JP7617537B2 - Invention hearing system, invention hearing device, invention hearing method, and invention hearing program - Google Patents

Description

The present invention relates to an invention hearing system, an invention hearing device, an invention hearing method, and an invention hearing program.

Obtaining a patent remains extremely important, and patent rights often have a significant impact on a business. Inventive activities are not limited to companies, but are also undertaken by individuals. In some cases, inventors themselves may apply for a patent, but generally, intellectual property officers or patent attorneys will hear about the details of the invention from the inventor, and then develop the business or apply for a patent for the invention.

After receiving an explanation of the invention from the inventor, detailed consideration is required regarding what business strategy to adopt, whether to file a patent application, and what the content of the patent application should be. At the same time, in today's fast-paced world, there is a demand to shorten the time required for consideration. To meet these demands, the inventor needs to be able to properly express what the invention is, and the listener needs to be able to properly understand what the invention is. However, while inventors are accustomed to research and development, they may not be accustomed to expressing inventions, for example, putting the invention into writing. It is also common for others to have difficulty understanding the content of the invention.

It takes time and effort for an inventor to put the details of an invention into writing in order to explain it. In addition, it can be difficult to decide what to document, which can take a long time. Similarly, it takes time and effort for patent attorneys and intellectual property officers to hear the details of the invention from the inventor and put the details of the invention into writing. In addition, understanding and summarizing an invention requires expertise as well as comprehension, reading comprehension, and expression skills. Understanding and summarizing the details of an invention is a highly intellectual and difficult task.

Patent Document 1 discloses an invention proposal information processing device that is configured to be connectable via communication to a user terminal used by a user and an expert terminal used by an expert. The invention proposal information processing device of Patent Document 1 performs the following processes.
After receiving invention proposal information from a user, the system asks questions regarding inventiveness, novelty, etc., and accepts the user's response. From the user's responses, question items to which the user has given a negative response regarding patentability, or has answer data indicating that they do not know or cannot make a judgment, are extracted, and an expert suitable to handle the case is selected. The system sends the invention proposal information to the selected expert to make an inquiry, and accepts the expert's response to the inquiry. If the expert approves, the expert's response and comments are sent to the user.
According to Patent Document 1, it is possible to easily obtain appropriate advice on the content of an invention, and to reflect the advice in the content of the invention to improve it.

JP 2004-46876 A

However, in Patent Document 1, a human expert needs to understand the invention, read the user's response, and comment on the invention. While it may not be necessary for the user and the expert to meet in person to talk, the human expert needs to respond, and the burden on the human expert remains the same. For this reason, it is desirable to reduce the time and effort required to hear about the contents of an invention.

The present invention aims to provide an invention hearing system that can hear the contents of an invention and eliminates the need for human listeners to do any work.

In order to solve the above problems, the hearing system of the present invention comprises:
a questioning section for asking the user questions about the invention;
An invention hearing system having a reception unit that receives a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The questioning unit asks a question by inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and transmitting the generated question to a user, and repeats generating and transmitting questions. When generating a question, the prompt for generating a question includes one or more questions asked before the question is generated, one or more answers to the questions asked before the question is generated, and an instruction for generating a question for the trained model.
It is characterized by:

The present invention provides an invention hearing system that can hear the contents of an invention and eliminates the need for human listeners to do the listening.

1 is a diagram showing an example of an invention hearing system and an invention hearing device of the present invention. 13 is an example of a screen. 1 is a screen example (1) of the first embodiment. 13 is a screen example (2) of the first embodiment. 13 is a screen example (3) of the first embodiment. 13 is a screen example (4) of the first embodiment. 13 is a screen example (5) of the first embodiment. 13 is a screen example (6) of the first embodiment. 1 is an example of an invention proposal. 1 is an example of a patent application specification. 13 is a screen example (1) of the second embodiment. 13 is a screen example (2) of the second embodiment. 13 is a screen example (3) of the second embodiment. 13 is a screen example (4) of the second embodiment. 13 is a screen example (5) of the second embodiment. 13 is a screen example (6) of the second embodiment. 13 is a screen example (7) of the second embodiment. 13 is a screen example (1) of the third embodiment. 13 is a screen example (2) of the third embodiment. 13 is a screen example (6) of the fourth embodiment. 13 is a screen example (1) of the reference example 1.

The invention hearing system, invention hearing device, invention hearing method, and invention hearing program according to the present invention will be described below with reference to the drawings. Note that the present invention is not limited to the embodiments shown below, and can be modified within the scope of what a person skilled in the art can imagine, including other embodiments, additions, modifications, deletions, etc., and any aspect is within the scope of the present invention as long as it provides the functions and effects of the present invention.

The invention hearing system of the present invention is
a questioning section for asking the user questions about the invention;
An invention hearing system having a reception unit that receives a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The questioning unit inputs a prompt for generating a question into a trained model, generates a question as an output of the trained model, and asks the question by sending the generated question to a user;
The prompt for generating a question is characterized by including at least an answer to the question and an instruction to generate a question for the trained model.

The hearing device of the present invention comprises:
a questioning section for asking the user questions about the invention;
A hearing device for an invention having a reception unit for receiving a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The questioning unit inputs a prompt for generating a question into a trained model, generates a question as an output of the trained model, and asks the question by sending the generated question to a user;
The prompt for generating a question is characterized by including at least an answer to the question and an instruction to generate a question for the trained model.

One embodiment of the present invention will be described. FIG. 1 is a diagram for explaining the invention hearing system 1 and invention hearing device 10 of this embodiment. In FIG. 1, the invention hearing system 1, the invention hearing device 10, a storage means 31, a user 40, and a terminal 41 are illustrated. Note that the arrows in the figure show a schematic representation of data flow, instructions, etc.

The user 40 is assumed to be an inventor, but is not limited to this and may be someone other than an inventor. The user 40 may be, for example, a corporation, an individual, an organization, etc., as appropriate, and examples of the user 40 include countries, companies, patent offices, and other offices. Although the term "user" is used, it may also be referred to as an employer, administrator, user, target person, recipient, etc. In the following explanation, the terms user and target person are used interchangeably. Examples of the terminal 41 include a PC, a smartphone, a tablet, etc.

The invention hearing device 10 of this embodiment has a questioning unit 11, a receiving unit 12, and, if necessary, a document creation unit 13, an invention creation unit 14, etc. The invention hearing device 10 of this embodiment has, for example, a CPU, RAM, ROM, HDD, IF, etc., and, if necessary, other parts, etc.

The communication method between the terminal 41 and the invention hearing device 10 is not particularly limited and can be changed as appropriate. Communication may be via the Internet, etc., or may be wired or wireless.

The questioning section 11 asks the user questions about the invention. The questions are questions about one or more of the name of the invention, the background of the invention, prior art related to the invention, the problem of the invention, the means for solving the problem of the invention, and the effects of the invention. Such questions ask about the contents of the invention, and by the user answering the questions, information for understanding the contents of the invention is accumulated. The questions may be closed questions or open questions. The means for solving the problem may be referred to as invention-specific matters, configuration, features, etc.

The invention is assumed to be an invention before a patent application is filed, but is not limited to this. In the case of an invention before a patent application is filed, the purpose of using the system and device of this embodiment can be, for example, to file a patent application, to create documents and materials to explain the invention, to deepen understanding of the invention, to organize the contents of the invention, and to improve the invention. In the case of an invention after a patent application is filed, the purpose can be, for example, to use it as material for consideration for making another invention, or to deepen understanding of an invention that has been filed by another person.

The questioning unit 11 asks a question by inputting a prompt for generating a question into the trained model, generating a question as an output of the trained model, and sending the generated question to the user. There are no particular limitations on the method of sending the question, and any known technology can be used. The sent question is displayed, for example, on the terminal used by the user.

The questioning unit 11 generates new questions and asks them, which makes the interview process more efficient. Furthermore, according to this embodiment, information about the invention can be collected through a combination of questions and answers. Furthermore, the time and effort required for a human to interview about the invention can be saved.

The trained model is not particularly limited and can be selected as appropriate. An example of the trained model used in this embodiment is a sentence generation language model. An example of the sentence generation language model is GPT (Generative Pre-trained Transformer). GPT is a sentence generation language model that learns based on a large amount of data on the web.

The trained model used in this embodiment outputs according to the input prompt. Prompts can be used to generate questions, create documents, expand the invention, etc., and the content can be changed as appropriate depending on the application. The prompt includes instructions for the trained model, etc.

The prompt for generating a question includes at least the answer to the question and an instruction for generating a question for the trained model, and may include other items as necessary. Prompts for creating documents and prompts for expanding the invention will be described later.

Examples of question generation instructions include an instruction to ask questions to understand the invention, an instruction to hold an interview to apply for a patent on the invention, an instruction to hold an interview to prepare a document explaining the invention, etc. A prompt for question generation can include multiple instructions, and it is preferable to include detailed instructions in the prompt for question generation in order to improve the accuracy of the questions.

Examples of instructions to ask questions to understand an invention include instructions to list items such as the problem the invention is intended to solve, the solution to that problem, and the effects of the invention, and to generate questions that will clarify these items.

In addition to the above, the question generation instructions may also include, for example, the number of characters in the question, specifications for the question text, etc. The expected listener may also be specified. For example, a patent attorney, an intellectual property officer, a developer, a manager, etc. may be specified as the listener. Specifying the listener in this way makes it easier to meet the user's intended use of the system. For example, if it is expected that a patent application will be filed, it is preferable to specify a patent attorney, etc. as the listener, and if documents and materials for considering a business plan are to be created, it is preferable to specify an intellectual property officer, a manager, etc. as the listener.

When specifying a listener in a question generation instruction, the method of specification can be selected appropriately, and may be, for example, an expression that would be conveyed to a human. For example, an expression that a human can understand, such as "You are a patent attorney," is possible. Specifying a listener in a question generation instruction does not require knowledge of programming languages and can be done using everyday conversation.

When a listener is specified, the trained model will output according to the specification. Although not particularly limited, for example, in the case of a patent attorney, questions for preparing a patent application specification may be generated, and in the case of an intellectual property officer or a manager, questions for the impact on the company's business, budget preparation, and patent strategy creation may be generated.

The question unit 11 may send the question output by the trained model to the user as is, or may edit the output question before sending it to the user.

It is preferable that the interrogation unit 11 asks questions multiple times. In other words, it is preferable that the interrogation unit 11 repeatedly generates and transmits questions. By asking questions multiple times, the invention can be better understood.

When the questioning unit 11 repeatedly generates and transmits questions, it is preferable not to transmit questions with the same content. In other words, it is preferable not to ask the same question. Asking the same question not only places a burden on the user, but also leads to a decrease in the reliability of the system.

To avoid asking the same question, for example, the prompt for question generation can include questions that have already been asked, and an instruction to not ask the same question can be added as a question generation instruction. The same question is not limited to questions that match in wording, but also includes questions that have the same meaning. A trained model can determine whether questions have the same content.

Preferably, the prompts for question generation include questions that have already been asked. In this case, the precision of the questions can be improved and also instructions can be given to avoid asking the same questions again.

The questioning unit 11 repeatedly generates and transmits questions, and when generating a question, it is preferable that the prompt for generating the question includes one or more questions that were asked before the question was generated and one or more answers to the questions that were asked before the question was generated. By including the question and the answer in the prompt for generating the question, the accuracy of the generated question can be improved.

When the question unit 11 repeatedly generates and transmits questions, the answer to the question included in the prompt for generating the question may be one or more of multiple answers, but it is preferable that all answers are included in the prompt for generating the question. In this case, the accuracy of the question can be improved.

When the questioning unit 11 repeatedly generates and transmits questions, the question included in the prompt for question generation may be one or more of the multiple questions, but it is preferable that all questions are included in the prompt for question generation. In this case, the accuracy of the questions can be improved.

When the questioning unit 11 repeatedly generates and transmits questions, it is more preferable that all questions and answers are included in the prompt for generating the question. In other words, it is more preferable that the questioning unit 11 repeatedly generates and transmits questions, and when generating a question, it is more preferable that all questions and their answers asked before generating the question are included in the prompt for generating the question. In this case, the accuracy of the questions can be further improved. It is also easier to prevent the same question from being asked.

When the questioning unit 11 repeatedly generates and transmits questions, it is preferable that the invention hearing device 10 has a function of automatically creating a history of questions and answers. This history can be used, for example, to determine whether a question has already been asked, or to weight specific questions and answers. As described below, when receiving ratings of questions from users, the weighting value may be increased for questions that are highly rated. For example, a question with a high weighting value may be considered to be a particularly important question, and questions related to this question may be generated.

When the question unit 11 repeatedly generates and sends questions, the question generation instruction included in the prompt for question generation may be changed each time a question is generated, or one question generation instruction may be used in common. In order to avoid complicating the process, it is preferable to use the latter, i.e., to use one question generation instruction in common to generate each question.

The reception unit 12 receives answers from the user. The method of receiving the answers is not particularly limited and can be selected as appropriate. Receiving an answer is not limited to cases where the user sends an answer to the invention hearing device 10, but also includes cases where the invention hearing device 10 retrieves an answer from an arbitrary storage location. The answer from the user may be in the form of text, an image, audio, or video.

The reception unit 12 may receive an evaluation of the question from the user. In this case, it is preferable that the questioning unit 11 includes the evaluation of the question in the prompt for generating the question. By generating a question by including the evaluation of the question in the prompt for generating the question, the accuracy of the question can be improved.

The format of the question evaluation is not particularly limited, and examples include a three-level evaluation or a five-level evaluation. Other examples include text such as "good," "average," and "bad." The reason for the evaluation may also be accepted. For example, a sentence such as "This question is a good question because of the following reason" may be accepted. Accepting the reason for the evaluation further improves the accuracy of question generation.

The reception unit 12 may receive an outline of the invention from the user before the questioning unit 11 asks the first question. In this case, when generating the first question, the questioning unit 11 preferably includes the outline of the invention in the prompt for generating the question as an answer to the question. By generating questions using the outline of the invention, the accuracy of the questions can be improved, and questions asking about the contents described in the outline of the invention can be eliminated, improving the efficiency of the hearing. The outline of the invention may be referred to as an explanation of the invention, a description of the invention, etc.

The summary of the invention may be in the form of text, images, audio, or video. The summary of the invention may be received in file format, or by inputting text onto a screen. The summary of the invention may be received only once, or multiple times.

The items to be described in the summary of the invention are not particularly limited and can be selected as appropriate. The summary of the invention may be written based on the inventor's free ideas. Assuming that a patent application will be filed, for example, the name of the invention, the problem that the invention is intended to solve, the means for solving the problem, and the effects of the invention may be described. In addition, a screen for inputting these items may be displayed on the user's terminal, and these items may be accepted and used as the summary of the invention.

When the summary is in text, the method of including it in the prompt for generating a question is not particularly limited and can be selected appropriately. For example, the text may be entered directly into the prompt for generating a question, or a file path may be specified.
When the outline of the invention is an image, audio, or video, the method of including it in the prompt for generating a question is not particularly limited and can be selected appropriately. For example, a file path may be specified in the prompt for generating a question. In addition, the characteristics of the image, audio, or video may be extracted as characters before being included in the prompt for generating a question, and the extracted characters may be included in the prompt for generating a question.
Characters include alphanumeric characters, symbols, sentences, and the like.

When generating a question by including the outline of the invention in the prompt for generating a question, the outline of the invention may be included in the prompt for generating a question only when the first question is generated, but it is preferable to use the outline of the invention for generating all questions. In other words, when generating a question, the question unit 11 preferably generates a question by including the outline of the invention in the prompt for generating a question.

In this embodiment, the outline of the invention does not have to be received before the first question is asked. If the outline of the invention is not received, the questioning unit 11 may ask, for example, a standard question as the first question.
This example will be described again. Before the questioning unit 11 asks the first question, the reception unit 12 does not accept an outline of the invention from the user, and when the questioning unit 11 asks the first question, it does not generate a question but asks a standard question. A user who is not accustomed to explaining inventions may not know what to write as the outline of the invention, so by asking a standard question, it is possible to prevent a user who is not accustomed to explaining inventions from becoming confused.

The content of the questions may be adjusted taking the user into consideration. For example, an ID may be assigned to the user, and information such as whether the user has used the invention hearing system in the past, whether the user has filed a patent application, etc. may be stored in a storage means. In this way, it is possible to determine whether the current user is familiar with using the system, and if the user is not familiar with the system, an explanation of the question or an explanation about patent applications may be provided when asking a question.

Figure 2 is a diagram showing an example of an invention hearing screen 50. For example, a question generated by the questioning unit 11 is displayed in a question field 51. When the user inputs an answer in an answer field 52 and presses an answer button 53, the input answer is sent and accepted by the acceptance unit 12. A skip button 54 may be provided on the invention hearing screen 50, and the skip button 54 is pressed when a question cannot be answered.

When a question is skipped, it is preferable to include the skipped question in the prompt for generating the question. In this case, the accuracy of question generation can be improved. Also, a message to that effect may be included in the prompt for generating the question so that the user can know that the question has been skipped.

The invention hearing system 1 of this embodiment may include a document creation unit 13 that creates documents explaining the contents of the invention.
The document creation unit 13 inputs a prompt for document generation into the trained model and generates the document as an output of the trained model. The prompt for document generation includes a user's answer to a question and a document generation instruction for the trained model.

Examples of documents created by the document creation unit 13 include invention proposals (which may also be referred to as invention descriptions, etc.), patent specifications, and patent claims. Documents created by the document creation unit 13 may also be referred to as materials, etc.

Preparing a document that explains the contents of an invention makes it easier to explain the invention to others, and makes it easier for people other than the inventor to understand the invention. In addition, preparing a patent specification reduces the workload of patent attorneys.

The items to be included in the invention proposal are not particularly limited and can be selected as appropriate. It is preferable that the invention proposal include items to be included in a patent specification, such as the name of the invention, background, the problem that the invention is intended to solve, the means for solving the problem, and the effects of the invention. Items that are considered important to the invention may also be included. The items to be included in the invention proposal can be adjusted by appropriately changing the document generation instructions included in the document generation prompt.

The document generation prompt may include other items in addition to the user's answers and document generation instructions. For example, it may include questions posed to the user. Including questions in the document generation prompt improves the accuracy of document generation. If a question is asked multiple times, it is preferable to include all questions and answers in the document generation prompt.

The document generation instructions may, for example, specify the type of document. They may also specify the items to be included in the document, or the format and style of the document. For example, the document generation instructions may specify the number of characters in the document, the number of characters for each item, the style, language, color, and font size of the document.

The prompt for document generation may include a sample separate document. For example, when creating a patent specification, the contents of the published specification may be included. Multiple sample separate documents may be included. Additionally, the document generation instructions may include an instruction to write as JP Patent Publication XXXX-XXXXX. Multiple sample separate documents may be specified.

The document creation unit 13 may edit the document output by the trained model. For example, the document creation unit 13 may create a document by inputting (applying) the content output by the trained model into a template (format) of the document. In this embodiment, the document creation unit 13 is referred to as a document creation unit, but may also be referred to as a document generation unit, etc.

The timing when the document creation unit 13 creates the document can be selected as appropriate; for example, the document is created when an instruction is received from the user. The degree of progress of the hearing on the invention may also be used. For example, when the degree of progress of the hearing on the invention exceeds a predetermined value, an instruction to create a document may be received from the user. The degree of progress of the hearing on the invention may be output, for example, to a trained model. The degree of progress of the hearing on the invention may be presented to the user together with a question. The degree of progress of the hearing on the invention may also be referred to as the degree of hearing progress, the degree of content of the invention, the degree of understanding of the invention, etc.

One way to quantify the progress of invention hearings is, for example, to convert the percentage of questions asked about the items in the target document into a numerical value.

The use of the document created by the document creation unit 13 is not particularly limited and can be selected as appropriate. For example, the document may be used by the inventor to explain the invention to others. It may also be used to convert the contents of the inventor's explanation into text data and store it. The document may also be used to file a patent application. The created document may also be used for the purpose of improving an invention, expanding the invention, etc. An example of using a document to expand an invention is described below.

The invention generation unit 14, for example, extends an invention described in a document created by the document creation unit 13. The invention generation unit 14 inputs a prompt for invention extension into the trained model and generates the extended invention as an output of the trained model. The prompt for invention extension includes the document and an invention extension instruction for the trained model.

An extended invention is an invention that extends an invention (also referred to as an original invention, etc.) described in a document created by the document creation unit 13. The type of extended invention to be generated can be adjusted by appropriately adjusting the prompt for invention extension. For example, the generated extended invention can be adjusted by changing the content specified in the invention extension instruction. An extended invention can also be said to be an invention that is output by inputting the prompt for invention extension into a trained model.

Although not particularly limited, the conditions for generating an extended invention in the invention extension instruction are, for example,
(1) An extended invention has the same problem and effect as the original invention, but the configuration does not have to be related to the original invention. (2) An extended invention does not have the same problem and effect as the original invention, but the configuration is related to the original invention. (3) It is possible to specify that an extended invention has the same problem and effect as the original invention, but the configuration is related to the original invention.

The number of extended inventions to be generated is not particularly limited and can be selected as appropriate. The number and scope of extended inventions can be specified in the invention extension instructions.

The extended invention may be an invention that includes the original invention, or may be an invention that does not include the original invention but is related to it. As mentioned above, the extended invention may be an invention that has problems and effects in common with the problems and effects of the original invention, or may be an invention that has problems and effects that are not in common with the problems and effects of the original invention. As mentioned above, the extended invention may be an invention that has a configuration in common with the configuration of the original invention (which may be referred to as a means, etc.), or may be an invention that has a configuration that is not in common with the configuration of the original invention. If the extended invention is an invention that has a configuration that is not in common with the configuration of the original invention, for example, it is an invention that has problems and effects in common with the problems and effects of the original invention.

Creating an extended invention can bring new insights to the inventor, give the inventor the opportunity to obtain additional patent rights, and increase the number of possible embodiments of the original invention.

The method of including a document in the invention extension prompt is not particularly limited and can be selected as appropriate. For example, the contents of the document can be converted into text format and included in the invention extension prompt, or the path to the file can be specified.

The invention creation section 14 may extend the invention without using the documents created by the document creation section 13. For example, the invention may be created using the questions asked by the questioning section 11 and the answers thereto.
For example, the invention generation unit 14 inputs a prompt for extending the invention into the trained model and generates the extended invention as an output of the trained model, where the prompt for extending the invention includes the question, an answer to the question, and an invention extension instruction for the trained model.
In this case, since there is no need to use the document created by the document creation unit 13, the number of operations required to generate the extended invention can be reduced.

According to the present invention, as described above, there is provided a computer implemented invention hearing method.
The invention hearing method of the present invention is an invention hearing method performed by a computer,
a questioning step of asking a user questions about the invention;
A receiving step of receiving a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The questioning step includes inputting a prompt for question generation into a trained model, generating a question as an output of the trained model, and sending the generated question to a user, thereby asking the question;
The prompt for generating a question is characterized by including at least an answer to the question and an instruction to generate a question for the trained model.

According to the present invention, as described above, an invention hearing program is provided.
The invention hearing program of the present invention is
On the computer,
a questioning process for asking a user questions about the invention;
An invention hearing program that executes an acceptance process for accepting an answer from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The question processing includes inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and sending the generated question to a user to ask a question;
The prompt for generating a question is characterized by including at least an answer to the question and an instruction to generate a question for the trained model.

The invention hearing method and invention hearing program of the present invention allow an inventor to be interviewed about the contents of the invention, eliminating the need for human labor on the part of the listener.

Example 1
Next, Example 1, which is an example of this embodiment, will be described with reference to Figures 3A to 3F, 4A, and 4B. Example 1 is an example of a case where an outline of an invention is accepted, and will be described using an invention of a drone with an umbrella as an example.

Figure 3A is an example screen (1) of the invention hearing screen 50, which is an example of a screen for inputting an outline of the invention. As shown in the figure, the outline of the invention is input in the answer field 52 of the screen 50. Although not limited to this, in the example shown, "a drone with an umbrella" can also be said to be the name of the invention. The descriptions "giving the drone an umbrella," "camera," and "program" can also be said to be means for solving the problem.

In addition, in the example screen (1), the area where the outline of the invention is entered is called an answer field 52. The reason for this is that in the first embodiment, the outline of the invention is treated as an answer to a question, and a question is generated using the outline of the invention. In the example screen (1), the area may be called an input field instead of an answer field.

In screen example (1), when the answer button 53 is pressed with the outline of the invention written, the outline of the invention is sent to the invention hearing device 10, and the reception unit 12 receives the outline of the invention. Next, the questioning unit 11 includes the outline of the invention in a prompt for generating a question as an answer to the question, inputs the prompt for generating a question into the trained model, and generates the first question as an output of the trained model.

Figure 3B is a screen example (2) of the invention hearing screen 50, which is an example of the first question. The first question asks a closed question about the effect of the invention. In response, the user answers that the presented effect of the invention is correct. The answer is entered in the answer field 52, and the answer button 53 is pressed. This causes the reception unit 12 to receive the answer from the user.

As shown in the figure, the question may be given an option to select an evaluation. In this example, the evaluation selection area 55 allows the selection of three options: good, average, or poor. In this case, the selected character (e.g., "good") may be sent as is, or may be converted into a number (e.g., "1") and sent.

In this embodiment 1, the question generation instruction included in the prompt for question generation is, for example, "Please refer to the outline of the invention and ask questions to understand the invention in detail." The language of the question generation instruction is not particularly limited.

Next, the questioning unit 11 includes the outline of the invention, the question, the answer to the question, and the question generation instruction in a prompt for question generation, inputs the prompt for question generation into the trained model, and generates a second question. Note that the question generation instructions in this Example 1 are common to each question, and specify that duplicate questions are to be excluded.

Figure 3C is a screen example (3) of the invention hearing screen 50, which is an example of the case where a second question is asked. In the second question, for example, an open-ended question is asked about the means for solving the problem (which may also be called the features or configuration of the invention). As with screen example (2), the answer is entered in the answer field 52 and the answer button 53 is pressed, whereby the receiving unit 12 receives the answer.

Next, the questioning unit 11 includes the outline of the invention, all questions (first and second questions), all answers to the questions (first and second answers), and question generation instructions in a prompt for question generation, inputs the prompt for question generation into the trained model, and generates a third question.

Figure 3D is a screen example (4) of the invention hearing screen 50, which is an example of the case where the third question is asked. In the third question, for example, an open-ended question is asked about the configuration and means of the invention. As with screen example (2) and the like, the answer is entered in the answer field 52 and the answer button 53 is pressed, whereby the receiving unit 12 receives the answer.

As shown in the figure, multiple items may be asked in one question. In the example shown, the question is about the method to start and stop using the drone, so two items are asked in one question. Note that whether multiple items can be asked in one question can be changed by adjusting the question generation instructions as appropriate. If an instruction to ask one item in one question is specified in the question generation instructions, the trained model will output a question that asks about one item.

Figure 3E is screen example (5) of invention hearing screen 50, which is an example of a screen that informs the user that the invention hearing has progressed and documents can be created. Screen example (5) displays the hearing progress, and indicates that documents can be created because it has reached a predetermined value or higher. By pressing document creation button 57, the documents shown in Figures 4A and 4B described below can be created.

When the document creation button 57 is pressed, for example, a transition to a screen for selecting the type of document may occur. Also, an area for selecting the type of document may be provided in the example screen (5), allowing the type of document to be selected in the example screen (5).

In screen example (5), the user can choose to create a document or continue with the question.
3F is a screen example (6) of the invention hearing screen 50, which shows an example of a case where the questioning continuation button 56 is pressed to continue the questioning. When the questioning continuation button 56 is selected, the questioning unit 11 generates the next question (here, the fourth question) and presents the generated question to the user. As shown in the figure, a document creation button 57 may be displayed on the screen. In this embodiment 1, if the progress of the invention hearing exceeds a predetermined threshold, the user can create documents at any time.

As shown in Figures 3A to 3F, in this Example 1, the system automatically generates questions and asks them based on the answers from the user. By having a computer hear about the invention, it is possible to reduce human effort by eliminating the need for a patent attorney, intellectual property officer, or other human being to hear about the invention.

Figure 4A shows an example of a document created when the document creation button 57 is selected, and is a diagram showing an example of a document created by the document creation unit 13. The document shown in the figure is an invention proposal.

The items to be written in the invention proposal are not particularly limited and can be selected as appropriate. It is preferable that the items to be written in a patent specification be written in the invention proposal, for example, as in the illustrated example, the name of the invention, background, the problem that the invention aims to solve, the means for solving the problem, and the effects of the invention. Also, as illustrated, items that are considered important in the invention may be written. The items to be written in the invention proposal can be adjusted by appropriately changing the document generation instructions included in the document generation prompt.

The created document may be made editable. For example, when the edit button 61 is selected, a document edit screen may be displayed and edits may be accepted from the user. The document creation unit 13 edits the document based on the edits accepted from the user. Here, it is described that the document is edited, but the document may also be regenerated and re-created.

The created document may also be made outputtable. For example, when the output button 62 is selected, the document is output as a file. The file format is not particularly limited, and examples include a text file, a PDF file, and a file that can be edited using word processing software.

Figure 4B is a diagram showing another example of a document created when the document creation button 57 is selected, and is a diagram showing another example of a document created by the document creation unit 13. The document shown in the figure is a patent application specification. By changing the document generation instructions included in the document generation prompt and specifying that a patent application specification be generated, a document can be generated in the format of a patent application specification.

It is also possible to generate only the claims by changing the document generation instructions included in the document generation prompt. The document creation unit 13 may also create drawings for a patent application. Drawings can be generated by changing the document generation instructions included in the document generation prompt and specifying that drawings for a patent application be generated.

Example 2
Next, Example 2, which is an example of this embodiment, will be described with reference to Figures 5A to 3G. Example 2 is an example of a case where an outline of the invention is not accepted, and will be described using an invention of a drone holding an umbrella as an example, as in Example 1. Explanations of matters similar to those in Example 1 will be omitted.

Figure 5A is an example screen (1) of the invention hearing screen 50, which is an example of the screen when the first question is asked. In Example 2, before the first question is asked, an outline of the invention is not accepted from the user, and when the first question is asked, no question is generated but a standard question is asked. In the example shown, the first standard question asks for the name of the invention. In response, the user enters the answer in the answer field 52 and presses the answer button 53.

Next, the reception unit 12 receives an answer from the user, and the questioning unit 11 generates the next question (second question) by including the answer from the user in a prompt for generating a question. In the second embodiment, when generating the second question, the prompt for generating a question includes the answer to the first question and an instruction to generate a question. Then, the prompt for generating a question is input to the trained model, and a question is generated as an output.

Figure 5B is an example screen (2) of the invention hearing screen 50, which is an example of the screen when a second question is asked. As shown in the figure, the second question is a question that takes into account the answer to the first question. In other words, in this embodiment, it is also possible to generate a question without using the outline of the invention.

The user answers the second question, and the answer to the question is used to generate the next question (third question). In this Example 2, when generating the third question, the first question and answer, the second question and answer, and the same question generation instruction as the first question are included in the prompt for generating the question, and this prompt for generating the question is input to the trained model. In other words, in Example 2, when generating a question, all questions and their answers asked before the question was generated, and the question generation instruction are included in the prompt for generating the question.

Figures 5C to 5E are screen examples (3) to (5) of the invention hearing screen 50, which are examples of the screen when the third to fifth questions are asked. The third to fifth questions are processed in the same way as the second question.

Figure 5F is screen example (6) of invention hearing screen 50, which is an example of a screen that informs the user that the invention hearing has progressed and documents can be created. Here, the same as in Figure 3E described above can be done. Screen example (6) displays the hearing progress, and indicates that documents can be created because it has reached a predetermined value or higher. By pressing document creation button 57, the documents shown in Figures 4A and 4B described above can be created.

Figure 5G is an example screen (7) of the invention hearing screen 50, which shows an example of the case where the questioning continuation button 56 is pressed to continue asking questions. Here, the same as in Figure 3F described above can be performed. When the questioning continuation button 56 is selected, the questioning unit 11 generates the next question (here, the sixth question) and presents the generated question to the user. As shown in the figure, a document creation button 57 may be displayed on the screen. In this Example 2, if the progress of the invention hearing exceeds a predetermined threshold, the user can create documents at any time.

As described above, in this second embodiment, it is possible to generate questions without using an outline of the invention. As shown in FIG. 5A, the first question does not ask about the outline of the invention, but, for example, asks about the name of the invention. Users who are unfamiliar with patents may be confused about what to write and explain as the outline of the invention, so in this second embodiment, even unfamiliar users can be smoothly interviewed about the invention.

Example 3
Next, an example of the present embodiment, Example 3, will be described with reference to Figures 6A and 6B. Example 3 is an example of a case where the invention is expanded using the invention proposal created in Example 1 (or Example 2).

Figure 6A is an example screen (1) of the invention expansion screen 70, which is an example of a screen for specifying an invention proposal that describes the contents of the invention. In the example shown, an electronic file of the invention proposal saved on the user's terminal is specified. For example, by pressing the file specification button 72 to specify a file, the file name is displayed in the file display field 71. After specifying the file, pressing the execute button 73 will expand the invention.

Figure 6B shows an example screen (2) of the invention extension screen 70, which is an example of a screen that displays an extended invention (extended invention). In the example shown, the name of the invention described in the invention proposal is displayed in the invention display field 74. In addition, in the example shown, three extended inventions (1) to (3) are displayed.

The invention generation unit 14 in this embodiment 3 extends the invention described in the invention proposal. Moreover, the invention generation unit 14 in this embodiment 3 inputs a prompt for invention extension into the trained model, and generates extended inventions (1) to (3) as the output of the trained model. The prompt for invention extension in this embodiment 3 includes the invention proposal (document) and an invention extension instruction for the trained model.

Although not particularly limited, the invention extension instruction in the third embodiment sets the following as the generation conditions of the extended invention:
(1) The problem and effect are common to the original invention, but the configuration does not have to be related to the original invention, (2) The problem and effect are not common to the original invention, but the configuration is related to the original invention, or (3) The problem and effect are common to the original invention, but the configuration is related to the original invention. By doing this, for example, the extended inventions (1) to (3) shown in the figure are generated.

As shown in the figure, a "balloon with an umbrella" is generated as extended invention (1) (symbol 75a), a "tracking drone" is generated as extended invention (2) (symbol 75b), and a "drone with an umbrella" is generated as extended invention (3) (symbol 75c). Note that extended invention (3) has the same name "drone with an umbrella" as the original invention, but its configuration is different from the original invention. For the sake of explanation, the names of extended inventions (1) to (3) (underlined parts in the figure) are displayed in the illustrated example, but this embodiment is not limited to this, and the names do not need to be displayed.

In this embodiment, a document may be created for a specific extended invention by pressing the document creation button 57. In addition, in this embodiment, extended invention evaluation areas 76a to 76c are provided, and the accuracy of the generation of extended inventions can be improved by evaluating the extended invention. In addition, in this embodiment, a hearing button 77 is provided, and a further hearing similar to that in embodiments 1 and 2 may be conducted for a specific extended invention. The content can be enriched by conducting a hearing for the extended invention.

Example 4
Next, a fourth embodiment will be described with reference to Fig. 7. The fourth embodiment is an example of extending the invention at the point of Fig. 5F in the second embodiment. Unlike the third embodiment, the fourth embodiment extends the invention without using the outline of the invention.

Figure 7 is an example of a screen displayed during Example 2, informing the user that the hearing for the invention has progressed and the invention can be expanded. Note that in Figure 7, a document creation button 57 similar to that in Figure 5F may be displayed.

In Example 4, when the invention expansion button 78 is pressed, the invention heard is expanded based on the questions and answers given so far. In Example 4, the question, the answer to the question, and the invention expansion instructions are included in a prompt for expanding the invention, and this prompt for expanding the invention is input to the trained model, generating an expanded invention as output. The expanded invention generated in this way can be, for example, expanded inventions (1) to (3) as shown in FIG. 6B, similar to Example 3.

(Reference Example 1)
Next, a first reference example will be described with reference to Fig. 8. A first reference example is an example of a case where the invention is expanded using the outline of the invention input as in Fig. 3A of the first embodiment.

Figure 8 shows another example of screen example (1) of the invention hearing screen 50, which is another example of the case where an outline of the invention is entered before asking the first question. In reference example 1, when the invention expansion button 78 is pressed, the invention described in the invention outline is expanded.

In Reference Example 1, the outline of the invention and an invention extension instruction are included in a prompt for extending the invention, and this prompt for extending the invention is input to a trained model, and an extended invention is generated as the output. The extended invention generated in this way can be, for example, extended inventions (1) to (3) as shown in Figure 6B.

REFERENCE SIGNS LIST 1 invention hearing system 10 invention hearing device 11 question section 12 reception section 13 document creation section 14 invention creation section 40 user 41 terminal 50 invention hearing screen 51 question section 52 answer section 53 answer button 54 skip button 55 question evaluation area 56 question continuation button 57 document creation button 60 document 61 correction button 62 output button 70 invention extension screen 71 file display section 72 file specification button 73 execute button 74 invention display section 75a to 75c extended inventions 76a to 76c extended invention evaluation area 77 hearing implementation button 78 invention extension button

Claims (17)

a questioning section for asking the user questions about the invention;
An invention hearing system having a reception unit that receives a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The questioning unit asks a question by inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and sending the generated question to a user, and repeatedly generates and sends questions, and when generating a question, includes in the prompt for generating a question one or more questions asked before the question was generated, one or more answers to the questions asked before the question was generated, and a question generation instruction for the trained model. a questioning section for asking the user questions about the invention;
An invention hearing system having a reception unit that receives a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
the reception unit receives an outline of the invention from the user before the questioning unit asks a first question,
The invention hearing system is characterized in that the questioning unit asks questions by inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and sending the generated question to a user, and when generating a first question, includes an overview of the invention and a question generation instruction for the trained model in the prompt for generating a question, and when generating a second or subsequent question, includes at least one or more answers to the question asked before the question was generated and a question generation instruction for the trained model in the prompt for generating a question. the reception unit receives an outline of the invention from the user before the questioning unit asks a first question,
The invention hearing system described in claim 1, characterized in that when generating a first question, the questioning unit generates the first question by including an outline of the invention and a question generation instruction for the trained model in the question generation prompt. the reception unit does not receive an outline of the invention from the user before the questioning unit asks a first question,
2. The invention hearing system according to claim 1, wherein the questioning section does not generate a question when asking a first question, but asks a standard question. The invention hearing system described in claim 2, characterized in that the question section repeatedly generates and sends questions, and when generating a question, includes one or more questions asked before generating the question and one or more answers to the questions asked before generating the question in the prompt for generating the question. The invention hearing system according to claim 1 or 2, characterized in that the questioning unit repeatedly generates and sends questions, and when generating a question, includes all questions and their answers asked before generating the question in the prompt for generating the question. The hearing system according to claim 1 or 2, characterized in that the questioning unit repeatedly generates and transmits questions, and does not transmit questions with the same content. The reception unit receives an evaluation of the question from a user,
3. The hearing system according to claim 1, wherein the questioning section includes an evaluation of the question in the prompt for generating the question. A document preparation department is provided to prepare documents explaining the contents of the invention;
The document creation unit inputs a prompt for document generation into the trained model and generates the document as an output of the trained model;
3. The invention hearing system according to claim 1, wherein the document generation prompts include a user's answer to a question and a document generation instruction for the trained model. an invention creation unit that creates an extended invention by expanding the invention described in the document;
The invention generation unit inputs a prompt for invention extension into the trained model and generates the extended invention as an output of the trained model;
The invention hearing system of claim 9, wherein the prompt for invention extension includes the document and invention extension instructions for the trained model. an invention generating unit for generating an extended invention;
The invention generation unit inputs a prompt for invention extension into the trained model and generates the extended invention as an output of the trained model;
The invention hearing system described in claim 1 or 2, characterized in that the prompt for expanding the invention includes a question generated by the questioning unit and sent to the user, an answer from the user to the question, and an invention expansion instruction for the trained model. a questioning section for asking the user questions about the invention;
A hearing device for an invention having a reception unit for receiving a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The questioning unit asks a question by inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and sending the generated question to a user, and repeatedly generates and sends questions, and when generating a question, includes in the prompt for generating a question one or more questions asked before the question was generated, one or more answers to the questions asked before the question was generated, and a question generation instruction for the trained model. a questioning section for asking the user questions about the invention;
A hearing device for an invention having a reception unit for receiving a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
the reception unit receives an outline of the invention from the user before the questioning unit asks a first question,
The questioning unit asks questions by inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and sending the generated question to a user, and when generating a first question, includes an overview of the invention and a question generation instruction for the trained model in the prompt for generating a question, and when generating a second or subsequent question, includes at least one or more answers to the question asked before the question was generated and a question generation instruction for the trained model in the prompt for generating a question, characterized in that A computer-implemented invention hearing method, comprising:
a questioning step of asking a user questions about the invention;
A receiving step of receiving a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The questioning process involves inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and sending the generated question to a user, thereby asking a question, and repeatedly generating and sending questions, and when generating a question, including in the prompt for generating a question one or more questions asked before the question was generated, one or more answers to the questions asked before the question was generated, and a question generation instruction for the trained model, characterized in that the questioning process includes in the prompt for generating a question. A computer-implemented invention hearing method, comprising:
a questioning step of asking a user questions about the invention;
A receiving step of receiving a response from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The receiving step receives an outline of the invention from a user before the questioning step asks a first question,
The questioning process involves inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and sending the generated question to a user to ask a question, and when generating a first question, including an overview of the invention and an instruction for generating a question for the trained model in the prompt for generating a question, and when generating a second or subsequent question, including at least one or more answers to questions asked before the question was generated and an instruction for generating a question for the trained model in the prompt for generating a question, characterized in that On the computer,
a questioning process for asking a user questions about the invention;
An invention hearing program that executes an acceptance process for accepting an answer from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The question processing involves inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and sending the generated question to a user, thereby asking a question, and repeatedly generating and sending questions, and when generating a question, including in the prompt for generating a question one or more questions asked before the question was generated, one or more answers to the questions asked before the question was generated, and a question generation instruction for the trained model, the hearing program of the invention characterized in that: On the computer,
a questioning process for asking a user questions about the invention;
An invention hearing program that executes an acceptance process for accepting an answer from a user,
The question is a question asking one or more of the following: the title of the invention, the background of the invention, prior art related to the invention, a problem to be solved by the invention, a means for solving the problem to be solved by the invention, and an effect of the invention;
The reception process receives an outline of the invention from a user before the questioning process asks a first question,
The question processing involves inputting a prompt for generating a question into a trained model, generating a question as an output of the trained model, and sending the generated question to a user to ask a question, and when generating a first question, including an overview of the invention and a question generation instruction for the trained model in the prompt for generating a question, and when generating a second or subsequent question, including at least one or more answers to the question asked before the question was generated and a question generation instruction for the trained model in the prompt for generating a question, said invention hearing program being characterized in that

Images