Explanatory text output device, explanatory text output program, and explanatory text output method
The explanatory text output device addresses the challenge of organizing information by category, enhancing understanding through structured sentence generation based on hierarchical classifications.
Patent Information
- Application Number
- JP2020158901
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2020-09-23
- Publication Date
- 2024-11-13
- Estimated Expiration
- 2040-09-23
Smart Images

Figure 0007581728000001 
Figure 0007581728000002 
Figure 0007581728000003
Abstract
Description
[Technical field]
[0001] The present invention relates to an explanatory text output device that outputs explanatory text about an object. [Background technology]
[0002] An explanatory text generation device has been proposed that uses a knowledge base to generate explanatory text about things. For example, Patent Document 1 discloses an explanatory text generation device that searches for concepts that have sensory features that are present in the concept to be explained but not included in similar concepts, and generates explanatory text by combining the features of the searched concepts. [Prior art documents] [Patent documents]
[0003] [Patent Document 1] Japanese Patent Application Publication No. 8-305708 Summary of the Invention [Problem to be solved by the invention]
[0004] However, depending on the object being explained, it may be necessary to include information related to the object in the explanatory text. For example, if the object being explained is a work of art, it may be necessary to explain not only the work of art, but also the artist of the work of art and the location where the work of art is stored. In this case, the generated explanatory text must be organized by category, such as the work of art itself, the artist of the work of art, and the location where the work of art is stored, otherwise the content of the explanatory text may be difficult to understand. This is particularly noticeable when listening to the explanatory text by audio.
[0005] The present invention has been made in view of the above circumstances, and has an object to provide an explanatory text output device and the like capable of generating explanatory text in which information about things is organized by category. [Means for solving the problem]
[0006] An explanatory text output device according to one aspect of the present application is characterized in that it includes an acquisition unit that acquires a plurality of pieces of related information for a plurality of objects to be extracted from a memory unit that stores a plurality of pieces of related information labeled with attributes related to the plurality of objects to be extracted and a hierarchical classification, a sentence generation unit that generates a plurality of sentences using the labeled attributes from the acquired plurality of pieces of related information, an explanatory text generation unit that combines the generated plurality of sentences based on the classification to generate an explanatory text, and an output unit that outputs the generated explanatory text. Effect of the Invention
[0007] According to one aspect of the present application, it is possible to generate explanatory text that allows viewers to easily understand the content. [Brief description of the drawings]
[0008] [Figure 1] FIG. 1 is an explanatory diagram illustrating an example of the configuration of an explanation providing system. [Diagram 2] FIG. 2 is a block diagram illustrating an example of a hardware configuration of an output device. [Diagram 3] FIG. 2 is a block diagram showing an example of a hardware configuration of a user terminal. [Figure 4] FIG. 1 is an explanatory diagram showing the structure of an artwork DB in the form of a directed graph. [Diagram 5] FIG. 13 is an explanatory diagram showing an example of an artwork DB. [Figure 6] FIG. 13 is an explanatory diagram showing another example of an artwork DB. [Figure 7] 13 is a flowchart illustrating an example of a procedure for generating and outputting a file. [Figure 8] FIG. 11 is an explanatory diagram showing an example of a priority DB. [Figure 9] 13 is a flowchart illustrating another example of a procedure of the generation and output process. [Figure 10] FIG. 11 is an explanatory diagram showing a process of generating an explanatory sentence. [Figure 11] FIG. 13 is an explanatory diagram showing another example of the priority DB. [Figure 12] 13 is a flowchart illustrating another example of a procedure of the generation and output process. [Figure 13]FIG. 11 is an explanatory diagram showing a process of generating an explanatory sentence. [Figure 14] FIG. 2 is a block diagram showing an example of a functional unit included in the output device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0009] (Embodiment 1) Hereinafter, an embodiment will be described with reference to the drawings. FIG. 1 is an explanatory diagram showing an example of the configuration of an explanatory text providing system. The explanatory text providing system 100 includes an output device (explanatory text output device) 1 and a user terminal 2. The output device 1 and the user terminal 2 are connected to each other via a network N so as to be able to communicate with each other. The output device 1 is configured with a server computer, a PC (Personal Computer), or the like. The output device 1 may also be configured with a multicomputer consisting of multiple computers, a virtual machine virtually constructed by software, or a quantum computer. Furthermore, the functions of the output device 1 may be realized by a cloud service. The user terminal 2 is configured with a notebook computer, a panel computer, a tablet computer, a smartphone, or the like.
[0010] 2 is a block diagram showing an example of a hardware configuration of an output device 1. The output device 1 includes a control unit 11, a main memory unit 12, an auxiliary memory unit 13, a communication unit 14, and a reading unit 15. The control unit 11, the main memory unit 12, the auxiliary memory unit 13, the communication unit 14, and the reading unit 15 are connected by a bus B.
[0011] The control unit 11 has one or more arithmetic processing devices such as a central processing unit (CPU), a micro-processing unit (MPU), a graphics processing unit (GPU), etc. The control unit 11 reads out and executes a control program 1P (explanation output program) stored in the auxiliary storage unit 13, thereby performing various information processing, control processing, etc. related to the output device 1, and realizes functional units such as an acquisition unit, a sentence generation unit, an explanation generation unit, and an output unit.
[0012] The main memory unit 12 is a static random access memory (SRAM), a dynamic random access memory (DRAM), a flash memory, etc. The main memory unit 12 mainly temporarily stores data necessary for the control unit 11 to execute arithmetic processing.
[0013] The auxiliary storage unit 13 is a hard disk or a solid state drive (SSD) or the like, and stores a control program 1P and various DBs (databases) necessary for the control unit 11 to execute processing. The auxiliary storage unit 13 stores an artwork DB 131, a priority DB 132, and templates TP. The auxiliary storage unit 13 may be an external storage device connected to the output device 1. The various DBs and the like stored in the auxiliary storage unit 13 may be stored in a database server or cloud storage different from the output device 1.
[0014] The communication unit 14 communicates with the user terminal 2 via the network N. In addition, the control unit 11 may use the communication unit 14 to download the control program 1P from another computer via the network N or the like, and store it in the auxiliary storage unit 13.
[0015] The reading unit 15 reads the portable storage medium 1a including a CD (Compact Disc)-ROM and a DVD (Digital Versatile Disc)-ROM. The control unit 11 may read the control program 1P from the portable storage medium 1a via the reading unit 15 and store it in the auxiliary storage unit 13. The control unit 11 may also read the control program 1P from the semiconductor memory 1b.
[0016] 3 is a block diagram showing an example of the hardware configuration of a user terminal 2. The user terminal 2 includes a control unit 21, a main memory unit 22, an auxiliary memory unit 23, a communication unit 24, an input unit 25, and a display unit 26. Each component is connected by a bus B.
[0017] The control unit 21 controls each part of the hardware according to a control program 2P stored in the auxiliary storage unit 23. The control unit 21 has one or more arithmetic processing devices such as a CPU, an MPU, a GPU, etc. The control unit 21 provides various functions by reading and executing the control program 2P stored in the auxiliary storage unit 23.
[0018] The main memory unit 22 is an SRAM, a DRAM, a flash memory, etc. The main memory unit 22 mainly temporarily stores data necessary for the control unit 21 to execute arithmetic processing.
[0019] The auxiliary storage unit 23 is a hard disk or an SSD, etc., and stores various data necessary for the control unit 21 to execute processing. The auxiliary storage unit 23 may be an external storage device connected to the user terminal 2. Each piece of data, etc. stored in the auxiliary storage unit 23 may be stored in a database server or cloud storage.
[0020] The communication unit 24 communicates with the output device 1 via the network N. In addition, the control unit 21 may use the communication unit 24 to download the control program 2P from another computer via the network N or the like, and store it in the auxiliary storage unit 23.
[0021] The input unit 25 is a keyboard and a mouse. The display unit 26 includes a liquid crystal display panel or the like. The display unit 26 displays the explanatory text and the like output by the output device 1. The display unit 26 may also be a touch panel display integrated with the input unit 25. The user terminal 2 may perform display on an external display device.
[0022] Furthermore, the output device 1 or the user terminal 2 may include a voice synthesis unit that converts text into voice. The user terminal 2 acquires the explanatory text from the output device 1 and converts the acquired explanatory text into voice. Alternatively, the output device 1 converts the explanatory text into voice and transmits it to the user terminal 2. The user terminal 2 outputs the explanatory text as voice.
[0023] FIG. 4 is an explanatory diagram showing the structure of an artwork DB in the form of a directed graph. Node N1 in the center of the diagram is one of the nodes, and is an image showing an artwork. The end node of an edge indicates related information of the start node. The edge is labeled with an attribute related to the start node and a hierarchical classification. For example, edge E1 from node N1 to node N2 is labeled with the artwork name, which is an attribute, and (1 → 2a) as a hierarchical classification. This indicates that the artwork name of node N1 is node N2 (Mona Lisa). The same applies to the other nodes and edges. The function of the hierarchical classification will be described later.
[0024] FIG. 5 is an explanatory diagram showing an example of an art DB. FIG. 5 shows an example in which the art DB 131 is a card-type database. A record is created for each art piece in the art DB 131. Each record is in the form of a table. Each table includes a label column, a classification column, and a data column. The label column stores attributes. The classification column indicates a hierarchical classification including the connection relationships between nodes. The data column stores related information. In FIG. 5, the artist of the work titled "Snow Scene with a Chinese Family" is unknown. Therefore, the data for "Artist," "Country of Birth," and "Year of Birth" are missing. "Place of Creation" is also unknown and missing.
[0025] Fig. 6 is an explanatory diagram showing another example of an art DB. Fig. 6 shows a schema when the art DB 131 is implemented in a relational database. Here, the art DB 131 is represented by three tables.
[0026] FIG. 6A shows an example of a work table. The work table 1311 stores information about works (artworks). The work table 1311 includes a work ID column, a work name column, an artist ID column, a production year column, and a collection museum ID column. The work ID column stores a work ID that can uniquely identify the work. The work name column stores the name of the work. The artist ID column stores an artist ID that indicates the artist of the work. The artist ID is a foreign key and is the primary key of artist table 1312 described below. The production year column stores the year the work was created. If the work was created over multiple years, the production year column stores the year production began and ended. The collection museum ID column stores a museum ID that indicates the museum that owns the work. The museum ID is a foreign key and is the primary key of museum table 1313 described below.
[0027] FIG. 6B shows an example of an author table. The author table 1312 stores information about the author of a work. The author table 1312 includes an author ID column, a name column, a country of origin column, a year of birth column, and a year of death column. The author ID column stores an author ID that can uniquely identify the author. The name column stores the name of the author. The country of origin column stores the country of origin of the author. The year of birth column stores the year of birth of the author. The year of death column stores the year of death of the author. If the author is still alive, the year of death column stores a blank or no value (NULL).
[0028] FIG. 6C shows an example of a museum table. Museum table 1313 stores information about museums that hold artworks. Museums here are not limited to facilities that have "museum" in their name. They also include art galleries, resource centers, memorial halls, etc. that hold artworks. Museum table 1313 includes a museum ID column, a museum name column, a location column, and a visitor column. The museum ID column stores a museum ID that can uniquely identify a museum. The museum name column stores the name of the museum. The location column stores the location of the museum. In FIG. 6C, the country of location is stored. The visitor column stores the number of visitors per year.
[0029] Next, the process performed by the explanatory text providing system 100 will be described. FIG. 7 is a flow chart showing an example of the procedure of the generating and outputting process. The control unit 11 of the output device 1 acquires the identification information of the work for which the explanatory text is to be generated, for example, the work ID, from the user terminal 2 or the like (step S1). The control unit 11 searches the artwork DB 131 using the identification information as a key to acquire the work information (step S2). The control unit 11 generates a simple sentence for each label included in the work information using the acquired work information and the template TP (step S3). Based on the classification of the labels, the control unit 11 rearranges the order of the generated simple sentences so that they are grouped according to the classification (step S4). The control unit 11 combines the simple sentences in the group (step S5). The control unit 11 outputs the combined explanatory text obtained (step S6). The control unit 11 transmits the explanatory text to the user terminal 2. The control unit 11 may store the explanatory text in the auxiliary storage unit 13, and transmit the stored explanatory text when an inquiry about the same work ID is received from the user terminal 2.
[0030] Next, an example of the generation and output process is shown. We will explain the case of generating a description of the Mona Lisa shown in Figure 5. Using the template TP from the artwork information, the following sentence is obtained. The character string in parentheses after the sentence indicates the classification on which the sentence is based. The classification is also shown in Figure 4.
[0031] "This work is the Mona Lisa" (1→2a) "The artist is Leonardo da Vinci" (1→2b) "I'm from Italy" (2b → 3a) "I was born in 1452" (2b→3b) "The production date is between 1503 and 1519" (1→2c) "It was made in Italy" (1→2d) "It is owned by the Louvre Museum" (1→2e) "The location is France" (2e → 3a) "The number of visitors is 7.4 million." (2e → 3b)
[0032] Next, the order is rearranged based on the classification. The order is rearranged so that each classification is grouped together. Some strings indicating classification contain an arrow. In such classifications, the classification at the base of the arrow is called the preceding classification, and the classification at the tip of the arrow is called the succeeding classification. The grouping is done as follows. If the string indicating the classification does not contain an arrow, grouping is done using the string indicating the classification. In the above case, the string is "1". If the string indicating the classification contains an arrow, grouping is done using the preceding classification. In the above case, the strings are "1", "2b", and "2e". If the succeeding classification is also the preceding classification, the succeeding classification is used. In the above case, classification "1→2b" is grouped using the string "2b". From the above algorithm, grouping is done using the strings "1", "2b", and "2e". The classifications grouped using each string are hereinafter referred to as classification 1, classification 2b, and classification 2e. First, grouping using classification 1 is as follows.
[0033] "This work is the Mona Lisa" (1→2a) "The production date is between 1503 and 1519" (1→2c) "It was made in Italy" (1→2d)
[0034] When grouped under category 2b, it looks like this: As mentioned above, when grouping "1 → 2b," the grouping is not grouped under category "1" but rather grouped under category "2b."
[0035] "The artist is Leonardo da Vinci" (1→2b) "I'm from Italy" (2b → 3a) "I was born in 1452" (2b→3b)
[0036] When classified as category 2e, it looks like this: As mentioned above, when classifying "1 → 2e", the classification is "2e" rather than "1".
[0037] "It is owned by the Louvre Museum" (1→2e) "The location is France" (2e → 3a) "The number of visitors is 7.4 million." (2e → 3b)
[0038] Next, combine the statements in the group to get the following:
[0039] "This work is the Mona Lisa, created between 1503 and 1519 in Italy." (Category 1) "The artist was Leonardo da Vinci, he was born in Italy in 1452." (Category 2b) "It is housed at the Louvre, is located in France, and has 7.4 million visitors" (category 2e)
[0040] The above three sentences are output as the explanation. This embodiment has the following effects. By rearranging the order of the generated sentences using classification, it is possible to generate an explanation with high readability. In addition, it is possible to flexibly respond to cases where there are labels that do not have data or labels that cannot be included in the explanation due to a character limit or the like.
[0041] (Embodiment 2) This embodiment relates to a mode in which importance is assigned to each label in the artwork DB 131, and a label to be included in the description is selected based on the importance. Here, the importance is determined based on how much data of that label is contained in data with the same structure. The ranking of labels based on importance is called priority. Note that the system configuration and hardware configuration are the same as those in the first embodiment, and therefore description thereof will be omitted.
[0042] FIG. 8 is an explanatory diagram showing an example of a priority DB. Priority DB 132 stores a priority based on importance for each label. Priority DB 132 includes a label column, a classification column, a data count column, a normalization column, and a priority column. The label column stores a label. The classification column stores a label classification. The data count column stores the number of data. The number of data corresponds to importance. The normalization column stores a normalized value of the number of data for a label. The normalization column stores a value obtained by dividing the number of data for a label by the number of data for the label that always or has the most data, for example, the number of data for a work. The normalized value may be used as importance. The priority column stores a ranking when labels are ranked by importance.
[0043] FIG. 9 is a flowchart showing another example of the procedure of the generation and output process. The control unit 11 of the output device 1 acquires identification information, for example, the work ID, of the work for which the description is to be generated from the user terminal 2 or the like (step S21). The control unit 11 searches the artwork DB 131 using the identification information as a key to acquire work information (step S22). The control unit 11 acquires the priority from the priority DB 132 (step S23). It is sufficient to acquire only the label having data. The control unit 11 assigns 1 to the loop variable n (step S24). The control unit 11 acquires the data of the label of the priority n (step S25). The control unit 11 generates a simple sentence using the template TP, and stores the generated sentence in a temporary storage area in association with the classification (step S26). The temporary storage area is provided in the main storage unit 12 or the auxiliary storage unit 13. The control unit 11 increments the loop variable n by 1 (step S27). The control unit 11 determines whether the loop variable n is greater than a threshold k (step S28). The value of k is set in advance. If k is 5, simple sentences for labels with priorities 1 to 5 are generated. If the control unit 11 determines that the loop variable n is less than the threshold k (NO in step S28), the process returns to step S25. If the control unit 11 determines that the loop variable n is greater than the threshold k (YES in step S28), the control unit 11 rearranges the order of the simple sentences stored in the temporary storage area so that they are grouped by label classification (step S29). The control unit 11 combines the rearranged simple sentences by classification (step S30). The control unit 11 outputs the explanation obtained by combining the simple sentences (step S31).
[0044] FIG. 10 is an explanatory diagram showing the process of generating a description. As above, this is an example of generating a description of the Mona Lisa. The threshold k is 5. From the priority DB 132 in FIG. 8, the label with priority 1 is the title of the work. The label with priority 2 is the artist. The label with priority 3 is the year of production. The label with priority 4 is the country of origin. The label with priority 5 is the year of birth. The control unit 11 generates a single sentence for each label. The generated results are the five sentences shown on the left side of FIG. 10. The control unit 11 rearranges the order so that the labels are grouped together according to classification. The control unit 11 moves the third sentence to the second sentence. This results in the sentence order shown in the center of FIG. 10. The control unit 11 combines the sentences for each group. As a result, the description shown on the right side of FIG. 10 is generated. The control unit 11 outputs the description.
[0045] In this embodiment, a simple sentence is generated for a label with a high priority based on importance, so that a compact and substantial explanatory sentence can be generated.
[0046] (Embodiment 3) This embodiment relates to a form in which the importance of each label is weighted. Also, it relates to a form in which an upper limit is set on the number of characters in the description. Note that the system configuration and hardware configuration are the same as those in the first embodiment, and therefore the description will be omitted.
[0047] FIG. 11 is an explanatory diagram showing another example of the priority DB. The priority DB 132 is shown in FIG. 11B. FIG. 11A shows the priority DB being created. In FIG. 11A, the label column, classification column, and normalization column are the same as those in FIG. 8. The distance column shows the distance (degree of relevance) of each label from the work. The distance can be found from the classification column. For example, the classification of the work is 1, and the classification of the work name is "1→2a", so the distance is 1. Similarly, the distance of the author is also 1. The classification of the country of origin is "2b→3a", while the classification of the author is "1→2b", so the relationship between the work and the country of origin can be expressed as work→author→country of origin. Therefore, the distance of the country of origin is 2. When the distance is found in the same way, it is as shown in FIG. 11A. The normalized importance value is weighted by the inverse of the distance. In other words, the value obtained by dividing the normalized importance value by the distance is set as the importance in this embodiment. The weighted value is the value of the importance column. Then, the importance is ranked. The order of the labels represents the priority. The table shown in Fig. 11B is an example of the priority DB 132 according to the present embodiment.
[0048] Next, the generation of the explanatory text will be described. FIG. 12 is a flowchart showing another example of the procedure of the generation and output process. The control unit 11 of the output device 1 acquires the identification information of the artwork for which the explanatory text is to be generated, such as the artwork ID and the upper limit number (upper limit number of characters) from the user terminal 2 or the like (step S41). The control unit 11 searches the artwork DB 131 using the identification information as a key to acquire the artwork information (step S42). The control unit 11 assigns 1 to the loop variable n (step S43). The control unit 11 acquires the data of the label with priority n (step S44). The control unit 11 acquires the data of the label with priority n by referring to the priority DB 132, and acquires the data of the acquired label from the artwork information. The control unit 11 generates a simple sentence using the template TP, and stores the generated sentence in a temporary storage area in association with the classification (step S45). The control unit 11 determines whether the total number of characters of the simple sentences already created is equal to or greater than the upper limit number (step S46). If the control unit 11 determines that the total number of characters in the already created simple sentences is less than the upper limit (NO in step S46), the control unit 11 increments the loop variable n by 1 (step S47) and returns the process to step S44. If the control unit 11 determines that the total number of characters in the already created simple sentences is equal to or greater than the upper limit (YES in step S46), the control unit 11 rearranges the order of the already created simple sentences so that they are grouped by category (step S48). The control unit 11 combines the rearranged sentences by category (step S49). The control unit 11 outputs the combined sentences (step S50). Note that if the number of characters in the explanation sentence exceeds the upper limit as a result of combining the simple sentences, the simple sentence based on the label with the lowest priority may be removed and steps S48 and after may be executed again.
[0049] FIG. 13 is an explanatory diagram showing the process of generating a description. As above, this is an example of generating a description of the Mona Lisa. The upper limit of the number of characters is set to 75. The control unit 11 references the priority DB 132 and generates a simple sentence for the label "Title of Artwork" with priority 1. Next, simple sentences are generated in the order of the labels "Artist", "Year of Production", "Collection", "Country of Origin", and "Year of Birth". The generated simple sentences are as shown on the left of FIG. 13.
[0050] The values in parentheses next to each sentence indicate the classification and number of characters. Here, the total number of characters up to "Country of Origin" is 69 characters. If the simple sentence about "Year of Birth" is added, the total becomes 79 characters, which exceeds the upper limit of characters, so the simple sentence about "Year of Birth" is discarded. Next, the control unit 11 rearranges the order according to the label classification. The result is shown in the center of Figure 13. Next, the control unit 11 combines the simple sentences for each classification. The result is the sentence to be output as the explanation, and is shown on the right side of Figure 13.
[0051] In this embodiment, since the importance of the labels is weighted, it is possible to generate a description in which the items more closely related to the artwork appear earlier in the order of appearance, making it possible to generate a description with sufficient content even when generating a description within the upper limit of the number of characters.
[0052] 14 is a block diagram showing an example of the functional units of the output device. The output device 1 includes, as its functional units, an acquisition unit 11a, a sentence generation unit 11b, an explanatory sentence generation unit 11c, and an output unit 11d. Each of these functional units is realized by the control unit 11 operating based on a control program 1P.
[0053] The acquisition unit 11a acquires a plurality of pieces of related information of the object to be extracted from a storage unit that stores a plurality of pieces of related information labeled with attributes related to the objects to be extracted and hierarchical classifications. The sentence generation unit 11b generates a plurality of sentences using the labeled attributes from the acquired plurality of pieces of related information. The explanatory sentence generation unit 11c combines the generated plurality of sentences based on the classification to generate an explanatory sentence. The output unit 11d outputs the generated explanatory sentence. Note that the explanatory sentence generation unit 11c may combine the generated sentences in an order based on the priority.
[0054] In addition, although not shown in FIG. 14, the output device 1 may include, as functional units, a priority acquisition unit that acquires a priority for each label, a selection unit that selects multiple simple sentences based on the acquired priority, and a determination unit that determines whether the number of characters in the explanatory sentence exceeds a predetermined value.
[0055] In the above embodiment, a simple sentence is generated from each label and its data, but this is not limited to the above. Before generating sentences, multiple labels may be grouped together, and one compound sentence may be generated for each group. In the above example, the labels "Artwork", "Year of Creation", and "Place of Creation" are grouped together, so a compound sentence such as "This work is the Mona Lisa, it was created between 1503 and 1519, and it was created in Italy" is generated without creating a simple sentence.
[0056] In addition, compound sentences may be generated by modifying the template TP. For example, a template TP that reads "This work was created by <author data>, who was born in <country of origin data> on <year of birth data>" can be prepared. This template TP makes it possible to generate compound sentences such as "This work was created by Leonardo da Vinci, who was born in Italy in 1452."
[0057] In the above-described embodiment, when specifying a work, the user may input the name of the work by voice. The output device 1 acquires identification information of the specified work by voice recognition. The user may also specify a work by a chatbot. The output device 1 acquires identification information of the work from a message of the chatbot. The output device 1 replies with the generated explanatory text by the chatbot.
[0058] In the above embodiment, a work is specified and an explanatory text about the work is generated, but this is not limited to the above. An explanatory text about the artist may also be generated by specifying an artist. To handle such a case, an artist DB is constructed and used by reconstructing the art DB 131 shown in Fig. 5 from the viewpoint of the artist. If the art DB 131 is implemented in a relational database as shown in Fig. 6, there is no need to reconstruct the database.
[0059] The generation of explanatory text is not limited to artworks. It can be applied to general products. In this case, the artwork DB 131 is used as a product DB, and each label and its data are changed to be suitable for the product. The assumed use is online shops, virtual shopping malls, etc.
[0060] The technical features (constituent elements) described in each embodiment can be combined with each other, and by combining them, new technical features can be formed. The embodiments disclosed herein are illustrative in all respects and should not be considered as limiting. The scope of the present invention is defined by the claims, not by the above meaning, and is intended to include all modifications within the scope and meaning equivalent to the claims. [Explanation of symbols]
[0061] 100 Description provision system 1 Output device 11 Control section 11a Acquisition part 11b Sentence generator 11c Description Generation Section 11d Output section 12 Main memory 13 Auxiliary storage 131 Art DB 1311 Works Table 1312 Author Table 1313 Museum Table 132 Priority DB TP Template 14 Communications Department 15 Reading unit 1P control program 1a Removable storage media 1b Semiconductor memory 2. User terminal B Bus
Claims
1. an acquisition unit that acquires a plurality of pieces of related information corresponding to a plurality of objects to be extracted from a storage unit that stores a plurality of pieces of related information in which labels indicating attributes related to the plurality of objects to be extracted are associated with hierarchical classifications; a priority acquisition unit that acquires a priority for each label from a storage unit that stores the priority assigned to each label; a sentence generator that generates a plurality of sentences using attributes indicated by labels included in each of the acquired plurality of pieces of related information; an explanatory sentence generation unit that divides the generated sentences into groups based on the classification, combines sentences in each group in an order based on the priority of labels indicating attributes used in generating each sentence, and generates an explanatory sentence; an output unit for outputting the generated explanatory text; An explanatory text output device comprising:
2. a selection unit that selects a plurality of the sentences based on the priority acquired by the priority acquisition unit; Equipped with The explanatory sentence generating unit generates the explanatory sentence from the selected sentence.
2. The explanatory text output device according to claim 1,
3. The priority is determined based on the importance of each label.
3. The explanatory text output device according to claim 2,
4. The importance is determined based on the number of data items for each of the labels stored in the storage unit.
4. The explanatory text output device according to claim 3,
5. The importance is weighted based on the degree of association between the number of data items and the object and the label.
5. The explanatory text output device according to claim 4.
6. A determination unit is provided for determining whether or not the number of characters exceeds a predetermined value, The selection unit selects the sentences one by one in order of decreasing priority; Each time the selection unit selects a sentence, the determination unit determines whether or not the total number of characters in the sentence exceeds the predetermined value, and when it determines that the total number of characters does not exceed the predetermined value, the determination unit continues the selection operation of the selection unit, and when it determines that the total number of characters exceeds the predetermined value, the determination unit discards the sentence last selected by the selection unit.
6. The explanatory text output device according to claim 2, wherein the explanatory text output device further comprises a display unit.
7. A computer that can access a storage unit that stores a plurality of pieces of association information in which labels indicating attributes related to a plurality of objects to be extracted are associated with hierarchical classifications, acquiring a plurality of pieces of related information corresponding to a plurality of the extraction targets; obtaining a priority for each of the labels from a storage unit that stores the priority assigned to each of the labels; generating a plurality of sentences using attributes indicated by labels included in each of the plurality of pieces of related information obtained; classifying the generated sentences into groups based on the classification, combining the sentences in each group in an order based on the priority of the labels indicating the attributes used in generating each sentence, and generating an explanatory sentence; Output the generated explanation An explanatory text output program that causes a process to be executed.
8. A computer that can access a storage unit that stores a plurality of pieces of association information in which labels indicating attributes related to a plurality of objects to be extracted are associated with hierarchical classifications, acquiring a plurality of pieces of related information corresponding to a plurality of the extraction targets; obtaining a priority for each of the labels from a storage unit that stores the priority assigned to each of the labels; generating a plurality of sentences using attributes indicated by labels included in each of the plurality of pieces of related information obtained; classifying the generated sentences into groups based on the classification, combining the sentences in each group in an order based on the priority of the labels indicating the attributes used in generating each sentence, and generating an explanatory sentence; Output the generated explanation 23. A method for outputting an explanatory text, comprising:
Citation Information
Patent Citations
Method and device for generating explanatory document
JP1996305708A
Step description device, program and recording medium recording the same
JP2012248126A
Constructing a Narrative Based on a Collection of Images
US20180357211A1
