KR102761105B1 - A cloud infrastructure automatic design system and a cloud business management platform comprising the saem - Google Patents

Abstract

According to one embodiment of the present invention, an infrastructure automatic design system for designing a cloud infrastructure that satisfies the requirements of a request for proposal announced by an arbitrary requesting agency may include: a receiving module that receives target project information, which is information on a target request for proposal, which is a request for proposal for a desired project that a target company wishes to support; an extraction module that extracts target requirement information, which is a requirement in the target request, based on the target project information; and a construction module that calculates a target infrastructure, which is a cloud infrastructure in which the target requirement information can be realized, using an infrastructure model learned by machine learning.

Description

{A CLOUD INFRASTRUCTURE AUTOMATIC DESIGN SYSTEM AND A CLOUD BUSINESS MANAGEMENT PLATFORM COMPRISING THE SAEM}

The present invention relates to a cloud automatic design system that automatically determines cloud infrastructure and a cloud business management platform including the same.

Government agencies, academia, and research institutes are announcing a variety of projects to obtain necessary services. The details of these projects are shared in the form of a request for proposal (RFP). A request for proposal (RFP) can mean a document that contains the requirements requested by the demand organization. The supplying organization reviews the request for proposal (RFP) of each organization and announces competitive projects. However, the supplying organization is experiencing the following problems when applying for the announced projects.

First, the supplier must search for the business announced by the demand agency every time, read the contents of the announced business, and determine whether the announced business is a business that the supplier can apply for. As a result, the supplier frequently misses the business that should be applied for, and a lot of manpower is spent on determining whether or not to apply.

Second, each time a supplier applies for a cloud business, it must calculate and decide on the cloud infrastructure required for the cloud business. As a result, it takes a lot of time to decide on the cloud infrastructure when applying for a business, and even when carrying out multiple businesses, the cloud infrastructure is set too heavily, making it difficult to carry out the business.

The present invention is intended to solve the above-described problems, and provides a cloud infrastructure automatic design system that conveniently searches for businesses and easily determines cloud infrastructure, and a cloud business management platform including the same.

According to one embodiment of the present invention, an infrastructure automatic design system for designing a cloud infrastructure that satisfies the requirements of a request for proposal announced by an arbitrary requesting agency may include: a receiving module that receives target project information, which is information on a target request for proposal, which is a request for proposal for a desired project that a target company wishes to support; an extraction module that extracts target requirement information, which is a matter required in the target request, based on the target project information; and a construction module that calculates a target infrastructure, which is a cloud infrastructure in which the target requirement information can be realized, using an infrastructure model utilizing machine learning.

In addition, the system further includes a resource module that calculates idle resources, which are cloud infrastructure resources that can be utilized on the server of the target company; and the construction module can calculate the target infrastructure within the idle resources.

In addition, the above-described construction module learns the infrastructure model through machine learning based on a request for proposal for a past project and a cloud infrastructure, and can re-learn the infrastructure model by comparing the expected cloud infrastructure using the request for proposal for a past project and the infrastructure model with the cloud infrastructure specified in the past project.

In addition, the method further includes a selection prediction module that calculates a selection probability, which is a possibility that the target company will be selected for the desired project; and a decision module that calculates a modified infrastructure by modifying the target infrastructure using a predetermined modification method based on the selection probability when a predetermined modification condition is satisfied; and the predetermined modification condition may be a condition in which, based on the time of the cloud infrastructure recommendation request for the desired project, the sum of the target infrastructure and the cloud infrastructure allocated to the pending project, which is a project for which the target company is determined to apply for the project, exceeds the idle resources.

Additionally, the above-determined modification method may be a method for determining a method of modifying the cloud infrastructure for each interval of the above-described selection probability.

In addition, the above-determined modification method may be a method of not modifying a cloud infrastructure corresponding to a business having the selection probability in the first section that is less than or equal to the first probability and greater than the second probability, which is a probability lower than the first probability.

In addition, the above-determined modification method may be a method of not modifying the cloud infrastructure corresponding to the project having the selection probability in the second section when the sum of the cloud infrastructures assigned to the desired project and the pending project having the selection probability in the first section and the second section exceeding the third probability which is lower than the second probability but less than or equal to the second probability is less than or equal to the idle resources.

In addition, the above-determined correction method may be a method of calculating a correction infrastructure by reducing the cloud infrastructure corresponding to the project having the selection probability in the third section based on the selection probability, when the sum of the cloud infrastructures assigned to the desired project and the pending project having the selection probability in the first section and the second section exceeds the third probability which is lower than or equal to the second probability and is less than or equal to the idle resources, and at the same time, the sum of the cloud infrastructures assigned to the desired project and the pending project having the selection probability in the first section, the second section, and the third section exceeds the idle resources.

In addition, the above-determined correction method may be a method of calculating a correction infrastructure by reducing the cloud infrastructure corresponding to the project having the above-described selection probability in the second section based on the above-described selection probability when the sum of the cloud infrastructures assigned to the above-described desired project and the pending project in the first section and the second section exceeds the above-described idle resources, which is a third probability that is lower than the above-described second probability and less than or equal to the above-described second probability.

In addition, the above-determined correction method may be a method of not allocating cloud infrastructure to a project having a rate of return below a predetermined standard, when the sum of cloud infrastructures allocated to the desired project and the pending project having the above-described selection probability in the first section and the second section exceeds the third probability which is lower than or equal to the second probability and has a probability exceeding the idle resources, and at the same time, a project having the above-described selection probability in the third section which is lower than or equal to the third probability and has a probability exceeding the fourth probability which is lower than the third probability exists.

An automatic infrastructure design method according to one embodiment of the present invention is implemented by an automatic infrastructure design system, and may include a step of receiving, by a receiving module, target project information, which is information on a target request, which is a request for proposal for a desired project that a target company wishes to support; a step of extracting, by an extraction module, target requirement information, which is a requirement in a target request, based on the target project information; and a step of generating, by a construction module, a target infrastructure, which is a cloud infrastructure that can realize the target requirement information, using an infrastructure model utilizing machine learning.

A business search system according to one embodiment of the present invention may include a collection module that collects requests for proposals posted on an online network; a processing module that derives extracted information from the contents of the collected requests for proposals; a function module that derives characteristic information representing the characteristics of a target company from target information, which is information about the target company; a response module that processes and compares the extracted information and the characteristic information using a predetermined method to determine a degree of similarity representing the degree of similarity; and an interface module that derives an interface on which the degree of comparative similarity is displayed so as to express whether the target company and the request for proposal are related.

Additionally, the extracted information may consist of requirements within the request for proposal.

In addition, the target information may be information about a service provided by the target company or a product produced by the target company, and the characteristic information may be information related to a function among the target information.

In addition, the above-determined method can determine the comparative similarity by embedding the extracted information and the characteristic information into the same space and then calculating the cosine similarity of the vectorized extracted information and the vectorized characteristic information.

In addition, the response module can calculate the comparative similarity for each product of the target company to indicate whether the product or service of the target company is related to the request for proposal.

In addition, the interface module can produce an interface so that the comparative similarity is displayed in the form of a bar for each product or service of the target company.

In addition, when the target company applies for the business that is the basis of the request for proposal, the method further includes a simulation module that simulates whether or not the cloud infrastructure can be allocated; and the interface module can produce an interface so that the possibility mark, which is an image indicating the possibility of whether or not the cloud infrastructure can be allocated, and the comparative similarity are simultaneously displayed.

In addition, the simulation module can perform simulation only for businesses whose comparative similarity is higher than a predetermined standard, and the interface module can produce an interface so that the possible indication is displayed only for businesses for which simulation was performed.

In addition, the above simulation module can conduct a simulation by considering cloud resources allocated to a pending business, which is a business period that overlaps at least part of the business period of the business being simulated, based on idle resources, which are cloud infrastructure resources that can be utilized on the server of the target company, and for which the target company is expected to apply for the business.

A business search method according to one embodiment of the present invention is a business search method implemented by a business search system, which may include a step of collecting request for proposals posted on an online network by a collection module; a step of calculating extracted information from the contents of the request for proposal by a processing module; a step of calculating characteristic information representing the characteristics of a target company from target information, which is information about the target company, by a function module; a step of processing and comparing the extracted information and the characteristic information by a predetermined method by a response module to determine a comparative similarity representing similar information; and an interface module for calculating an interface in which the comparative similarity is displayed so as to express whether the target company and the request for proposal are related by an interface module.

The cloud infrastructure automatic design system according to the present invention and the cloud business management platform including the same can increase work convenience.

Additionally, it can effectively minimize the time required to apply for a business.

Additionally, it can maximize business progress efficiency.

However, the effects of the present invention are not limited to the effects described above, and effects not mentioned can be clearly understood by a person having ordinary skill in the art to which the present invention pertains from this specification and the attached drawings.

Figure 1 is a relationship diagram of a cloud business management platform according to one embodiment of the present invention.
Figure 2 is a configuration diagram of a cloud business management platform according to one embodiment of the present invention.
FIG. 3 is a flowchart of a business search method implemented by a business search system of a cloud business management platform according to one embodiment of the present invention.
Figure 4 is a drawing for explaining the operation process of a business search system according to one embodiment of the present invention.
FIG. 5 is a flowchart of an infrastructure automatic design method implemented by an infrastructure automatic design system of a cloud business management platform according to one embodiment of the present invention.
Figure 6 is a drawing for explaining the operation process of an infrastructure automatic design system according to one embodiment of the present invention.
Figure 7 is a flowchart illustrating the process of calculating modified infrastructure by an infrastructure automatic design system according to one embodiment of the present invention.

Hereinafter, specific embodiments of the present invention will be described in detail with reference to the drawings. However, the spirit of the present invention is not limited to the presented embodiments, and those skilled in the art who understand the spirit of the present invention will be able to easily propose other regressive inventions or other embodiments included within the scope of the spirit of the present invention by adding, changing, deleting, etc. other components within the scope of the same spirit, but this will also be considered to be included within the scope of the spirit of the present invention.

Figure 1 is a relationship diagram of a cloud business management platform according to one embodiment of the present invention.

Referring to FIG. 1, the cloud business management platform (100) can be connected to a cloud server (P10) of the target company, the target company, and/or an external server (P20) via a wired/wireless network.

The wireless network referred to in the present invention may be a core network integrated with a wired public network, a wireless mobile communication network, or a portable Internet, or may mean a worldwide open computer network structure that provides various services existing in the TCP/IP protocol and its upper layer, such as HTTP (Hyper Text Transfer Protocol), HTTPS (Hyper Text Transfer Protocol Secure), Telnet, FTP (File Transfer Protocol), DNS (Domain Name System), SMTP (Simple Mail Transfer Protocol), and the like, and is not limited to these examples, but comprehensively means a data communication network capable of transmitting and receiving data in various forms.

The target company (P30) may refer to a company that provides cloud services.

Additionally, the target company (P30) may be a user of the cloud business management platform (100).

The target company (P30) may include all corporations and/or individuals using the cloud business management platform (100).

Here, the target company (P30) may mean a computing device managed, operated, and controlled by the target company, and hereinafter, the word 'target company' may include both the meaning of 'target company' and 'computing device used by the target company.'

The computing device referred to in the present invention may mean a device capable of processing information processing operations.

For example, computing devices may include desktop computers, laptops, smartphones, mobile terminals including Personal Digital Assistants (PDAs), Portable Multimedia Players (PMPs), handheld terminals, and/or smart TVs.

An external server (P20) may refer to a server that publishes a bid announcement and request for proposal for a business on an online network.

For example, external western (P20) can refer to a server managed and operated by any demand agency such as a public institution or government agency.

However, without limitation thereto, the external server may include all servers that handle and disclose business bid announcements and requests for proposals.

A cloud server (P10) may mean a server that provides cloud services through virtualization to an arbitrary number of people.

The server referred to in the present invention may include other configurations for performing a server environment of the server. The server may include any type of device.

For example, a server may be a digital device having computing power, such as a laptop computer, a notebook computer, a desktop computer, a web pad, or a mobile phone, and having a processor and memory.

For example, the server may be a web server. However, the type of server is not limited to this, and can be changed in various ways at a level that is obvious to a person skilled in the art.

A cloud business management platform (100) according to one embodiment of the present invention may include a business search system (110) that searches for a request for proposal posted on an online network and calculates and displays whether a company and a business are related, an infrastructure automatic design system (120) that automatically designs and presents a cloud infrastructure for a desired business that a target company wishes to support, and a management system (130) that manages a business applied for by a target company.

The management system (130), the business search system (110) and the infrastructure automatic design system (120) can exchange information with each other to proceed with necessary processes.

Business search systems, infrastructure automation design systems and management systems may refer to computing devices.

Below, each system will be described in detail.

Figure 2 is a configuration diagram of a cloud business management platform according to one embodiment of the present invention.

Referring to FIG. 2, a business search system (110) according to one embodiment of the present invention may include a collection module (111) that collects requests for proposals posted on an online network, a processing module (112) that derives extracted information from the contents of the collected requests for proposals, a function module (113) that derives characteristic information representing the characteristics of a target company from target information, which is information about the target company, a response module (114) that processes and compares the extracted information and the characteristic information using a predetermined method to determine a degree of similarity representing the degree of similarity, and an interface module (115) that derives an interface on which the degree of comparative similarity is displayed so as to express whether the target company and the request for proposal are related.

In addition, the business search system (110) may further include a simulation module (116) that simulates whether the target company can be allocated a cloud infrastructure when applying for the business that is the basis of the request for proposal.

In addition, the business search system (110) may further include a selection module (117) that selects a request for proposal presented to a target company through an interface using a predetermined selection method.

In addition, the business search system (110) may further include a storage module (118) in which all information necessary for implementing a business search method, such as target information and collected proposal requests, is stored.

The collection module (111) can collect a request for proposal published (announced) on an online network from an external server.

A request for proposal is a document requesting proposals for a business, and may contain information about the support requirements, schedule, support methods, and benefits for the business.

In addition, the collection module (111) can receive a request for business search and search information from the target company.

Search information is information about the business that the target company is trying to search for, and may be keywords.

Search information is information generated by the target company entering information into the target company's computing device and may be a type of search word.

For example, it could be ‘cloud’, ‘CMP’, ‘virtualization’, etc.

However, without limitation thereto, the type of the above search information can be varied in various ways at a level that is obvious to a person skilled in the art.

The collection module (111) can receive target information, which is information about the target company, from the target company. In addition, the collection module (111) can collect the target information from an external server.

For example, the target information may include information about services provided by the target company or products produced by the target company.

For example, the above target information may be information about the company itself, including information about the company's sales, location, business partners, and employee history/career.

For example, target information may include product/service features, specific performance, manuals, etc.

The corporate information of the target company can be defined as target information.

The processing module (112) can extract information from the contents of the above request for proposal.

Here, the extracted information may be information about requirements in the above request for proposal.

For example, requirements may mean information about what needs to be done in the business requested in the request for proposal.

For example, requirements may mean information about requirements that are to be realized through business, such as "Provide auto scaling function for flexibly responding to changes in usage", "High-speed Input/Output processing through 10G network configuration", "Internal network separation between external services and virtual servers", "Disks provide stable disk volume with RAID configuration", "NFS (Network File System) type Shared Storage for DB and content storage", and "Security-related H/W and S/W must be configured as CC certified products."

For example, requirements may include support requirements, participation restrictions, desired business outcomes, selection evaluation criteria, etc.

For example, extracted information may mean information about the requirements for cloud infrastructure requested in the RFP.

However, without limitation thereto, specific examples of the above requirements can be varied in various ways at a level that would be obvious to a person skilled in the art.

For example, the processing module (112) can extract information from the contents of a request for proposal based on natural language processing.

For example, the processing module (112) can produce an extraction information model, which is a model that selects/extracts extraction information from a request for proposal through machine learning, and can produce extraction information from a request for proposal through the extraction information model.

To explain specifically, the system administrator can set the input data as a request for proposal, and set the output data of the input data as information indicating (designating) which part of the request for proposal is a requirement, and the processing module can fine-tune the input data and output data labeled by the administrator with a pre-trained language model to produce an extracted information model.

For example, the pre-trained language model may be SBERT, GPT, or RoBerTa, but the present invention is not limited thereto.

An extraction information model can be a model that analyzes and calculates which part of a request for proposal is a requirement by analyzing the pattern between input data and output data.

Unlike the above, the extraction information model can select extraction information based on keywords. Keywords can be preset by the administrator, and can be a general model that selects sentences containing keywords as extraction information. However, the present invention is not limited thereto.

Below, a detailed description of the extraction method of the extracted information may be omitted within the scope of known technology.

The function module (113) can select characteristic information from target information.

The characteristic information may be information related to function among the above target information.

Specifically, characteristic information may be information related to functions among information about services/products within the target information.

For example, characteristic information may be information about the type of product and/or service, the service method, recommended specifications for the computing device of the product/service user, the content of the product and/or service, and the functions of the product and/or service.

However, without limitation, specific examples of characteristic information can be modified in various ways at a level that is obvious to those skilled in the art.

For example, the function module (113) can produce a characteristic information model, which is a model that selects/extracts characteristic information from target information through machine learning, and can produce characteristic information from target information through the characteristic information model.

To explain specifically, a system administrator can set input data as target information and set output data of the input data as information indicating (designating) which part of the target information is characteristic information, and the function module (113) can fine-tune input data and output data labeled by the administrator with a pre-trained language model to produce a characteristic information model.

For example, the pre-trained language model may be SBERT, GPT, or RoBerTa, but the present invention is not limited thereto.

A feature information model can be a model that analyzes the correlation between target information and feature information within the target information by analyzing the pattern between input data and output data.

For example, if the target information is "OKESTRO has sales of approximately 50 billion won and is located in Yeongdeungpo-gu. The products released by OKESTRO are 'SYMPHONY A.I.', 'CONTRABASS', 'OKESTRO CMP', and 'TROMEBONE'. 'SYMPHONY A.I.' is an AIOps (Artificial Intelligence for IT Operations) solution that provides auto scaling between containers and provides a separate network integrated management service for external services and virtual servers. (hereinafter, omitted).", then the characteristic information could be "'SYMPHONY A.I.' is an AIOps (Artificial Intelligence for IT Operations) solution that provides auto scaling between containers and provides a separate network integrated management service for external services and virtual servers."

For example, the feature information model can select feature information from target information based on keywords, but the present invention is not limited thereto.

Below, a detailed description of the method for extracting target information may be omitted within the scope of known technology.

The response module (114) can process and compare the extracted information and the characteristic information in a predetermined manner to calculate and determine the comparative similarity.

For example, a predetermined method can determine the comparative similarity by embedding the extracted information and the characteristic information into the same space and then calculating the cosine similarity of the vectorized extracted information and the vectorized characteristic information.

The response module (114) can embed the extracted information and the characteristic information into the same space using a natural language processing model.

For example, the response module (114) can embed the extracted information and/or characteristic information into the same space using BERT and/or a fine-tuned model of BERT.

However, without limitation thereto, the type of model used by the corresponding module (114) for embedding can be varied in various ways at a level that is obvious to a person skilled in the art.

The corresponding module (114) may have a larger comparative similarity value as the cosine similarity increases, and conversely, a smaller comparative similarity value as the cosine similarity decreases.

The response module (114) can calculate the comparative similarity for each product of the target company to indicate whether the product or service of the target company is related to the request for proposal.

The response module (114) can compare characteristic information and extracted information about a product and/or service of a target company using a predetermined method to calculate comparative similarity.

That is, the response module (114) can calculate the degree of similarity with the request for proposal for each product or service of the target company.

The selection module (117) can select a request for proposal corresponding to a product of a target company using a predetermined selection method.

The selection module (117) can select keywords corresponding to the target company based on the target information.

The selection module (117) can produce a keyword model that produces keywords corresponding to corporate information based on past corporate information and keywords through machine learning/deep learning.

To explain specifically, the system administrator can set the input data as corporate information and label the output data of the input data by setting the keywords for the corporate information of the input data. The above selection module can create a keyword model by fine-tuning (additional learning) the input data and output data that are labeled with each other to a pre-trained language model.

For example, the pre-trained language model may be SBERT, GPT, or RoBerTa, but the present invention is not limited thereto.

A keyword model can be a model that analyzes the correlation between corporate information and keywords within that corporate information by analyzing patterns between input data and output data.

For example, if the corporate information is "This company provides multi-cloud services using OKESRTO CMP software. OKESRTO CMP provides multi-cloud integrated management services by managing infrastructure in the form of services on the platform and operating software-based integration.", then the keyword may be "multi-cloud." However, this is only an example and the present invention is not limited thereto.

However, without being limited to this, the keyword model may be a model that derives keywords from corporate information based on word frequency. The keyword model can divide corporate information into phrase units, then vectorize each phrase to determine the phrase with the highest frequency. Here, the keyword model can determine the 10 phrases with the highest frequency from corporate information as keywords.

A pre-determined selection method may be a method of selecting requests for proposals based on keywords corresponding to the target company.

A predetermined selection method may be a method of selecting a request for proposal having keywords derived from a keyword model in which target information is input.

The interface module (115) can select a request for proposal, which is a request for proposal presented to a target company through an interface in a predetermined display method, and produce an interface on which the request for proposal is displayed.

The interface module (115) can select a display request form, which is a request for proposal, to be displayed to the target company based on a pre-determined display method.

The processing module (112) and the response module (114) can extract only the display request form and calculate the comparative similarity, but the present invention is not limited thereto.

The predetermined display method may include a first display method for selecting a request for proposal containing a keyword provided by a target company as a request for display, and a second display method for automatically generating a keyword corresponding to a target company and selecting a request for proposal containing the generated keyword as a request for display.

The target company can choose how the interface module (115) selects the display request form.

When the target company selects the first display method, the interface module (115) can select a request for proposal that includes a keyword of search information from among the request for proposals collected through the collection module (111) as a request for display.

If the target company selects the second display method, the interface module (115) can designate the request for proposal selected by the selection module (117) as the request for display.

The interface module (115) can generate and transmit to the target company a user interface that can display summary information of the display request form, an image that can distinguish each display request form, and a comparison and similarity between the display request form and the target company's products/services.

For example, the interface module (115) can produce an interface so that the comparative similarity is displayed in the form of a bar for each product or service of the target company.

However, without limitation to this, the method of indicating comparative similarity can be modified in various ways at a level that is obvious to those skilled in the art.

For example, the interface module (115) can express the comparative similarity as a number from 0 to 100 for each product or service.

The interface module (115) can produce a user interface so that a possibility mark, which is an image indicating the possibility of being selected for a cloud infrastructure business as a simulation result, and the comparative similarity are simultaneously displayed.

For example, the interface module (115) can generate a user interface so that the above possible indication is displayed only for businesses for which simulation has been performed.

Through the user interface provided by the interface module (115), the target company can search for a request for proposal and review the details of the required request for proposal.

The simulation module (116) can simulate whether the target company can be assigned a cloud infrastructure managed by the target company when the target company applies for the business that is the basis of the request for proposal.

The simulation module (116) can conduct simulation only for businesses whose comparative similarity is above a predetermined standard.

As a specific example, the simulation module (116) can conduct a simulation only for a request for proposal/request for indication that has a similarity level with at least one of the target company's products or services that exceeds a predetermined standard.

When a target company searches for a request for proposal, the simulation module (116) can simulate whether cloud infrastructure can be supported when applying for a business based on the request for proposal.

The simulation module (116) treats a business corresponding to an arbitrary request for display as a desired business described later, and can simulate whether cloud infrastructure can be supported when applying for a business corresponding to a request for display.

For example, if a target infrastructure or modified infrastructure can be assigned to a business corresponding to a request for display, the simulation module (116) can determine that the business corresponding to the request for display can be supported by cloud infrastructure.

Conversely, if only the target infrastructure can be assigned to the business corresponding to the request for display, the simulation module (116) may determine that the business corresponding to the request for display can be supported by the cloud infrastructure.

The simulation module (116) can transmit and receive information necessary for simulation from the infrastructure automatic design system.

For example, the simulation module (116) can conduct a simulation by considering cloud resources allocated to a pending business, which is a business period that overlaps at least part of the business period of the business being simulated, based on idle resources, which are cloud infrastructure resources that can be utilized on the server of the target company, and for which the target company is expected to apply for the business.

For this purpose, the simulation module (116) can receive information on the pending business from the assignment module (127).

A detailed description of how the target infrastructure or modified infrastructure is derived can be provided later.

The storage module (118) may store all information necessary for implementing a business search method, such as corporate information (target information) and collected proposal requests.

For example, the storage module may include internal memory and/or external memory.

For example, the embedded memory may include at least one of volatile memory (e.g., DRAM, SRAM, or SDRAM), nonvolatile memory (e.g., one time programmable ROM (OTPROM), PROM, EPROM, EEPROM, mask ROM, flash ROM, flash memory, a hard drive, or a solid state drive (SSD).

External memory may include a flash drive, such as CF (compact flash), SD (secure digital), Micro-SD, Mini-SD, xD (extreme digital), MMC (multi-media card), or memory stick.

The collection module (111), processing module (112), function module (113), response module (114), interface module (115), simulation module (116), selection module (117) and/or storage module (118) can be connected to each other in a way that enables information communication.

An infrastructure automatic design system (120) according to one embodiment of the present invention may include a receiving module (121) that receives target project information, which is information on a target request, which is a request for proposal for a desired project that a target company wishes to support, in an infrastructure automatic design system (120) for designing a cloud infrastructure that satisfies the requirements of a request for proposal announced by an arbitrary requesting agency, an extraction module (122) that extracts target requirement information, which is a requirement in the target request, based on the target project information, and a construction module (123) that generates a target infrastructure, which is a cloud infrastructure in which the target requirement information can be realized, using an infrastructure model learned by machine learning.

In addition, the above-described infrastructure automatic design system (120) may further include a resource module (124) that calculates idle resources, which are cloud infrastructure resources that can be utilized on the server of the target company, a selection prediction module (125) that calculates a selection probability, which is a possibility that the target company will be selected for the desired project, and a decision module (126) that calculates a modified infrastructure by modifying the target infrastructure using a predetermined modification method based on the selection probability when a predetermined modification condition is satisfied.

In addition, the above-described infrastructure automatic design system (120) may further include an assignment module (127) that stores and manages information on cloud infrastructure assigned to pending projects necessary for satisfying predetermined modification conditions.

The receiving module (121) can receive information about a target request form, which is a request for proposal for a desired project that the target company wishes to support.

The target company can request a recommendation for a cloud infrastructure for the desired business it wishes to apply for, and the receiving module (121) can receive this request signal.

The target company can transmit information about the target request for the desired business to the receiving module (121) in the request signal.

Here, information about the target request may include information such as a request for proposal for the desired project or a link to access the request for proposal.

In contrast, when the receiving module (121) receives a request signal from a target company, the receiving module (121) can request and receive target task information for a desired project from the collection module (111) or the storage module (118) of the business search system (110) based on information that can specify the desired project included in the request signal.

The extraction module (122) can extract target requirement information, which is a requirement of the target request form, based on target task information.

For example, target requirement information refers to information about requirements for cloud infrastructure requested in the target request, and may refer to information about target requirements.

For example, target requirement information may mean information on the performance and functions of the cloud infrastructure to be realized through the business, such as "providing auto scaling function for flexibly responding to changes in usage", "high-speed input/output processing through 10G network configuration", "separation of internal networks between external services and virtual servers", "disks provide stable disk volume with RAID configuration", "NFS (Network File System) type shared storage for DB and content storage", and "security-related H/W and S/W must be configured as CC certified products".

However, without limitation thereto, specific examples of the above target requirement information can be modified in various ways at a level that is obvious to a person skilled in the art.

For example, the extraction module (122) can extract target request information from the contents of the target request form based on natural language processing.

For example, the extraction module (122) can produce an extraction information model, which is a model that selects/extracts target requirement information from a request for proposal through machine learning, and can produce target requirement information from a request for proposal through the extraction information model.

To explain specifically, the system administrator can set the input data as a request for proposal (RFP), and set the output data of the input data as information indicating (designating) which part of the RFP is a requirement, and the extraction module can produce an extraction information model by fine-tuning (additional learning) the input data and output data labeled by the administrator on a pre-trained language model.

For example, the pre-trained language model may be SBERT, GPT, or RoBerTa, but the present invention is not limited thereto.

The extraction information model can be a model that analyzes the correlation between the target request and the requirements within the target request by analyzing the pattern between the input data and the output data.

For example, the extraction information model can select target requirement information based on keywords, but the present invention is not limited thereto.

Below, a detailed description of the method for extracting target requirement information may be omitted within the scope of known technology.

The extraction information model is learned in the processing module (112) and the extraction module (122) and can be used in each module.

Alternatively, the extraction information model may be learned in the processing module (112) or the extraction module (122) and transferred to a module that did not participate in the learning.

The requested information in the target request form can be defined as target request information.

Unlike the above, the extraction information model can select extraction information based on keywords. Keywords can be preset by the administrator, and can be a general model that selects sentences containing keywords as extraction information. However, the present invention is not limited thereto.

The construction module (123) can use an infrastructure model learned through machine learning (machine learning utilized) to produce a target infrastructure, which is a cloud infrastructure in which the target requirement information can be realized.

Cloud infrastructure can refer to cloud resources that a target company must provide when proposing a business or conducting a business.

For example, cloud infrastructure may include all components required for cloud computing, including the types of hardware devices (servers, disks, CPUs, RAMs, GPUs, network equipment, etc.), virtualization types (virtual machines, containers), virtualized RAM/CPU specifications, virtual storage specifications, and the amount of network resources (network speed).

The construction module (123) can learn the infrastructure model through machine learning based on information about a request for proposal for a project implemented in the past and the cloud infrastructure used when performing the project based on the request for proposal.

That is, the construction module (123) can produce an infrastructure model by performing machine learning/deep learning based on the requirements information of past projects and the cloud infrastructure utilized in the selected project.

To explain specifically, a system administrator can set requirement information, which is information about requirements described in a request for proposal for a project implemented in the past, as input data, and set information about cloud infrastructure used when the project that is the basis of the request for proposal is implemented as output data of the input data, and the construction module (123) can create an infrastructure model by fine-tuning (additional learning) the input data and output data labeled by the administrator on a pre-trained language model.

For example, the pre-trained language model may be SBERT, GPT, or RoBerTa, but the present invention is not limited thereto.

To this end, the infrastructure model can extract required information from previously published requests for proposals in conjunction with the extraction information model.

An infrastructure model can be a model that analyzes the correlation between the requirements of a request for proposal and the cloud infrastructure to realize the requirements by analyzing the patterns between input and output data.

For example, if the requirements stated in the past request for proposal were "Disk provides stable disk volume with RAID configuration" and "Auto scaling function is provided to flexibly respond to changes in usage," the infrastructure model can determine the target infrastructure to be IBM's Storwize series or Dell EMC's PowerVault series disks used in the cloud infrastructure, and determine the target infrastructure to have software that provides the Auto scaling function installed.

The construction module (123) can re-learn the infrastructure model by comparing the expected cloud infrastructure with the cloud infrastructure specified in the past project using the request for proposal for the past project and the infrastructure model.

Specifically, the construction module (123) can compare the cloud infrastructure specified in a past project with the cloud infrastructure produced by inputting the request for proposal for the past project into the infrastructure model, and re-learn the infrastructure model based on the difference, thereby increasing the accuracy of the infrastructure model.

Through this re-learning process, the construction module (123) can learn the infrastructure model more accurately.

The construction module (123) can produce the target infrastructure within the idle resources.

The construction module (123) can receive idle resources from the resource module (124).

The resource module (124) can calculate idle resources, which are cloud infrastructure resources that can be utilized on the server of the target company.

To this end, the resource module (124) can receive information about available resources of a cloud server of the target company or a cloud server from the target company.

Available resources can be calculated as the difference between the total resources of the cloud server and the resources that the cloud server is using at the time of automatic design of the cloud infrastructure.

For example, in terms of storage, if the total resources of the cloud server are 10,000 TB and the resources of the cloud server being used at the time of automatic design of the cloud infrastructure are 5,000 TB, the available resources can be 5000 TB.

In addition, the resource module (124) can receive information on resources scheduled for use that have been applied for business from the allocation module (127) but have not been reflected in the cloud server.

In other words, if a business has already been applied for and selected, it may be desirable to exclude it from idle resources because the target company must build the cloud infrastructure presented at the time of application.

The resource module (124) can calculate idle resources as the difference between available resources and resources to be used.

The selection prediction module (125) can calculate the selection probability, which is the possibility that the target company will be selected for the desired project.

For example, the selection prediction module can calculate the selection probability based on the selection criteria and the satisfaction level, which is a value indicating how well the target company satisfies the conditions required for the desired project compared to the selection criteria.

To explain specifically, the selection prediction module can calculate the above satisfaction level by comparing the requirements stated in the request for proposal for the desired business with the functions of the product/service provided by the company.

To this end, the selection prediction module can extract requirements from the request for proposal by utilizing the extraction information model of the extraction module. In addition, the selection prediction module can extract characteristic information from the target information of the company by utilizing the characteristic information model of the function module.

In addition, the selection prediction module can determine whether the details of the requirements match the details of the characteristic information, and calculate a satisfaction level as the ratio of requirements similar to the characteristic information among all requirements.

Selection criteria can be pre-specified by the system administrator.

For example, the selection criterion may be 70%, but the present invention is not limited thereto, and the specific value of the selection criterion may be varied in various ways at a level that is obvious to a person skilled in the art.

In contrast, the selection prediction module can derive selection criteria based on past data.

As a specific example, the selection prediction module can calculate how well the products/services of the companies selected for the project satisfy the requirements of the request for proposal by comparing the company information of the past request for proposal and the company selected for the project based on the request for proposal.

In addition, selection criteria can be derived by calculating the level of satisfaction with each company's product/service for each past RFP and taking the average.

For example, if the conditions required in the RFP are 1) provision of auto scaling function, 2) high-speed input/output processing through 10G network configuration, and 3) security-related H/W and S/W must be configured as CC-certified products, but the service/product provided by the company provides auto scaling function and can perform high-speed input/output processing through 10G network configuration but has not received CC certification, the selection prediction module will determine that the company can satisfy only two of the three conditions, and the satisfaction level can be calculated as 66.6%.

The selection prediction module can evaluate the selection probability as higher the further (larger) the satisfaction level is from the selection criteria, if the satisfaction level is higher than the selection criteria.

Conversely, the selection prediction module can evaluate the selection probability as lower when the satisfaction level is lower than the selection criterion.

The selection prediction module can calculate the selection probability with a probability between 0% and 100%.

The decision module (126) can produce a modified infrastructure by modifying the target infrastructure using a predetermined modification method based on the selection probability when a predetermined modification condition is satisfied.

The predetermined modification condition may be that the sum of the target infrastructure and the cloud infrastructure allocated to the pending project for which the target company has been determined to apply for the project exceeds the idle resources at the time of the request for cloud infrastructure recommendation for the above desired project.

Here, a pending business is a business for which the target company has decided to apply for a business by deciding on a cloud infrastructure, and may mean a business with a business period that overlaps at least some dates with the business period of the desired business.

To this end, the decision module (126) can receive the cloud infrastructure assigned to the pending business from the assignment module (127).

A predetermined modification method may be a method of determining a method for modifying the cloud infrastructure of a predicted project by section of the selection probability by considering both the target infrastructure and the cloud infrastructure assigned to the pending project based on the time of the request for cloud infrastructure recommendation for the above desired project.

As a specific example, the above-determined modification method may be a method of not modifying a cloud infrastructure corresponding to a business (a desired business and a pending business) having the above-described selection probability in a first section that is less than or equal to the first probability and greater than the second probability, which is a probability lower than the first probability.

For example, the first probability might be 100%, the second probability might be 85%, and so on.

However, without limitation thereto, the specific numerical values of the first probability and/or the second probability can be varied in various ways at a level that is obvious to a person skilled in the art.

In addition, the above-determined modification method may be a method of not modifying the cloud infrastructure corresponding to the project having the above-described selection probability in the second section when the sum of the cloud infrastructures assigned to the desired project and the pending project having the above-described selection probability in the first section and the second section exceeding the third probability which is lower than the second probability but less than or equal to the second probability is less than or equal to the idle resources.

For example, the third probability might be 70%.

However, without limitation to this, the specific numerical value of the third probability can be varied in various ways at a level that is obvious to ordinary technicians.

In addition, the predetermined correction method may be a method of calculating a correction infrastructure by reducing the cloud infrastructure corresponding to the project having the selection probability in the third section based on the selection probability when the sum of the cloud infrastructures assigned to the desired project and the pending project having the selection probability in the first section and the second section exceeds the third probability which is lower than or equal to the second probability and is less than or equal to the idle resources, and at the same time, the sum of the cloud infrastructures assigned to the desired project and the pending project having the selection probability in the first section, the second section, and the third section exceeds the idle resources.

In addition, the predetermined correction method may be a method of calculating a correction infrastructure by reducing the cloud infrastructure corresponding to the project having the selection probability in the second section based on the selection probability when the sum of the cloud infrastructures assigned to the desired project and the pending project having the selection probability in the first section and the second section exceeds the third probability which is lower than the second probability and less than the second probability exceeds the idle resources.

In addition, in a case where the sum of cloud infrastructures allocated to the desired business and the pending business having the above selection probability in the first section and the second section exceeds the third probability which is lower than or equal to the second probability and less than or equal to the second probability, and at the same time exceeds the idle resources, and a business having the above selection probability exists in the third section which is lower than or equal to the third probability and greater than the fourth probability which is lower than the third probability, there may be a method of not allocating cloud infrastructure to a business having a rate of return below a predetermined standard.

Below, we will describe in detail the predetermined modification method.

The assignment module (127) can manage cloud infrastructure information for pending projects and provide it to necessary modules.

The target company can receive recommendations for target infrastructure through the construction module (123) or modified infrastructure through the decision module (126) for the business it wishes to support. The target company can receive recommendations for cloud infrastructure and proceed with an internal payment process before actually applying for the business.

The allocation module (127) can produce an interface that displays the target infrastructure or modified infrastructure for the desired project and provide it to the target company.

In other words, when a cloud infrastructure is allocated to a business, the business may be on hold until the business application is submitted, and such a business can be defined as a pending business.

The target company can transmit information about the pending business and information about the cloud infrastructure assigned to the pending business to the assignment module (127).

If there is a change in the cloud infrastructure assigned to a pending business, the assignment module (127) can change the information to the changed information and notify the target company of the changed information.

The allocation module (127) can produce an interface that displays specific information about the target infrastructure or modified infrastructure and provide it to the target company.

Specifically, if a pre-determined modification condition is not satisfied, the allocation module (127) can generate an interface in which the target infrastructure is displayed and provide it to the target company.

In contrast, if a pre-determined modification condition is satisfied, the allocation module (127) can generate an interface displaying the modification infrastructure and provide it to the target company.

Through this, target companies can intuitively understand the cloud infrastructure required for their desired business, easily write a business plan, and apply for the desired business.

The management system (130) can manage information on whether a business has been selected or not by application of a target company, and can produce an interface that displays whether or not it has been selected and provide it to the target company.

In addition, the management system (130) can manage the schedule for the selected business, and can provide the target company with an interface that displays a deadline for the schedule and an image indicating the schedule.

In addition, the management system (130) can store information on personnel of the selected business and store a list of businesses in which each personnel participated in a database.

The management system (130) can, in response to a request from a target company, produce and provide a list of business participation careers for any personnel.

Below, the business search method implemented by the business search system (110) will be described in detail.

FIG. 3 is a flowchart of a business search method implemented by a business search system of a cloud business management platform according to one embodiment of the present invention, and FIG. 4 is a diagram for explaining the operation process of a business search system according to one embodiment of the present invention.

Referring to FIGS. 3 and 4, a business search method according to an embodiment of the present invention may include, in a business search method implemented by a business search system (110), a step of collecting request for proposals posted on an online network by a collection module (111), a step of calculating extracted information from the contents of the request for proposal by a processing module (112), a step of calculating characteristic information representing the characteristics of a target company from target information, which is information about the target company, by a function module (113), a step of processing and comparing the extracted information and the characteristic information in a predetermined manner by a response module (114) to determine a comparative similarity indicating similar information, and an interface module (115) for calculating an interface in which the comparative similarity is displayed so as to express whether the target company and the request for proposal are related.

The target company can request a search of a request for proposal using a cloud business management platform (100) (business search system (110)).

At this time, the target company can select a display method and send keywords together as search information.

The collection module (111) can collect proposal requests made public through an external server.

The processing module (112) can select and produce extracted information from the contents of a request for display or a request for proposal.

In addition, the function module (113) can select and produce characteristic information from the target information of the target company.

The response module (114) can compare characteristic information and extracted information in a predetermined manner to determine comparative similarity.

Here, the response module (114) can calculate the comparative similarity only for the display request form, but the present invention is not limited thereto.

To this end, the request for proposal, the request of the target company and the search information collected by the collection module (111) can be directly transferred to the interface module (115) or the interface module (115) through the storage module (118), and the interface module (115) can select a request for display from among the request for proposal.

Additionally, lists of display requests selected by the interface module (115) can be transmitted to the processing module (112) and/or the response module (114).

The interface module (115) can display a user interface (Q10) that displays an image (Q11), comparative similarity (Q12), and/or possible display (Q13) that can distinguish each display request form selected by a pre-determined display method, and provide it to the target company.

FIG. 5 is a flowchart of an infrastructure automatic design method implemented by an infrastructure automatic design system of a cloud business management platform according to one embodiment of the present invention, and FIG. 6 is a drawing for explaining an operation process of an infrastructure automatic design system according to one embodiment of the present invention.

Referring to FIGS. 5 and 6, an infrastructure automatic design system according to an embodiment of the present invention is implemented by an infrastructure automatic design system, and may include a step of receiving, by a receiving module (121), target project information, which is information on a target request form, which is a request for proposal for a desired project that a target company wishes to support, a step of extracting, by an extraction module (122), target requirement information (R12), which is a requirement in a target request form (R11), based on the target project information, and a step of generating, by a construction module (123), a target infrastructure, which is a cloud infrastructure that can realize the target requirement information, using an infrastructure model (M10) learned by machine learning.

The target company can request a cloud infrastructure recommendation by specifying a business (desired business) to receive a cloud infrastructure recommendation, and the request signal can be received by the receiving module (121).

The receiving module (121) can receive target task information for a desired project from the project search system (110) or an external source.

The extraction module (122) can extract target requirement information from target task information, and the construction model can input target requirement information into a pre-learned infrastructure model to produce target infrastructure, which is a cloud infrastructure required for a desired project.

The allocation module (127) can produce an interface in which the target infrastructure (U11) is displayed and provide it to the target company.

Through this, target companies can easily identify the cloud infrastructure that must be included when applying for a business.

However, here, the decision module (126) can produce a modified infrastructure by modifying the target infrastructure using a predetermined modification method when a predetermined modification condition is satisfied.

This can be modified or maintained depending on the range in which the probability of selection of the desired project falls, but a detailed explanation of the predetermined modification method is provided later.

Figure 7 is a flowchart illustrating a process of producing a modified infrastructure by an infrastructure automatic design system (120) according to one embodiment of the present invention.

Referring to FIG. 7, the decision module (126) determines whether a predetermined modification condition is satisfied, and if the predetermined modification condition is satisfied, the modification infrastructure can be calculated using a predetermined modification method.

The predetermined modification method may be a method for determining a method of modifying the target infrastructure for each interval of the above selection probability.

The decision module (126) can proceed with a predetermined modification method based on the sum of the target infrastructure and the cloud infrastructure assigned to the pending project for which the target company has been determined to apply (hereinafter referred to as the planned infrastructure).

The decision module (126) can determine whether the planned infrastructure of a project with a selection probability of the first section among the desired projects and pending projects exceeds idle resources.

If the planned infrastructure of a project with a selection probability of the first section among the desired project and pending project exceeds idle resources, the decision module (126) can transmit an interface that recommends to the target company to select and support the project with a selection probability of the first section. Through this, the target company can select some projects and withdraw applications for the remaining projects so that the planned infrastructure of the project with a selection probability of the first section among the desired project and pending project does not exceed idle resources.

For example, if the selection probability of hope project A is 95%, the selection probability of pending project B is 93%, the selection probability of pending project C is 98%, the storage of the cloud infrastructure of hope project A is 100 TB, the storage of the cloud infrastructure of pending project B is 150 TB, the storage of the cloud infrastructure of pending project C is 200 TB, and the storage of idle resources is 400 TB, the decision module (126) can transmit an interface containing a message recommending that the target company cancel the application for some of the projects among projects A, B, and C.

This may be to prevent the target company's cloud server from becoming overloaded when businesses with a high probability of being selected are applied for and selected at the same time.

If the planned infrastructure of a project with a selection probability of the first section among the desired projects and pending projects is less than the idle resources, the decision module (126) may not modify the cloud infrastructure for the project with a selection probability of the first section.

For example, if the selection probability of hope project A is 95%, the selection probability of pending project B is 93%, the selection probability of pending project C is 98%, the storage of the cloud infrastructure of hope project A is 100 TB, the storage of the cloud infrastructure of pending project B is 150 TB, the storage of the cloud infrastructure of pending project C is 200 TB, and the storage of idle resources is 600 TB, the decision module (126) may not modify the cloud infrastructures assigned to projects A, B, and C.

Next, the decision module (126) can determine whether the planned infrastructure of a project with a selection probability of the first section and the second section among the desired project and the pending project is less than the idle resource.

If the planned infrastructure of a project with a selection probability of the first and second sections among the desired and pending projects is less than the idle resources, the decision module (126) may not modify the cloud infrastructure of the project with a selection probability of the first and second sections.

For example, if the selection probability of hope project A is 95%, the selection probability of pending project B is 93%, the selection probability of pending project C is 80%, the selection probability of pending project D is 60%, and the storage of the cloud infrastructure of hope project A is 100 TB, the storage of the cloud infrastructure of pending project B is 150 TB, the storage of the cloud infrastructure of pending project C is 200 TB, the storage of the cloud infrastructure of pending project D is 300 TB, and the storage of idle resources is 500 TB, the decision module (126) may not modify the cloud infrastructures assigned to projects A, B, and C.

Next, the decision module (126) can determine whether the planned infrastructure of a project having a selection probability of the first section, second section, and third section among the desired project and pending project is less than or equal to idle resources.

If the planned infrastructure of a project with a selection probability of the first, second and third sections among the desired and pending projects is less than the idle resources, the decision module (126) may not modify the cloud infrastructure of the project with a selection probability of the first to third sections.

On the other hand, if the planned infrastructure of a project having the selection probability of the first section, the second section, and the third section among the hoped-for projects and the pending projects exceeds idle resources, the decision module (126) can reduce the cloud infrastructure corresponding to the project having the selection probability of the third section based on the selection probability to produce a modified infrastructure.

As a specific example, the decision probability can lower the rate of reduction of target infrastructure as the selection probability increases, and the rate of reduction of target infrastructure can increase as the selection probability decreases.

Additionally, the decision module (126) can produce a modified infrastructure by reducing the cloud infrastructure having the selection probability of the third section.

For example, if the selection probability of hope project A is 95%, the selection probability of pending project B is 93%, the selection probability of pending project C is 80%, the selection probability of pending project D is 60%, the selection probability of pending project E is 55%, and the storage of the cloud infrastructure of hope project A is 100 TB, the storage of the cloud infrastructure of pending project B is 150 TB, the storage of the cloud infrastructure of pending project C is 200 TB, the storage of the cloud infrastructure of pending project D is 300 TB, the storage of the cloud infrastructure of pending project E is 300 TB, and the storage of idle resources is 800 TB, the decision module (126) can set the storage assigned to pending project D to be reduced to 180.43 TB and the storage assigned to pending project E to be reduced to 169.57 TB.

Here, the decision module (126) can reduce the cloud infrastructure of the business having the selection probability of the third section so that the sum of the cloud infrastructure having the selection probability of the first to third sections is equal to or less than the idle resources.

Next, the decision module (126) may not modify the cloud infrastructure corresponding to a business having a selection probability of the fourth section that is lower than or equal to the fifth probability, which is lower than the fourth probability.

This may be because, since the selection probability is low, it is good for the company's image to apply for a business by being designated as the first assigned or recommended infrastructure.

For example, the fifth probability might be 50%.

However, without limitation to this, the specific numerical value of the fifth probability can be varied in various ways at a level that is obvious to ordinary technicians.

Unlike the above, if the planned infrastructure of a project with a selection probability of the first section among the hoped-for projects and pending projects is less than or equal to idle resources, but the planned infrastructure of a project with a selection probability of the first and second sections among the hoped-for projects and pending projects exceeds idle resources, the following process may be changed depending on whether there is a hoped-for project or pending project with a selection probability of the third section.

For example, if the sum of the cloud infrastructures assigned to the desired business and the pending business having the selection probability in the first section and the second section exceeds the third probability which is lower than the second probability but less than the second probability, exceeds the idle resources and at the same time there is no business having the selection probability in the third section, the decision module (126) can reduce the cloud infrastructure corresponding to the business having the selection probability in the second section based on the selection probability to produce a modified infrastructure.

A detailed description of how to reduce infrastructure based on the selection probability may be omitted to the extent that it overlaps with the above-mentioned content.

In contrast, if the sum of the cloud infrastructures assigned to the desired business and the pending business having the selection probability in the first section and the second section exceeding the third probability which is lower than or equal to the second probability and less than or equal to the second probability exceeds the idle resources, and if there is a business having the selection probability in the third section exceeding the fourth probability which is lower than or equal to the third probability and less than or equal to the third probability, the decision module (126) may not assign cloud infrastructure to the business having a rate of return below a predetermined standard.

The decision module (126) can calculate the rate of return for desired projects and pending projects.

For example, the decision module (126) can calculate the rate of return by dividing the difference between the allocated project cost and the cost incurred in constructing the target infrastructure and the fixed cost incurred in order to proceed with the project by the allocated project cost.

For example, the decision module (126) can calculate a modified infrastructure so as not to allocate cloud infrastructure to a desired business or a pending business with a negative rate of return.

The decision module (126) can change the criteria for the excluded rate of return.

For example, the decision module (126) can calculate a rate of return criterion in which the sum of cloud infrastructures assigned to businesses with a selection probability in the first to third sections is less than idle resources, and can withdraw cloud infrastructures assigned to businesses with a selection probability in the second and third sections that do not satisfy the calculated rate of return criterion.

If the target company confirms that it plans to apply for the desired project as the target infrastructure or modified infrastructure, the target company can send information about the planned application for the desired project to the allocation module (127).

If there is a change in the cloud infrastructure of a pending project during the process of calculating the target infrastructure or modified infrastructure recommended for the desired project, the allocation module (127) can transmit information on the changed cloud infrastructure of the pending project to the target company.

Through this, target companies can effectively prevent conflicts with the cloud infrastructure resources of existing pending projects when applying for new projects.

In order to more clearly express the technical idea of the present invention, the attached drawings briefly express or omit components that are not related to or have little to do with the technical idea of the present invention.

Although the configuration and features of the present invention have been described above based on embodiments according to the present invention, the present invention is not limited thereto, and it will be apparent to those skilled in the art that various changes or modifications can be made within the spirit and scope of the present invention, and therefore it is made clear that such changes or modifications fall within the scope of the appended patent claims.

100: Cloud Business Management Platform 110: Business Search System
120: Infrastructure automatic design system 130: Management system

Claims (11)

In an infrastructure automatic design system for designing a cloud infrastructure that satisfies the requirements of a request for proposal published by an arbitrary demand agency,
A receiving module that receives target project information, which is information about a target request form, which is a request for proposal for a desired project that the target company hopes to support;
An extraction module that extracts target requirement information, which is a requirement in a target request form, based on the above target task information;
A resource module that calculates idle resources, which are cloud infrastructure resources that can be utilized on the servers of the above target company;
A construction module that generates a target infrastructure, which is a cloud infrastructure that can realize the target requirement information, within the idle resources using an infrastructure model utilizing machine learning;
A selection prediction module that calculates the selection probability, which is the possibility that the above target company will be selected for the above desired project; and
Including a decision module that produces a modified infrastructure by modifying the target infrastructure using a predetermined modification method based on the selection probability when a predetermined modification condition is satisfied;
The above decision module is,
The above modified infrastructure is calculated so that the sum of cloud infrastructures assigned to businesses with a selection probability in sections 1 to 3 is less than idle resources,
The above pre-determined modification conditions are:
As of the time of requesting a cloud infrastructure recommendation for the above-mentioned desired project, the sum of the above-mentioned target infrastructure and the cloud infrastructure allocated to the pending project for which the above-mentioned target company has been determined to apply for the project is the condition of exceeding the above-mentioned idle resources.
The above predetermined modification method is,
The above modified infrastructure calculated is different depending on the existence of a business having a selection probability of the third section, and the sum of the cloud infrastructures allocated to the desired business and the pending business having the above selection probability in the first section exceeding the second probability which is lower than the first probability and lower than the first probability and in the second section exceeding the third probability which is lower than the second probability exceeds the idle resources, and in the third section exceeding the fourth probability which is lower than the third probability and lower than the third probability, a method of not allocating cloud infrastructure to a business having a rate of return below a predetermined standard, when a business having the above selection probability exists in the third section exceeding the fourth probability which is lower than the third probability and lower than the third probability.
Automated infrastructure design system.
In the first paragraph,
The above resource module is,
The available resources are calculated as the difference between the resources used by the cloud server at the time of automatic design of the cloud infrastructure from the total resources of the cloud server, and the idle resources are calculated as the difference between the resources to be used for which the application for the business has been completed but not reflected in the cloud server.
Automated infrastructure design system.
In the first paragraph,
The above construction module is,
The infrastructure model is learned through machine learning based on the request for proposal for the past implemented business and the cloud infrastructure.
By comparing the expected cloud infrastructure with the cloud infrastructure specified in the past project using the request for proposal for the past project and the infrastructure model, the infrastructure model is re-learned.
Automated infrastructure design system.
In the first paragraph,
The above selection prediction module is,
By calculating whether the company's products and services satisfy the requirements of the request for proposals for each past request for proposals and averaging them, the selection criteria are derived, and the satisfaction level, which is the value indicating whether the target company satisfies the conditions required for the desired project compared to the selection criteria, and the selection probability is derived based on the selection criteria.
Automated infrastructure design system.
In the first paragraph,
The above predetermined modification method is,
A method for determining how to modify cloud infrastructure for each interval of the above selection probability.
Automated infrastructure design system.
In paragraph 5,
The above predetermined modification method is,
A method in which the cloud infrastructure corresponding to the business having the above selection probability in the above first section is not modified.
Automated infrastructure design system.
In Article 6,
The above predetermined modification method is,
A method of not modifying the cloud infrastructure corresponding to the project having the above selection probability in the second section when the sum of the cloud infrastructure assigned to the above desired project and the above pending project having the above selection probability in the first section and the above second section is less than or equal to the idle resources.
Automated infrastructure design system.
In Article 7,
The above predetermined modification method is,
The sum of the cloud infrastructure allocated to the above desired project and the above pending project with the above selection probability in the above first section and the above second section is less than or equal to the above idle resources,
A method of calculating a modified infrastructure by reducing the cloud infrastructure corresponding to the project with the above selection probability in the third section based on the above selection probability, when the sum of the cloud infrastructures assigned to the above desired project and the pending project with the above selection probability in the first section, the second section and the third section exceeds the idle resources.
Infrastructure automated design system.
In Article 6,
The above predetermined modification method is,
When there is no business with the above selection probability in the above 3rd section, if the sum of the cloud infrastructure allocated to the desired business and the pending business with the above selection probability in the above 1st section and the above 2nd section exceeds the idle resources,
A method of calculating a modified infrastructure by reducing the cloud infrastructure corresponding to a business having the above selection probability in the above second section based on the above selection probability.
Automated infrastructure design system.
In the first paragraph,
If the above predetermined modification conditions are satisfied, an allocation module is further included that generates an interface in which the above modification infrastructure is displayed and provides it to the target company;
The above allocation module is,
If there is a change in the cloud infrastructure assigned to the above-mentioned pending business, the information will be changed and the changed information will be notified to the above-mentioned target company.
Automated infrastructure design system.
An infrastructure automatic design method for designing a cloud infrastructure that satisfies the requirements of a request for proposal published by an arbitrary demand agency, implemented by an infrastructure automatic design system.
A step in which target project information, which is information on a target request form, which is a request for proposal for a desired project that the target company wishes to support, is received by the receiving module;
A step in which target requirement information, which is a matter required in a target request form, is extracted based on the target task information by an extraction module;
A step in which idle resources, which are cloud infrastructure resources that can be utilized on the server of the target company, are produced by the resource module;
A step of generating a target infrastructure, which is a cloud infrastructure in which the target requirement information can be realized, within the idle resources by using an infrastructure model utilizing machine learning through a construction module;
A step of calculating the selection probability, which is the possibility that the target company will be selected for the desired project, by the selection prediction module; and
A step of producing a modified infrastructure by modifying the target infrastructure using a predetermined modification method based on the selection probability when a predetermined modification condition is satisfied by a decision module;
The steps for calculating the above modified infrastructure are:
The above modified infrastructure is calculated so that the sum of the cloud infrastructure assigned to the business with the above selection probability in the first to third sections is less than the idle resources.
The above pre-determined modification conditions are:
As of the time of requesting a cloud infrastructure recommendation for the above-mentioned desired project, the sum of the above-mentioned target infrastructure and the cloud infrastructure allocated to the pending project for which the above-mentioned target company has been determined to apply for the project is the condition of exceeding the above-mentioned idle resources.
The above predetermined modification method is,
The above modified infrastructure calculated is different depending on the existence of a business having a selection probability of the third section, and the sum of the cloud infrastructures allocated to the desired business and the pending business having the above selection probability in the first section exceeding the second probability which is lower than the first probability and lower than the first probability and in the second section exceeding the third probability which is lower than the second probability exceeds the idle resources, and in the third section exceeding the fourth probability which is lower than the third probability and lower than the third probability, a method of not allocating cloud infrastructure to a business having a rate of return below a predetermined standard, when a business having the above selection probability exists in the third section exceeding the fourth probability which is lower than the third probability and lower than the third probability.
A method for automated infrastructure design.

Images