Information processing device, information processing method, and recording medium

Abstract

To provide an information processing device or the like capable of presenting suitable processing flow including processing using AI.SOLUTION: In an information processing device, registration means registers data obtained by associating an AI flow representing processing flow including processing using one or more AIs with an explanatory sentence being a text describing a summary of processing to be executed by the AI flow in a database. Flow acquisition means acquires a first AI flow associated with one explanatory sentence representing processing being the most similar to processing for achieving a user's desired purpose on the basis of the data registered in the database. Display information generation means generates display information to present at least one AI flow from the first AI flow or a second AI flow generated on the basis of a summary sentence being a sentence describing the summary of the processing for achieving the user's desired purpose as recommendation flow to the user.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present disclosure relates to the technology of AI (Artificial Intelligence). [Background Art]

[0002] In recent years, AI constructed by machine learning models and the like has been used in various fields such as video analysis.

[0003] Specifically, for example, Patent Document 1 discloses a technology for generating an object detection identification model used for object detection through observation under unknown observation conditions by machine learning using known observation images as training data. [Prior Art Documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Unexamined Patent Application Publication No. 2022-76675 [Summary of the Invention] [Problem to be Solved by the Invention]

[0005] Here, when creating a suitable processing flow including AI-based processing for a user to achieve a desired object, the user needs to have at least a certain level of knowledge about the functions of AI included in the processing flow.

[0006] In contrast, Patent Document 1 does not disclose a method related to creating a processing flow including AI-based processing. Therefore, according to the technology disclosed in Patent Document 1, there arises a problem that a suitable processing flow including AI-based processing cannot be presented to a user.

[0007] An object of the present disclosure is to provide an information processing apparatus capable of presenting a suitable processing flow including AI-based processing to a user. [Means for Solving the Problem]

[0008] In one aspect of this disclosure, the information processing device includes: registration means for registering data in a database that associates an AI flow representing a processing flow that includes processing using one or more AIs, and an explanatory text which is a sentence describing the outline of the processing performed by the AI ​​flow; flow acquisition means for acquiring a first AI flow associated with an explanatory text which represents processing that is most similar to processing to achieve the user's desired objective, based on the data registered in the database; and display information generation means for generating display information to present to the user, as a recommended flow, at least one of the following AI flows: the first AI flow and a second AI flow generated based on an outline text which is a sentence describing the outline of processing to achieve the user's desired objective. The display information generation means generates the display information for presenting the second AI flow to the user as the recommended flow when it determines, based on the distance between the feature vector extracted from the first explanatory text and the feature vector extracted from the summary text, that the processing of the first explanatory text and the processing of the summary text are dissimilar. In other aspects of this disclosure, the information processing device includes: registration means for registering data in a database that associates an AI flow representing a processing flow that includes processing using one or more AIs with an explanatory text which is a sentence describing an outline of the processing performed by the AI ​​flow; flow acquisition means for acquiring a first AI flow associated with an explanatory text that represents processing that is most similar to processing to achieve a user's desired objective, based on the data registered in the database; and display information generation means for generating display information to present to the user at least one of the first AI flow and a second AI flow generated based on the first AI flow and an outline text which is a sentence describing an outline of processing to achieve the user's desired objective, as a recommended flow. The display information generation means generates the display information to present to the user the first AI flow and the second AI flow as the recommended flow when it determines that the processing of the first explanatory text and the processing of the outline text are similar based on the distance between a feature vector extracted from the explanatory text and a feature vector extracted from the outline text.

[0009] This disclosure moreover From another perspective, The computer executes The information processing method involves registering data in a database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with an explanatory text, which is a sentence describing the outline of the processing performed by the AI ​​flow; obtaining a first AI flow associated with an explanatory text that represents the processing most similar to the processing that achieves the user's desired objective, based on the data registered in the database; and generating display information to present at least one of the AI ​​flows, which is selected from the first AI flow and a second AI flow generated based on the explanatory text describing the outline of the processing that achieves the user's desired objective, as a recommended flow to the user. Based on the distance between the feature vector extracted from the first explanatory text and the feature vector extracted from the summary text, if it is determined that the processing of the first explanatory text and the processing of the summary text are dissimilar, the display information is generated to present the second AI flow to the user as the recommended flow. . In yet another aspect of this disclosure, a computer-based information processing method includes registering data in a database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with an explanatory text, which is a sentence describing the outline of the processing performed by the AI ​​flow; obtaining a first AI flow associated with an explanatory text that represents processing most similar to processing to achieve a user's desired objective based on the data registered in the database; generating display information to present at least one AI flow from the first AI flow and a second AI flow generated based on the explanatory text and an outline text, which is a sentence describing the outline of processing to achieve the user's desired objective, to the user as a recommended flow; and generating the display information to present the first AI flow and the second AI flow to the user as a recommended flow when it is determined that the processing of the explanatory text and the processing of the outline text are similar based on the distance between a feature vector extracted from the explanatory text and a feature vector extracted from the outline text.

[0010] In yet another aspect of this disclosure, the recording medium registers data in a database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with a descriptive text, which is a sentence describing the outline of the processing performed by the AI ​​flow; retrieves a first AI flow associated with a descriptive text that represents the processing most similar to the processing that achieves the user's desired objective, based on the data registered in the database; and generates display information to present to the user at least one of the AI ​​flows, which is a recommended flow, from the first AI flow and a second AI flow generated based on the descriptive text, which is a sentence describing the outline of the processing that achieves the user's desired objective. Based on the distance between the feature vector extracted from the first explanatory text and the feature vector extracted from the summary text, if it is determined that the processing of the first explanatory text and the processing of the summary text are dissimilar, the display information is generated to present the second AI flow to the user as the recommended flow. Record the program that instructs the computer to perform the process. In yet another aspect of this disclosure, the recording medium records a program that causes a computer to execute a process to generate the display information for presenting the first AI flow and the second AI flow to the user as the recommended flow when it is determined that the processing of the first description and the processing of the summary are similar based on the distance between a feature vector extracted from the first description and a feature vector extracted from the summary, and when it is determined that the processing of the first description and the processing of the summary are similar. [Effects of the Invention]

[0011] According to this disclosure, it will be possible to present users with a suitable processing flow that includes AI-based processing. [Brief explanation of the drawing]

[0012] [Figure 1] A diagram showing the schematic configuration of an information processing system including a server device according to the first embodiment. [Figure 2] A block diagram showing the hardware configuration of the server device according to the first embodiment. [Figure 3] A block diagram showing the functional configuration of the server device according to the first embodiment. [Figure 4] A diagram illustrating a concrete example of an AI flow. [Figure 5] A diagram illustrating a concrete example of an AI flow. [Figure 6] A diagram illustrating a concrete example of an AI flow formula. [Figure 7] A diagram illustrating a concrete example of an AI flow formula. [Figure 8] A diagram illustrating a concrete example of table data. [Figure 9]A flowchart illustrating an example of processing performed in the server device according to the first embodiment. [Figure 10] A block diagram showing a functional configuration of the information processing apparatus according to the second embodiment. [Figure 11] A flowchart for explaining processing performed in the information processing apparatus according to the second embodiment. MODES FOR CARRYING OUT THE INVENTION

[0013] Hereinafter, preferred embodiments of the present disclosure will be described with reference to the drawings.

[0014] <First Embodiment> [Schematic Configuration] FIG. 1 is a diagram showing a schematic configuration of an information processing system including a server device according to the first embodiment. As shown in FIG. 1, an information processing system 1 includes a server device 100, a provider-side terminal device 200, and a user-side terminal device 300.

[0015] The server device 100 is configured to be capable of communicating with the provider-side terminal device 200 and the user-side terminal device 300. The server device 100 also performs processing for registering an AI flow included in registration information transmitted from the provider-side terminal device 200. Further, based on query information transmitted from the user-side terminal device 300, the server device 100 acquires an AI flow suitable for achieving a user's desired objective, and outputs the AI flow to the user-side terminal device 300.

[0016] An AI flow is a graphical representation of a processing flow including processing using one or more AIs. A specific example of the AI flow will be described later.

[0017] The provider-side terminal device 200 has the function of communicating with the server device 100, the function of inputting information to be sent to the server device 100, and the function of displaying information received from the server device 100. In addition, the provider-side terminal device 200 has the function of displaying information in accordance with the operation of the provider that provides the AI ​​flow. Specifically, the provider-side terminal device 200 is composed of devices such as a personal computer, a smartphone, and a tablet computer.

[0018] The provider-side terminal device 200 has a configuration that allows input of an AI flow and an explanatory text, which is a sentence describing the outline of the processing performed by the AI ​​flow in natural language, in a predetermined GUI (Graphical User Interface). The aforementioned predetermined GUI may be, for example, displayed when accessing the server device 100, or displayed when launching a predetermined application installed on the provider-side terminal device 200. The provider-side terminal device 200 also transmits registration information, which includes the AI ​​flow and explanatory text entered in the aforementioned predetermined GUI, to the server device 100.

[0019] The user-side terminal device 300 has the function of communicating with the server device 100, the function of inputting information to be sent to the server device 100, and the function of displaying information received from the server device 100. Furthermore, the user-side terminal device 300 has the function of displaying information corresponding to the user's actions, such as those of a user who wishes to use a suitable AI flow to achieve a desired objective. Specifically, the user-side terminal device 300 is composed of devices such as a personal computer, a smartphone, and a tablet computer.

[0020] The user-side terminal device 300 has a configuration that allows the user to input a summary statement, which is a sentence describing the outline of the process to achieve the user's desired objective, in natural language, in a predetermined GUI. The aforementioned predetermined GUI may be, for example, one that is displayed when accessing the server device 100, or one that is displayed when launching a predetermined application installed on the user-side terminal device 300. The user-side terminal device 300 also sends query information, which is information including the summary statement entered in the aforementioned predetermined GUI, to the server device 100.

[0021] [Hardware configuration] Figure 2 is a block diagram showing the hardware configuration of a server device according to the first embodiment. As shown in Figure 2, the server device 100 includes an interface (IF) 111, a processor 112, a memory 113, a recording medium 114, and a database (DB) 115.

[0022] IF111 performs data input and output with external devices. For example, information transmitted from the provider-side terminal device 200 and information transmitted from the user-side terminal device 300 are input to the server device 100 through IF111.

[0023] The processor 112 is a computer such as a CPU (Central Processing Unit) and controls the entire server device 100 by executing pre-prepared programs. Specifically, the processor 112 performs processing related to the generation of display information for presenting the recommended flow described later to the user.

[0024] Memory 113 consists of ROM (Read Only Memory), RAM (Random Access Memory), and other components. Memory 113 is also used as working memory while the processor 112 is executing various processes.

[0025] The recording medium 114 is a non-volatile, non-temporary recording medium such as a disk-shaped recording medium or semiconductor memory, and is configured to be detachable from the server device 100. The recording medium 114 stores various programs that the processor 112 executes. When the server device 100 performs various processes, the programs stored in the recording medium 114 are loaded into the memory 113 and executed by the processor 112.

[0026] DB115 stores, for example, information input through IF111 and processing results obtained from the processor 112.

[0027] [Functional Configuration] Figure 3 is a block diagram showing the functional configuration of a server device according to the first embodiment. As shown in Figure 3, the server device 100 includes a registration processing unit 11, a data storage unit 12, a feature extraction processing unit 13, a similar flow acquisition unit 14, a flow generation unit 15, and a display information generation unit 16.

[0028] The registration processing unit 11 functions as a registration means. The registration processing unit 11 converts the AI ​​flow included in the registration information obtained from the provider-side terminal device 200 into an AI flow formula and outputs data associating the AI ​​flow formula with the explanatory text to the data storage unit 12. In other words, the registration processing unit 11 performs processing to register data in the database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with an explanatory text, which is a sentence describing the outline of the processing executed by the AI ​​flow.

[0029] An AI flow formula is a string of characters that uniquely represents the processing flow using one or more AIs by applying natural language processing to the AI ​​flow and summary text, etc., and arranging the extracted words according to predetermined rules. In other words, an AI flow formula can be treated as a string of characters that represents the same processing flow as the processing flow represented by the AI ​​flow. The registration processing unit 11 converts the AI ​​flow into an AI flow formula by performing a process of arranging the words extracted from the AI ​​flow contained in the registration information according to predetermined rules. Furthermore, according to this embodiment, for example, the AI ​​flow formula can be converted back into an AI flow by using the words extracted from the AI ​​flow formula based on the predetermined rules mentioned above. Specific examples of AI flow formulas will be explained later.

[0030] The data storage unit 12 functions as a database. The data storage unit 12 also stores table data, which is data recording the association status of AI flow formulas and explanatory texts based on the data obtained from the registration processing unit 11. In other words, the AI ​​flow formulas and explanatory texts based on the data obtained from the registration processing unit 11 are registered in the table data of the data storage unit 12. A specific example of the table data will be explained later.

[0031] The feature extraction processing unit 13 extracts a feature vector that quantifies the features of the summary sentence by applying predetermined natural language processing to the summary sentence contained in the query information obtained from the user terminal device 300. The feature extraction processing unit 13 also outputs data relating the summary sentence and the feature vector extracted from the summary sentence to the similar flow acquisition unit 14.

[0032] The feature extraction processing unit 13 extracts n feature vectors, which are numerical representations of the individual features of the n description sentences, by individually applying predetermined natural language processing to the n description sentences (registered) contained in the table data stored in the data storage unit 12. The feature extraction processing unit 13 also outputs data to the similarity flow acquisition unit 14 that associates one description sentence with one feature vector extracted from that description sentence, for each of the n description sentences and feature vectors.

[0033] In the following explanation, unless otherwise specified, the case in which feature vectors are individually extracted from all the descriptive sentences contained in the table data stored in the data storage unit 12 will be used as an example. Furthermore, according to this embodiment, the process related to the extraction of feature vectors by the feature extraction processing unit 13 may include, for example, natural language processing using a Transformer model, which is a deep learning model related to natural language analysis. As the aforementioned Transformer model, for example, models such as BERT, ELECTRA, and T5 can be used.

[0034] The Similar Flow Acquisition Unit 14 functions as a flow acquisition means. The Similar Flow Acquisition Unit 14 calculates the distance between the feature vector associated with the summary text and the n feature vectors associated with each of the n description texts, based on the data obtained from the feature extraction processing unit 13. Specifically, the Similar Flow Acquisition Unit 14 calculates the aforementioned distance as, for example, the Euclidean distance or the cosine distance. The Similar Flow Acquisition Unit 14 then identifies a description text associated with the feature vector for which the minimum distance was calculated among the n distances calculated for each of the n feature vectors. The Similar Flow Acquisition Unit 14 also acquires an AI flow expression associated with the description text identified as described above by referring to the table data in the data storage unit 12. In other words, the Similar Flow Acquisition Unit 14 acquires an AI flow associated with a description text that represents the process most similar to the process for achieving the user's desired objective, based on the data registered in the table data of the data storage unit 12. The similar flow acquisition unit 14 then outputs data including the AI ​​flow formula acquired as described above and the minimum distance value described above to the display information generation unit 16.

[0035] The flow generation unit 15 generates an AI flow expression corresponding to the processing of the summary sentence by applying predetermined natural language processing to the summary sentence included in the query information obtained from the user terminal device 300. The flow generation unit 15 also outputs data including the AI ​​flow expression generated from the summary sentence to the display information generation unit 16.

[0036] Furthermore, according to this embodiment, the process related to the generation of the AI ​​flow formula by the flow generation unit 15 may include, for example, natural language processing using a Transformer model, which is a deep learning model related to natural language analysis. In addition, as the aforementioned Transformer model, models such as BERT, ELECTRA, and T5 can be used.

[0037] The display information generation unit 16 functions as a means for generating display information. Based on the data obtained from the similar flow acquisition unit 14 and the data obtained from the flow generation unit 15, the display information generation unit 16 generates display information to present at least one AI flow as a recommended flow to the user, and transmits the generated display information to the user terminal device 300. In addition, when generating the display information, the display information generation unit 16 performs a process to convert the AI ​​flow formula into an AI flow using words extracted from the AI ​​flow formula.

[0038] The aforementioned recommended flow should include an AI flow suitable for achieving the user's desired objective as described in the summary. Preferably, the aforementioned recommended flow should include an AI flow corresponding to the data obtained from the flow generation unit 15. Furthermore, the aforementioned recommended flow should include at least one AI flow from among the AI ​​flow corresponding to the data obtained from the similar flow acquisition unit 14 and the AI ​​flow corresponding to the data obtained from the flow generation unit 15. Alternatively, the aforementioned recommended flow may include an AI flow corresponding to the data obtained from the similar flow acquisition unit 14 and an AI flow corresponding to the data obtained from the flow generation unit 15.

[0039] The display information generation unit 16 may decide whether or not to present the AI ​​flow corresponding to the data as a recommended flow to the user based on the comparison result obtained by comparing the minimum distance value included in the data obtained from the similar flow acquisition unit 14 with a predetermined value. In other words, the display information generation unit 16 may decide whether or not to present the AI ​​flow corresponding to the data as a recommended flow to the user based on the minimum distance value included in the data obtained from the similar flow acquisition unit 14.

[0040] Specifically, for example, the display information generation unit 16 may not present the AI ​​flow corresponding to the data obtained from the similar flow acquisition unit 14 as a recommended flow to the user if the aforementioned minimum distance value is greater than a predetermined value. In such cases, the display information generation unit 16 may present the AI ​​flow corresponding to the data obtained from the flow generation unit 15 as a recommended flow. That is, when the display information generation unit 16 determines, based on the aforementioned minimum distance, that the processing of a description and the processing of a summary identified by the similar flow acquisition unit 14 are dissimilar, it may present the AI ​​flow corresponding to the data obtained from the flow generation unit 15 as a recommended flow.

[0041] Furthermore, for example, the display information generation unit 16 may present to the user as a recommended flow an AI flow corresponding to the data obtained from the similar flow acquisition unit 14 when the aforementioned minimum distance value is less than or equal to a predetermined value. In such a case, the display information generation unit 16 may further present as a recommended flow an AI flow corresponding to the data obtained from the flow generation unit 15. That is, when the display information generation unit 16 determines, based on the aforementioned minimum distance, that the processing of a description sentence identified by the similar flow acquisition unit 14 and the processing of a summary sentence are similar, it may present as a recommended flow an AI flow corresponding to the data obtained from the similar flow acquisition unit 14 and an AI flow corresponding to the data obtained from the flow generation unit 15.

[0042] In this embodiment, the similar flow acquisition unit 14 may, for example, acquire k distances in ascending order from the n distances calculated as described above, acquire k feature vectors used to calculate the k distances, and identify k explanatory sentences associated with each of the k feature vectors. Alternatively, the similar flow acquisition unit 14 may acquire k AI flow formulas associated with the aforementioned k explanatory sentences by referring to the table data in the data storage unit 12, and output data including the k AI flow formulas and the aforementioned k distances to the display information generation unit 16. In such a case, the display information generation unit 16 may determine which of the k AI flows corresponding to the data to present to the user as a recommended flow based on the comparison result obtained by comparing the values ​​of each of the k distances included in the data obtained from the similar flow acquisition unit 14 with a predetermined value.

[0043] [Specific example] Next, we will explain some specific examples of the elements in the configuration described above.

[0044] (Specific examples of AI workflows) Figure 4 is a diagram illustrating a specific example of an AI flow. The AI ​​flow AFA in Figure 4 graphically represents the process of detecting scenes containing people from video footage. The AI ​​flow AFA includes six blocks: "Video (VIDEO)", "[YOLO General Object Detection (YOLO)]", "Category Filter (FILTER)", "Person (CATE:Person)", "Object Existence Detect (EXIST)", and "End of Processing (END)". In the AI ​​flow AFA, all blocks except "Person (CATE:Person)" are connected using arrows indicating the order of processing. The "Person (CATE:Person)" block in the AI ​​flow AFA represents the parameter set as the processing target of the "Category Filter (FILTER)" block. The parameters in the AI ​​flow AFA processing should be set as needed.

[0045] In AI flow AFA, the "VIDEO" block indicates that a video is input from the outside. Further, the "[General Object Detection (YOLO)]" block in AI flow AFA indicates that processing using an object detection AI called YOLO (You Only Look Once) is performed on the video input from the outside. Further, the "Category Filter (FILTER)" and "Person (CATE:Person)" blocks in AI flow AFA indicate that a filter for human detection is applied to the processing result obtained by the processing using the object detection AI. Further, the "Object Presence Determination (EXIST)" block in AI flow AFA indicates that determination processing on whether a person is present is performed on the processing result obtained through the filter for human detection. Further, the "End of Processing (END)" block in AI flow AFA indicates that a series of processing is terminated.

[0046] FIG. 5 is a diagram for explaining a specific example of an AI flow. AI flow AFB in FIG. 5 graphically represents the flow of processing for detecting a scene containing an umbrella from a video. Further, AI flow AFB includes six blocks: "VIDEO", "[General Object Detection (YOLO)]", "Category Filter (FILTER)", "Umbrella (CATE:Umbrella)", "Object Presence Determination (EXIST)", and "End of Processing (END)". Further, each block other than the "Umbrella (CATE:Umbrella)" block in AI flow AFB is connected using arrows indicating the order of processing. Further, the "Umbrella (CATE:Umbrella)" block in AI flow AFB represents a parameter set as the processing target of the "Category Filter (FILTER)" block. Note that parameters in the processing of AI flow AFB may be set as needed.

[0047] In the AI ​​flow AFB, the "Video" block indicates that video is input from an external source. The "[YOLO Object Detection]" block in the AI ​​flow AFB indicates that the externally input video is processed using an AI called YOLO for object detection. The "FILTER" and "CATE:Umbrella" blocks in the AI ​​flow AFB indicate that an umbrella detection filter is applied to the processing result obtained using the object detection AI. The "EXIST" block in the AI ​​flow AFB indicates that the processing result obtained after the umbrella detection filter is processed to determine whether or not an umbrella exists. The "END" block in the AI ​​flow AFB indicates that the series of processes has ended.

[0048] The provider can create (input) an AI flow like the one shown in the above example by operating a predetermined GUI displayed on the provider-side terminal device 200. In this embodiment, it is sufficient that the predetermined GUI displayed on the provider-side terminal device 200 has a configuration that allows setting at least one element included in the AI ​​flow shown in the above example, such as the string within a block, the number of blocks, the number of arrows, and the direction of the arrows. The display information generation unit 16 can also generate display information to present the AI ​​flow shown in the above example as a recommended flow to the user, and transmit the generated display information to the user-side terminal device 300.

[0049] (Specific example of an AI flow formula) Figure 6 is a diagram illustrating a specific example of an AI flow formula. The AI ​​flow formula ASA in Figure 6 is a string representing the processing flow for detecting scenes in which people are present in video. The AI ​​flow formula ASA is created by arranging words extracted from the AI ​​flow AFA by applying natural language processing to it, according to predetermined rules. Specifically, the AI ​​flow formula ASA is generated by arranging words extracted from the AI ​​flow AFA according to the rule that the first argument of the name representing the processing of a block in the AI ​​flow AFA specifies the name representing the output destination of the processing of that block, and the parameters for executing the processing of that block. Furthermore, according to the AI ​​flow formula ASA, the parameter "CATE:person" is specified in the "FILTER" processing, while no parameters are set in the other processing steps.

[0050] Figure 7 is a diagram illustrating a specific example of an AI flow formula. The AI ​​flow formula ASB in Figure 7 is a string representing the processing flow for detecting scenes containing umbrellas from video footage. The AI ​​flow formula ASB is created by applying natural language processing to the AI ​​flow AFB and arranging the extracted words according to predetermined rules. Specifically, the AI ​​flow formula ASB is generated by arranging the words extracted from the AI ​​flow AFB according to the rule that the first argument of the name representing the processing of a block in the AI ​​flow AFB specifies the name representing the output destination of the processing of that block, and the parameters for executing the processing of that block. Furthermore, according to the AI ​​flow formula ASB, the parameter "CATE: umbrella" is specified in the "FILTER" processing, while no parameters are set in the other processing steps.

[0051] The registration processing unit 11 can convert the AI ​​flow contained in the registration information obtained from the provider-side terminal device 200 into an AI flow formula by using the method described in the above specific example. Furthermore, the flow generation unit 15 can generate an AI flow formula corresponding to the summary sentence contained in the query information obtained from the user-side terminal device 300 by applying the method described in the above specific example. Furthermore, by applying the method described in the above specific example, the AI ​​flow formula can be converted into an AI flow.

[0052] (Specific example of table data) Figure 8 is a diagram illustrating a specific example of table data. The table data TD in Figure 8 contains three data entries that record the association status of AI flow formulas and explanatory texts based on data obtained from the registration processing unit 11. Specifically, the table data TD includes three explanatory texts: "Detect scenes with people from video," "Detect scenes with umbrellas from video," and "Detect scenes with both people and umbrellas simultaneously from video," along with three AI flow formulas associated with each of these three explanatory texts. Note that the table data TD only needs to contain at least one data entry that records the association status of AI flow formulas and explanatory texts based on data obtained from the registration processing unit 11.

[0053] [Processing flow] Next, the processing flow performed in the server device according to the first embodiment will be described. Figure 9 is a flowchart showing an example of the processing performed in the server device according to the first embodiment.

[0054] First, the registration processing unit 11 converts the AI ​​flow included in the registration information obtained from the provider-side terminal device 200 into an AI flow formula, and registers the data relating the AI ​​flow formula and the explanatory text into the table data of the data storage unit 12 (step S11).

[0055] Next, when query information is input to the user terminal device 300, the feature extraction processing unit 13 extracts a feature vector corresponding to the summary sentence contained in the query information (step S12). In this processing flow, if no query information is input, the processing from step S12 onward will not be performed.

[0056] Next, when query information is input to the user terminal device 300, the feature extraction processing unit 13 extracts a feature vector corresponding to each of the explanatory sentences registered in the table data of the data storage unit 12 (step S13).

[0057] Next, the similar flow acquisition unit 14 calculates the distance between the feature vector obtained in step S12 and each feature vector obtained in step S13 (step S14).

[0058] Next, the similar flow acquisition unit 14, based on the distance calculation result in step S14, acquires one AI flow expression from among the AI ​​flow expressions registered in the table data of the data storage unit 12 that is associated with one explanatory sentence that represents the process most similar to the processing of the summary sentence (step S15).

[0059] Next, when query information is input to the user-side terminal device 300, the flow generation unit 15 generates an AI flow expression corresponding to the summary text included in the query information (step S16).

[0060] Next, the display information generation unit 16 generates display information to present at least one AI flow as a recommended flow to the user, based on the AI ​​flow formula acquired in step S15 and the AI ​​flow formula generated in step S16 (step S17). Then, the display information generation unit 16 transmits the display information generated in step S17 to the user-side terminal device 300.

[0061] As described above, according to this embodiment, at least one AI flow can be presented as a recommended flow to a user who has entered a summary text in the user-side terminal device 300. Furthermore, as described above, according to this embodiment, an AI flow corresponding to the summary text entered in the user-side terminal device 300 can be presented to the user as a recommended flow. Furthermore, as described above, according to this embodiment, an AI flow representing processing similar to the processing of the summary text entered in the user-side terminal device 300 can be presented to the user as a recommended flow.Therefore, according to this embodiment, a suitable processing flow including processing using AI can be presented to the user.

[0062] <Second Embodiment> Figure 10 is a block diagram showing the functional configuration of the information processing device according to the second embodiment.

[0063] The information processing device 500 according to this embodiment has the same hardware configuration as the server device 100. The information processing device 500 also includes a registration means 501, a flow acquisition means 502, and a display information generation means 503.

[0064] Figure 11 is a flowchart illustrating the processing performed in the information processing device according to the second embodiment.

[0065] The registration means 501 registers data in the database 600 that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with a descriptive text, which is a sentence describing the outline of the processing performed by the AI ​​flow (step S51).

[0066] The flow acquisition means 502 acquires a first AI flow associated with a description that represents the process most similar to the process for achieving the user's desired objective, based on the data registered in the database 600 (step S52).

[0067] The display information generation means 503 generates display information to present to the user, as a recommended flow, at least one of the following: a first AI flow and a second AI flow generated based on an outline statement which is a sentence describing the outline of the process to achieve the user's desired objective (step S53).

[0068] According to this embodiment, a suitable processing flow, including processing using AI, can be presented to the user.

[0069] Some or all of the above embodiments may also be described as follows, but are not limited to the following:

[0070] (Note 1) A registration means for registering data in a database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with a descriptive text, which is a sentence describing the outline of the processing performed by the AI ​​flow. A flow acquisition means that acquires a first AI flow associated with a descriptive sentence representing the process most similar to the process that achieves the user's desired objective, based on the data registered in the aforementioned database. Display information generation means for generating display information to present to the user, as a recommended flow, at least one of the following AI flows: the first AI flow and the second AI flow generated based on a summary statement which is a statement describing the outline of the process for achieving the user's desired objective; An information processing device having

[0071] (Note 2) The display information generation means is an information processing device for which the first AI flow and the second AI flow are presented as the recommended flow, as specified in Appendix 1.

[0072] (Note 3) The display information generation means is an information processing device for Appendix 1 that determines whether or not to present the first AI flow to the user as the recommended flow, based on the distance between the feature vector extracted from the first explanatory text and the feature vector extracted from the summary text.

[0073] (Note 4) The display information generation means is an information processing device for which, when it is determined that the processing of the first explanatory text and the processing of the summary text are dissimilar based on the distance, the display information is generated to present the second AI flow to the user as the recommended flow.

[0074] (Note 5) The information processing device described in Appendix 3 generates the display information for presenting the first AI flow and the second AI flow to the user as the recommended flow when it is determined, based on the distance, that the processing of the first explanatory text and the processing of the summary text are similar.

[0075] (Note 6) Data is registered in a database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with a descriptive text, which is a sentence describing the overview of the processing performed by the AI ​​flow. Based on the data registered in the aforementioned database, a first AI flow is obtained that is associated with a description of the process that most closely resembles the process for achieving the user's desired objective. An information processing method for generating display information to present to the user, as a recommended flow, at least one of the following AI flows: the first AI flow and the second AI flow generated based on a summary statement which is a statement describing the outline of the process for achieving the user's desired objective.

[0076] (Note 7) Data is registered in a database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with a descriptive text, which is a sentence describing the overview of the processing performed by the AI ​​flow. Based on the data registered in the aforementioned database, a first AI flow is obtained that is associated with a description of the process that most closely resembles the process for achieving the user's desired objective. A recording medium that records a program that causes a computer to execute a process to generate display information for presenting to the user, as a recommended flow, at least one of the following AI flows: the first AI flow and the second AI flow generated based on a summary statement which is a statement describing the outline of the process for achieving the user's desired objective.

[0077] Although the present disclosure has been described above with reference to embodiments, the present disclosure is not limited to the above embodiments and examples. Various modifications to the structure and details of the present disclosure are possible, as can be understood by those skilled in the art within the scope of the present disclosure. [Explanation of Symbols]

[0078] 11. Registration Processing Unit 12 Data storage unit 13 Feature Extraction Processing Unit 14 Similar Flow Acquisition Unit 15 Flow generation unit 16 Display information generation section 100 Server Devices

Claims

1. A registration means for registering data in a database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with an explanatory text, which is a sentence describing the outline of the processing performed by the AI ​​flow. A flow acquisition means that acquires a first AI flow associated with a descriptive sentence representing the process most similar to the process that achieves the user's desired objective, based on the data registered in the aforementioned database. Display information generation means for generating display information to present to the user, as a recommended flow, at least one of the following AI flows: the first AI flow and the second AI flow generated based on a summary statement which is a statement describing the outline of the process for achieving the user's desired objective; It has, The display information generation means is an information processing device that generates the display information for presenting the second AI flow to the user as the recommended flow when it is determined that the processing of the first description and the processing of the summary are dissimilar, based on the distance between the feature vector extracted from the first description and the feature vector extracted from the summary.

2. A registration means for registering data in a database that associates an AI flow representing a processing flow that includes processing using one or more AIs, and an explanatory text which is a sentence describing the outline of the processing executed by the AI ​​flow, A flow acquisition means that acquires a first AI flow associated with a descriptive sentence representing the process most similar to the process that achieves the user's desired objective, based on the data registered in the aforementioned database. Display information generation means for generating display information to present to the user, as a recommended flow, at least one of the following AI flows: the first AI flow and the second AI flow generated based on a summary statement which is a statement describing the outline of the process for achieving the user's desired objective; It has, The display information generation means is an information processing device that generates the display information for presenting the first AI flow and the second AI flow to the user as the recommended flow when it determines that the processing of the first explanation and the processing of the summary are similar, based on the distance between the feature vector extracted from the first explanation and the feature vector extracted from the summary.

3. A computer-based information processing method, Data is registered in a database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with a descriptive text, which is a sentence describing the outline of the processing performed by the AI ​​flow. Based on the data registered in the aforementioned database, a first AI flow is obtained that is associated with a description of the process that most closely resembles the process for achieving the user's desired objective. Display information is generated to present to the user, as a recommended flow, at least one of the following AI flows: the first AI flow and the second AI flow generated based on a summary statement which is a statement describing the outline of the process for achieving the user's desired objective. An information processing method for generating display information to present the second AI flow to the user as the recommended flow, when it is determined that the processing of the first explanatory text and the processing of the summary text are dissimilar, based on the distance between the feature vector extracted from the first explanatory text and the feature vector extracted from the summary text.

4. A computer-based information processing method, Data is registered in a database that associates an AI flow, which represents a processing flow that includes processing using one or more AIs, with a descriptive text, which is a sentence describing the outline of the processing performed by the AI ​​flow. Based on the data registered in the aforementioned database, a first AI flow is obtained that is associated with a description of the process that most closely resembles the process for achieving the user's desired objective. Information processing method for generating display information to present to the user at least one of the following AI flows as a recommended flow: the first AI flow and the second AI flow generated based on a summary statement which is a sentence describing the outline of the process for achieving the user's desired objective; and when it is determined that the processing of the first explanation statement and the processing of the summary statement are similar based on the distance between the feature vector extracted from the first explanation statement and the feature vector extracted from the summary statement, the method generates the display information to present to the user the first AI flow and the second AI flow as the recommended flow.

5. Data is registered in a database that associates an AI flow representing a processing flow that includes processing using one or more AIs, and an explanatory text which is a sentence describing the outline of the processing executed by the AI ​​flow. Based on the data registered in the aforementioned database, a first AI flow is obtained that is associated with a description of the process that most closely resembles the process for achieving the user's desired objective. Display information is generated to present to the user, as a recommended flow, at least one of the following AI flows: the first AI flow and the second AI flow generated based on a summary statement which is a statement describing the outline of the process for achieving the user's desired objective. A recording medium that records a program that causes a computer to execute a process to generate display information for presenting the second AI flow to the user as the recommended flow, when it is determined that the processing of the first explanatory text and the processing of the summary text are dissimilar, based on the distance between the feature vector extracted from the first explanatory text and the feature vector extracted from the summary text.

6. Data is registered in a database that associates an AI flow representing a processing flow that includes processing using one or more AIs, and an explanatory text which is a sentence describing the outline of the processing executed by the AI ​​flow. Based on the data registered in the aforementioned database, a first AI flow is obtained that is associated with a description of the process that most closely resembles the process for achieving the user's desired objective. Display information is generated to present to the user, as a recommended flow, at least one of the following AI flows: the first AI flow and the second AI flow generated based on a summary statement which is a statement describing the outline of the process for achieving the user's desired objective. A recording medium that records a program that causes a computer to execute a process to generate display information for presenting the first AI flow and the second AI flow to the user as the recommended flow, when it is determined that the processing of the first explanatory text and the processing of the summary text are similar, based on the distance between the feature vector extracted from the first explanatory text and the feature vector extracted from the summary text.

Images