Product selection device and program

Abstract

To provide a product selection device etc., capable of selecting a proper robot product on a vague demand from a user.SOLUTION: A product selection device 1 comprises: a storage part 2 which stores a product performance word conversion DB 5 including a product performance list 3 including performance information on each robot product and a conversion rule for conversion into product performance words 52 representing performance of robot products; an information output part 6 which outputs information to a product selection UI; an input reception part 7 which receives an input document input as a user's demand from the product selection UI; and a product selection part 8 which extracts a word from the input document and converts it into a product performance word 52 by the use of the product performance word conversion DB 5, and then refers to the product performance list 4 to select a robot product having performance matching the product performance word 52 or the document including the product performance word 52.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a product selection device for selecting robot products and the like.

Background Art

[0002] Industrial robots have a wide variety of products, and since catalogs and the like contain technical terms, it is difficult for users to select products by themselves. For example, Patent Document 1 discloses a technique in which an interactive agent receives information from a user in an interactive manner and provides information on medical products. If such a service using this technique can be provided on the Internet, in smartphone apps, etc., anyone can easily consider purchasing a robot product.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, when selecting a product using an interactive agent as in Patent Document 1, an issue is that the number of interactions for question-and-answer exchanges increases. In particular, when a user inputs ambiguous requests, not only does the number of interactions increase, but in some cases, an appropriate product cannot be selected.

[0005] The present invention has been made in view of the above-described problems, and an object thereof is to provide a product selection device and the like that can select an appropriate robot product without increasing the burden on the user in response to ambiguous requests from the user.

Means for Solving the Problems

[0006] A first invention for achieving the above-described object is a product selection device comprising: a storage unit that stores a product performance list including performance information for each robot product and a database including conversion rules from one or more words to product performance words expressing the performance of the robot product; an information output unit that outputs information to a product selection user interface; an input reception unit that receives an input sentence input as a user's request from the product selection user interface; a product selection unit that extracts words from the input sentence, converts them into the product performance words using the database, and selects the robot product having performance that matches the product performance words or a sentence including the product performance words with reference to the product performance list.

[0007] The product selection unit may transmit a prompt including the input sentence, the product performance words, and an instruction to generate a paraphrased sentence of the input sentence using the product performance words to an external generative AI server, receive an answer to the prompt from the generative AI server, and the information output unit may output the answer to the prompt as a paraphrased sentence of the input sentence to the product selection user interface.

[0008] Further, the storage unit stores a product list including at least one of the portable weight and the movable range of the robot product, the information output unit outputs a selection object of at least one of the portable weight and the movable range to the product selection user interface, the input reception unit receives a selection content by the selection object, and the product selection unit may select the robot product having at least one of the portable weight and the movable range that matches the selection content and having performance that matches the product performance words or a sentence including the product performance words.

[0009] The second invention is a program for causing a computer to function as a product selection device including a storage unit that stores a product performance list including performance information for each robot product and a database including conversion rules from one or more words to product performance words that express the performance of the robot product, an information output unit that outputs information to a product selection user interface, an input reception unit that receives an input sentence input as a user's request from the product selection user interface, extracts words from the input sentence, converts them into the product performance words using the database, and selects the robot product having performance that matches the product performance words or a sentence including the product performance words with reference to the product performance list.

Advantages of the Invention

[0010] According to the present invention, it is possible to provide a product selection device or the like that can select an appropriate robot product without increasing the burden on the user for ambiguous requests from the user.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Embodiments for Carrying Out the Invention

[0012] Based on the following drawings, this embodiment will be described in detail. FIG. 1 is a block diagram showing an example of the hardware of a computer constituting the product selection device of the present invention. As shown in FIG. 1, the product selection device 1 is a computer having hardware in which a CPU (Central Processing Unit) 11, a memory 12, an auxiliary storage device 13, an input / output device 14, etc. are connected via a bus 15.

[0013] The CPU 11 executes the processes described later by reading out the programs stored in the auxiliary storage device 13 in advance into the memory 12 and executing them. The auxiliary storage device 13 is a hard disk drive (HDD) or a solid state drive (SSD: Solid State Drive), etc. The input / output device 14 is an input device such as a mouse, keyboard, microphone, an output device such as a display, speaker, and a communication device that performs data transmission and reception with an external computer, etc. The input / output device 14 performs data input and output under the control of the CPU 11. The connection form for data input and output may be wireless or wired.

[0014] The computer constituting the product selection device 1 is a server installed in a data center, cloud, etc., or a user terminal such as a PC (Personal Computer), smartphone, tablet terminal, etc. The product selection device 1 may be configured by one unit or a plurality of units.

[0015] FIG. 2 is a block diagram showing the configuration of the product selection device of FIG. 1. The product selection device 1 is a device that selects robot products and provides services to users who are considering purchasing robot products. As shown in FIG. 2, the product selection device 1 includes a storage unit 2, an information output unit 6, an input reception unit 7, a product selection unit 8, etc. by the cooperation of software and hardware resources.

[0016] When realizing the service according to the present invention as a web application, the product selection device 1 is a web server. In this case, a program for causing the web server to function as the product selection device 1 is installed. The user accesses the web server via the Internet using a web browser on the user terminal. Also, when realizing the service according to the present invention as a stand-alone type user terminal application, the product selection device 1 is the user terminal. In this case, a program for causing the user terminal to function as the product selection device 1 is installed. Further, when realizing the service according to the present invention as a client-server type user terminal application, the product selection device 1 is a computer system composed of a web server and a user terminal. In this case, a program related to a part of the functions of the product selection device 1 is installed on the web server, and a program related to other functions is installed on the user terminal.

[0017] The storage unit 2 stores a product list 3 including at least one of the portable weight and the movable range of the robot product. The product list 3 includes data items such as a product name 31, a portable weight 32, and a movable range 33. The product name 31 is the name of the robot product. The portable weight 32 is the weight of the work that can be transported. The movable range 33 is the range (= maximum reach length) in which the arm can be moved. The product list 3 may include an image of the robot product, a product description article, and the like.

[0018] In addition, the memory unit 2 stores a product performance list 4 that includes performance information for each robot product. The product performance list 4 includes a product name 41 and data items related to the performance information of the robot product. The product name 41 is the same as the product name 31. Spot welding 42, substrate transfer 43, operating speed 44, collaborative robot 45, and clean room 46 are the performance information of the robot product. Spot welding 42 indicates whether it is a robot for spot welding. Substrate transfer 43 indicates whether it is a robot for substrate transfer. The operating speed 44 qualitatively indicates the speed of the operating speed. In the example shown in FIG. 2, the operating speed 44 is fast, normal, slow, etc. The collaborative robot 45 indicates whether it is a collaborative robot. The clean room 46 indicates whether it is a robot for the clean room. Note that the performance information of the robot product is not limited to these.

[0019] In addition, the memory unit 2 stores a product performance word conversion database 5 (hereinafter referred to as "product performance word conversion DB 5") that includes a conversion rule from one or more extracted words 51 to product performance words 52 that represent the performance of the robot product. The correspondence between the extracted word 51 and the product performance word 52 is one-to-one or many-to-one. The extracted word 51 assumes a word included in a sentence indicating the user's request. The product performance word 52 is a word that represents the performance of the robot product and is associated with the performance information in the product performance list 4. For example, the product performance word 52 is the item name of the performance information (e.g., "spot welding"), a word that includes the item name (e.g., "welding" that includes "spot welding"), a related word (e.g., "fast" for "operating speed"), etc. In the example shown in FIG. 2, a conversion rule in which two extracted words 51 of "substrate" and "carry" are converted into a product performance word 52 of "substrate transfer" is illustrated. The product performance word conversion DB 5 may be constructed by pre-learning and the conversion rule may be added at any time. The extracted word 51 is not limited to Japanese and may be a foreign language such as English.

[0020] The storage unit 2 stores the data of the product list 3 and the product performance list 4 while associating them with each other. For example, the storage unit 2 assigns product identification information for identifying robot products to the product list 3 and the product performance list 4, and associates the data of the product list 3 and the product performance list 4 by the product identification information. Alternatively, the storage unit 2 may store the data of the product list 3 and the product performance list 4 integrally as one file or one table in a database management system.

[0021] In addition, the storage unit 2 stores, in the product performance word conversion DB 5, the association information with each item of the performance information of the product performance list 4 for each product performance word 52. For example, the storage unit 2 assigns item identification information for identifying each item of the performance information of the product performance list 4, and stores the item identification information associated with the product performance word 52. By this association information, each item of the product performance list 4 and the product performance word 52 of the product performance word conversion DB 5 are associated with each other.

[0022] The information output unit 6 outputs information to a product selection user interface (hereinafter referred to as "product selection UI"). The product selection UI may be a screen displayed on a computer display or one using sound output from a computer speaker. The information output unit 6 outputs information necessary for product selection to the product selection UI via a display, a speaker, or the like. When outputting information by sound, the information output unit 6 converts text data into sound data by a known text-to-speech technology.

[0023] The input reception unit 7 receives an input sentence input as a user's request from the product selection UI. The input reception unit 7 receives the input sentence via an input device such as a mouse, a keyboard, or a microphone. When receiving the input sentence by sound, the input reception unit 7 converts the sound data into text data by a known speech recognition technology.

[0024] The product selection unit 8 extracts words from the input text, converts the extracted words 51 into product performance words 52 using the product performance word conversion DB 5, and selects a robot product having a performance that matches the product performance word 52 or a text including the product performance word 52 with reference to the product performance list 4.

[0025] FIG. 3 is a flowchart showing an example of the processing flow in the product selection device of FIG. 2. Hereinafter, a case where the service according to the present invention is realized as a web application will be described as an example. As shown in FIG. 3, the product selection device 1 outputs a product selection UI (step S1). Specifically, the product selection device 1 as a web server transmits the data of the product selection UI to the user terminal, and the user terminal displays the product selection UI on the display via the web browser.

[0026] FIG. 4 is an example of the product selection user interface in the product selection device of FIG. 2. As shown in FIG. 4, the product selection UI 9 has a selection column 91, an entry column 93, and a selection button 95. The selection column 91 includes a selection object 92 used for selecting at least one of the portable weight and the movable range of the robot product. In the example shown in FIG. 4, the selection object 92 is a checkbox. The selection object 92 may be a text box for inputting a number with upper and lower limits, a slider bar that can continuously select numbers, or the like. Thus, the information output unit 6 of the product selection device 1 outputs a selection object 92 for at least one of the portable weight and the movable range to the product selection UI 9. The entry column 93 includes an entry object 94 for inputting the user's request. The entry object 94 is preferably a text box in which text can be freely input. The selection button 95 is pressed after selection in the selection column 91 or after input in the entry column 93.

[0027] Return to the description of FIG. 3. When a selection operation is performed in the selection column 91 of the product selection UI 9 and the selection button 95 is pressed, the input reception unit 7 of the product selection device 1 receives the selection content by the selection object 92 of the product selection UI 9 (step S2). Specifically, the user terminal transmits the selection content to the product selection device 1, and the product selection device 1 receives the selection content.

[0028] Next, the information output unit 6 of the product selection device 1 outputs a robot product and the product selection UI 9 that satisfy the selection content (step S3). Specifically, the product selection device 1 searches the product list 3 using the selection content as a search condition, and transmits the product name 31, image, and data of the product selection UI of the robot product that satisfy the selection content to the user terminal. The user terminal displays the product name 31, image, and the product selection UI on the display. For example, when "~50" of the portable weight is selected in the selection column 91 of the product selection UI 9, the product selection device 1 acquires information on robot products whose portable weight 32 in the product list 3 is 50 kg or less, and the user terminal displays the information on the display. Thereby, the user can confirm a robot product that satisfies the conditions of the desired portable weight and movable range.

[0029] Next, the input reception unit 7 of the product selection device 1 receives the input text in the product selection UI 9 (step S4). Specifically, when an input text is input into the entry object 94 of the product selection UI 9 on the user terminal and the selection button 95 is pressed, the user terminal transmits the input text to the product selection device 1, and the product selection device 1 receives the input text. Here, the user inputs the desired usage and functions into the entry object 94.

[0030] Next, the product selection unit 8 of the product selection device 1 extracts words from the input text (step S5), and converts the extracted words 51 into product performance words 52 using the product performance word conversion DB 5 (step S6).

[0031] Figure 5 shows an example of the input sentence, product performance words, and paraphrased sentence in the product selection device of Figure 2. The product selection device 1 extracts words from the input sentence 21 using known natural language processing techniques. Then, the product selection device 1 searches the product performance word conversion DB5 using one or more of the extracted words 51 as search conditions. If there is a matching conversion rule, it acquires the product performance word 52 related to that conversion rule. In the example of Figure 5, the product selection device 1 extracts "substrate", "carry", "person", "approach", "slow down", etc. as the extracted words 51 from the input sentence 21 of "Carry the substrate. I want a function that makes the robot slow down when a person of current interest approaches." Then, the product selection device 1 searches the product performance word conversion DB5 using any combination of the extracted words 51 as search conditions. As a result of the search, the product selection device 1 converts the extracted words 51 of "substrate" and "carry" into the product performance word 52 of "substrate transfer". Similarly, the product selection device 1 converts the extracted words 51 of "person", "approach", and "slow down" into the product performance word 52 of "collaborative robot".

[0032] Returning to the description of Figure 3. The product selection unit 8 of the product selection device 1 generates a paraphrased sentence of the input sentence 21 using the product performance word 52 (step S7). Known generation AI (Artificial Intelligence) technology is used for the generation process of the paraphrased sentence. Specifically, the product selection device 1 sends a prompt (question sentence) including the input sentence 21, the product performance word 52, and an instruction to generate a paraphrased sentence of the input sentence 21 using the product performance word 52 to an external generation AI server. Then, the product selection device 1 receives an answer to the prompt from the external generation AI server and uses the answer as the paraphrased sentence. In the example of Figure 5, the product selection device 1 obtains an answer of "For substrate transfer. I want a collaborative robot." as the paraphrased sentence 22 from the external generation AI server. Note that the paraphrased expression is not limited to a concise expression, and can also be a polite expression, a translation into a foreign language, etc., and is not particularly limited.

[0033] Next, the product selection unit 8 of the product selection device 1 selects a product that matches the product performance word 52 or the paraphrased sentence 22 (= a sentence including the product performance word 52) (step S8).

[0034] For example, the product selection device 1 searches the product performance list 4 using the product performance word 52 as a search condition. In the example of FIG. 5, the product selection device 1 searches for a robot product in which the substrate transfer 43 in the product performance list 4 is "〇" and the collaborative robot 45 is "〇", and selects the search result as a product that matches the product performance word 52. In the example of the product performance list 4 in FIG. 2, the product selection device 1 selects a robot product whose product name 41 is "A" as a matching product.

[0035] Also, for example, the product selection device 1 may store a product description sentence in advance for each product name 41 in the product performance list 4, and select a product that matches the paraphrased sentence 22 based on the similarity between the paraphrased sentence 22 and the product description sentence.

[0036] Since the input sentence 21 allows free description, it may contain words that are not appropriate for selecting a product. On the other hand, since the product performance word 52 is a word associated with the performance information in the product performance list 4, the product selection device 1 can appropriately select a product by using the product performance word 52. Also, since the paraphrased sentence 22 is also a sentence including the product performance word 52, the product selection device 1 can appropriately select a product by using the paraphrased sentence 22. In this way, the product selection device 1 can select an appropriate product for the ambiguous request from the user without increasing the user's burden.

[0037] In particular, when accepting the selection by the selection object 92, the product selection unit 8 selects a robot product that has at least one of the portable weight and the movable range that matches the selection content, and has a performance that matches the product performance word 52 or a sentence including the product performance word 52. This is suitable for a user who is considering purchasing a robot product for the first time.

[0038] The selection criteria for robot products can be broadly classified into quantitative information such as portable weight and movable range, and qualitative information such as usage and level of operating speed. It is more convenient to select quantitative information from options rather than input it as text. On the other hand, since qualitative information may contain technical terms, it is difficult for users who are considering purchasing a robot product for the first time to select from options. Therefore, the product selection device 1 in the present embodiment presents options for quantitative information, accepts the selection, accepts free-form input as text for qualitative information, and selects a robot product based on both pieces of information.

[0039] Next, the information output unit 6 of the product selection device 1 outputs the selected product selected in step S8 and the paraphrased sentence 22 obtained in step S7 (step S9). Specifically, the product selection device 1 as a web server transmits the data of the information of the selected product and the paraphrased sentence 22 to the user terminal, and the user terminal displays the information of the selected product and the paraphrased sentence 22 on the display via a web browser.

[0040] The paraphrased sentence 22 is the answer of the generation AI server to the prompt in step S7 and includes the product performance word 52. By providing the paraphrased sentence 22 to the user, the user can understand how the product selection device 1 interpreted the input sentence 21 and selected the robot product, improving the user's acceptance.

[0041] FIG. 6 is a flowchart showing an example of the processing flow of a modified example of the product selection device in FIG. 3. The modified product selection device 1 selects a product without using the generation AI technology. Steps S11 to S14 shown in FIG. 6 are the same as steps S1 to S4 shown in FIG. 3, and thus the description thereof is omitted.

[0042] The product selection unit 8 of the product selection device 1 extracts product performance words 52 from the input sentence 21 (step S15). In a modified example, the storage unit 2 stores a list of product performance words 52 in advance. Then, the product selection device 1 extracts words that match the list of product performance words 52 stored in the storage unit 2 from the input sentence 21.

[0043] FIG. 7 is an example of an input sentence and product performance words in the product selection device of FIG. 6. In the example of FIG. 7, the product selection device 1 extracts "substrate transfer", "operating speed", "faster", and "clean room" as product performance words 52 from the input sentence 21: "I want to use it for substrate transfer. Since I want to improve productivity, please make the operating speed faster. It will be used in a clean room."

[0044] Returning to the description of FIG. 6, the product selection unit 8 of the product selection device 1 selects a product that matches the product performance word 52 (step S16). For example, the product selection device 1 searches the product performance list 4 using the product performance word 52 as a search condition. Then, the information output unit 6 of the product selection device 1 outputs the selected product selected in step S16 (step S17).

[0045] Since the product selection device 1 of the modified example does not use the generative AI technology, the response speed is improved. Also, since the product selection device 1 of the modified example stores the product performance word 52 instead of the product performance word conversion DB5, learning of the product performance word conversion DB5 is not required, and the preparation in advance becomes easy. However, if the input sentence 21 does not include the product performance word 52, the product cannot be selected. Therefore, the product selection device 1 of the modified example is suitable for a user who has experience in purchasing robot products.

[0046] As described above, the preferred embodiments of the product selection device and the like according to the present invention have been described with reference to the accompanying drawings, but the present invention is not limited to such examples. It is obvious that those skilled in the art can come up with various modification examples or correction examples within the scope of the technical idea disclosed in the present application, and it is naturally understood that those also belong to the technical scope of the present invention.

Description of Reference Numerals

[0047] 1………Product selection device 2………Memory unit 3………Product list 4………Product performance list 5………Product performance word conversion database (DB) 6………Information output unit 7………Input reception unit 8………Product selection unit 9………Product selection user interface (UI) 21………Input text 22………Rewritten text 51………Extracted word 52………Product performance word 92………Selected object

Claims

1. A storage unit that stores a product performance list including performance information for each robot product, and a database including conversion rules from one or more words to product performance words that express the performance of the robot product; An information output unit that outputs information to a product selection user interface; An input reception unit that receives an input sentence input as a user's request from the product selection user interface; A product selection unit that extracts words from the input sentence, converts them into the product performance words using the database, and selects the robot product having performance that matches the product performance words or a sentence including the product performance words with reference to the product performance list; A product selection device, characterized by comprising the above.

2. The product selection unit transmits a prompt including the input sentence, the product performance words, and an instruction to generate a paraphrased sentence of the input sentence using the product performance words to an external generative AI server, and receives an answer to the prompt from the generative AI server; The information output unit outputs the answer to the prompt to the product selection user interface as a paraphrased sentence of the input sentence. The product selection device according to claim 1, characterized by the above.

3. The storage unit stores a product list including at least one of the portable weight and the movable range of the robot product; The information output unit outputs a selection object of at least one of the portable weight and the movable range to the product selection user interface; The input reception unit receives the selection content by the selection object; The product selection unit selects the robot product having at least one of the portable weight and the movable range that matches the selection content and having performance that matches the product performance words or a sentence including the product performance words. The product selection device according to claim 1 or claim 2, characterized by the above.

4. A computer, A storage unit that stores a product performance list including performance information for each robot product, and a database including conversion rules from one or more words to product performance words that express the performance of the robot product; An information output unit that outputs information to a product selection user interface; An input reception unit that receives an input sentence input as a user's request from the product selection user interface; A product selection unit that extracts words from the input text, converts them into the product performance words using the database, and selects the robot product having a performance that matches the product performance word or the text including the product performance word with reference to the product performance list; A program for causing a product selection device including the same to function.

Images