WO2021157832A1

WO2021157832A1 - System and method for mediating project

Info

Publication number: WO2021157832A1
Application number: PCT/KR2020/017100
Authority: WO
Inventors: 강성일
Original assignee: 강성일
Priority date: 2020-02-05
Filing date: 2020-11-27
Publication date: 2021-08-12
Also published as: KR20230079306A

Abstract

The present invention relates to a method and a system for mediating a cloud-based incremental intelligence improvement deep-learning project by using system collaboration with an expert device and, more specifically, generates a statistics and probability DB and a machine learning DB through statistical and probability calculation and machine learning from the contents of a project participant recruitment notice posted by a project organization and the project results received by project participants, so as to find, from among the project participants, a project participant suitable for a new project participant recruitment notice by means of a statistics and probability prediction model and a machine learning inference model, thereby informing the corresponding project participant and the project organization, and thus quickly sends project information suitable for the project participant to the project participant, so that the project organization quickly finds a project participant suitable for a planned project.

Description

Project Brokerage System and Method

The present invention relates to a method and system for highly intelligent project brokerage by gradually improving intelligence on a cloud-based basis by collaboration with expert devices and systems.

A project refers to a group of tasks such as research, business, test, experiment, test, etc. carried out to achieve a certain purpose within a certain period of time. One project can be carried out in a way that performs various requirements under certain constraints such as a set period, an allotted amount, and an input manpower. As an example, a project may include a clinical research project. In connection with the progress of the clinical research project, an app is provided so that clinical trial project participants can read recruitment announcements and participate in the clinical trial project.

However, since clinical trial project participants have to read recruitment announcements one by one in order to find a project suitable for them, and carefully review and understand the contents of the recruitment announcements, there is a serious problem of wasting time.

In addition, since the project organization (or project announcer) that recruits clinical trial participants has to wait without an appointment until clinical trial project participants apply for participation, it is difficult to proceed with the clinical trial project at the desired time, or it is difficult to recruit participants. Failure to do so causes the project to end prematurely.

As an example, according to the paper of '2016 Clinical Trial Recruitment Improvement Plan Study: From the Sponsor's Perspective_Young-Hwan Yang', a domestic research data, 35.3% of the subjects experienced early termination of the project due to the failure or delay in subject recruitment, and the clinical trial was not completed according to the planned schedule. 41.2% of the cases did not progress. In addition, according to the overseas research data IBM Global Business Services, it is reported that more than 75% of clinical trial projects failed to recruit participants within the set deadline. As described above, recruiting project participants within a set deadline is an important issue that can determine the success or failure of a project.

In accordance with the above-mentioned problem recognition, the clinical trial information integration server provides a questionnaire to the clinical trial project participant terminal, compares and judges the eligibility questionnaire among the questionnaires prepared by the clinical trial project participant, and notifies the eligibility to the clinical trial project participant terminal A test project information matching method has been developed. (Refer to KR Patent Publication 2018-0133301) However, this conventional method generates a questionnaire and provides it to the clinical trial project participant terminal, compares and determines the eligibility among the clinical trial project participant questionnaire to determine whether or not the patient is eligible to the clinical trial project participant terminal There is a drawback that it is difficult to use in reality because it includes a cumbersome process of notifying.

The present invention has been proposed to solve the problems of the prior art as described above, and more specifically, by comparing and analyzing the recruitment announcement contents of the project organization and the project participation history of the project participants, to detect a suitable project participant, and the degree of suitability It is to provide a project mediation method and system that allows project participants to easily and quickly perform recruitment by providing the project organization and project participant terminals.

The project brokerage system for achieving this purpose collects the project announcement contents from the project agency device that publishes the project notification contents through the website, the project agency device, configures the project DB, and proceeds with the project in the project DB Extract and store the project key information related to the project, collect project participant information from the project participant device related to the participant who will participate in the project, or collect project results from the project organization device related to the previously participated project. By applying an inference model according to machine learning to the main project information and the main project participant information while selecting the subject based on statistics and probability calculations and predictions for the project main information and the project participant main information After selecting the final target by screening and performing inspection and correction of the selected subjects, the project brokerage device that provides information on the final target to at least one of the project organization device and the project participant device, the project When registering as a participant in the mediation device and receiving new project recommendation information from the project mediation device, the project participant device that transmits a new project participation acceptance message to the project mediation device according to user input, an expert performing the inspection and correction device, in relation to the project participant recruitment, comprising a payment server supporting at least one payment service of cost processing of the project institution device, cost processing of the project participant device, and fee deduction of the project brokerage device do it with

A project brokerage device according to an embodiment of the present invention includes a crawler that collects project participation recruitment announcements from a website operated by a project agency device, a data management cloud server that receives the data collected by the crawler and stores it in a project DB, the above The project participant main information DB that stores the main information of the participants who will participate in the project, the statistics and probability prediction module that performs statistical and probability calculation and prediction on the main project information, and the project participant main information stored in the project participant main information DB A primary subject selection module for selecting a primary subject by the statistics and probability prediction module as a target, a machine learning inference module for performing machine learning inference on the main project information, the primary subject among the project participant main information A secondary subject selection module that selects secondary subjects from the rest of the project participants except for the project participants selected as and a correction module, a final target selection and notification module for selecting a final target by combining all of the primary and secondary subjects, and notifying the project participants and the project organization of the selection results.

In addition, the project mediation device may include the age, gender, height, weight, medical history, family history, blood type, address, location, career, educational background, specialization, field of interest, self-introduction information, certification of the participant in relation to the project participant device. A member login and information management module that supports input and management of at least one of information, project progress history and results, hobbies, specialties, preferences, food, clothing, and shelter information, and supports confirmation of the result history of subject selection, the contents of the above project participant recruitment notice and corresponding information An information collection module for receiving project participant information actually participating in clinical trials from the project institution device and storing it in the project participant DB, the project participant main information extracting the project participant main information from the project participant DB and storing it in the project participant main information DB Information extraction module, a statistics and probability calculation module for generating statistics and probability DB by statistical and probability calculation from the project participant main information DB, a machine learning module for generating a machine learning DB by machine learning from the project participant main information DB can be composed of

The project brokerage method according to an embodiment of the present invention performs statistical and probability calculation and prediction for pre-stored project main information based on the pre-stored project participant main information and the announcement provided by the project organization device in relation to the project participant device. performing the first target selection through, after the first target selection, the second target selection based on machine learning for the main project information and the remaining main information of the participants except for the first subject in the project participant main information performing, providing the first subject and the second subject selection result to an expert device, performing inspection and correction by the expert device, selecting a final subject based on the inspection and correction; and providing the final target selection information to at least one of the project organization device and the project participant device.

In addition, the project mediation method generates a project DB by crawling the content of the recruitment announcement for new project participation posted on a website operated by the project institution device, and extracts the main project information from the project DB to extract the main project information DB generating, updating the results according to the inspection and correction to the main project participant information, generating statistics and probability DB by statistical and probability calculation based on the updated project participant main information, and the updated project participant main information The method may further include at least one of the steps of generating a machine learning DB by machine learning based on the information.

In addition, the project mediation method includes, through a payment server, deducting a fee according to the agreement with the project participant device, and paying the cost to the project participant device participating in the project according to the compensation information described in the project main information. may include more.

Finally, the project mediation method may further include the step of paying a certain fee to the project mediation device whenever the project participant device participates in a project through a payment server.

The project brokerage method and system of the present invention described above exhibits the following useful results.

1) The project organization device learns the results of the previous project to make a model, and when a new project announcement is made, the project participant device suitable for the announcement is selected and the project participant device and the project agency device are notified to proceed with the project. The project can be carried out in a short time without the time delay related to the recruitment of participants.

2) In the project organization device, the project participants suitable for the newly announced project contents can be notified quickly, so the overall project planning and operation can be designed in a short period of time.

3) By registering project participant information, project participants can quickly receive notification of a suitable project for themselves without searching for project information.

4) The system performance is guaranteed by the inspection and calibration of the expert device, and the prediction and inference accuracy of the system can be improved by allowing the inspection and calibration results of the expert device to be immediately reflected in the system.

5) Project participant devices can be notified of selection through various communication interfaces, such as SMS, PUSH, SNS (KakaoTalk, Facebook), and instance app link, available without the need to install a separate app, avoiding unnecessary communication service subscription. there is.

1 and 2 are an example of a configuration diagram of a project brokerage system according to an embodiment of the present invention.

Figure 3 is a view showing an example of a project brokerage device operating method according to an embodiment of the present invention.

4 is a diagram illustrating an example of a DB creation and operation method of a project brokerage device according to an embodiment of the present invention.

5 is a diagram illustrating an example of a DB update and operation method of a project brokerage device according to an embodiment of the present invention.

6 is a diagram illustrating a signal flow between each component of the project mediation system according to an embodiment of the present invention.

7 is a view showing an example of the configuration of a project brokerage device according to an embodiment of the present invention.

8 is a diagram illustrating an example of a method for selecting project participants of a project brokerage device according to an embodiment of the present invention.

Hereinafter, preferred embodiments in which the object of the present invention can be specifically realized will be described in detail with reference to the accompanying drawings. In describing the present embodiment, the same names are used for the same components, and an additional description thereof will be omitted.

1 and 2 are an example of a configuration diagram of a project brokerage system according to an embodiment of the present invention. Here, A, B, C, D, E, and F of FIG. 1 are respectively connected to A', B', C', D', E', and F' of FIG. 2 and are expressed as one figure.

1 and 2 , the project mediation system 10 of the present invention is a project agency device 200 , a project participant device 300 , a project mediation device 100 , an expert device 400 , and a payment server 500 . ) may be included. The project mediation system 10 of the present invention is based on cloud-based gradual intelligence enhancement deep learning by the expert device 400 and system collaboration, and the project participant device 300 suitable for the announcement content By retrieving and notifying them, you can support rapid progress in project operation and management of appropriate project participants.

The project agency device 200 may design a project, and after recruiting the project participant device 300 through the project intermediary device 100 in relation to the designed project progress, the project may be carried out. For example, when the project is a specific clinical trial, the project institution device 200 provides a clinical trial announcement to the project mediation device 100, and through the project mediation device 100, the project participant device suitable for the clinical trial announcement ( 300) can be collected. The project organization device 200 is a storage device for storing the designed project, a communication interface for communicating with the project mediation device 100, the storage of the project and the delivery of the project announcement content, and the project participant device participating in the project ( 300) may include a processor for performing information management.

As an example, the project organization device 200 may register the project announcement related to the project 201 progress on the website 202 (or web page) operated by the project 201 after designing the project 201 . The project organization device 200 may operate by allocating each web site 202 to each of the projects 201 . In response, the project brokerage device 100 crawls the new project participation recruitment announcement posted on each website 202 with the crawler 109, and the crawled contents are the data management cloud server 110 through the project DB (111) is generated.

As another example, the project agency device 200 may access the website (or web page) provided by the project intermediary device 100, and write the project announcement content on the website. The contents of the project announcement entered through the website include, for example, the purpose of the project (research), the main information of the participants (age, gender, height, weight, disease history, family history, blood type, address, location, work experience, academic background, specialized field) , fields of interest, self-introduction information, certification information, project progress history and results, hobbies, specialties, preferences, food, clothing, and shelter information, etc.) information, predictable side effects, application and inquiry information, and the like. When the project agency device 200 writes the project announcement content through the website, the project mediation device 100 searches for the project participant device 300 based on the entered project announcement content, and can suggest there is.

The project organization device 200 may pay a fee related to the project participant device 300 recruitment through the payment server 500 . For example, the project institution device 200 pays a cost 1:1 with the project participant devices 300 participating in the project through the payment server 500 , and the project participant device 300 to the project intermediary device 100 . A separate fee may be paid for recruiting them. Alternatively, the project agency device 200 may execute the total cost related to the recruitment of the project participant devices 300 to the project mediation device 100 through the payment server 500 at the time the project participants complete participation. . In this case, the project intermediary device 100 may provide a cost excluding a fee among the total received costs to each of the project participant devices 300 through the payment server 500 . Here, the payment server in which the project agency device 200 pays a cost to the project mediation device 100 and the payment server in which the project mediation device 100 pays costs to the project participant devices 300 are the same financial company's payment server or different financial payment servers. As another example, when the project organization device 200 pays a cost related to the project progress to each of the project participant devices 300 participating in the project, the project participant device 300 pays a fee related to the project brokerage to the payment server 500 It is also possible to separately pay the project brokerage device 100 through.

The project participant device 300 receives the project participation recommendation information from the project brokerage device 100 in relation to the project that the project organization device 200 proceeds, and after checking the project announcement content, determining whether to participate in the project It may include the participant's terminal. The project participant device 300 may include various electronic devices such as a computing device or a mobile electronic device. The project participant device 300 may access a website operated by the project mediation device 100 and perform member registration and information entry according to the operation of the participant. In this process, the project participant device 300 provides project participant main information (eg, age, gender, height, weight, medical history, family history, blood type, address, location, career, academic background, specialization, field of interest, self-introduction) information, certification information, project progress history and results, hobbies, special skills, preferences, food, clothing, and shelter information), or the project 201 name of the project organization device 200 may be entered. Project participant main information written by the project participant device 300 may be used when selecting a project participant. The project participant device 300 may pay a membership fee to the project mediation device 100 through the payment server 500 in relation to the use of project mediation, and receive a project mediation service. Alternatively, the project participant device 300 may provide a predetermined fee to the project brokerage device 100 through the payment server 500 every time the project 201 participates. In relation to cost processing, the project participant device 300 may select one of various items provided to the project intermediary device 100 (eg, payment of membership fees or payment of fees).

The expert device 400 may contribute to the final target selection by providing a professional opinion in relation to the participant selection of the project mediation device 100 . The expert device 400 may provide information on the suitability of project announcement content and project participant main information. In this regard, the expert device 400 receives the project announcement contents, the project participant main information, and the selection grounds from the project mediation device 100, and checks and corrects the results of the received information to the project mediation device 100. can provide In this regard, the expert device 400 may provide an expert operating environment related to verification and calibration, or may build a database of expert knowledge related to verification and calibration of the expert. The expert apparatus 400 may generate project participant suitability inspection information and correction information for a project based on the built-up expert knowledge database, and may provide the inspection and correction information to the project brokerage apparatus 100 . The expert knowledge database may be updated according to the project progress results of the project participant devices 300 . In this regard, the project intermediary device 100 or the project agency device 200, with the consent of the project participant device 300, without exposing personal information (eg, providing only body information and results related to project progress), the project The result may be provided to the expert device 400 to support updating the expert knowledge database. The expert device 400 may be included as a part of the project mediation device 100, or may be provided and operated as a separate configuration. The expert device 400 may receive a predetermined cost from the payment server 500 according to the inspection and correction according to the request of the project mediation device 100 from the project mediation device 100 .

The payment server 500 is at least one of the cost payment of the project agency device 200 , the cost processing of the project brokerage device 100 , the cost processing of the project participant device 300 , and the cost processing of the expert device 400 . can represent one. The payment server 500 may include a bank server or a card company server operated by at least one financial company. The payment server 500 may support, for example, account transfer of the project institution device 200 or the project brokerage device 100 . In addition, the payment server 500 may support the fee processing of the project brokerage device 100 and payment of the cost of the project participant device 300 . In this regard, the payment server 500 may be entrusted with processing costs and costs related to the project progress from the project institution device 200, and may pay the project participation cost and the project brokerage service cost according to the project progress status. . For example, the payment server 500 receives the cost associated with the project 201 progress, the project mediation device 100 information, the project participant device 300 information from the project institution device 200, and the project participant receives the corresponding project When participating in 201 , or upon completion of the project, or during project Jinhae, while paying the project participation cost to the project participant device 300 , a related fee may be provided to the project brokerage device 100 . In relation to the payment of the cost of the expert device 400 , the payment server 500 acts on behalf of the payment of the cost of the project mediation device 100 , or when the project agency device 200 designates the expert device 400 , the project agency device It is possible to substitute the payment for the expert device 400 of (200).

The project mediation device 100 performs the management of the project organization device 200 and the project participant devices 300, and when a new project announcement occurs, the search and selection, selection, and notification of the appropriate project participant device 300 is processed. can do. To explain this in more detail,

The project mediation device 100 collects the project announcement content presented by the project organization device 200, searches for and selects the project participant device 300 suitable for the project announcement content, and optionally selects the expert device 400 After undergoing inspection and correction, the final target can be selected. The project mediation device 100 provides project participation recommendation information to the project participant device 300 corresponding to the selected final target, and depending on whether the project participant device 300 is accepted, the project participant device 300 information ( For example, participant key information) may be provided to the project organization device 200 . In this process, the project mediation device 100, from the project participant recruitment announcement content and the project participant device 300 previously experienced by the project agency device 200, statistics and probability calculation and machine learning from the previous experience. Statistics and probability DB and machine learning DB are created by the statistics and probability prediction model and machine learning inference model to find the project participant device 300 suitable for the announcement of new project participation recruitment and the corresponding project participant device 300 and the project organization By notifying the device 200 , the project organization device 200 may support the project participant device 300 to quickly proceed with the project.

In this regard, the project mediation device 100 crawls the new project participation recruitment announcement posted on each website 202 with the crawler 109, and the crawled information is transferred to the project DB through the data management cloud server 110. (111) can be stored. The project main information extraction module 113 extracts project main information from the project DB 111 to generate the project main information DB 112 . The main project information may include, for example, the purpose of the project (research), the subject of participation, the characteristics of the drug for the project, the main selection criteria, the number of people recruited, the project (research) method, the information provided at the time of participation, predictable side effects, applications and inquiries. can

The project mediation device 100 uses the project participant member login and information management module 120, and the participant's personal information (age, gender, height, weight, disease history, family history, blood type) related to the project participant device 300 , address, location, career, academic background, specialization, interests, self-introduction information, certification information, project progress and results, hobbies, specialties, preferences, food, clothing, and shelter information, etc.) can confirm. In this regard, the project participant member login and information management module 120 provides a website through which the project participant device 300 can access and input personal information, and perform personal information storage management with participant consent. there is. The project participant member login and information management module 120 may store information of the subscribed project participant device 300 in the project participant DB 102 .

As another example, the project mediation device 100 may include an information collection module 101 . The information collection module 101 may receive the project participant recruitment announcement content and the project participant information who actually participated in the clinical trial from the project institution device 200 and store it in the project participant DB 102 . The project participant main information extraction module 103 extracts the project participant main information from the project participant DB 102 and stores it in the project participant main information DB 104 . The project participant main information DB (104) contains the name of the clinical trial subject disease, age, gender, height, weight, disease history, family history, blood type, address, location, career, academic background, specialization, and interest in the project participation recruitment notice. , self-introduction information, certification information, project progress history and results, hobbies, special skills, preferences, food, clothing, clothing, and shelter information, selection criteria, etc. are extracted based on rules and the participant's personal information is stored.

The statistics and probability calculation module 105 performs statistical and probability calculations on the main information stored in the project participant main information DB 104, and generates the statistics and probability DB 107 as a result of the performance. The statistics and probability calculation module 105 provides a specific index (eg, age, gender, height, weight, medical history, family history, blood type, address, location, career, educational background, specialization, interest) for the participant's main information. , self-introduction information, certification information, project progress history and results, hobbies, specialties, preferences, food, clothing, and shelter information) can perform statistics and probability calculations on the fitness of previous project results. For example, in the statistics and probability DB 107, participant personal information (eg, disease name: brain tumor, current status: post-surgery standard treatment) for key project information (e.g.: disease name: glioblastoma, condition: autologous blood mononuclear cells can be collected, etc.) middle, etc.) are stored according to the frequency of occurrence.

The machine learning module 106 performs machine learning on the main information stored in the project participant main information DB 104, and creates a machine learning DB 108 based on the execution result. For example, the machine learning module 106 performs data evaluation on main information, extracts attributes for main information based on deep learning, and calculates (or infers) predictive information based on the extracted attributes. can For example, similar age, gender, height, weight, medical history, family history, blood type, address, location, experience, education, specialization, interest, self-introduction information, certification information, project progress history and results, hobbies, specialties, preferences Based on the project history of the participant who has at least one of , food, clothing, and shelter information, the suitability for a new project can be inferred through machine learning. As an example, the machine learning DB 108 stores a learning model in which a word extracted from personal information of a project participant is an input value, and a word extracted from a clinical test subject disease name is a target value.

The project main information extraction module 113 may extract main project information from the project DB 112 and generate the main project information DB 112 based on the extracted main project information. The main project information stored in the project main information DB 112 may be transmitted to the statistics and probability prediction module 114 and the machine learning inference module 116 . The project main information DB 112 contains the name of the clinical trial subject disease, age, gender, height, weight, disease history, family history, blood type, address, location, career, academic background, and specialization extracted from the project participation recruitment notice based on the rules. It stores major information such as field, field of interest, self-introduction information, certification information, project progress history and results, hobbies, special skills, preferences, food, clothing, and shelter information, and selection criteria. At this time, the method of extracting information based on rules from the list of sentences included in the announcement is as follows.

"Age 20 or older, male and female under 70" -> Rule: "Age #number #age or older, #under age #under age" -> Age: 20~69

"New patient confirmed with glioblastoma" -> Rule: "+patient confirmed with #disease#" -> Disease name: glioblastoma

The statistics and probability prediction module 114 performs statistics and probability prediction on the information stored in the statistics and probability DB 107 and the new project main information stored in the project main information DB 112, and the result is the primary target may be transmitted to the selection module 115 . For example, the statistics and probability prediction module 114 is a clinical trial main information DB (age, gender, height, weight, disease history, family history, blood type, address, location, career, academic background, professional field) extracted from the announcement of new project participation recruitment , fields of interest, self-introduction information, certification information, project progress history and results, hobbies, special skills, preferences, food, clothing, and shelter information), and select participants by frequency of occurrence (probability). The primary subject selection module 115 may select a primary subject greater than or equal to a specified reference value from among the results predicted by the statistics and probability prediction module 114 , and transmit the selection result to the inspection and correction module 118 .

The machine learning inference module 116 performs machine learning based on the machine learning results for the participant main information stored in the machine learning DB 108 and the main project information stored in the project main information DB 112, and infers accordingly results can be produced. The machine learning inference module 116 may transmit the calculated inference result to the secondary subject selection module 117 . As an example, the machine learning inference module 116 is used to select a participant suitable for a new project target disease name from existing member information. In the machine learning inference model, words are vectorized by the Word2vec model and used. The secondary subject selection module 117 may select subjects satisfying a specified condition as secondary subjects from the transmitted reasoning results, and transmit the selection result to the inspection and correction module 118 . The secondary subject selection module 117 may select secondary subjects from the rest of the project participants except for the project participants selected as primary subjects from among the main information of the project participants. Alternatively, the secondary subject selection module 117 may perform secondary subject selection through machine learning independently of the primary subject selection.

The inspection and correction module 118 provides the primary subject information transmitted by the primary subject selection module 115 and the secondary subject information transmitted by the secondary subject selection module 117 to the expert device 400, and the expert device It is possible to receive the inspection and calibration results from 400 . The primary subject information is the age, gender, height, weight, disease history, family history, blood type, address, location, career, academic background, specialization, field of interest, self-introduction information, and certification information of the participant selected primarily from the participant main information. , project progress history and results, hobbies, specialties, preferences, food, clothing, and shelter information, and may include condition information selected as the primary participant according to statistics and probability calculations for the participant information and main project information. The secondary subject information is the age, gender, height, weight, medical history, family history, blood type, address, location, career, academic background, specialization, field of interest, self-introduction information, certification information, project of the participant selected from the participant main information. It includes at least one of progress history and results, hobbies, specialties, preferences, food, clothing, and shelter information, and may include information on condition of a second selected participant based on inferences calculated according to machine learning for the participant information and main project information. . The expert device 400 checks whether the content automatically selected by the inspection and correction module 118 is correct, and if it is incorrect, it undergoes a correction process of correcting it. For example, the expert device 400 includes primary subject selection conditions, project key information, primary subject's age, gender, height, weight, disease history, family history, blood type, address, location, career, academic background, specialization, and interest Check at least one of fields, self-introduction information, certification information, project progress history and results, hobbies, specialties, preferences, food, clothing, and shelter information, statistics and probability information, and evaluate whether the result is valid as the primary target selection; Subject selection conditions, main project information, age, gender, height, weight, medical history, family history, blood type, address, location, career, educational background, specialization, interests, self-introduction information, certification information, project of the secondary target It is possible to evaluate whether the selection as a secondary subject is valid by checking at least one of progress history and results, hobbies, specialties, preferences, food, clothing, and shelter information, and inference results according to machine learning. The expert device 400 may transmit the primary subject inspection and correction result and the secondary subject inspection and correction result to the inspection and correction module 118 . The inspection and correction module 118 may transmit the result received from the expert device 400 to the final target selection and notification module 119 . In another aspect, the inspection and correction module 118 may update the information on which inspection and correction of the expert device 400 are performed as new information in the project participant main information DB 104 . Based on the updated information, the statistics and probability DB 107 may be updated by the statistics and probability calculation module 105 . Also, the machine learning DB 108 may be updated by the machine learning module 106 based on the updated information. The statistics and probability prediction module 114 operates by the updated statistics and probability DB 107 , and the machine learning inference module 116 operates by the updated machine learning DB 108 .

The final target selection and notification module 119 selects the final target by adding up both the primary target and the secondary target, and SMS, PUSH, SNS (Kakao Talk, Facebook), while delivering the selection result to the project participant device 300 through the instance app link, and notifies the project organization device 200 of the result. Here, the final target selection and notification module 119 first transmits the result to the project participant device 300 , and when presenting the intention to participate in a new project from the project participant device 300 , it corresponds to the project organization device 200 . It can be delivered to the project participant device (300).

Referring to FIGS. 1 to 3 , in relation to a method of operating a project brokerage device according to an embodiment of the present invention, in step 501 , the project brokerage device 100 is a website related to projects 1 to N 201 1 to The project DB 111 is created by crawling the content of the new project participation recruitment notice posted on the N 202 using the crawler 109 . In relation to the crawling operation, the project intermediary device 100 may check the website 202 belonging to the project agency device 200 in a certain period or in real time, and check whether there is a new project announcement. Alternatively, the project mediation device 100 may prepare a website and receive a new project announcement from the project agency device 200 based on the website.

In step 502, the project mediation device 100 may create a project main information DB (112). In relation to Io, the project intermediary device 100 may extract the main project information from the information stored in the project DB 111 to generate the main project information DB.

In step 503, the project brokerage device 100 may perform the primary target selection based on statistics and probability prediction. To this end, the project mediation device 100 obtains the project participant main information from the project participant main information DB 104 in which the project participant main information is stored, and performs statistics and probability calculations for the project participant main information, statistics and probability DB 107 can be created. The project mediation device 100 may perform the primary target selection by performing statistics and probability prediction for the information stored in the statistics and probability DB (107) and the main project information DB (112).

In step 504, the project mediation apparatus 100 may perform secondary target selection using the machine learning DB. In this regard, the project mediation device 100 may perform machine learning on the project participant main information stored in the project participant main information DB 104, and create a machine learning DB 108 based on the execution result. The project mediation device 100 may perform secondary target selection based on information stored in the project main information DB 112 and the machine learning DB 108 . When selecting the secondary target, the project mediation device 100 may select the secondary target for the rest of the project participant except for the project participant selected as the primary target among the main information of the project participant.

In step 505, the project brokerage device 100 may perform the inspection and calibration, and perform the final target selection. In relation to inspection and correction, the project mediation device 100 may form a communication channel with the expert device 400 , and transmit the primary target selection information and the secondary target selection information to the expert device 400 . The expert device 400 may include a computing device of an institution having specialized knowledge related to the content of the corresponding project announcement. The expert device 400 may check the received primary target selection information and secondary target selection information, and may determine whether selection is appropriate according to the project announcement content. Alternatively, the expert device 400 may check whether the selected target does not meet the specific conditions required in the project announcement content, and provide the result to the project brokerage device 100 . For example, it is possible to filter subjects with body information or disease history information that should not participate in the project that presents specific announcement content, and provide the filtering result (filtering reason and exclusion subject information) to the project mediation device 100 .

In step 506, the project mediation device 100 may provide notification content for the final target. The project mediation device 100 may provide the final target to the project participant device 300 and the project organization device 200 after checking the inspection and correction results, and selecting the final target. Here, the project mediation device 100 provides project participation recommendation information in preference to the project participant device 300 corresponding to the final target, and when the project participant device 300 approves the participation, the approved project participant device ( 300) of information may be provided to the project organization device 200 . Alternatively, the project mediation device 100 provides the final target information to the project organization device 200 first, and may provide project participation recommendation information only to the project participant devices 300 approved by the project organization device 200 . there is.

1, 2 and 4, in relation to the project mediation device DB creation and operation method of the present invention, the project mediation device 100 may store participant information in the project participant DB 102 in step 601. there is. In this operation, the project intermediary device 100 receives the project participant recruitment notice from the project organization device 200 and the project participant information who actually participated in the clinical trial, and stores the received participant information in the project participant DB 102 . there is.

In step 602, the project intermediary device 100 may extract the project participant main information from the project participant DB (102), and store the extracted information in the project participant main information DB (104). The main information of project participants is the name of the clinical trial subject disease, age, gender, height, weight, medical history, family history, blood type, address, location, experience, educational background, field of specialization, field of interest, self-introduction information in the project participation recruitment notice. , certification information, project progress history and results, hobbies, special skills, preferences, food, clothing, clothing, and shelter information, selection criteria, etc. may include information extracted based on rules and the participant's personal information.

In step 603, the project mediation device 100 performs statistical and probability calculations on the project participant main information stored in the project participant main information DB 104, and generates statistics and probability DB 107 based on the result. can

In step 604, the project mediation device 100 may perform machine learning on the project participant main information stored in the project participant main information DB 104, and create a machine learning DB 108 based on the machine learning result. . On the other hand, although the order of the generation of the machine learning DB 108 after the generation of the statistics and probability DB 107 has been disclosed, the present invention is not limited thereto. The project brokerage device 100 performs generation of the statistics and probability DB 107 after the creation of the machine learning DB 108, or the generation of the machine learning DB 108 and the generation of the statistics and probability DB 107 in parallel (or simultaneously).

In step 605, the project brokerage device 100 may operate a statistical and probability prediction model based on the statistics and probability DB 107, and operate a machine learning inference model based on the machine learning DB 108 . Project main information stored in the project main information DB 112 may be used for the statistical and probabilistic prediction model operation and the machine learning inference model operation. The project main information, for example, may include at least the purpose of the project (research), the subject of participation, the characteristics of the drug for the project, the main selection criteria, the number of recruits, the project (research) method, and predictable side effects.

Through the above-described method, the project brokerage device 100 can generate and operate the statistics and probability DB and machine learning DB of the cloud-based gradual intelligence enhancement deep learning project brokerage system by system collaboration with an expert device.

Referring to FIG. 5 , in relation to the DB update operation method of the project brokerage apparatus 100 according to an embodiment of the present invention, in step 701 , the project brokerage apparatus 100 may update the project participant main information. In this regard, the project mediation device 100 may obtain the inspection and correction results for the primary target selection result and the secondary target selection result, and update the project participant main information based on the inspection and correction results. For example, the main information of the project participants may be updated based on the result of which participant is selected as the final target for a specific project. In the main information of the participants corresponding to the final target, the items selected as the final target of a specific project may be registered and updated.

In step 702, the project brokerage device 100 generates statistics and probability DB by statistical and probability calculation based on the updated information, or by updating the previous statistics and probability DB 107, an updated statistics and probability DB is generated. can do.

In step 703, the project mediation device 100 may generate a machine learning DB by machine learning based on the updated information, or, by updating the previous machine learning DB 108, may generate an updated machine learning DB. The order of

steps

702 and 703 described above may be changed. For example, the project mediation device 100 may update the statistics and probability DB after performing the update of the machine learning DB first when the update of the project participant main information is performed. Alternatively, the project brokerage device 100 may simultaneously perform the update of the machine learning DB and the update of the statistics and probability DB in parallel.

In step 704, the project brokerage apparatus 100 may operate a statistical and probability prediction model based on the updated statistics and probability DB, and may operate a machine learning inference model based on the updated machine learning DB. In this operation, the project brokerage device 100 obtains the main project information about the new new project, operates the statistical and probabilistic prediction model based on the new project main information and the updated statistics and probability DB, the new project Based on the main information and the updated machine learning DB, the machine learning inference model can be operated.

Based on the above-described method, the project brokerage device 100 of the present invention may perform a new information re-learning procedure of the cloud-based gradual intelligence enhancement deep learning project brokerage system by system collaboration with an expert device.

1, 2 and 6, in relation to the project brokerage, in step 801, the project participant device 300 accesses the project brokerage device 100 to input, modify, and inquire its own information. . In this regard, the project mediation device 100 may create and operate a website that the project participant device 300 can access. In step 803 , the project agency device 200 may provide the previous project results to the project intermediary device 100 . Here, the project agency device 200 providing the previous project results and the project agency device 200 for designing and operating a new new project may be the same or different. The project results may be provided according to the request of the project agency device 200 , or the project agency device 200 for designing and operating a new project may provide the previous project results to the project intermediary device 100 . The project result includes contact information with the project participant device 300 (eg, SMS, SNS, phone number, etc.), the project participant's age, gender, height, weight, disease history, family history, blood type, address, location, career, educational background. , professional field, field of interest, self-introduction information, certification information, project progress history and results, hobbies, specialties, preferences, food, clothing, and shelter information, etc. may be included. Steps 801 and 803 described above may be performed irrespective of the order.

The project mediation device 100 builds the project participant DB 102 based on the project participant's input information or previous project results, and extracts the participant main information from the project participant DB 102 to extract the project participant main information DB 104 ) can be created. The project mediation device 100 may generate the statistics and probability DB 107 and the machine learning DB 108 in steps 805 and 807 based on the project participant main information DB 104 .

Meanwhile, in step 807 , the project organization device 200 may crawl the content of the new project participation recruitment announcement from the website 202 of the project organization device 200 . The project brokerage device 100 builds the project DB 111 based on the new project announcements obtained through crawling, and extracts the main project information from the project DB 111 to generate the project main information DB 112. can

In step 809, the project mediation device 100 performs statistical and probability prediction using the project main information stored in the project main information DB 112, and the statistics and probability DB 107 to select the primary target while , in step 811, by performing machine learning inference using the main project information and the machine learning DB stored in the project main information DB 112, it is possible to perform secondary target selection.

In step 813, the expert device 400 performs inspection and correction of the primary target selection information and the secondary target selection information provided by the project mediation device 100, and transmits information about the inspection and correction to the project mediation device (100) can be provided.

In step 815, the project mediation device 100 selects the final target based on the inspection and calibration information, and in step 817, notifies the project participant information to the project organization device 200, and in step 819, to the selected final target Project recommendation information may be provided to the corresponding project participant device 300 . In step 819 , the project mediation device 100 may notify the selection information through SMS, PUSH, SNS (KakaoTalk, Facebook), and an instance app link related to the project participant device 300 .

1, 2 and 7 , the project brokerage device 100 according to an embodiment of the present invention includes a communication circuit 110 , a memory 120 , an input device 130 , a display device 140 , and a processor (150).

The communication circuit 110 may support the communication channel forming function of the project brokerage device 100 . The communication circuit 110 may be configured with at least one communication module capable of using various mobile communication networks such as 3G, 4G, 5G, or the like, or an Internet network. For example, the communication circuit 110 forms a communication channel with the project organization device 200, and in response to the control of the processor 150 to receive the previous project performance result and the new project announcement content from the project organization device 200. can In addition, the communication circuit 110 forms a communication channel with at least one project participant device 300, and from the project participant device 300, participant information (eg, the participant's age, gender, height, weight, medical history, Family history, blood type, address, location, career, academic background, specialization, interests, self-introduction information, certification information, project progress history and results, hobbies, specialties, preferences, food, clothing, and shelter information) can be received. In addition, the communication circuit 110 forms a communication channel with the expert device 400 in response to the control of the processor 150 , and in response to the control of the processor 150 , the expert device 400 provides target selection information and may receive inspection and correction information from the expert device 400 . Alternatively, the communication circuit 110 provides a communication channel through which the expert device 400 can access and check target selection information, and the expert device 400 provides target selection information posted on the web page of the project intermediary device 100 . You can enter inspection and calibration information for In addition, the communication circuit 110 may form a communication channel with the payment server 500 in response to the control of the processor 150 , and may transmit/receive payment related information.

The memory 120 may store various applications and data related to the operation of the project mediation device 100 . For example, the memory 120 is the project participant DB 102, the project participant main information DB 104, the project DB 111, the project main information DB 112, statistics and It may include a probability DB 107, a machine learning DB 108, and the like. The database included in the memory 120 may be updated according to the addition or deletion of new project participants, and may also be updated according to the creation of a new project and completion of an ongoing project. In addition, the memory 120 may store inspection and correction information on the selected subjects provided by the expert apparatus 400 . The stored inspection and calibration information may be later used for evaluation of the expert device 400 .

The input device 130 may support the input of the manager required for the operation of the project mediation device 100 . Alternatively, the input device 130 may provide a hardware interface capable of inputting data. Accordingly, the input device 130 may support manual input of data related to the project participant device 300 or the project organization device 200 or data input through a wired connection of another external storage device.

The display device 140 may provide at least one screen related to the operation of the project mediation device 100 . The display device 140 may output a screen interface with an external device that has formed a communication channel (eg, a wired or wireless communication channel) through the communication circuit 110 in response to the control of the processor 150 . For example, the display device 140 is a web page to which the project organization device 200 accesses, a web page to which the project participant device 300 accesses, a web page related to the use of the payment server 500 , and the expert device 400 . At least one of the inspection and calibration support web page may be output. As an example, the display device 140 according to the manager input, a list of project agency devices 200 supporting project brokerage, a list of project participant devices 300 participating in a specific project, and a history of participating in a previous project A list of project participant devices 300 with, and at least one expert device 400 list may be displayed.

The processor 150 may transmit and process data related to the operation of the project mediation device 100 . For example, the processor 150 includes the information collection module 101 described in FIGS. 1 and 2 , the project participant main information extraction module 103 , the statistics and probability calculation module 105 , the machine learning module 106 , and at least one of the crawler 109, data management cloud server module 110, project main information extraction module 113, statistics and probability prediction module 114, primary target selection module 115, machine learning inference module 116, It may include at least one of the secondary target selection module 117 , the inspection and correction module 118 , the final target selection and notification module 119 , and the project participant login and information management module 120 . Each of the above-described modules may be implemented as at least one hardware processor, or may be configured as a software module executed by one processor. The processor 150 may perform automatic selection and verification of participants suitable for the project announcement content according to an embodiment of the present invention based on each of the above-described modules, and may notify.

In addition, the processor 150 may support payment processing between devices (eg, the project organization device 200 , the project participant device 300 , and the expert device 400 ). For example, when the processor 150 is entrusted with payment of expenses to the project participants from the project institution device 200, payment information for receiving account information or expenses from the project participant main information of the participants participating in the project It can collect automatic payments and make payments (eg, account transfers) based on the collected information. In this operation, the processor 150 may guide the payment of expenses according to project participation, and automatically guide information on the related fee deduction. The processor 150 may support payment of costs related to the inspection and calibration of the expert device 400 . For example, the processor 150 calculates the cost according to a predefined rule for the inspection and correction of the primary screening target and the secondary screening target (eg, calculating the cost based on the number of people, or calculating the cost based on the difficulty of the project, urgency cost calculation, etc.) may be performed, and the calculated cost may be provided to the expert apparatus 400 . Along with the cost provision, the evaluation score for the inspection and correction provided by the expert device 400 may be provided to the expert device 400 . The evaluation score may be calculated by the project organization apparatus 200 after providing the data selected as the final target to the project organization apparatus 200 . For example, the evaluation score may be determined by the number of project participant devices 300 excluded from the final target information by the project organization device 200 . The processor 150 may provide a cost discount or increase (eg, an incentive) to the expert device 400 based on the evaluation score according to prior agreement.

Referring to Figures 7 and 8, in relation to the method of selecting project participants of the project brokerage device according to an embodiment of the present invention, the processor 150 of the project brokerage device 100 is a statistical and probability-based first target in step 901 selection can be performed. In this regard, the memory 120 of the project mediation device 100 may pre-store the project participant main information DB 104 and the project main information DB 112 . The participant main information stored in the project participant main information DB 104 may be obtained by input of the project participant device 300 or may be included in a previously performed project result provided by the project organization device 200 . The main project information stored in the project main information DB 112 is collected through crawling for the website 202 operated by the project institution device 200, or the project institution device 200 is the project brokerage device 100 ) can be obtained by passing The processor 150 may perform the first target selection through statistical and probability calculation and prediction for the obtained main information of the project and the main information of the project participants. Alternatively, after the first subject is selected, the second subject may be selected in the project participant information except for the first subject.

Meanwhile, the processor 150 may perform second target selection through machine learning in step 902 while step 901 is being performed or before or after step 901 . In this regard, the processor 150 may perform machine learning on the main project information and the participant main information, and operate an inference model for the machine learning to select the second target.

The processor 150 may inspect and correct the selection result in step 903 . In this regard, the project brokerage apparatus 100 may include an AI (Artificial Intelligence) module capable of performing inspection and correction of the selection result. The AI module may collect expert knowledge related to the project announcement content in relation to the conformity check for the project announcement content and participant information, and perform an inspection on the participant's suitability result based on the expert knowledge. Alternatively, the processor 150 of the project brokerage device 100 forms a communication channel with the expert device 400 capable of inspecting and correcting the screening result, and collaborating with the expert device 400 for the screening result. Inspection and calibration results can be collected.

The processor 150 may perform payment processing for the final target in step 904 . In this operation, the processor 150 provides project recommendation information to the project participant device 300 corresponding to the final target, and when receiving an acceptance message from the project participant device 300, the reward described in the project announcement content According to the cost criterion, a predetermined cost may be provided to the project participant device 300 through the payment server 500 . In this operation, the processor 150 may provide the project participant device 300 with a fee for participating in a specified fee according to a policy defined by the project intermediary device 100 , or according to a prior contract with the project participant device 300 . .

Meanwhile, in the above description, the subject selection based on statistics and probability prediction and the subject selection through machine learning are performed in parallel as an example, but the present invention is not limited thereto. For example, the project mediation device 100 of the present invention selects a primary subject to perform statistics and probability prediction based on the main project information and project participant information, and then, within the primary subject, an inference model through machine learning can be applied to select secondary subjects. Alternatively, the project brokerage device 100 of the present invention selects the primary target by applying an inference model through machine learning based on the main project information and the project participant information, and then for the project participants belonging to the primary target. Secondary subjects can be selected by performing statistics and probability prediction. In addition, the project intermediary device 100 of the present invention may select the second target for the rest of the project participant, except for the project participant selected as the primary target among the main information of the project participant when the secondary target is selected.

Therefore, all project participants selected from only one of the 'statistics and probability prediction module' and the 'machine learning inference module' can be recruited. It has the effect of complementing the existing content. If a project participant registers only project participant information once, he or she can quickly receive project participation information suitable for them without searching. As the present invention proceeds in a manner that performs various requirements under certain constraints such as a set period, an allocated amount, and an input manpower for a single project regardless of the project field, the project period is shortened, costs are reduced, and input It can be expected to be effective in various project requirements such as efficient recruitment of manpower.

As described above, in the cloud-based gradual intelligence enhancement deep learning project mediation system by system collaboration with an expert device according to an embodiment of the present invention, the project participant device recruitment notice announced by the project institution device and the project participant device are transferred Statistics and probability DB and machine learning DB are generated by statistical and probability calculation and machine learning from the project results experienced in Find and notify the participant devices and project organization devices so that the project can proceed directly from the project organization devices to the project participant devices.

In addition, the cloud-based gradual intelligence enhancement deep learning project mediation system by system collaboration with the expert device of the present invention is a crawler that collects project participation recruitment announcements posted on the project agency device website, and the data collected by the crawler is stored in the project DB Data management cloud server stored in the project DB, project key information extraction module that creates project key information DB by extracting project key information, statistics and probability that predicts and calculates statistics and probability from project key information DB and statistics and probability DB Prediction module, primary target selection module that selects primary target by statistical and probability prediction module for previously secured project participant devices, machine learning inference module that infers machine learning from project main information DB and machine learning DB, A secondary subject selection module that selects secondary subjects for the rest of the project participant devices selected as the primary target among the secured project participant devices, a correction process that checks whether the selected content is correct and corrects if it is incorrect Inspection and correction module that goes through a process, the primary and secondary subjects are combined and selected as final subjects, and the result of selection is to the device of the project participant and the device of the project organization through SMS, PUSH, SNS (KakaoTalk, Facebook), and the instance app link. Selection and notification module to notify the final target, personal information (age, gender, height, weight, disease history, family history, blood type, address, location, career, educational background, specialization, field of interest, self-introduction information, certification information) , project progress history and results, hobbies, special skills, preferences, food, clothing, and shelter information, etc.) An information collection module for receiving project participant device information from the project organization device and storing it in the project participant device DB, extracting the project participant device major information from the project participant device DB and project participant device major information Project participant device main information extraction module to store in DB, statistics and probability calculation module to generate statistics and probability DB by statistical and probability calculation from project participant device main information DB, machine learning by machine learning from project participant device main information DB It can be composed of a machine learning module that creates a learning DB.

In addition, the cloud-based gradual intelligence enhancement deep learning project mediation method by system collaboration with an expert device according to an embodiment of the present invention crawls the content of the new project participation recruitment announcement posted on the website provided by the project institution device. Creating a DB, extracting the project main information from the project DB to generate the project main information DB, selecting the primary target by statistical and probability prediction from the project main information DB and the statistics and probability DB, the project Secondary target selection by the main information DB and machine learning DB, the final target selection and notification by expert device inspection and correction, and the step of providing notification contents to the project participant device or the project institution device .

In addition, the cloud-based gradual intelligence enhancement deep learning project mediation method by system collaboration with an expert device according to an embodiment of the present invention is to retrieve the project participant device recruitment notice and the project participant device information that actually participated in the clinical trial from the project institution device. receiving the data and storing it in the project participant device DB, extracting the project participant device main information from the project participant device DB and storing it in the project participant device main information DB; And generating a probability DB, generating a machine learning DB by machine learning in the project participant device main information DB, a statistical and probability prediction model is operated by the statistics and probability DB, and machine learning by the machine learning DB It may include a step in which the inference model operates.

In addition, the cloud-based gradual intelligence enhancement deep learning project mediation method by system collaboration with an expert device according to an embodiment of the present invention is a step in which new information is accumulated and updated in the project participant device main information DB by expert device inspection and correction , generating a statistics and probability DB by statistical and probability calculation based on the updated information, generating a machine learning DB by machine learning based on the updated information, and predicting statistics and probability by the updated statistics and probability DB It may include the step of operating the model, and operating the machine learning inference model by the updated machine learning DB.

In addition, the cloud-based gradual intelligence enhancement deep learning project mediation method by system collaboration with an expert device according to an embodiment of the present invention includes the steps of a project participant device accessing the project mediation system to input, modify, and inquire its own information , the project agency device transmits the project results (recruitment announcements for project participation and information on the device of actual participating project participants) to the project brokerage system, and the project brokerage system generates statistics and probability DB and machine learning DB based on the project results. Step, to crawl the recruitment notice for new project participation from the website of the project institution device, select the primary target by statistical and probability prediction using statistics and probability DB, and select the secondary target by machine learning inference using the machine learning DB Step, it may include the step of inspecting and correcting the data selected by the expert device inspection and calibration, and selecting and notifying the final target.

In addition, the present invention may provide systems, apparatuses, and methods according to various other embodiments.

Claims

a project agency device that publishes project announcements through a website;

The project announcement contents are collected from the project institution device to configure the project DB, and project key information related to the project progress is extracted from the project DB and stored, and project participant information from the project participant device related to the participant who will participate in the project is retrieved. Collect or collect project results according to project performance in which project participants have previously participated from the project institution device and store them as project participant main information, and perform statistical and probability calculations and predictions for the project main information and the project participant main information After selecting the final target by applying an inference model according to machine learning to the main project information and the main project participant information while selecting the subject based on the a project intermediary device that provides information about the final target to at least one of the project organization device and the project participant device;

a project participant device for registering as a participant in the project mediation device, and receiving new project recommendation information from the project mediation device, transmitting a new project participation acceptance message to the project mediation device according to a user input;

an expert device for performing the inspection and calibration;

In relation to the project participant recruitment, a payment server supporting at least one payment service of cost processing of the project organization device, cost processing of the project participant device, and fee deduction of the project brokerage device; characterized by comprising: project brokerage system.
a crawler that collects project participation recruitment announcements from a website operated by a project agency device;

a data management cloud server that receives the data collected by the crawler and stores it in the project DB;

Project participant main information DB for storing main information of the participants to participate in the project;

a statistics and probability prediction module for performing statistical and probability calculation and prediction on the project main information;

a primary subject selection module for selecting a primary subject by the statistical and probability prediction module based on the main project participant information stored in the project participant main information DB;

a machine learning inference module for performing machine learning inference on the main project information;

a secondary subject selection module for selecting secondary subjects from the rest of the project participants except for the project participants selected as the primary subjects from among the main information of the project participants;

an inspection and correction module that inspects whether the primary subject and the secondary subject selection content is correct, and undergoes a correction process to correct it if incorrect;

Project brokerage device comprising a; final target selection and notification module for selecting the final target by combining both the first target and the second target, and notifying the project participants and the project organization of the selection result.
3. The method of claim 2,

The age, gender, height, weight, medical history, family history, blood type, address, location, career, educational background, specialization, field of interest, self-introduction information, certification information, project progress history and results in relation to the project participant device; a member login and information management module that supports input and management of at least one of hobbies, specialties, preferences, and food, clothing, and shelter information and supports confirmation of a target selection result history;

an information collection module for receiving the project participant recruitment notice and project participant information actually participating in the clinical trial from the project institution device and storing the information in the project participant DB;

a project participant main information extraction module for extracting the project participant main information from the project participant DB and storing it in the project participant main information DB;

a statistics and probability calculation module for generating a statistics and probability DB by statistical and probability calculation from the project participant main information DB;

A machine learning module for generating a machine learning DB by machine learning from the project participant main information DB; Project brokerage device, characterized in that it consists of.
Performing the first target selection through statistical and probability calculation and prediction for the pre-stored main project information based on the project participant main information stored in advance in relation to the project participant device and the announcement provided by the project organization device;

after selecting the first subject, performing second subject selection based on machine learning for the main project information and remaining main information of the participants except for the first subject in the main project participant information;

providing the first subject and the second subject selection result to an expert device, and performing inspection and correction by the expert device;

selecting a final subject based on the inspection and correction;

Providing the final target selection information to at least one of the project organization device and the project participant device; Project brokerage method comprising a.
5. The method of claim 4,

generating a project DB by crawling new project participation recruitment announcements posted on a website operated by the project institution device, and extracting the main project information from the project DB to generate a project main information DB;

The results according to the inspection and correction are updated in the main project participant information, statistics and probability DB are generated by statistical and probability calculation based on the updated main project participant information, and machine learning is performed based on the updated main project participant information creating a machine learning DB by; Project brokerage method, characterized in that it further comprises at least one step of.
5. The method of claim 4,

Through the payment server, deducting the fee according to the agreement with the project participant device, and paying the cost to the project participant device participating in the project according to the compensation information described in the project main information; characterized by further comprising How to broker a project.
5. The method of claim 4,

Through the payment server, the project participant device paying a certain fee to the project brokerage device every time the project participates; Project brokerage method further comprising a.