JP6950362B2 - Information processing systems and programs - Google Patents

Description

The present invention relates to information processing systems and programs that simulate others using artificial intelligence.

In recent years, ingenuity has been made to solve various problems related to artificial intelligence.

For example, the technology disclosed in Patent Document 1 is a system capable of constructing a body of knowledge based on the correct knowledge accumulated by humans and displaying the processing status in all processing processes in a form that humans can understand. Therefore, it is possible to realize "artificial intelligence that does not run away" that solves the "symbol grounding problem".

Japanese Unexamined Patent Publication No. 2016-40730

However, while artificial intelligence has been able to collect user information and provide assistance tailored to the needs of individual users, it has not been able to simulate others.

In view of the above circumstances, an object of the present invention is to provide an information processing system and a program capable of simulating another person by using artificial intelligence.

In order to achieve the above object, the information processing system according to one embodiment of the present invention includes a network-connected server and a terminal device, and the server is on a cloud in which general data can be widely collected. Communication unit that can communicate with the terminal device, general-purpose data learning result that is the result of collecting general data, big data analysis and machine learning, and personal data of a specific individual are collected, big data analysis and machine learning Using the general-purpose data learning result, a storage unit that stores the personal data learning result as a result, a trigger processing unit that detects an action from the user as a trigger, and the content of the action detected as a trigger are used. A task recognition unit that recognizes tasks to be performed by inference, a general-purpose data learning unit that generates the general-purpose data learning result, and a general-purpose, general processing that makes inferences using the general-purpose data learning result and does not depend on individuals. The general processing execution unit that executes the above, the personal processing execution unit that performs inference using the personal data learning result, and executes the personal processing that depends on the specific individual, and the personal data learning result from the outside. It is equipped with a personal data learning result acquisition unit to be acquired.

Therefore, it is possible to simulate others using artificial intelligence.

In order to achieve the above object, the program according to one embodiment of the present invention is a program that operates in an information processing system including a server connected to a network and a terminal device, and is a trigger that detects an action from a user as a trigger. Processing unit, task recognition unit that recognizes the content of the action detected as a trigger as a task to be performed by inference using general-purpose data learning results that are the results of collecting general data, analyzing big data, and machine learning. , A general-purpose data learning unit that generates the general-purpose data learning result, a general processing execution unit that performs inference using the general-purpose data learning result and executes general-purpose and general processing that does not depend on an individual, and a specific individual The personal processing execution unit that collects personal data, analyzes big data, performs inference using the personal data learning result that is the result of machine learning, and executes personal processing that depends on the specific individual, and the above. The server is operated as a personal data learning result acquisition unit that acquires personal data learning results from the outside.

As described above, according to the present invention, it is possible to simulate others by using artificial intelligence.

It is an overall block diagram of the information processing system 100 which concerns on one Embodiment of this invention. It is a block diagram when the server 10 is composed of a general computer. It is a flowchart for demonstrating the flow of the general-purpose data learning process in an information processing system 100. It is a flowchart for demonstrating the flow of the personal data learning process performed outside the information processing system 100. It is a flowchart for demonstrating the general process flow when a user works in the information processing system 100. It is a flowchart for demonstrating the process at the time of answering the question which the user asked to the information processing system 100. It is a flowchart for demonstrating the process at the time of asking a question for the report made by a user. It is a flowchart for demonstrating the process at the time of reviewing the received document.

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

[Overview]
First, an overview will be given.

Since the conventional AI (Artificial Intelligence) assistant is provided as a cloud service, it is possible to learn the behavior of many people, and improve the accuracy and function of the service provided without user updates. You can do it. On the other hand, it was not possible to behave according to the behavior and experience of a specific individual (other).

Similarly, in the case of a bot, it does not behave according to the behavior or experience of a specific individual.

That is, since both the conventional AI assistant and the bot are treated as different personalities from the specific individual, it is not possible to judge what to do if it is a specific individual, and it is not a substitute for the specific individual.

In the present invention, the AI system machine-learns the behaviors and experiences of a specific individual so far, and in addition to the learning results by machine learning using general-purpose data, the learning results of machine learning based on personal data are obtained. By acquiring from others and making inferences using the acquired personal data learning results, it is possible to act as if it were a specific individual (another person) and substitute for a specific individual.

The outline has been described above.

[overall structure]
Next, the overall configuration of the information processing system 100 according to the embodiment of the present invention will be described. FIG. 1 is an overall configuration diagram of an information processing system 100 according to an embodiment of the present invention.

The information processing system 100 includes a network-connected server 10 on the cloud, a personal computer (personal computer, terminal device) 20 used by each user, and the like.

The server 10 is on the cloud, and general data is widely collected as general-purpose data from each PC 20 of user A to user E and other devices, and the collected data is analyzed by big data and machine-learned to obtain general-purpose data. Holds the learning results.

Users A to E access the server 10 via their own PC 20.

The server 10 can act as a substitute for the user F by performing inference using both the general-purpose data learning result collected by itself and the personal data learning result of the user F acquired from the outside. In other words, it can also be called a user F simulator.

As described above, in the information processing system 100, at the time of inference, general semantic inference is performed using the general-purpose data learning result, and semantic inference according to a specific individual is performed using the personal data learning result.

The overall configuration has been described above.

[Server Configuration]
Next, the configuration of the server 10 will be described. The server 10 may be configured by dedicated hardware or software, or may be configured by a general computer. FIG. 2 shows a configuration diagram when the server 10 is composed of a general computer.

As shown in the figure, the server 10 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, an operation input unit 14, a communication unit 15, a display unit 16, and a storage unit. It has a unit 17, and each of these blocks is connected via a bus 18.

The ROM 12 stores a plurality of programs and data such as firmware for executing various processes. The RAM 13 is used as a work area of the CPU 11, and temporarily holds an OS (Operating System), various applications being executed, and various data being processed.

The storage unit 17 is, for example, an HDD (Hard Disk Drive), a flash memory, or other non-volatile memory. The storage unit 17 stores the OS, various applications, various data, general-purpose data learning results 17a, and personal data learning results 17b.

The general-purpose data learning result 17a stores the result of big data analysis and machine learning of the general-purpose data collected by the server 10.

The personal data learning result 17b is a result of big data analysis and machine learning of the personal data of the user F acquired from the outside.

The communication unit 15 is connected to a network for exchanging information with a PC 20 or the like on the network.

The CPU 11 expands the program corresponding to the instruction given from the operation input unit 14 among the plurality of programs stored in the ROM 12 and the storage unit 17 into the RAM 13, and the display unit 16 and the storage unit according to the expanded program. 17 is appropriately controlled.

The operation input unit 14 is, for example, a pointing device such as a mouse, a keyboard, a touch panel, or other operation device.

The display unit 16 is, for example, a liquid crystal display, an EL (Electro-Luminescence) display, a plasma display, or the like.

Next, a functional block realized by executing a program in the CPU 11 will be described.

The functional blocks realized in the CPU 11 of the server 10 are a trigger processing unit 11a, a task recognition unit 11b, a general-purpose data learning unit 11c, a general processing execution unit 11d, a personal processing execution unit 11e, and a personal data learning result acquisition unit 11f. Is.

The trigger processing unit 11a detects an action from the user, specifically, the start of a conversation, etc. as a trigger.

The task recognition unit 11b recognizes the content of the action from the user detected as a trigger as a task to be performed by inference using the general-purpose data learning result 17a.

The general-purpose data learning unit 11c analyzes widely collected general data as big data and performs machine learning to generate a general-purpose data learning result 17a.

The general processing execution unit 11d makes an inference using the general-purpose data learning result 17a, and executes general-purpose and general processing that does not depend on the individual.

The personal processing execution unit 11e makes an inference using the personal data learning result 17b, and executes personal processing depending on the individual.

The personal data learning result acquisition unit 11f acquires the personal data learning result 17b relating to a specific individual from the outside of the information processing system 100.

The configuration of the server 10 has been described above.

[Processing flow (general-purpose data learning processing)]
Next, the flow of general-purpose data learning processing in the information processing system 100 will be described. FIG. 3 is a flowchart for explaining the flow of general-purpose data learning processing in the information processing system 100.

First, the general-purpose data learning unit 11c on the server 10 widely collects general-purpose general data (step S1).

Next, the general-purpose data learning unit 11c performs big data analysis on the collected general data (raw data) (step S2).

Next, the general-purpose data learning unit 11c performs machine learning on the big data analysis result (step S3).

Next, the general-purpose data learning unit 11c updates the general-purpose data learning result using the machine learning result (step S4).

Next, the process returns to step S1 and the process is continued.

As described above, the general-purpose data learning process described above is continuously executed independently of other processes. Therefore, the comprehension ability for understanding the content of the request to the information processing system 100 can be continuously improved.

The flow of general-purpose data learning processing in the information processing system 100 has been described above.

[Processing flow (personal data learning process)]
Next, the flow of personal data learning processing performed outside the information processing system 100 will be described. FIG. 4 is a flowchart for explaining the flow of personal data learning processing performed outside the information processing system 100.

First, the user's personal data is collected (step S5). The personal data referred to here is assumed to be various data such as a document created by the user, the text of an email, the data of conversations made by the user, and the tendency of the user's behavior.

Next, big data analysis is performed on the collected personal data (raw data) (step S6).

Next, machine learning is performed on the big data analysis result (step S7).

Next, the machine learning result is saved as the personal data learning result 17b (step S8).

Next, the process returns to step S5 and the process is continued.

The flow of personal data learning processing performed outside the information processing system 100 has been described above. The flow of the personal data learning process described above is an example, and the personal data learning result may be obtained by another method.

[Processing flow (general processing)]
Next, the general flow of processing when the information processing system 100 is approached by the user will be described. FIG. 5 is a flowchart for explaining a general processing flow when a user works on the information processing system 100. A more specific process will be described later.

First, the trigger processing unit 11a determines whether or not the action from the user who is the trigger is detected (step S10).

If the trigger is not detected (N in step S10), the process returns to step S10 and waits.

When the trigger is detected (Y in step S10), the task recognition unit 11b then uses the general-purpose data learning result 17a to understand the content of the user's action and recognize the task to be performed (step S11). ..

Next, the general processing execution unit 11d executes general processing that does not depend on the individual by inference using the general-purpose data learning result 17a (step S12).

Next, the personal processing execution unit 11e executes an individual-dependent personal processing by inference using the personal data learning result 17b (step S13).

Next, the personal processing execution unit 11e outputs the deliverable (step S14).

The general flow of processing when the information processing system 100 is approached by the user has been described above.

[Process flow (for answering questions)]
Next, a more specific processing flow will be described. Here, a process for answering a question asked by the user to the information processing system 100 will be described. FIG. 6 is a flowchart for explaining a process when the user answers a question asked to the information processing system 100.

First, the trigger processing unit 11a receives a question from a conversation with the user or the like (step S20).

Next, the task recognition unit 11b infers using the general-purpose data learning result 17a, understands the contents of the received question, and recognizes the answer task to the question (step S21).

Next, the personal processing execution unit 11e makes an inference using the personal data learning result 17b, and creates an answer that the user F will give (step S22).

Next, the personal processing execution unit 11e outputs the created answer (step S23).

Next, the trigger processing unit 11a determines whether or not the conversation with the user has ended (step S24).

If the conversation with the user has not been completed yet (N in step S24), the process returns to step S20 and the process is continued.

When the conversation with the user is completed (Y in step S24), the process is terminated.

As described above, by using the information processing system 100, the user can obtain an answer from the knowledge of another person simulated by the information processing system 100.

The processing when the user answers the question asked to the information processing system 100 has been described above.

[Process flow (in the case of question task for reporting)]
Next, the processing when asking a question for the report made by the user will be described. FIG. 7 is a flowchart for explaining a process when asking a question for a report made by a user.

First, the trigger processing unit 11a receives a report by the user (step S30).

Next, the task recognition unit 11b makes an inference using the general-purpose data learning result 17a, understands the report content, and recognizes the question task for the report content (step S31).

Next, the personal processing execution unit 11e infers using the personal data learning result 17b, and creates a question that the user F will ask for the report content (step S32).

Next, the personal processing execution unit 11e outputs the created question (step S33).

As described above, the information processing system 100 asks the user a question according to the characteristics of the user F. Therefore, it is possible to obtain a hypothetical question before actually reporting to the user F and conducting a question and answer session.

The processing when asking a question for a report made by a user has been described above.

[Process flow (for document review task)]
Next, the process for reviewing the received document will be described. FIG. 8 is a flowchart for explaining a process when reviewing the received document.

First, the trigger processing unit 11a receives the document to be reviewed (step S40).

Next, the task recognition unit 11b infers using the general-purpose data learning result 17a, understands the contents of the received document, and recognizes the review task (step S41).

Next, the personal processing execution unit 11e infers using the personal data learning result 17b and creates a review comment that the user F will perform (step S42).

Next, the personal processing execution unit 11e outputs the created review comment (step S43).

As described above, when the document to be reviewed by the information processing apparatus 100 is given, the review along with the review tendency performed by the user F, for example, the tendency to strictly check typographical errors and the tendency to question the essential meaning of the document. Is done.

The process for reviewing the received document has been described above.

[Supplementary information]
As described above, the information processing system 100 according to the present invention includes the server 10 connected to the network and the terminal device 20, and the server 10 is located on a cloud capable of widely collecting general data, and the terminal. A communication unit 15 that can communicate with the device 20, a general-purpose data learning result 17a that is the result of collecting general data, analyzing big data, and performing machine learning, and a machine that collects personal data of a specific individual and analyzes big data. The general-purpose data learning includes a storage unit 17 in which the personal data learning result 17b, which is the learning result, is stored, a trigger processing unit 11a that detects an action from the user as a trigger, and the content of the action detected as a trigger. Using the result 17a, the task recognition unit 11b that recognizes the task as a task to be performed by inference, the general-purpose data learning unit 11c that generates the general-purpose data learning result 17a, and the general-purpose data learning result 17a are used to perform inference, and each individual A personal who executes inferences using the general processing execution unit 11d that executes general-purpose and general processing that does not depend on the personal data and the personal data learning result 17b, and executes personal processing that depends on the specific individual. It includes a processing execution unit 11e and a personal data learning result acquisition unit 11f that acquires the personal data learning result 17b from the outside.

Therefore, it is possible to simulate others using artificial intelligence.

In addition, the present invention is not limited to the above-described embodiment, and it goes without saying that various modifications can be made without departing from the gist of the present invention.

10 ... Server 11 ... CPU
11a ... Trigger processing unit 11b ... Task recognition unit 11c ... General-purpose data learning unit 11d ... General processing execution unit 11e ... Personal processing execution unit 11f ... Personal data learning result acquisition unit 12 ... ROM
13 ... RAM
14… Operation input unit 15… Communication unit 16… Display unit 17… Storage unit 17a… General-purpose data learning result 17b… Personal data learning result 18… Bus 100… Information processing system

Claims (2)

Including networked servers and terminal equipment
The server
Located in the cloud where general data can be widely collected
A communication unit capable of communicating with the terminal device,
The general-purpose data learning result, which is the result of collecting general data, analyzing big data, and machine learning, and the personal data learning result, which is the result of collecting personal data of a specific individual, analyzing big data, and machine learning are stored. With the stored memory
A trigger processing unit that detects the user's action as a trigger,
A task recognition unit that recognizes the content of the action detected as a trigger as a task to be performed by inference using the general-purpose data learning result.
A general-purpose data learning unit that generates the general-purpose data learning result,
A general processing execution unit that performs inference using the general-purpose data learning result and executes general-purpose and general processing that does not depend on an individual.
A personal processing execution unit that performs inference using the personal data learning result and executes personal processing that depends on the specific individual, and a personal data learning result acquisition unit that acquires the personal data learning result from the outside. Information processing system equipped. A program that runs in an information processing system that includes a network-connected server and terminal equipment.
Trigger processing unit that detects user's action as a trigger,
A task recognition unit that recognizes the content of the action detected as a trigger as a task to be performed by inference using general-purpose data learning results that are the results of collecting general data, analyzing big data, and machine learning.
The general-purpose data learning unit that generates the general-purpose data learning result,
A general processing execution unit that performs inference using the general-purpose data learning result and executes general-purpose and general processing that does not depend on an individual.
Personal data execution that collects personal data of a specific individual, analyzes big data, makes inferences using the personal data learning result that is the result of machine learning, and executes personal processing that depends on the specific individual. A program that operates the server as a unit and a personal data learning result acquisition unit that acquires the personal data learning result from the outside.

Images