing device 100 simulates the experiment based on the coded research record data through the cloud lab (eg, the experimental materials included in the coded research record data and the structure of the material derived according to the experiment process). and composition and properties of the material, etc.) can be matched and stored with the coded research record data used in the simulation.

In step S520, the computing device 100 simulates the experiment based on the coded research record data through the cloud lab (e.g., the experimental materials included in the coded research record data and the structure of the material derived according to the experimental process). and composition and properties of the material, etc.) can be matched and stored with the coded research record data used in the simulation.

At this time, the simulation result may be added to and stored in the research result information included in the coded research record data in the cloud lab, and the computing device 100 receives and stores coded research record data in which the simulation result is added and stored from the cloud lab. and can be managed. Hereinafter, with reference to FIGS. 16 to 20, a method for providing a synthetic material development process using an artificial intelligence model performed by a computing device will be described.

16 is a diagram for explaining a method of extracting a synthetic material and guide information for synthesizing the synthetic material through a first artificial intelligence model, in various embodiments.

Referring to FIG. 16 , the computing device 100 may extract and provide information on a synthetic material having a specific structure and characteristics and guide information for synthesizing the synthetic material. In FIGS. 16 to 18, the synthetic material is described as a candidate material for new drug development, but is not limited thereto.

Hereinafter, although new drug development is described as an example to explain the synthetic material development process according to the disclosed embodiment, the synthetic material development process according to the disclosed embodiment can be used for the development of synthetic materials in various fields other than new drugs.

For example, the synthetic material development process according to the disclosed embodiments can be utilized in various fields, such as developing synthetic materials for increasing heat resistance of semiconductor materials and developing synthetic materials for increasing electrical conductivity of battery materials, and is not limited to specific fields. don't

In step S710, the computing device 100 may generate and train a first artificial intelligence model.

Here, the first artificial intelligence model (or computational model, neural network, network function, neural network) is composed of one or more network functions, and one or more network functions are interconnected, which can generally be referred to as 'nodes'. It can consist of a set of computational units. These 'nodes' may also be referred to as 'neurons'. One or more network functions include at least one or more nodes. Nodes (or neurons) that make up one or more network functions can be interconnected by one or more 'links'.

In the first artificial intelligence model, one or more nodes connected through a link may form a relationship of a relatively input node and an output node. The concept of an input node and an output node is relative, and any node in an output node relationship with one node may have an input node relationship with another node, and vice versa. As described above, the input node to output node relationship can be created around the link. More than one output node can be connected to one input node through a link, and vice versa.

In a relationship between an input node and an output node connected through one link, the value of the output node may be determined based on data input to the input node. Here, a node interconnecting an input node and an output node may have a weight. The weight may be variable, and may be variable by a user or an algorithm so that the first artificial intelligence model performs a desired function. For example, when one or more input nodes are interconnected by respective links to one output node, the output node is set to a link corresponding to values input to input nodes connected to the output node and respective input nodes. An output node value may be determined based on the weight.

As described above, in the first artificial intelligence model, one or more nodes are interconnected through one or more links to form an input node and output node relationship in the first artificial intelligence model. Characteristics of the first artificial intelligence model may be determined according to the number of nodes and links in the first artificial intelligence model, a relationship between the nodes and links, and a value of a weight assigned to each link. For example, when there are two first artificial intelligence models having the same number of nodes and links and different weight values between the links, the two first artificial intelligence models may be recognized as different from each other.

Some of the nodes constituting the first artificial intelligence model may constitute one layer based on distances from the first input node. For example, a set of nodes having a distance of n from the first input node may constitute n layers. The distance from the first input node may be defined by the minimum number of links that must be passed through to reach the corresponding node from the first input node. However, the definition of such a layer is arbitrary for explanation, and the order of the layer in the first artificial intelligence model may be defined in a method different from that described above. For example, a layer of nodes may be defined by a distance from a final output node.

The initial input node may refer to one or more nodes to which data is directly input without going through a link in relation to other nodes among nodes in the first artificial intelligence model. Alternatively, in the relationship between nodes based on links in the first artificial intelligence model network, it may refer to nodes that do not have other input nodes connected by links. Similarly, the final output node may refer to one or more nodes that do not have an output node in relation to other nodes among nodes in the first artificial intelligence model. Also, the hidden node may refer to nodes constituting the first artificial intelligence model other than the first input node and the last output node. The first artificial intelligence model according to an embodiment of the present disclosure may have more nodes of an input layer than nodes of a hidden layer close to an output layer, and the number of nodes decreases as the number of nodes increases from the input layer to the hidden layer. It can be an artificial intelligence model.

The first artificial intelligence model may include one or more hidden layers. A hidden node of a hidden layer may use outputs of previous layers and outputs of neighboring hidden nodes as inputs. The number of hidden nodes for each hidden layer may be the same or different. The number of nodes of the input layer may be determined based on the number of data fields of the input data and may be the same as or different from the number of hidden nodes. Input data input to the input layer may be operated by a hidden node of the hidden layer and may be output by a fully connected layer (FCL) that is an output layer.

In various embodiments, the computing device 100 may build learning data for learning the first artificial intelligence model, and use the built learning data to perform supervised learning, unsupervised learning, The first artificial intelligence model may be trained using at least one of semi-supervised learning and semi-supervised learning.

Learning of the first artificial intelligence model is to minimize output errors. In the learning of the first artificial intelligence model, the learning data is repeatedly input into the first artificial intelligence model, the output of the first artificial intelligence model for the learning data and the error of the target are calculated, and the first artificial intelligence is directed to reduce the error. This is a process of updating the weight of each node of the first artificial intelligence model by backpropagating the error of the model from the output layer of the first artificial intelligence model to the input layer.

In the case of teacher learning, the learning data in which the correct answer is labeled is used for each learning data (ie, the labeled learning data), and in the case of comparative teacher learning, the correct answer may not be labeled in each learning data. That is, for example, learning data in the case of teacher learning regarding data classification may be data in which each learning data is labeled with a category. Labeled training data may be input to the first artificial intelligence model, and an error may be calculated by comparing an output (category) of the first artificial intelligence model with a label of the training data.

As another example, in the case of comparative history learning for data classification, an error may be calculated by comparing input learning data with an output of the first artificial intelligence model. The calculated error is back-propagated in the first artificial intelligence model in a reverse direction (ie, from the output layer to the input layer), and connection weights of each node of each layer of the first artificial intelligence model may be updated according to the back-propagation. The amount of change in the connection weight of each updated node may be determined according to a learning rate.

The computation of the first artificial intelligence model on the input data and the backpropagation of errors may constitute a learning cycle (epoch). The learning rate may be applied differently according to the number of iterations of the learning cycle of the first artificial intelligence model. For example, a high learning rate is used at the beginning of the learning of the first artificial intelligence model to ensure that the first artificial intelligence model quickly achieves a certain level of performance to increase efficiency, and a low learning rate is used in the later stage to increase accuracy. .

In the learning of the first artificial intelligence model, generally, the training data may be a subset of actual data (ie, data to be processed using the learned first artificial intelligence model), and thus, errors in the training data are reduced, but For real data, there may be learning cycles in which the error increases. Overfitting is a phenomenon in which errors for actual data increase due to excessive learning on training data. For example, a phenomenon in which the first artificial intelligence model, which learned a cat by showing a yellow cat, does not recognize that the cat is a cat when it sees a cat other than yellow may be a type of overfitting.

Overfitting can act as a cause of increasing the error of machine learning algorithms. Various optimization methods can be used to prevent such overfitting. In order to prevent overfitting, methods such as increasing training data, regularization, and omitting some nodes of a network in the process of learning may be applied.

In various embodiments, the computing device 100 may include attribute information (eg, structure, property information) of each of a plurality of synthetic materials and information for synthesizing each of a plurality of synthetic materials (eg, information for synthesizing each of a plurality of synthetic materials). The first artificial intelligence model may be trained using materials, equipment, information on synthesis methods, etc.) as learning data. However, the present invention is not limited thereto, and the computing device 100 may use yield information for each of a plurality of synthesis methods as learning data to learn the first artificial intelligence model.

In step S720, the computing device 100 may obtain one or more conditions from the user. Here, the one or more conditions may mean properties that the synthetic material should have (eg, it should have an effect on a specific disease, etc.), but is not limited thereto.

In step S730, the computing device 100 predicts a synthetic material having a structure and characteristics that satisfy one or more conditions by using the one or more conditions obtained through step S720 as input values of the first artificial intelligence model, and the predicted synthesis Result values including materials, equipment and synthesis methods for synthesizing the substance can be extracted.

For example, when the condition input from the user is "effective for disease z and yield is 90% or more", the computing device 100 inputs the condition to the first artificial intelligence model, so that the optimized disease z is effective. A synthetic material having a structure and characteristics can be predicted, a material for generating the synthetic material can be determined according to the predicted synthetic material, and a synthesis method capable of synthesizing the synthetic material with a yield of 90% or more It can be extracted as a result value by determining the equipment required to perform this.

The first artificial intelligence model includes attribute information (e.g., structure, property information) of each of a plurality of synthetic materials and information for synthesizing each of a plurality of synthetic materials (e.g., materials, equipment, and synthesis for synthesizing each of a plurality of synthetic materials). method) and yield information for each synthesis method as learning data in the form of “generating a specific synthetic substance effective for a specific disease with a yield of N% when synthesized with specific materials, specific equipment, and specific synthesis processes” As a model that learns the correlation between information for synthesizing property information synthetic materials and yield information, it is possible to generate a synthetic material that satisfies the condition and a corresponding synthetic material only by inputting a specific condition into the first artificial intelligence model. Guide information can be extracted.

In addition, the first artificial intelligence model is a model that learns the correlation between the property information of synthetic materials, information for synthesizing synthetic materials, and yield information, and, contrary to the above operation, specific materials, equipment, and synthesis methods are first Information on synthetic materials predicted to be synthesized according to specific materials, equipment, and synthesis methods can be extracted as input values of the artificial intelligence model as result values.

In various embodiments, the computing device 100 may provide the result value to the user, and in response to providing the result value, feedback information (eg, for synthesizing the predicted synthetic material) from the user. materials, equipment, and information on synthetic substances generated according to synthesis methods) may be input, and the first artificial intelligence model may be re-learned using the input feedback information as learning data.

At this time, the computing device 100 uses the information as learning data only when the synthetic material predicted to be generated based on the result value provided to the user is different from the synthetic material actually generated based on the feedback information input from the user. 1 AI models can be retrained.

In various embodiments, when the computing device 100 intends to perform scale-up on the synthetic material predicted through the first artificial intelligence model, the yield according to the plurality of synthesis methods is used as learning data. A synthesis method for synthesizing a synthetic material can be re-extracted through the first artificial intelligence model.

More specifically, the computing device 100 may generate a first artificial intelligence model that has learned yield information for each synthesis method and a first artificial intelligence model that has not learned yield information for each synthesis method, and generates yield information for each synthesis method. By using the derivation of the result value through the first unlearned artificial intelligence model as a default, priority is given to predicting synthetic materials that satisfy specific conditions, but in some cases, the development of synthetic materials is approved and approved. Only when the scale-up of the corresponding synthetic material is completed, the synthesis method may be re-extracted through the first artificial intelligence model that has learned the yield information for each synthesis method.

17 is a diagram for explaining a method of setting test conditions for a synthetic material through a second artificial intelligence model, in various embodiments.

Referring to FIG. 17 , the computing device 100 may set conditions for a test (eg, preclinical test or clinical test) of a specific synthetic material.

In step S810, the computing device 100 may generate and train a second artificial intelligence model.

Here, the structure and operation form of the second artificial intelligence model may be the same as that of the first artificial intelligence model, but is not limited thereto.

In various embodiments, the computing device 100 may provide information about a plurality of synthetic materials and information about tests (eg, conditions set to conduct a preclinical test or clinical test on a plurality of synthetic materials, and test results according to the set conditions). etc.) can be used as learning data to learn the second artificial intelligence model.

In step S820, the computing device 100 may obtain information about a synthetic material to be subjected to a preclinical or clinical test.

In step S830, the computing device 100 may use the information on the synthetic material acquired through step S820 as an input value of the second artificial intelligence model to extract a result value for setting conditions for testing the synthetic material. .

The second artificial intelligence model is a model learned using information and test information on a plurality of synthetic materials as learning data, that is, a model that learns test results according to test conditions for each learning material, information on a specific learning material Test conditions that can derive optimal test results can be extracted only by the operation of inputting to the second artificial intelligence model.

In addition, since the second artificial intelligence model is a model that learns the test results according to the test conditions for each learning material, the test result of the synthetic material is predicted by using the specific conditions for the test of the synthetic material as the input value of the second artificial intelligence model Values can be extracted as result values.

In various embodiments, the computing device 100 may re-learn the second artificial intelligence model by using the result of testing the learning material according to the conditions set according to the above method as learning data.

18 is a view for explaining a method of deriving a patient and dosage suitable for a new drug when developing a new drug through a third artificial intelligence model, in various embodiments.

Referring to FIG. 18 , the computing device 100, through the operations of FIGS. 17 and 18 , when a new drug development approval and completion process for a specific synthetic material is completed, when a new drug containing a specific synthetic material is developed, a patient suitable for the corresponding new drug and suitable dosages can be extracted.

In step S910, the computing device 100 may generate and train a third artificial intelligence model.

Here, the structure and operation form of the third artificial intelligence model may be the same as those of the first artificial intelligence model and the second artificial intelligence model, but are not limited thereto.

In various embodiments, the computing device 100 uses information on a plurality of test subjects, information on a plurality of synthetic materials developed as new materials, and test results of each of a plurality of synthetic materials developed as new materials as learning data to obtain a third AI models can be trained.

For example, the computing device 100 uses information about a plurality of patients, information about a plurality of synthetic materials developed as new drugs, and preclinical test or clinical test results of each of a plurality of synthetic materials developed as new drugs as learning data. A third artificial intelligence model can be trained.

In step S920, the computing device 100 may acquire information about the synthetic material for which the test has been completed.

For example, the computing device 100 may obtain information about synthetic materials for which new drug development has been approved and completed.

In step S930, the computing device 100 uses the information on the synthetic material obtained through step S920 as an input value of the third artificial intelligence model, and results in information about a target suitable for using the new material including the synthetic material and the amount used. values can be extracted.

For example, the computing device 100 uses the information on the synthetic material obtained through step S920 as an input value of the third artificial intelligence model, and provides information on a patient suitable for administering a new drug containing the synthetic material and a dosage amount as a result. values can be extracted.

In an embodiment related to drug development, the third artificial intelligence model provides information on a plurality of patients, information on a plurality of synthetic substances developed as new drugs, and preclinical test or clinical test results of each of a plurality of synthetic substances developed as new drugs. It is a model learned with learning data, that is, a model that learns the correlation between the patient's attributes (age, gender, underlying disease, etc.) and the effect of the new drug, and the correlation between the dosage and the effect of the new drug. Provides information on specific learning substances 3 It is possible to extract which patients are suitable for the new drug and what the appropriate dosage is just by inputting the motion to the artificial intelligence model.

In addition, the computing device 100 not only directly guides the development of new drugs through the artificial intelligence model as described above, but also guides indirectly, such as providing helpful literature for research fields or matching researchers studying similar fields. can do.

More specifically, as shown in FIG. 19, the computing device 100 analyzes pre-generated standardized research record data corresponding to a specific user (researcher) to provide important detailed conditions (substances used as materials, equipment, synthesis). process, etc.) can be extracted, and data optimized for research contents can be provided by searching and providing unstructured data (eg, papers, patents) corresponding to the extracted detailed conditions.

In addition, as shown in FIG. 20, the computing device 100 provides a function of posting information (eg, personal information, research field, and portfolio) for a specific user (researcher) or sharing research paper data and research contents. Based on this, it is possible to provide a research sharing and collaboration platform that provides a function to match other researchers studying similar fields.

A method for automatically generating standardized research record data for learning the aforementioned artificial intelligence model and a method for providing a synthetic material development process using the artificial intelligence model have been described with reference to the flow chart shown in the drawings. For a brief description, the method for automatically generating standardized research record data for learning of an artificial intelligence model and the method for providing a synthetic material development process using an artificial intelligence model have been illustrated and described as a series of blocks, but the present invention is based on the order of the blocks. Without limitation, some blocks may be performed concurrently or in a different order than shown and performed herein. In addition, new blocks not described in the present specification and drawings may be added, or some blocks may be deleted or changed. Hereinafter, the UI 10 provided by the computing device 100 will be described with reference to FIGS. 21 to 24 .

21 to 24 are diagrams illustrating forms of user interfaces applicable to various embodiments by way of example.

Referring to FIGS. 21 to 24 , the UI 10 includes a first button 11 for inputting information on experimental materials, a second button 12 for inputting information on experimental equipment, and experiments and research processes. It may include a third button 13 for inputting information about.

In response to a selection of at least one of the first button 11, the second button 12, and the third button 13 by the user, the computing device 100 provides information about the experiment material and information about the experiment equipment. A UI for receiving any one of information and information on experiments and research processes may be provided. For example, when the user selects the first button 11 on the UI 10 (eg, inputting a mouse click), the UI 14 for inputting information on experimental materials in a pop-up form may be provided. .

In addition, although not shown in the drawings, as described above, a UI for inputting information on experimental equipment may be provided in response to the user selecting the second button 12, and the user may use the third button 13 In response to selecting ), a UI for inputting information on experiments and research processes may be provided. At this time, each of the UIs may be output in a pop-up form similarly to the UI 14 for inputting information on the experiment material.

The computing device 100 includes a UI 14 for inputting information on experimental materials, a UI for inputting information on experimental equipment, and experimental materials input through the UI for inputting information on experiments and research processes; It is possible to create experimental material information, research and experiment process information, and research result information by automatically completing and interlocking keywords for research and experiment processes and research results, and UI (10) according to each generated information according to a preset template. can be placed and registered on the For example, each piece of information may be implemented in the form of an individual box as shown in FIG. 23, but information belonging to the same category may be arranged in the same column.

At this time, the card including specific information disposed (registered) on the UI 10 may be freely copied and disposed in another area on the UI 10 according to a user's request. For example, the computing device 100 requests copying of the first card 16 including information on a specific experimental material (eg, selection of the first card 16, a preset key (eg, Ctrl) + drag input), a second card 16', which is a copy of the first card 16, can be created, displayed, and registered in an area designated by the user (eg, an area where the mouse pointer is located after a drag input). there is.

In addition, a plurality of cards including specific information disposed (registered) on the UI 10 may be freely mixed according to a user's request. For example, the computing device 100 may request a combination of a first card including information on a first test material and a second card including information on a second test material (eg, selecting and selecting a first card). In response to obtaining an action of dragging and placing the second card on top of the second card, selecting and dragging the second card and placing it on the first card, etc.), combining the information included in the first card with the information included in the second card A new third card (eg, including information on the first experimental material and information on the second experimental material) may be created.

When all information is placed (registered) on the UI 10 and generation of the finally standardized research record data is completed, the computing device 100 provides information about the standardized research record data through the summary information providing UI 17. Summary information can be provided.

Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art to which the present invention pertains can be implemented in other specific forms without changing the technical spirit or essential features of the present invention. you will be able to understand Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

100: synthetic material development process providing device (or computing device)
200: user terminal
300: external server
400: Network

Claims (10)

In a method performed by a computing device,
obtaining one or more conditions; and
Extracting information about a synthetic material that satisfies the obtained one or more conditions, and extracting guide information for synthesizing the synthetic material based on the extracted information about the synthetic material,
Method for providing synthetic material development process using artificial intelligence model. According to claim 1,
The step of extracting the guide information,
Generating a first artificial intelligence model;
Attribute information including structure and characteristic information of each of a plurality of synthetic materials, information for synthesizing each of the plurality of synthetic materials - information for synthesizing each of the plurality of synthetic materials is information for synthesizing each of the plurality of synthetic materials Learning the generated first artificial intelligence model using - as learning data; and
A material for predicting a synthetic material having a structure and characteristics satisfying the one or more acquired conditions by using the obtained one or more conditions as an input value of the learned first artificial intelligence model, and synthesizing the predicted synthetic material. , Including the step of extracting the result value including the equipment and synthesis method,
Method for providing synthetic material development process using artificial intelligence model. According to claim 2,
The step of learning the generated first artificial intelligence model,
In response to providing the extracted result value to the user, feedback information from the user - the feedback information is information about materials, equipment, and synthesized materials produced according to the synthesis method for synthesizing the predicted synthetic material. Including - Receiving an input; and
Re-learning the learned first artificial intelligence model using the input feedback information as learning data,
Method for providing synthetic material development process using artificial intelligence model. According to claim 2,
The step of extracting the guide information,
Extracting information on synthetic materials predicted to be synthesized according to the specific materials, equipment, and synthesis methods as input values of the learned first artificial intelligence model as a result value of specific materials, equipment, and synthesis methods. including,
Method for providing synthetic material development process using artificial intelligence model. According to claim 2,
The step of learning the generated first artificial intelligence model,
Learning the learned first artificial intelligence model using yields according to a plurality of synthesis methods as learning data;
The step of extracting the result value is,
When scaling-up of the predicted synthetic material is to be performed, synthesizing the predicted synthetic material through a first artificial intelligence model learned using the yield according to the plurality of synthesis methods as learning data Including the step of re-extracting the synthesis method for
Method for providing synthetic material development process using artificial intelligence model. According to claim 1,
The step of extracting the guide information,
Generating a second artificial intelligence model;
Information on a plurality of synthetic materials and information on tests - the information on tests includes conditions set for conducting preclinical tests or clinical tests on the plurality of synthetic materials, and test results under the set conditions - learning the generated second artificial intelligence model using learning data; and
Setting conditions for testing the synthetic material by using the information about the synthetic material as an input value of the learned second artificial intelligence model,
The step of learning the generated second artificial intelligence model,
Re-learning the learned second artificial intelligence model using the results of testing the synthetic material according to the set conditions as learning data,
Method for providing synthetic material development process using artificial intelligence model. According to claim 6,
The step of extracting the guide information,
Extracting a predicted value of the test result of the synthetic material as a result value by using a specific condition for the test of the synthetic material as an input value of the learned second artificial intelligence model,
Method for providing synthetic material development process using artificial intelligence model. According to claim 1,
The step of extracting the guide information,
Generating a third artificial intelligence model;
Learning the generated third artificial intelligence model using information on a plurality of test subjects, information on a plurality of synthetic materials developed as new materials, and test results of each of the plurality of synthetic materials developed as new materials as learning data ; and
When developing a new material containing the synthetic material by using the information on the synthetic material as an input value of the third artificial intelligence model, the new material containing the synthetic material is suitable for use and information on the amount of use is extracted as a result value Including steps,
Method for providing synthetic material development process using artificial intelligence model. processor;
network interface;
Memory; and
A computer program loaded into the memory and executed by the processor,
The computer program,
instructions that obtain one or more conditions; and
Including instructions for extracting information on a synthetic material that satisfies the obtained one or more conditions, and extracting guide information for synthesizing the synthetic material based on the extracted information on the synthetic material.
Device for providing synthetic material development process using artificial intelligence model. Combined with a computing device,
obtaining one or more conditions; and
Extracting information on a synthetic material that satisfies the obtained one or more conditions, and extracting guide information for synthesizing the synthetic material based on the extracted information on the synthetic material using an artificial intelligence model comprising the step of extracting Stored in a computer-readable recording medium to execute the synthetic material development process providing method,
A computer program recorded on a computer-readable recording medium.

Images