JP7477698B1 - Information processing device, information processing method, and program. - Google Patents

Abstract

[Problem] To easily generate desired code. [Solution] An information processing device comprising: an acquisition unit that acquires a document including a method for using a library in a specified programming language and sample code for the library, and a task definition that instructs the user to generate sample code using the library; and a processing unit that acquires sample code information corresponding to the task definition in a manner following the method for using the library, the sample code information being obtained by inputting the document and the task definition into an AI chatbot that is trained to receive text instructions including program code and output information according to the text instructions. [Selected Figure] Figure 2

Description

The present invention relates to an information processing device, an information processing method, and a program.

A program generation device has been disclosed that includes a generation unit that generates a composite program based on one or more program parts, a visualization unit that visualizes the contents of the composite program, and a regeneration unit that regenerates the composite program based on the one or more program parts when a user inputs a predetermined instruction for the visualized composite program (see, for example, Patent Document 1).

JP 2022-17685 A

With conventional technology, it was sometimes difficult to generate the desired code.

The present invention has been made in consideration of these circumstances, and one of its objectives is to provide an information processing device, an information processing method, and a program that can easily generate desired code.

One aspect of the present invention is an information processing device that includes an acquisition unit that acquires a document including a method for using a library in a specified programming language and sample code for the library, and a task definition that instructs the system to generate sample code using the library, and a processing unit that acquires information about the sample code corresponding to the task definition in a manner that conforms to the method for using the library by inputting the document and the task definition into an AI chatbot that is trained to receive text instructions including program code as input and output information in accordance with the text instructions.

According to one aspect of the present invention, it is possible to provide an information processing device, an information processing method, and a program that can easily generate desired code.

FIG. 1 is a diagram illustrating an example of a configuration of an information processing system 1. FIG. 1 is a diagram illustrating a configuration of an information processing device 100. 2 is a diagram for explaining an overview of the processing of a summarizing unit 110. FIG. 10 is a diagram for explaining an overview of processing by a processing unit 130. FIG. 1A and 1B are diagrams illustrating an example of sample code (1) generated without using this embodiment and sample code (2) generated using this embodiment. A sequence diagram showing an example of the flow of processing performed by the information processing device 100 and the AI chatbot server 200.

Below, an embodiment of a control method for an information processing device of the present invention will be described with reference to the drawings. Various devices for providing services to users and performing internal analysis, such as the "information processing device" that will be described below, may be realized by a distributed group of devices, and each device may be operated by a different business operator. In addition, the owner of the hardware for the device (the provider of the cloud server) and the business that actually operates it may also be different.

Figure 1 is a diagram showing an example of the configuration of an information processing system 1. The information processing system 1 includes, for example, an information processing device 100 and an AI chatbot server 200. These communicate via a network NW. The network NW includes, for example, the Internet, a LAN (Local Area Network), a wireless base station, a provider device, etc.

[AI Chatbot]
The AI chatbot server 200 is a model that periodically obtains information from an information source 300 such as an Internet site by crawling, and is trained to reply according to the text when a text is input based on common knowledge. The AI chatbot server 200 is known under the names of ChatGPT (GPT-4), Alpaca, Vicuna, Raven RWKV, Marvin, etc., for example. The AI chatbot server 200 is a model that is trained to receive, for example, a text instruction including a program code as an input, and output information according to the text instruction. When a document and a task definition are input, the AI chatbot server 200 outputs sample code information corresponding to the task definition in a manner according to the usage of the library. For example, the AI chatbot server 200 is a model that is trained to receive, for example, a document, a task definition, and a text instruction, and output sample code information corresponding to the task definition in a manner according to the usage of the library according to the instruction. When a document, a task definition, and an instruction are input to the AI chatbot server 200, the AI chatbot server 200 outputs information on a sample code corresponding to the task definition in a manner that conforms to the usage of the library according to the instruction. In the above, sample data may be input, and sample code incorporating the sample data may be output.

[Information processing device]
FIG. 2 is a configuration diagram of the information processing device 100. The information processing device 100 includes, for example, a summarizing unit 110, an acquiring unit 120, a processing unit 130, a providing unit 140, and a storage unit 170. The components other than the storage unit 170 are realized by, for example, a hardware processor such as a CPU (Central Processing Unit) executing a program (software). Some or all of these components may be realized by hardware (including circuitry) such as an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a GPU (Graphics Processing Unit), or may be realized by cooperation between software and hardware. The program may be stored in advance in a storage device such as an HDD (Hard Disk Drive) or a flash memory (a storage device having a non-transient storage medium), or may be stored in a removable storage medium such as a DVD or a CD-ROM (a non-transient storage medium), and may be installed in the storage device by mounting the storage medium in a drive device.

The storage unit 70 is a HDD, a flash memory, a RAM (Random Access Memory), etc. The storage unit 70 may be a NAS (Network Attached Storage) device that the information processing device 100 can access via a network.

The storage unit 70 stores, for example, document information 172, summary information 174, definition information (task definition, schema definition) 176, generation information 178, and the like. The document information 172 is a document in a specific programming language. For example, the document information 172 is information in Structured Query Language (SQL). Details of this information will be described later.

(Summary)
The summary unit 110 divides the document information 172 and inputs each of the divided documents in turn to the AI chatbot server 200, thereby obtaining summary information that summarizes the input divided documents.

The document information 172 includes, for example, document information related to a specific version of the library (e.g., how to use the library), sample code (e.g., code such as Java), points to note in a specific version, FAQ, a list of changes up to the specific version, and the like. Some of the above information may be omitted. The document information 172 may also include, for example, information such as sample data. The sample code is sample code for implementing the functions of the library. The sample data is sample data used in the sample code. There is a limit to the amount of data (e.g., the number of tokens) that the AI chatbot server 200 can process at one time. If the amount of data in the document information 172 exceeds the limit, the summarizing unit 110 divides the document so that the amount of data does not exceed the limit. The summarizing unit 110 inputs the divided documents in order to the AI chatbot server 200 and obtains a summary of the document output by the AI chatbot server 200.

3 is a diagram for explaining an overview of the processing of the summarizing unit 110. For example, the summarizing unit 110 divides a document into three parts, inputs the document into the AI chatbot server 200 in three separate inputs, and obtains a summary for each input. The summarizing unit 110 generates summary information 174 that is summarized and encoded based on the summarized document, and stores the generated summary information 174 in the storage unit 170. This is a preprocessing step for re-processing in the AI chatbot server 200, as described later. The method of summarizing and encoding is determined by the internal implementation of the AI chatbot server 200. The summary information 174 is, for example, information that integrates summaries. For example, the summary information 174 is a single document that is formed by sequentially connecting summaries corresponding to the divided documents output by the AI chatbot server 200. The above document includes, for example, document information related to a specific version of the library (for example, how to use the library), sample code, points to note in a specific version, FAQ, a list of changes up to the specific version, and the like. Some of the above information may be omitted. The document may also include information such as sample data.

(Acquisition Department)
The acquiring unit 120 acquires the summary information 174 (document information 172) and the definition information 176 from the storage unit 170. When the summary information 174 or the definition information 176 is stored in another storage unit or provided from another device, the acquiring unit 120 acquires it from the other storage unit or from another device.

Definition information 176 is a task definition that instructs the generation of sample code using a library (document information 172). Definition information 176 is, for example, information that defines the structure and properties of database data, the relationships between data, and rules for operating the database. Definition information 176 is, for example, information that defines how data is specifically stored. Definition information 176 is, for example, information that indicates the framework of the database.

(Processing section)
The processing unit 130 obtains information on sample code corresponding to the task definition in a manner conforming to the usage of the library according to the instructions, which is obtained by inputting a document, a task definition, and an instruction to the AI chatbot server 200. For example, the processing unit 130 inputs summary information 174 and definition information 176 to the AI chatbot server 200, and instructs the AI chatbot server 200 to output sample code that applies sample data of the document conforming to the usage of the library input to the definition information 176.

FIG. 4 is a diagram for explaining an overview of the processing of the processing unit 130. The processing unit 130 acquires the generation information 178 output by the AI chatbot server 200. The generation information 178 is sample code that applies sample data of a document in accordance with the usage of the library input to the definition information 176. The sample code is sample code in which sample data is incorporated into the framework of the code. For example, sample code in which sample data in accordance with the usage of the library is incorporated is generated in "i0 (AR1 in the figure)" ... "i8 (AR2 in the figure)" of the definition information 176 in FIG. 4. The sample code is code that can test whether the function of the library is realized using sample data, or code that supports the administrator in creating code. The sample code may be partially missing or partially defective. In this case, the administrator checks and corrects the sample code to eliminate the missing or defective parts. Specifically, the processing unit 130 instructs the AI chatbot server 200 to automatically generate program code using the library. In FIG. 4, as an example, the processing unit 130 passes definition information of an SQL table and instructs the processing unit 130 to generate a usage example of the library code accordingly. Note that in the above example, instead of the sample code into which the sample data has been imported, the sample code before the sample data was imported may be generated.

(Providing Department)
The providing unit 140 displays the sample code acquired by the processing unit 130 on a display unit (not shown) or provides it to another terminal device. Furthermore, the providing unit 140 may execute the sample code to check whether it operates normally.

In this embodiment, the AI chatbot server 200 has been described as having a limit on the amount of data it can handle, but if there is no limit on the amount of data it can handle or if the target document can be processed without being divided, the processing performed by the summarization unit 110 and the functional configuration of the summarization unit 110 may be omitted. In this case, the processing unit 130 inputs the document information 172 and the definition information 176 to the AI chatbot server 200, requests the AI chatbot server 200 to output sample code that uses sample data in accordance with how to use the library of the input document, and obtains the above sample code from the AI chatbot server 200.

The processing of each of the above functional units may be performed based on the user's operation. For example, the summarizing unit 110, the acquiring unit 120, the processing unit 130, or the providing unit 140 may perform processing in response to the user's operation, or each processing may be performed automatically.

As described above, the information processing device 100 automatically generates sample code using documents, making it easy to generate the desired code.

For example, you may be given a document and asked to quickly create sample code using the library in the document. It may be a lot of work to go through the entire document and create the sample code.

Even when trying to use an AI chatbot, it may be difficult to create the desired sample code. For example, the AI chatbot server 200 may have learned information before a specific document was made public, but may not have learned about the specific document. In this case, even if the AI chatbot server 200 is used, it may not be possible to create sample code that uses the library of the specific document.

In this embodiment, as described above, even if the AI chatbot has not studied a specific document, the information processing device 100 can easily generate sample code that conforms to the contents of the specific document by inputting summary information 174 (document information 172) and definition information 176 to the AI chatbot server 200.

Furthermore, even if the AI chatbot cannot process the entire document at once, in this embodiment, the information processing device 100 can easily create sample code by dividing the document and using the divided documents.

FIG. 5 is a diagram showing an example of sample code (1) generated without using this embodiment and sample code (2) generated using this embodiment. Sample code (1) is sample code generated without inputting a specific document. Sample code (1) does not reflect the contents of a specific (latest) document. That is, sample code (1) does not reflect the contents of a recent document, and code of a non-existent library generated independently by the AI chatbot server 200 is described. In contrast, in sample code (2), a specific document is provided to the AI chatbot server 200, so sample code reflecting the specific document is generated. As described above, the information processing device 100 can easily create sample code reflecting a specific document.

[Sequence Diagram]
FIG. 6 is a sequence diagram showing an example of the flow of processing executed by the information processing device 100 and the AI chatbot server 200. Note that some or all of the processing of the sequence may be performed manually. For example, the processing of dividing the document in S10 may be performed by a work instructor. First, the information processing device 100 divides the document information 172 (S10). Next, the information processing device 100 transmits the divided document to the AI chatbot server 200 and requests a summary (S12). Next, the AI chatbot server 200 summarizes the transmitted document (S14). Next, the AI chatbot server 200 transmits the summary to the information processing device 100 (S16). The information processing device 100 and the AI chatbot server 200 execute the processing from S12 to S16 for each divided document.

Next, the information processing device 100 integrates the summaries summarized by the AI chatbot server 200 (S18). Next, the information processing device 100 transmits the integrated summary, summary information 174, and definition information 176, to the AI chatbot server 200, and requests the generation of sample code using sample data in accordance with the usage of the document library (S20). Next, the AI chatbot server 200 generates sample code in response to the request (S22), and transmits the generated sample code to the information processing device 100 (S24). Next, the information processing device 100 acquires the transmitted sample code and stores the acquired sample code in the storage unit 170 as generation information 178 (S26). The transmitted sample code is provided to, for example, a work instructor. For example, it is displayed on a display unit of the work instructor's terminal device.

As described above, the information processing device 100 can automatically generate the desired sample code using the AI chatbot server 200.

In this embodiment, the use of a database schema has been described as an example, but it may also be applied to a task of acquiring sample code. In this case, the document information 172 is information about the sample code, and the definition information 176 includes the code definition when creating the code.

The above describes the form for carrying out the present invention using an embodiment, but the present invention is not limited to such an embodiment, and various modifications and substitutions can be made without departing from the spirit of the present invention.

100 Information processing device 110 Summary unit 120 Acquisition unit 130 Processing unit 172 Document information 174 Summary information 176 Definition information 178 Generation information

Claims (6)

an acquisition unit that acquires a document including a method for using a library in a predetermined programming language and sample code for the library, and a task definition that instructs the user to generate sample code corresponding to the task definition in a manner conforming to the method for using the library ;
a processing unit that acquires sample code corresponding to the task definition in a manner conforming to the usage of the library acquired by the acquisition unit, the sample code being obtained by inputting a document including a method for using a library in a predetermined programming language and sample code of the library, and a task definition instructing the AI chatbot to generate sample code using the library, the document and the task definition being input to the AI chatbot that has been trained to output sample code corresponding to the task definition in a manner conforming to the usage of the library;
An information processing device comprising: a summarizing unit that divides the document from a reference position in order so that the data capacity of the document does not exceed a limit , and inputs each of the divided documents in order to the AI chatbot, thereby acquiring summary information that summarizes the input divided documents;
The processing unit acquires the sample code obtained by inputting, into the AI chatbot, a document in which the plurality of pieces of summary information acquired by the summarizing unit are linked together in the order in which the pieces of summary information were output , and the task definition.
The information processing device according to claim 1 . The AI chatbot is a model trained to reply in response to the input by learning a document including a method for using a library of a programming language and a sample code of the library, and a sample code corresponding to a task definition in a manner according to the method for using the library using the library ;
The model is a model that has not learned a document corresponding to the predetermined programming language .
The information processing device according to claim 1 . One or both of the acquisition unit and the processing unit executes processing in response to a user's operation.
The information processing device according to claim 1 . Obtaining a document including a method for using a library in a predetermined programming language and sample code for the library, and a task definition that instructs the user to use the library to generate sample code corresponding to the task definition in a manner that conforms to the method for using the library ;
A document including a method for using a library in a predetermined programming language and sample code for the library, and a task definition instructing to generate sample code using the library are input to an AI chatbot that has been trained to output sample code corresponding to the task definition in a manner conforming to the method for using the library, and the document and task definition are input to the AI chatbot, the AI chatbot being trained to output sample code corresponding to the task definition in a manner conforming to the method for using the library, thereby obtaining sample code corresponding to the task definition in a manner conforming to the method for using the library obtained .
Information processing methods. On the computer,
obtaining a document including a method for using a library in a predetermined programming language and sample code for the library, and a task definition that instructs the user to use the library to generate sample code corresponding to the task definition in a manner that conforms to the method for using the library ;
An AI chatbot is trained to receive as input a document including a method for using a library in a predetermined programming language and sample code for the library, and a task definition instructing the AI chatbot to generate sample code using the library in a manner conforming to the method for using the library, and to output sample code corresponding to the task definition in a manner conforming to the method for using the library. The AI chatbot is caused to obtain sample code corresponding to the task definition in a manner conforming to the method for using the library, the sample code being obtained by inputting the document and the task definition.
program.

Images