KR102635363B1 - Code generation system based on interactive chatbot - Google Patents

Abstract

Disclosed is a conversational chatbot-based AI code generation system that includes a control server that sets development content through interaction with a developer using a conversational chatbot and generates code for the development content.

Description

Chatbot-based AI code generation system for conversation {CODE GENERATION SYSTEM BASED ON INTERACTIVE CHATBOT}

The present invention relates to a conversation chatbot-based AI code generation system.

Recently, in order to secure corporate competitiveness through digital transformation, companies are pursuing the integration of automation and management of corporate workflow application development using new technologies and environments such as cloud environments, artificial intelligence, robotics, agile, and DevOps. In particular, when developing applications, companies are seeking ways to prototype, develop, test, and deploy new workflow applications that improve productivity and quality in a short period of time. One of the core needs of these companies is an automatic software development method that can save as much time and cost as possible when developing software needed for the company's workflow applications. It is widely known that the coding method of the program required for automatic software development is to save time and cost by choosing a low coding or no coding method from the existing developer's manual coding method. These low-coding or no-coding methods are a type of program automatic coding method and are emerging as the core of automatic software development methods.

As a program automatic coding method, the ODBC (Open Database Connectivity) program automatic coding method is disclosed in Korean Patent Publication No. 10-2014-0052459 (publication date: May 7, 2014).

In the above disclosed invention, since all ODBC program coding processes according to the prior art are performed manually, first, there are problems in that coding of ODBC programming becomes complicated, coding of ODBC programming takes a lot of time, and developer mistakes occur when coding of ODBC programming. Considering the high possibility of errors occurring due to ODBC programming coding errors and the difficulty in finding the part where the error occurred, a method of automating the ODBC programming coding process using database structures and SQL query statements is adopted during ODBC programming. In the ODBC program automatic coding method that reduces the development period by reducing coding time and prevents errors during coding, a database table structure is used to code a header file to declare the type and length of data to be stored in the database. Declaration coding stage; A first coding automation step of generating a source code file for easily creating a declaration for coding automation by using the file coding the database table structure declaration as an input file; Creating a declaration file for secondary coding automation using the generated source code file; It is a configuration that includes a second coding automation step of executing result files for new coding automation using the written declaration files as input files.

In addition, as an invention related to a specific-purpose software development platform, the invention of a power software development platform disclosed in Republic of Korea Patent Publication No. 10-2019-0044981 (publication date: May 2, 2019).

In the above disclosed invention, software platform element technologies are being developed overlappingly, technology sharing is difficult, and individually constructed ICT infrastructure has the problem of maintaining the infrastructure at the time of construction even if the users of the system or the scale of the business are reduced. Considering that this may cause the utilization rate of infrastructure to decrease and costs to increase, a platform adapter that provides a connection environment for a plurality of communication devices each installed in a plurality of power facilities; a platform gateway that collects data through a plurality of power systems to which the plurality of communication devices are linked; a platform interface that provides distribution services for data collected by the platform gateway; A platform data hub that receives data through the distribution service, verifies the received data, and provides an access environment for data verified by the application developer; An analysis abstraction service provided to the application developer by converting a plurality of physical analysis engines into an abstracted service form; and an application gateway that receives the application from the application developer and provides the application to the application user. An invention including is disclosed.

In addition, the invention of a method and computing system for automatically generating embedded software on a virtualization system is disclosed in Republic of Korea Patent Publication No. 10-2019-0075508 (publication date: July 1, 2019).

The disclosed invention includes the steps of providing a user interface tool that allows a user to edit source code in a web-based cloud development environment; Running a simulation of the source code edited using the user interface tool on a first virtual system, and feeding back the simulation results of the source code to the user using the user interface tool; providing prebuilt software created in advance based on an operating system for operating embedded software on a target board on a second virtual system; and combining the source code with the prebuilt software previously created for the target board to generate a binary result that can be executed on the target board on a second virtual system.

The first disclosed invention is an invention to automatically code a company's software to eliminate the inefficiency of manual coding, the second disclosed invention is an invention regarding a platform for integrating and developing various power-related software, and the third is A public invention relates to an invention that seeks to develop software by providing a software development interface to users in a cloud environment.

However, the above disclosed inventions only automatically generate coding, develop various workflows on an integrated platform, and provide a development interface in a cloud environment, but it is difficult to generate the entire code for the project or operate code development through interaction. there was.

One aspect of the present invention discloses an interactive AI code generation system based on a chatbot for conversation.

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

The system of the present invention includes a control server that sets development content through interaction with a developer using a chatbot for conversation and generates code for the development content.

Meanwhile, the control server receives messages about at least one function to be implemented in the software from the developer through an interactive chatbot, analyzes the messages received from the developer, creates a development project, and performs each function in the software. a multi-file creation unit that generates a multi-file by creating at least one subfolder to do so; An additional function that receives a message from a developer about additional functions to be included in the development project created by the multi-file creation unit and sets detailed functions for the additional functions by sending a reverse question about the additional functions through an interactive chatbot. Function setting unit; and a code generation unit that generates code for the development project created by the multi-file generation unit and the additional function setting unit using a code generation artificial intelligence neural network.

Meanwhile, the multi-file generator stores the code generation history through the conversation chatbot for each developer to determine the average length of the message for each developer, at least one identification word frequently used by the developer, and the number of identification words used by the developer in the chatbot conversation. Extract, request and receive developer information before executing the conversation chatbot, analyze the conversation message through the developer's conversation chatbot through Equation 1 below, and calculate the transmission message similarity,

[Equation 1]

(In Equation 1, m _ab is the transmitted message similarity, M _a is the average length of the message transmitted by the developer in the current conversation, M _a ' is the number of developer-identified words included in the current conversation, and M _b is the transmitted message of the developer The average length of, M _b 'means the number of identifying words that developers use in chatbot conversations.)

According to the transmission message similarity, authentication is performed as to whether the developer currently participating in the conversation chatbot is the developer corresponding to the developer information. Developer authentication is approved only when the transmission message similarity is 0.5 or more, and the transmission message similarity is 0.5 or more. If it is less than 0.5, the execution of the conversation chatbot can be paused, additional authentication is performed through email authentication or mobile authentication, and the conversation chatbot can be re-run if developer authentication is approved.

According to one aspect of the present invention described above, the development content is confirmed through interaction with the developer using an interactive chatbot, and an appropriate code is generated using a code generation artificial intelligence neural network, thereby providing high quality IT services without a developer. In that it can be developed, it has the advantage of saving time and cost compared to existing development methods.

1 is a conceptual diagram of a system according to an embodiment of the present invention.
Figure 2 is a conceptual diagram of a control server according to an embodiment of the present invention.

The advantages and features of the present invention and methods for achieving them will become clear by referring to the embodiments described in detail below along with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below and will be implemented in various different forms. The present embodiments only serve to ensure that the disclosure of the present invention is complete and that common knowledge in the technical field to which the present invention pertains is not limited. It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout the specification.

The terminology used herein is for describing embodiments and is not intended to limit the invention. As used herein, singular forms also include plural forms, unless specifically stated otherwise in the context. As used in the specification, “comprises” and/or “comprising” does not exclude the presence or addition of one or more other components, steps and operations.

1 is a conceptual diagram of a system according to an embodiment of the present invention.

Referring to FIG. 1, a system 1 according to an embodiment of the present invention may include a control server 100 and a mobile terminal 200.

The control server 100 can provide a conversation chatbot-based AI code generation service according to the present invention.

The control server 100 may be its own server, a cloud server, or a p2p (peer-to-peer) set of distributed nodes for providing the service according to the present invention.

The control server 100 may perform one or more of the calculation, storage, reference, input/output, and control functions of a general computer, and may include an artificial neural network to be described later based on input data.

Control server 100 may include a processor and memory. The processor may perform the conversation chatbot-based AI code generation service according to the present invention, and may include devices capable of performing this. The processor may execute a program or control the control server 100. Program code executed by the processor may be stored in memory. The memory may store related information for performing a service according to the present invention or a program for implementing a method. The memory may be volatile memory or non-volatile memory.

The control server 100 may send data to or receive data from an external device using a network.

The control server 100 can train an artificial neural network or use a trained artificial neural network. The processor can train or execute an artificial neural network stored in memory, and the memory can store a trained artificial neural network. The electronic device that trains the artificial neural network and the electronic device that uses it may be the same, but may also be separate.

Artificial intelligence is a computer system that implements some of the functions of the human brain and can learn, guess, and make decisions on its own. As learning progresses, the probability of extracting an answer may increase. Artificial intelligence can be composed of learning and elemental technologies using it. Artificial intelligence learning is an algorithmic technology that classifies and learns features based on input data, and elemental technologies may be technologies that implement some of the functions of the human brain using learning algorithms.

Artificial intelligence is a technology that makes it easy to approach problems that can have multiple answers probabilistically, and can logically and probabilistically infer the optimal cycle, method, and plan according to any input data. Artificial intelligence's inference technology can include judging input data, optimization predictions, knowledge and probability-based reasoning, and preference-based planning.

Artificial neural network is one of the learning algorithms in the machine learning field and is a program that implements the connection between neurons and synapses in the brain. Artificial neural networks can be created through a program to create a neural network structure and then learn it to have the desired function. Although there may be errors, it is possible to learn based on huge data and output appropriate output data with input data. It has the advantage of being able to obtain output data with statistically good results and being similar to human reasoning.

Artificial intelligence chatbot (chatbot) is based on an artificial intelligence chatbot API providing platform. Types of artificial intelligence chatbot API providing platforms include, for example, IBM Watson AI API, SK Avril, conversation service (Conversation), and natural language understanding. It includes application program interfaces (APIs) such as services (Natural Language Understanding) and natural language classification services (Natural Language Classifier).

The type of artificial intelligence chatbot API providing platform used in this embodiment is not limited, and it is possible to use a variety of artificial intelligence chatbot API providing platforms.

The interactive chatbot-based AI code generation system according to this embodiment can determine the details of the function and generate multi-files by interacting with software development based on the artificial intelligence chatbot API providing platform, and can generate multi-files through an artificial intelligence neural network. You can generate code for multiple files.

The network is a communication network that is a high-speed backbone network of a large communication network capable of high-capacity, long-distance voice and data services, and may be a next-generation wired or wireless network to provide the Internet or high-speed multimedia services.

If the network is a mobile communication network, it may be a synchronous mobile communication network or an asynchronous mobile communication network. An example of an asynchronous mobile communication network may be a WCDMA (Wideband Code Division Multiple Access) type communication network. In this case, although not shown in the drawing, the network may include an RNC (Radio Network Controller). Meanwhile, although the WCDMA network was used as an example, it can be a 3G LTE network, a 4G network, other next-generation communication networks such as 5G, and other IP-based IP networks.

The mobile terminal 200 may include any terminal that can send and receive data through a network, such as a desktop computer, laptop, tablet, or smartphone.

The mobile terminal 200 may include one or more of the calculation function, storage function, reference function, input/output function, and control function of a computer in order to perform the service according to the present invention.

The mobile terminal 200 can access a website or install an application to receive services according to the present invention. The mobile terminal 200 can exchange data through a website or application.

The system (1) according to an embodiment of the present invention uses an interactive chatbot to confirm development details through interaction with the developer, and generates appropriate code using a code generation artificial intelligence neural network to create high-quality In that IT services can be developed without developers, it has the advantage of saving time and costs compared to existing development methods.

Figure 2 is a conceptual diagram of a control server according to an embodiment of the present invention.

Referring to FIG. 2, the control server 100 according to an embodiment of the present invention may include a multi-file creation unit 110, an additional function setting unit 120, and a code generation unit 130.

The multi-file creation unit 110 can create a software development project through interaction with a developer based on an interactive chatbot and generate multi-files to perform at least one function to be implemented in the software.

Artificial intelligence chatbot (chatbot) is based on an artificial intelligence chatbot API providing platform. Types of artificial intelligence chatbot API providing platforms include, for example, IBM Watson AI API, SK Avril, conversation service (Conversation), and natural language understanding. It includes application program interfaces (APIs) such as services (Natural Language Understanding) and natural language classification services (Natural Language Classifier).

The type of artificial intelligence chatbot API providing platform used in this embodiment is not limited, and it is possible to use a variety of artificial intelligence chatbot API providing platforms.

The multi-file creation unit 110 may receive a message about at least one function to be implemented in software from a developer through an interactive chatbot.

The multi-file creation unit 110 can create a project by analyzing a message received from a developer through an interactive chatbot, and create a multi-file by creating at least one subfolder to perform each function in the software.

For example, the multi-file creation unit 110 can create a development project called “Simple Shopping Mall” based on a message written by a developer through an interactive chatbot, and includes “auto”, “config”, “user”, Models such as “migrations” and “venv”, test files, view files, etc. can be created organically.

The additional function setting unit 120 can receive a message from the developer about additional functions to be included in the development project created by the multi-file creation unit 110, and sends a reverse question about the additional functions through an interactive chatbot. You can set detailed functions for additional features.

For example, the additional function setting unit 120 may receive a message about an additional function called “email authentication when signing up” from the developer, and through an interactive chatbot, it may ask “Can I just forward the email link and click it?” It provides reverse questions about “response, reaction when failure, whether you can log in until email authentication, email validity functions,” etc., and by receiving a response message to the reverse question, you can set detailed functions for additional functions.

The additional function setting unit 120 may create at least one subfolder to perform the set additional function and add it to the development project.

The code generation unit 130 may generate code for the development project created by the multi-file generation unit 110 and the additional function setting unit 120 using a code generation artificial intelligence neural network.

The code generator 130 may include an artificial intelligence neural network for code generation.

In this embodiment, the code generator 130 may include a plurality of pre-trained artificial neural networks to perform a machine learning algorithm. With machine learning, you can output output data based on input data and use the results to learn on your own, which can improve your own data processing ability. Artificial neural networks can extract features based on input data, infer regularities, and output result data. As this process accumulates, the reliability of the result data increases.

In this embodiment, the artificial neural network may be an algorithm that outputs code for a development project. Artificial neural networks can infer the best output data using big data as input data or after going through a processing process to organize unnecessary data.

ve Bayes Classifier, Hidden Markov Model, K-Means Clustering 등이 사용될 수 있다.Depending on the type of learning, artificial intelligence machine learning models include Super Viser Learning, UnSuper Viser Learning, Semisupervised Learning, and Reinforcement Learning. And machine learning algorithms include Decision Tree, K-Nearest Nearest Bor, Artificial Neural Network, Support Vector Machine, Ensemble Learning, Gradient Descent, Na

ve Bayes Classifier, Hidden Markov Model, K-Means Clustering, etc. can be used.

Artificial neural networks can stack and connect numerous artificial neurons in several layers. The artificial neural network may be pre-trained on various input values that may be included in the input data. Artificial neural networks can infer weak points according to data input by the user and output output data.

An artificial neural network may be an artificial neural network that is learned according to reinforcement learning, one of the learning methods. Reinforcement learning is a method of gradually increasing the probability of obtaining the correct result by setting rewards and limits.

Artificial neural networks can also be modeled based on Convelutional Neural Network (CNN) or Recurrent Neural Network (RNN).

Meanwhile, the multi-file creation unit 110 can perform developer authentication by analyzing the developer's interaction through the conversation chatbot.

For example, the multi-file generator 110 may request and receive developer information before executing the chatbot for conversation, and may authenticate whether the developer corresponding to the developer information is actually a developer participating in the chatbot for conversation.

For example, the multi-file generator 110 can calculate the similarity of the developer's transmitted message using Equation 1 below. The multi-file creation unit 110 can store and analyze the code creation history through the conversation chatbot for each developer, the average length of the message for each developer, at least one identification word frequently used by the developer, and the developer's identification in the chatbot conversation. The number of words used can be extracted and stored as developer identification information.

[Equation 1]

In Equation 1, m _ab is the transmitted message similarity, M _a is the average length of the message transmitted by the developer in the current conversation, M _a ' is the number of developer identification words included in the current conversation, and M _b is the length of the developer's transmitted message. The average length, M _b ', refers to the number of identifying words that developers use in chatbot conversations.

For example, if M _a =53, M _a '=5, M _b =73, and M _b '=6, m _ab =1 can be calculated, and in this case, developer authentication can be approved. At this time, the multi-file generator 110 can approve developer authentication only when the transmitted message similarity is 0.5 or more. If the transmitted message similarity is less than 0.5, the execution of the conversation chatbot is temporarily suspended and email authentication, mobile authentication, etc. After performing additional authentication, if the developer authentication is approved, the chatbot for conversation can be re-run.

In this way, the multi-file generator 110 can increase the accuracy and reliability of development content by identifying the developer currently having a chatbot conversation as the developer who requested authentication.

Meanwhile, the additional function setting unit 120 may obtain target user information from the developer through the mobile terminal 200 and set the number of focused conversation categories for setting detailed functions according to the target user information.

For example, the additional function setting unit 120 provides the average age of the target user, the average number of times the target user uses the software per day, the average time the target user uses the software per day, and the expected size of the target user through the mobile terminal 200. It may include etc.

The intensive conversation category may include user response waiting time, user response request method, user access restriction method, reaction method to user response, etc.

The additional function setting unit 120 can calculate a category identification number by applying target user information to Equation 2 below.

[Equation 2]

In Equation 2, Ci is the category identification number, n is a constant set in proportion to the expected size of target users, u _n is the average age of target users, t _c is the average number of software uses by target users, and c _n is the average number of uses of the software by target users. This refers to the average usage time of the software.

로 산출될 수 있다. 이와 같은 경우, 추가 기능 설정부(120)는 적어도 5개 이상의 카테고리에 대하여 챗봇 대화가 이루어질 수 있도록 한다. 즉, 추가 기능 설정부(120)는 소프트웨어의 타겟 사용자 정보에 맞추어 세부 기능 설정을 위한 챗봇 대화 카테고리를 설정함으로써 소프트웨어의 개발 완성도를 높일 수 있다.For example, if u _n =30, n=5, t _c =10, c _n =120, C _i =0.48

It can be calculated as In this case, the additional function setting unit 120 allows chatbot conversations to occur for at least five or more categories. In other words, the additional function setting unit 120 can improve the development completeness of the software by setting the chatbot conversation category for setting detailed functions according to the target user information of the software.

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

1: Chatbot-based AI code generation system for conversation
100: control server
200: mobile terminal

Claims (3)

Includes a control server that sets development content through interaction with the developer using a chatbot for conversation and generates code for the development content;
The control server is,
Receive messages about at least one function to be implemented in the software from the developer through an interactive chatbot, create a development project by analyzing the messages received from the developer, and at least one subfolder to perform each function in the software. A multi-file creation unit that generates a multi-file;
An additional function that receives a message from a developer about additional functions to be included in the development project created by the multi-file creation unit and sets detailed functions for the additional functions by sending a reverse question about the additional functions through an interactive chatbot. Function setting unit; and
It includes a code generation unit that generates code for the development project created by the multi-file generation unit and the additional function setting unit using a code generation artificial intelligence neural network,
The multi-file creation unit,
By storing the code creation history through conversation chatbots for each developer, we extract the average length of each developer's message, at least one identification word frequently used by the developer, and the number of times the developer uses the identification word in chatbot conversations, and run the conversation chatbot. Request and receive developer information beforehand, calculate the transmission message similarity by analyzing conversation messages through the developer's conversation chatbot using Equation 1 below,
[Equation 1]

(In Equation 1, m _ab is the transmitted message similarity, M _a is the average length of the message transmitted by the developer in the current conversation, M _a ' is the number of developer-identified words included in the current conversation, and M _b is the transmitted message of the developer The average length of, M _b 'means the number of identifying words that developers use in chatbot conversations.)
According to the transmission message similarity, authentication is performed as to whether the developer currently participating in the conversation chatbot is the developer corresponding to the developer information. Developer authentication is approved only when the transmission message similarity is 0.5 or more, and the transmission message similarity is 0.5 or more. A conversation chatbot-based AI code generation system that temporarily suspends the execution of the conversation chatbot if it is less than 0.5, performs additional authentication through email authentication or mobile authentication, and then re-runs the conversation chatbot if developer authentication is approved.
delete delete