Information processing device, information processing method, and program

Abstract

To make it easy to present the key learning points in the content. [Solution] The information processing device 1 includes a storage unit 12 that stores a trained model that has been trained as training data, which associates base keywords, which are keywords related to learning items, with base learning points, which are text indicating the learning points when learning about the learning items; an acquisition unit 131 that acquires content to be used as teaching material; a keyword extraction unit 133 that analyzes the content and extracts keywords contained in the content; and a learning point generation unit 134 that inputs the keywords contained in the content into the trained model and generates learning points, which are text indicating the learning points when learning about the content.

Description

Technical Field

[0006] , ,

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program.

Background Art

[0002] In recent years, VR (Virtual Reality) content has been used as a training material. In Patent Document 1, an imaging device for imaging the surgical techniques of a doctor is disclosed (for example, refer to Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] When conducting education using VR content, it is desirable to present the learning points in the VR content to the students receiving the education before viewing the VR content, and to present points for conducting appropriate discussions after viewing the VR video. For this purpose, it is necessary to secure instructor staff who understand the learning points, or to improve the learning guidelines. As a result, the burden on the content producer or educator side after shooting the VR content has increased, which has been a factor inhibiting the expansion of education using VR content.

[0005] Therefore, the present invention has been made in view of these points, and an object thereof is to be able to easily present the learning points in the content.

Means for Solving the Problems

[0006] An information processing apparatus in a first aspect of the present invention includes: a storage unit that stores a trained model trained as training data that associates reference keywords, which are keywords related to learning items, with reference learning points, which are text indicating the learning points when learning about the learning items; an acquisition unit that acquires content to be used as teaching material; a keyword extraction unit that analyzes the content and extracts keywords contained in the content; and a learning point generation unit that inputs the keywords contained in the content into the trained model and generates learning points, which are text indicating the learning points when learning about the content.

[0007] In the aforementioned content, the title of the content is further associated with it, the keyword extraction unit may extract keywords related to the content by analyzing the audio contained in the content, and the learning point generation unit may input the title of the content and the keywords related to the content into the trained model to generate the learning points.

[0008] The keyword extraction unit may further extract keywords related to the content that are estimated based on the objects that appear in the content.

[0009] The pre-trained model has learned training data that associates the attributes of the person learning the learning items, the reference keywords, and the reference learning points, which are text indicating the learning points when learning about the learning items. The learning point generation unit may generate different learning points for each of the attributes of the person learning.

[0010] The acquisition unit may further acquire point information indicating points pointed out by the content provider, and the learning point generation unit may further generate learning points that include text indicating learning points related to the points indicated by the point information.

[0011] The acquisition unit may further acquire point information indicating points pointed out by the content provider, and the learning point generation unit may generate the learning points configured to display the text related to the points pointed out by the content provider from among the text included in the learning points in a different manner from other text.

[0012] The storage unit stores a plurality of contents, the acquisition unit further acquires information indicating the learning status of a user who has viewed any of the plurality of contents, and the information processing device may further include a display control unit that displays information indicating the content recommended to the user from among the plurality of contents on the user's information terminal, based on the content viewed by the user, the user's learning status, and the plurality of learning points.

[0013] In a second aspect of the present invention, the information processing method includes the steps of: acquiring content to be used as teaching material, which is executed by a computer; an extraction step of analyzing the content and extracting keywords contained in the content; and inputting the keywords contained in the content extracted in the extraction step into a trained model that has been trained as training data relating base keywords, which are keywords related to learning items, and base learning points, which are text indicating the points to be learned when learning about the learning items, and generating learning points, which are text indicating the points to be learned when learning about the content.

[0014] In a program according to a third aspect of the present invention, the computer is made to perform the following steps: acquire content to be used as teaching material; analyze the content and extract keywords contained in the content; and input the keywords contained in the content extracted in the extraction step into a trained model that has been trained as training data relating base keywords, which are keywords related to learning items, and base learning points, which are text indicating the points to be learned when learning about the learning items, and generate learning points, which are text indicating the points to be learned when learning about the content. [Effects of the Invention]

[0015] According to the present invention, learning points in content can be easily presented. [Brief explanation of the drawing]

[0016] [Figure 1] This is a diagram illustrating the overview of the information processing system S according to the embodiment. [Figure 2] This is a block diagram showing the configuration of the information processing device 1. [Figure 3] This figure shows an example of training data used by a point learning model. [Figure 4] This figure shows an example of learning points generated by the learning point generation unit. [Figure 5] This figure shows an example of training data used by a point learning model. [Figure 6] This figure shows an example of learning points generated by the learning point generation unit. [Figure 7] This is a flowchart showing the processing flow in the information processing device 1. [Modes for carrying out the invention]

[0017] [Overview of Information Processing System S] FIG. 1 is a diagram for explaining the outline of the information processing system S. The information processing system S includes an information processing apparatus 1, a producer terminal 2, and a user terminal 3.

[0018] The information processing apparatus 1 is a computer for managing content produced for learning purposes. The content handled by the information processing apparatus 1 may be a moving image or a VR video. The information processing apparatus 1 receives the upload of content and distributes the content in response to a request. The information processing apparatus 1 generates learning points including text summarizing the purpose of learning, the method of reflection after viewing the content, the key points of learning, or learning items for the learning items targeted by the content. The information processing apparatus 1 stores and manages the generated learning points in association with the content.

[0019] The producer terminal 2 is a terminal used by the producer UA who produces content. As an example, the producer UA operates the producer terminal 2 to edit the captured video and generate content. The user terminal 3 is a terminal used by the user UB who watches the content for learning. The user UB may be an instructor who instructs students in a classroom or the like. The user terminal 3 is a smartphone, a tablet, a personal computer, a head-mounted display, or the like.

[0020] The processing in the information processing system S will be described. The information processing apparatus 1 acquires content from the producer terminal 2 ((1) in FIG. 1). The content may be provided with metadata indicating the title, the producer, the learning item, the captured action, the playback time, the shooting time, the production time, the shooting location, and the like. The information processing apparatus 1 analyzes the content and extracts keywords ((2) in FIG. 1). As an example, the keywords are important words, phrases, or sentences uttered in the content. The keywords may indicate the name of the subject such as the location where the content was shot, the person, animal, machine, or instrument reflected in the content. Further, the keywords may indicate the state or action of the subject reflected in the content.

[0021] The information processing device 1 generates learning points based on keywords (Figure 1, (3)). Specifically, the information processing device 1 inputs keywords into a point generation model and outputs learning point information. The details of the point generation model will be described later, but the point generation model is a trained model that accepts keywords as input and outputs learning points corresponding to those keywords. The point generation model is, for example, a large-scale language model. The information processing device 1 may further input metadata into the point generation model to generate learning point information.

[0022] The information processing device 1 stores the generated learning points in association with the content for which the learning points were generated. The information processing device 1 transmits the content and the learning points corresponding to that content to the user terminal 3 in response to a request from the user terminal 3 (Figure 1, (4)).

[0023] By configuring the information processing system S to generate learning points for learning items corresponding to the acquired content, it can easily present the learning points in the content. As a result, the information processing system S can reduce the burden on instructors who provide learning guidance and creators who produce the content.

[0024] [Configuration of Information Processing Device 1] Figure 2 is a block diagram showing the configuration of the information processing device 1. The information processing device 1 includes a communication unit 11, a storage unit 12, and a control unit 13. The control unit 13 includes an acquisition unit 131, a learning unit 132, a keyword extraction unit 133, a learning point generation unit 134, and a display control unit 135.

[0025] The communication unit 11 is a communication interface for sending and receiving data with other devices via a network. The storage unit 12 is a storage medium including ROM (Read Only Memory), RAM (Random Access Memory), SSD (Solid State Drive), hard disk drive, etc. The storage unit 12 pre-stores programs to be executed by the control unit 13.

[0026] The memory unit 12 stores the point generation model. The point generation model accepts keywords extracted from the content as input and outputs text indicating the learning points corresponding to the input keywords. The learning of the point generation model will be described later.

[0027] Furthermore, the memory unit 12 stores one or more contents. The memory unit 12 stores, for example, multiple contents transmitted from the creator terminal 2. The memory unit 12 stores the learning points generated by the learning point generation unit 134 for each of the multiple contents, associating them with each of the multiple contents.

[0028] The control unit 13 is a processor, such as a CPU (Central Processing Unit). By executing the program stored in the memory unit 12, the control unit 13 functions as an acquisition unit 131, a learning unit 132, a keyword extraction unit 133, a learning point generation unit 134, and a display control unit 135.

[0029] The learning process for the point generation model is described below. As an example, the acquisition unit 131 acquires training data from an administrator terminal (not shown) used by an administrator who manages the information processing device 1, and inputs it into the learning unit 132. The learning unit 132 trains a machine learning model based on the training data and generates a point generation model. Figure 3 shows an example of training data used by the learning unit 132 for learning. In the training data, reference keywords, which are keywords related to the learning items, and reference learning points, which are text indicating the learning points when learning about the learning items, are associated. The learning unit 132 stores the generated point generation model in the storage unit 12.

[0030] Returning to Figure 2, let's explain the process of generating learning points. The acquisition unit 131 acquires content to be used as teaching material. For example, the acquisition unit 131 may acquire content from the creator terminal 2, or it may acquire content from another device. The acquisition unit 131 may also acquire metadata in association with the content.

[0031] The keyword extraction unit 133 analyzes the content and extracts keywords contained within it. For example, the keyword extraction unit 133 extracts keywords from the content by analyzing the audio contained within it. Specifically, the memory unit 12 stores a keyword extraction model that has been trained to extract keywords from the audio that makes up the content when the content is input. The keyword extraction unit 133 inputs the content into the keyword extraction model and outputs the keywords contained in the audio of the content.

[0032] The keyword extraction unit 133 may extract keywords based on the video footage contained in the content. For example, the keyword extraction unit 133 extracts keywords related to the content that are estimated based on the objects shown in the content. In this case, the keyword extraction model is trained to output keywords related to the video footage in the content. As an example, if the video footage shows a person lying in bed wearing an oxygen mask, the keyword extraction unit 133 extracts the keyword "wearing an oxygen mask".

[0033] The keyword extraction unit 133 may extract keywords based on both the audio and video contained in the content. For example, the keyword extraction model may consist of a trained model that extracts keywords based on the video contained in the content and a trained model that extracts keywords based on the audio contained in the content.

[0034] The learning point generation unit 134 inputs keywords contained in the content into the point generation model and generates learning points, which are text indicating points to remember when learning the content. The learning point generation unit 134 stores the generated learning points in the storage unit 12, associating them with the content that the learning points are targeting.

[0035] The display control unit 135 may display a screen containing learning points generated in response to a request from the user terminal 3. Figure 4 shows an example of a screen displayed by the display control unit 135. In the screen shown in Figure 4, the learning theme, learning time, and learning points are displayed. As an example, the display control unit 135 refers to the metadata of the content acquired by the acquisition unit 131 to identify the learning time and displays the identified learning time on the screen. The display control unit 135 configures the information to be displayed for the learning theme and learning points based on the information generated by the point generation model based on keywords. If the metadata includes information indicating the learning theme (e.g., the title of the content), the display control unit 135 may display information indicating the learning theme identified based on the metadata.

[0036] Furthermore, if the learning theme identified based on metadata and the learning points generated by the point generation model are unrelated, it may cause confusion for the user. Therefore, the display control unit 135 may calculate the similarity between the learning theme identified based on metadata and the learning points generated by the point generation model, and display an alert on the user terminal 3 to draw the user's attention if the similarity is below a predetermined threshold.

[0037] The information processing device 1 is configured to generate learning points based on keywords extracted from the audio and video of the content, making it easy to present the learning points in the content. As a result, the burden on instructors who guide learning and creators who produce the content can be reduced.

[0038] Metadata such as titles may contain information that clearly indicates the content's learning characteristics, and by further inputting this metadata into the point generation model, the accuracy of the output learning points can be improved. Therefore, the learning point generation unit 134 may be configured to generate learning points using this metadata.

[0039] In this case, the acquisition unit 131 further acquires metadata of the content in association with the content. The metadata includes, for example, the title of the content. The learning point generation unit 134 inputs the title of the content and keywords related to the content into the point generation model and generates learning points. By configuring the content generation model to generate learning points based on the metadata, the accuracy of the learning points generated by the information processing device 1 can be improved.

[0040] Incidentally, just as doctors and nurses may look at different points when viewing the same scene, the learning objectives for a single piece of content may differ depending on the learner's proficiency level, position, occupation, etc. Therefore, the information processing device 1 may be configured to output learning points according to the attributes of the learner, based on a point generation model that has been trained to output different learning points according to the attributes of the learner. Examples of the learner's attributes include, but are not limited to, proficiency level, position, occupation, etc.

[0041] Figure 5 shows an example of training data that the learning unit 132 uses to train the point generation model in this case. In the training data shown in Figure 5, a reference keyword, the attributes of the learning target, and a reference learning point corresponding to the reference keyword and the learning target are associated. The learning unit 132 trains a machine learning model and generates a point generation model based on the training data acquired by the acquisition unit 131. As an example, the learning unit 132 is configured to accept a keyword as input and generate learning points corresponding to the attributes of the learning target and the keyword and the attributes of the target.

[0042] The learning point generation unit 134 generates different learning points for each subject's attributes. Specifically, the learning point generation unit 134 inputs the keywords extracted by the keyword extraction unit 133 into the point generation model and outputs the attributes of multiple subjects and the learning points corresponding to each of the attributes of the multiple subjects.

[0043] The point generation model may be configured to accept keywords and the attributes of the learner as input and generate learning points corresponding to the keywords and the attributes of the learner. In this case, the acquisition unit 131 acquires content and the attributes of the learner in association with each other, and the learning point generation unit 134 inputs the keywords and the characteristics acquired by the acquisition unit 131 into the point generation model and generates learning points.

[0044] With the information processing device 1 configured in this way, it becomes possible to generate learning points tailored to various target audiences from a single piece of content. As a result, the burden on instructors and others to create content that matches the attributes of the target audience can be reduced.

[0045] Content creators and providers (hereinafter simply referred to as "providers, etc.") often produce or provide content with the intention of highlighting specific points within the content that users should focus on. Therefore, by generating learning points based on the points pointed out by these content providers, etc., it becomes possible to generate learning points that reflect the intentions of the providers, etc.

[0046] Therefore, the acquisition unit 131 may further acquire point information indicating points pointed out by the content provider, etc. The point information may include, for example, text indicating things that should be taken into consideration and things that should not be taken into consideration in the video included in the content. For example, if the content concerns practical skills for treating patients in a medical setting, the point information may include information indicating "confirmation of name".

[0047] The learning point generation unit 134 generates learning points that further include text indicating learning points related to the points indicated by the pointed-out point information. The learning point generation unit 134 further inputs the pointed-out point information into the point generation model and generates learning points.

[0048] By configuring the information processing device 1 to generate learning points that reflect the points pointed out by content providers, it becomes possible to provide learners with learning points that are structured to make them aware of the points that content providers consider important.

[0049] Points highlighted by content creators or providers may be intended to be studied with particular attention compared to other points in the learning items. Therefore, the learning point generation unit 134 may be configured to generate learning points that are displayed with emphasis on items related to the points indicated by the highlighted points information, compared to other points.

[0050] For example, the point generation model in this case may be trained to further output information indicating the degree of relevance between the points indicated by the identified point information. The learning point generation unit 134 generates learning points configured to display text related to the points identified by the content provider from among the text included in the learning points in a different manner from other text. A different manner from other text might be, for example, displaying it in a different font (e.g., red) or a different font (e.g., bold, underlined).

[0051] The learning point generation unit 134 generates learning points configured to display in a different manner from other text the portion of the text output by the point generation model whose correlation with the points indicated by the pointed-out point information is above a predetermined threshold. Figure 6 shows an example of a learning point generated by the learning point generation unit 134 in this case. In the learning point shown in Figure 6, when the pointed-out point information includes information indicating "confirmation of name", the portion corresponding to the pointed-out point, "Confirm the patient's name," is highlighted with an underline.

[0052] By configuring the information processing device 1 to display the parts related to the points indicated by the points information in a different manner from other parts, users referring to learning points can more easily focus their attention on the points that the content creators or providers consider important.

[0053] When the information processing device 1 manages a large number of content items, it is conceivable that users may be confused about which content to refer to. Therefore, the information processing device 1 may be configured to manage the learning stage of users who have learned the distributed content and to recommend content according to their learning stage. In this case, the storage unit 12 stores data indicating the user who learned the content and their learning stage, associated with each of the multiple content items.

[0054] The acquisition unit 131 acquires information indicating the learning status of a user who has viewed one of the multiple contents stored in the storage unit 12. Examples of learning status include the user's heart rate changes while viewing the content, the changes in the position the user focused on in the video, and the progress of learning. The acquisition unit 131 acquires the content viewed by the user and the information indicating the user's learning status who viewed that content from the user terminal 3, associating them together.

[0055] The display control unit 135 displays information on the user terminal 3 indicating content recommended to the user from among multiple content items, based on the content the user has viewed, the user's learning status, and multiple learning points. For example, the storage unit 12 stores a content table that associates content, the learning status that serves as the completion criterion for that content, and content recommended upon completion of that content. The display control unit 135 identifies content recommended to the user based on the information acquired by the acquisition unit 131, which indicates the content the user has viewed and the user's learning status, and displays information indicating the identified content on the user terminal 3.

[0056] With the information processing device 1 configured in this way, it can recommend content that is appropriate to the learner's learning stage, and as a result, the burden on learners and their instructors to select learning content can be reduced.

[0057] [Processing flow in information processing device 1] Figure 7 is a flowchart showing the processing flow in the information processing device 1. The flowchart shown in Figure 7 starts from the point where content is acquired.

[0058] The acquisition unit 131 acquires content (S01). The acquisition unit 131 may also acquire the attributes of the target person to generate learning points in association with the content. The keyword extraction unit 133 analyzes the objects reflected in the video that constitutes the content and extracts keywords (S02). The keyword extraction unit 133 analyzes the audio that constitutes the content and extracts keywords (S03).

[0059] The keyword extraction unit 133 determines whether or not a keyword has been extracted (S04). If a keyword has been extracted (YES in S04), the learning point generation unit 134 determines whether or not there is a limitation on the target person for whom learning points are generated (i.e., a specification of the target person's attributes) (S05).

[0060] If there are restrictions on the target audience (YES in S05), the learning point generation unit 134 inputs the target audience's attributes and the extracted keywords into the point generation model and generates learning points (S06). If there are no restrictions on the target audience (NO in S05), the learning point generation unit 134 inputs the extracted keywords into the point generation model and generates learning points (S07). The learning point generation unit 134 stores the generated learning points in association with the content (S08), and then the information processing device 1 terminates the process.

[0061] If no keywords are extracted in S04, it is considered that there is a defect in the content, such as unclear audio or video contained in the content. Therefore, if no keywords are extracted (NO in S04), the display control unit 135 displays a message prompting the user to check the content (S09). Then, the information processing device 1 terminates processing.

[0062] [Effects of Information Processing Device 1] The information processing device 1 stores a point generation model that generates learning points based on keywords, and is configured to input keywords extracted from the audio and video of the content into the point generation model to generate learning points. This allows learners using the content to easily see the learning points within the content. As a result, the burden on instructors to plan lesson content can be reduced, and the burden on content creators to generate information to explain the content in detail can be reduced. In addition, by presenting the learning points within the content, learners can easily select content. Furthermore, by having the information processing device 1 generate learning points within VR content, further spread of VR learning can be expected.

[0063] Although the present invention has been described above using embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments, and various modifications and changes are possible within the scope of its gist. For example, all or part of the apparatus can be configured by functionally or physically distributing and integrating in any unit. Furthermore, new embodiments resulting from any combination of multiple embodiments are also included in the embodiments of the present invention. The effects of the new embodiments resulting from the combinations are combined with the effects of the original embodiments. [Explanation of symbols]

[0064] 1. Information Processing Device 2. Developer's terminal 3. User terminals 11 Communications Department 12 Storage section 13 Control Unit 131 Acquisition Department 132 Learning Department 133 Keyword Extraction Section 134 Learning Point Generation Unit 135 Display Control Unit

Claims

1. A storage unit that stores a trained model that has been trained as training data, which associates reference keywords, which are keywords related to the learning items, with reference learning points, which are text indicating the learning points when learning about the learning items. An acquisition unit that acquires content to be used as teaching material, A keyword extraction unit analyzes the aforementioned content and extracts keywords contained in the aforementioned content, A learning point generation unit inputs keywords contained in the content into the trained model and generates learning points, which are text indicating the points to remember when learning the content. An information processing device having

2. In the aforementioned content, the title of the content is further associated with: The keyword extraction unit extracts keywords related to the content by analyzing the audio contained in the content. The learning point generation unit inputs the title of the content and keywords related to the content into the trained model and generates the learning points. The information processing apparatus according to claim 1.

3. The keyword extraction unit further extracts keywords related to the content that are estimated based on the objects that appear in the content. The information processing apparatus according to claim 2.

4. The aforementioned trained model has learned training data that associates the attributes of the person learning the learning item, the reference keywords, and the reference learning points, which are text indicating the learning points when learning the learning item. The learning point generation unit generates different learning points for each of the subject's attributes. The information processing apparatus according to any one of claims 1 to 3.

5. The acquisition unit further acquires the points of reference information that indicates the points pointed out by the content provider, The learning point generation unit generates the learning points which further include text indicating learning points related to the points indicated by the pointed-out point information. The information processing apparatus according to any one of claims 1 to 3.

6. The acquisition unit further acquires the points of reference information that indicates the points pointed out by the content provider, The learning point generation unit generates learning points configured to display, among the text included in the learning points, text related to the points pointed out by the content provider in a different manner from other text. The information processing apparatus according to any one of claims 1 to 3.

7. The aforementioned storage unit stores multiple contents, The acquisition unit further acquires information indicating the learning status of a user who has viewed any of the multiple contents, The information processing device further includes a display control unit that displays information on the user's information terminal indicating content recommended to the user from among the multiple pieces of content, based on the content viewed by the user, the user's learning status, and the multiple learning points. The information processing apparatus according to claim 1.

8. A computer executes Steps to acquire content to be used as teaching materials, An extraction step of analyzing the aforementioned content and extracting keywords contained in the aforementioned content, A trained model, which has been trained using training data that associates base keywords, which are keywords related to learning items, with base learning points, which are text indicating the key points to learn when learning about the learning items, is given keywords included in the content extracted in the extraction step, and a learning point is generated, which is text indicating the key points to learn about the content. An information processing method having

9. On the computer, Steps to acquire content to be used as teaching materials, An extraction step of analyzing the aforementioned content and extracting keywords contained in the aforementioned content, A trained model, which has been trained using training data that associates base keywords, which are keywords related to learning items, with base learning points, which are text indicating the key points to learn when learning about the learning items, is given keywords included in the content extracted in the extraction step, and a learning point is generated, which is text indicating the key points to learn about the content. A program to execute.

Images