GENERATION APPARATUS, GENERATION METHOD, AND PROGRAM
The generation device supports shops in creating promotional information for electronic markets by using a large-scale language model to analyze similar items, identify common expressions, and edit for uniqueness, addressing the challenge of creating effective explanatory information.
Patent Information
- Application Number
- JP2024146992
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2024-08-28
- Publication Date
- 2024-12-09
- Estimated Expiration
- 2044-08-28
AI Technical Summary
Shops struggle to create effective explanatory information for items on electronic markets, which is crucial for promoting sales, and there is a need to reduce the burden of creating such information.
A generation device and method that utilizes a large-scale language model to generate explanatory information based on specification information, selects similar items from the market, identifies common and unique expressions, and allows editing and differential information accumulation to support the creation of appealing and non-plagiarized content.
Automatically generates explanatory information that effectively conveys the appeal of items, reduces the need for repetitive editing, and ensures originality by incorporating common expressions from diverse sources while avoiding duplication.
Smart Images

Figure 0007596589000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to a generating device, a generating method, and a program. [Background technology]
[0002] Conventionally, there is known a technology that supports a shop that sells items on an online market in creating information to be registered for selling the items on the online market. For example, Patent Document 1 discloses an information processing device that analyzes an image of an item uploaded by a shop and presents information on the attributes of the item that should be registered on the online market. [Prior art documents] [Patent documents]
[0003] [Patent Document 1] JP 2018-106284 A Summary of the Invention [Problem to be solved by the invention]
[0004] Meanwhile, the information that a shop registers to sell an item on an electronic marketplace includes explanatory information that plays an advertising role to promote the sale of the item. The explanatory information appeals to viewers about the appeal of the item, such as scenarios in which the item can be used and its advantages, and is important information for promoting sales. However, it is not easy for a shop to create appropriate explanatory information, and there is a demand to reduce the burden involved in creating explanatory information.
[0005] The present invention has been made to solve the above-mentioned problems, and has an object to provide a generating device, a generating method, and a program capable of supporting shops in an online market in creating explanatory information that serves an advertising role. [Means for solving the problem]
[0006] A generating device according to a first aspect of the present invention comprises: a reception unit that receives input of specification information to be attached to a target item to be put up for sale on an electronic marketplace that allows a shop to sell the item by disclosing specification information for identifying the item and explanatory information for promoting the purchase of the item from the shop; a selection unit that selects a plurality of similar items being sold in the online market by disclosing specification information similar to the specification information related to the received input; an identification unit that identifies a common expression that appears in a plurality of pieces of explanatory information attached to the plurality of selected similar items; a generating unit that generates a prompt for generating explanatory information based on the specification information related to the received input and the identified common expression; a generating unit that generates explanatory information to be attached to the target item by providing the generated prompt to a large-scale language model; The present invention is characterized by comprising:
[0007] In addition, in the generating device according to the above aspect, an editing unit that allows a shop that intends to sell the target item to edit the generated explanatory information about the target item; a storage unit that stores difference information representing a difference between description information generated for the target item and description information published for the target item in association with a shop that is going to put up the target item; Further equipped with The creation unit creates a prompt for generating explanatory information based on difference information accumulated in the past in association with the shop where the target item is to be put up for sale. It is characterized by:
[0008] In addition, in the generating device according to the above aspect, The selection unit selects the plurality of similar items based on the number of views, the number of sales, or the sales amount in the electronic marketplace. It is characterized by:
[0009] In addition, in the generating device according to the above aspect, The selection unit selects the plurality of similar items such that the plurality of similar items are sold at different shops. It is characterized by:
[0010] In addition, in the generating device according to the above aspect, The selection unit selects the plurality of similar items from the shops selling the plurality of similar items, excluding a shop that is going to sell the target item. It is characterized by:
[0011] In addition, in the generating device according to the above aspect, The identification unit further identifies a unique expression appearing in any one of a plurality of pieces of explanatory information attached to the selected plurality of similar items, The creating unit creates a prompt for generating the explanatory information that does not include the identified specific expression. It is characterized by:
[0012] In addition, in the generating device according to the above aspect, The selection unit selects the plurality of similar items as items that are already being sold in the online market by the shop that is going to put up the target item and that have a high number of views, sales volume, or sales amount in the online market. It is characterized by:
[0013] In addition, in the generating device according to the above aspect, the selection unit selects the plurality of similar items to include higher-ranked items that are items already being sold on the online market by the shop that is going to put up the target item, and that are items with higher views, sales numbers, or sales amounts on the online market, and lower-ranked items that are lower-ranked items; the identification unit identifies a higher-level expression that appears in explanatory information attached to a higher-level item included in the selected plurality of similar items but does not appear in explanatory information attached to a lower-level item included in the selected plurality of similar items, and a lower-level expression that appears in explanatory information attached to a lower-level item included in the selected plurality of similar items but does not appear in explanatory information attached to a higher-level item included in the selected plurality of similar items, The generating unit generates a prompt for generating the explanatory information, the prompt including the identified higher-level expression and not including the identified lower-level expression. It is characterized by:
[0014] A generating device according to a second aspect of the present invention comprises: a reception unit that receives input of specification information to be attached to a target item to be put up for sale on an electronic marketplace that allows a shop to sell the item by disclosing specification information for identifying the item and explanatory information for promoting the purchase of the item from the shop; a generating unit that generates a prompt for generating explanatory information based on the specification information related to the received input; a generating unit that generates explanatory information to be attached to the target item by providing the generated prompt to a large-scale language model; an editing unit that allows a shop that intends to sell the target item to edit the generated explanatory information about the target item; a storage unit that stores difference information representing a difference between description information generated for the target item and description information published for the target item in association with a shop that is going to put up the target item; Equipped with The creation unit creates a prompt for generating explanatory information based on difference information accumulated in the past in association with the shop where the target item is to be put up for sale. It is characterized by:
[0015] A generating method according to a third aspect of the present invention comprises: A generation method executed by a generation device, comprising: a receiving step of receiving, from a shop, specification information for identifying an item and explanatory information for promoting purchase of the item, to be attached to a target item to be put up for sale on an electronic marketplace that allows the shop to sell the item by disclosing the specification information; a selection step of selecting a plurality of similar items sold in the electronic market by disclosing specification information similar to the specification information related to the received input; A step of identifying a common expression that appears in a plurality of pieces of explanatory information attached to the plurality of selected similar items; a creating step of creating a prompt for generating explanatory information based on the specification information related to the received input and the identified common expression; a generating step of generating explanatory information to be attached to the target item by applying the generated prompt to a large-scale language model; The present invention is characterized by comprising:
[0016] A generating method according to a fourth aspect of the present invention comprises: A generation method executed by a generation device, comprising: a receiving step of receiving, from a shop, specification information for identifying an item and explanatory information for promoting purchase of the item, to be attached to a target item to be put up for sale on an electronic marketplace that allows the shop to sell the item by disclosing the specification information; a generating step of generating a prompt for generating explanatory information based on the specification information related to the received input; a generating step of generating explanatory information to be attached to the target item by applying the generated prompt to a large-scale language model; an editing step of having a shop that intends to sell the target item edit the generated explanatory information about the target item; a storing step of storing difference information representing a difference between description information generated for the target item and description information published for the target item in association with a shop which is going to put up the target item; Equipped with In the creating step, a prompt is created for generating explanatory information based on difference information accumulated in the past in association with the shop where the target item is to be put up for sale. It is characterized by:
[0017] A program according to a fifth aspect of the present invention comprises: Computer, a reception unit that receives input of specification information to be attached to a target item to be put up for sale on an electronic marketplace that allows a shop to sell the item by disclosing specification information for identifying the item and explanatory information for promoting the purchase of the item from the shop; a selection unit that selects a plurality of similar items being sold in the online market by disclosing specification information similar to the specification information related to the received input; an identification unit that identifies a common expression that commonly appears in a plurality of pieces of explanatory information attached to the plurality of selected similar items; a generating unit that generates a prompt for generating explanatory information based on the specification information related to the received input and the identified common expression; a generating unit that generates explanatory information to be attached to the target item by applying the generated prompt to a large-scale language model; The present invention is characterized in that it functions as a
[0018] A program according to a sixth aspect of the present invention comprises: Computer, a reception unit that receives input of specification information to be attached to a target item to be put up for sale on an electronic marketplace that allows a shop to sell the item by disclosing specification information for identifying the item and explanatory information for promoting the purchase of the item from the shop; a generating unit for generating a prompt for generating explanatory information based on the specification information related to the received input; a generating unit that generates explanatory information to be attached to the target item by applying the generated prompt to a large-scale language model; an editing unit that allows a shop that is going to sell the target item to edit the generated explanatory information about the target item; a storage unit that stores difference information representing a difference between explanatory information generated for the target item and explanatory information published for the target item in association with a shop that is going to put up the target item; Function as a The creation unit creates a prompt for generating explanatory information based on difference information accumulated in the past in association with the shop where the target item is to be put up for sale. It is characterized by:
[0019] The above program may be recorded on a non-transitory recording medium. The non-transitory recording medium can be distributed and sold independently of the computer. Here, the non-transitory recording medium refers to a tangible recording medium. Examples of the non-transitory recording medium include a compact disc, a flexible disc, a hard disk, a magneto-optical disk, a digital video disk, a magnetic tape, and a semiconductor memory. In addition, the transitory recording medium refers to the transmission medium (propagation signal) itself. Examples of the transitory recording medium include an electric signal, an optical signal, and an electromagnetic wave. In addition, the temporary storage area is an area for temporarily storing data and programs, and is, for example, a volatile memory such as a RAM (Random Access Memory). Effect of the Invention
[0020] According to the present invention, it is possible to provide a generating device, a generating method, and a program capable of supporting the creation of explanatory information that plays an advertising role for shops in an online market. [Brief description of the drawings]
[0021] [Figure 1] 2 is a diagram showing the relationship between a generation device, a terminal, an online market server, and an external server according to an embodiment. FIG. [Diagram 2] FIG. 2 is a diagram illustrating a hardware configuration of a generating device according to an embodiment. [Diagram 3] FIG. 2 is a diagram illustrating a functional configuration of a generating device according to an embodiment. [Figure 4] FIG. 13 is a diagram illustrating an example of a creation page according to the embodiment. [Diagram 5] FIG. 13 illustrates an example of a prompt according to an embodiment. [Figure 6] FIG. 13 is a diagram illustrating an example of a creation page according to the embodiment. [Figure 7] FIG. 11 is a diagram illustrating an example of a difference information table according to the embodiment. [Figure 8] 1 is a diagram for explaining a generation process executed by a generation device according to an embodiment; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0022] (1. Overall Configuration) As shown in FIG. 1, a generation device 100 according to an embodiment of the present invention is communicatively connected to a plurality of shop terminals 200, an online market server 300, and an external server 400 via a computer communication network 500 such as the Internet.
[0023] The generating device 100 is a device that supports a shop in creating explanatory information that plays an advertising role for an item sold in an electronic market. For example, in response to a request from a shop terminal 200, the generating device 100 generates explanatory information for a specified item, provides the generated explanatory information to the shop terminal 200, and accepts edits to the generated explanatory information from the shop terminal 200.
[0024] The shop terminal 200 is a device used by a shop that sells items in the online market. The shop terminal 200 is, for example, a smartphone, a tablet computer, etc. A shop that uses the shop terminal 200 uses the generation device 100 to create description information for an item that the shop wishes to put up for sale on the online market.
[0025] The online market server 300 is a server that manages the online market. For example, the online market accepts registration of information for selling items from the shop terminal 200. Also, for example, the online market server 300 provides item information, processes item purchases, and the like in response to requests from shopper terminals (not shown) used by shoppers.
[0026] Information for selling an item includes specification information and description information. Specification information is information for identifying an item. For example, specification information is an item name, an item image, etc. Description information is information for promoting the purchase of an item, and plays an advertising role. For example, description information is information that describes the usage scenarios of the item, and information that describes the advantages of the item.
[0027] The online market server 300 causes shops to sell items on the online market by publishing specification information for identifying an item and description information for promoting the purchase of the item. For example, the online market server 300 causes shops to sell items on the online market by publishing a sales page on the online market that describes the specification information and description information of the item.
[0028] The external server 400 is a server that has a large-scale language model and outputs information in response to an input prompt, as described below. For example, the external server 400 receives a prompt input from the generation device 100 and outputs information in response to the prompt to the generation device 100.
[0029] (2. Hardware configuration of the generation device) FIG. 2 is a block diagram showing a hardware configuration of generation device 100. As shown in FIG.
[0030] 2, the generating device 100 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM 13, a recording medium 14, an output device 15, a communication device 16, and an input device 17. The components are connected to each other via a bus 18.
[0031] The CPU 11 controls the overall operation of the generating device 100, and is connected to each of the components to exchange control signals and data.
[0032] In the ROM 12, operating programs and various data necessary for controlling the overall operation of the generating device 100 are recorded.
[0033] The RAM 13 is for temporarily storing data and programs, and holds programs and data read from the recording medium 14, as well as other data necessary for communication.
[0034] The recording medium 14 is composed of a hard disk, a flash memory, or the like, and records the data to be processed by the generating device 100.
[0035] The output device 15 includes a display device such as an LCD (Liquid Crystal Display) and an audio output device such as a speaker. The output device 15 outputs, for example, data output from the CPU 11 under the control of the CPU 11.
[0036] The communication device 16 includes a communication interface for connecting the generation apparatus 100 to a computer communication network such as the Internet, and communicates with other information processing apparatuses via the communication device 16.
[0037] The input device 17 includes an input device such as a button, a keyboard, a touch panel, a microphone, etc. The input device 17 receives an operation input from a user of the generation device 100, and outputs a signal corresponding to the received operation input to the CPU 11.
[0038] (3. Functional configuration of the generating device according to the embodiment) The functional configuration of the generating device 100 will be described with reference to FIG.
[0039] The generating device 100 functionally includes a receiving unit 101, a selecting unit 102, a specifying unit 103, a creating unit 104, a generating unit 105, an editing unit 106, and a storage unit 107. In this embodiment, the CPU 11 and the communication device 16 function as the receiving unit 101, the selecting unit 102, the specifying unit 103, the generating unit 105, the editing unit 106, and the storage unit 107, and the CPU 11 functions as the creating unit 104.
[0040] In the following, it is assumed that when the generating device 100 receives a request from the shop terminal 200 to start creating explanatory information, it provides the shop terminal 200 with a creation page for creating explanatory information. For example, when "shop S1" is creating a sales page for "item P1-1" that it plans to put up for sale on an online market and issues an instruction to start creating explanatory information in the shop terminal 200, the shop terminal 200 transmits a request to start creating explanatory information to the generating device 100. In response to the request to start creating explanatory information, the generating device 100 provides the shop terminal 200 with a creation page.
[0041] FIG. 4 shows an example of a creation page. The creation page 600-1 in FIG. 4 is provided when “Shop S1” requests the creation of description information for “Item P1-1”. The creation page 600-1 includes a field 601 in which the name of the item that the shop intends to put up on the online market is entered, a region 602 for selecting an image of the item to be referred to when generating description information, a field 603 for entering appealing points of the item to be referred to when generating description information, a button 604 for instructing the generation of description information, a field 605 in which the generated description information is displayed, and a button 606 for copying the description information displayed in field 605 to the clipboard. The item name displayed in field 601 is the name of the item entered by the shop while creating the sales page, and includes the attributes, features, etc. of the item in addition to the name of the item. The image of the item displayed in field 602 is an image uploaded by the shop while creating the sales page. By selecting a radio button 607 included in field 602, the shop can select an image of the item to be referred to when generating description information, and image 1 is selected in FIG. 4. Input into field 603 is optional, and description information will be generated even if selling points are not entered. The description information displayed in field 605 can be edited by the shop by filling it in directly. When button 606 is selected, the description information displayed in field 605 is copied to the clipboard. The shop can use the description information by copying the description information copied to the clipboard to the sales page they are creating.
[0042] The reception unit 101 receives input of specification information to be attached to a target item to be put up for sale on an online market from a shop.
[0043] The target item is an item that a shop is about to put up for sale on the online market, for example, the target item is "item P1-1" that "shop S1" is about to put up for sale on the online market. The specification information to be attached to the target item is information for identifying the item entered by the shop, for example, the item name, item image, selling points, etc. of the target item.
[0044] For example, when “Shop S1” selects button 604 on creation page 600-1 of FIG. 4, the reception unit 101 receives the item name “Item P1-1, Color C1, Size M1” displayed in field 601 and “Image 1” selected in area 602 as specification information.
[0045] The selection unit 102 selects a plurality of similar items being sold in the online market by disclosing specification information similar to the specification information related to the received input.
[0046] The criteria for determining similarity of the specification information are arbitrarily set by the user of the generating device 100. For example, when the number of words in the item name of an item sold on the online market that match the words included in the item name accepted as the specification information is equal to or exceeds a predetermined threshold, the selection unit 102 selects the item sold on the online market as a similar item. Also, for example, the selection unit 102 calculates the similarity between an image of an item sold on the online market and an image of an item accepted as the specification information using a learning device that has been subjected to machine learning, and selects an item sold on the online market with a similarity equal to or exceeds a predetermined threshold as a similar item. Note that the method for calculating the similarity of images is not limited to using such a learning device, and the similarity may be calculated using other methods, such as extracting characteristic parts from each image and performing pattern matching. In the following, "item P1-2" to "item P1-N (N: a positive number equal to or greater than 2)" are items sold on the online market that have specification information similar to the specification information of "item P1-1".
[0047] Here, the selection unit 102 selects a plurality of similar items based on the number of views, the number of sales, or the sales amount in the online market.
[0048] For example, the selection unit 102 selects, as a plurality of similar items, items "P1-2" to "P1-N" that have specification information similar to that of "Item P1-1" and are being sold on the online market, whose number of views, number of sales, or sales amount on the online market is equal to or exceeds a predetermined threshold. The predetermined thresholds for the number of views, number of sales, and sales amount are arbitrarily set by the user of the generation device 100.
[0049] Alternatively, the selection unit 102 selects a plurality of similar items such that the plurality of similar items are sold at different shops.
[0050] For example, the selection unit 102 selects a similar item from a plurality of items "Item P1-2" to "Item P1-N" that have specification information similar to that of "Item P1-1" and are sold in the online marketplace, by limiting the items sold by the same shop to one. The criterion for limiting to one is set arbitrarily by the user of the generation device 100. For example, the selection unit 102 selects the item with the largest number of views, number of sales, or sales amount among the items sold by the same shop. In the following, "Item P1-2" to "Item P1-M (M is a positive number equal to or greater than 2)" are selected as similar items, and "Item P1-2" to "Item P1-M" are respectively sold in the online marketplace by "Shop S1" to "Shop ST (T is a positive number equal to M-1)".
[0051] Furthermore, the selection unit 102 selects a plurality of similar items from the shops selling a plurality of similar items, excluding the shop that is going to put up the target item.
[0052] For example, the selection unit 102 selects, from among "Item P1-2" to "Item P1-M", "Item P1-3" to "Item P1-M" excluding "Item P1-2" sold by "Shop S1", as similar items.
[0053] The identifying unit 103 identifies a common expression that commonly appears in a plurality of pieces of explanatory information attached to the plurality of selected similar items.
[0054] A common expression is an expression that is commonly used among expressions used in a plurality of pieces of explanatory information attached to a plurality of similar items, For example, a common expression is a word, a phrase, a statement, or the like.
[0055] For example, the identification unit 103 accesses the online market server 300, refers to the description information included in the sales pages of "Item P1-3" to "Item P1-M," and identifies common expressions commonly used in the description information of "Item P1-3" to "Item P1-M." In the following, it is assumed that "Common Expression 1" to "Common Expression 3" are identified.
[0056] In addition, the identifying unit 103 further identifies a unique expression that appears in any one of the multiple pieces of explanatory information attached to the multiple selected similar items.
[0057] The unique expression is an expression used in any one of a plurality of pieces of explanatory information attached to a plurality of similar items, For example, the unique expression is a word, a phrase, a wording, or the like.
[0058] For example, the identifying unit 103 further identifies a unique expression used in any one of the description information of "item P1-3" to "item P1-M." In the following, it is assumed that "unique expression 1" to "unique expression 3" are identified.
[0059] The creation unit 104 creates a prompt for generating explanatory information based on the specification information related to the received input and the identified common expression.
[0060] Furthermore, the creation unit 104 creates a prompt for generating explanatory information that does not include the identified specific expression.
[0061] A prompt indicates instructions to be given to a large-scale language model. For example, a template of the prompt is stored in the recording medium 14 of the generation device 100, and the creation unit 104 creates the prompt by referring to the template. Specifically, the creation unit 104 creates a prompt including an instruction to generate explanatory information based on the item name and image accepted as the specification information, the explanatory information including "common expression 1" to "common expression 3" identified by the identification unit 103, and not including "unique expression 1" to "unique expression 3" identified by the identification unit 103.
[0062] An example of the created prompt is shown in Fig. 5. Prompt 700 in Fig. 5 includes an instruction 701 to generate explanatory information based on the item name, item image, etc. input as specification information, a constraint 702 that specifies that the explanatory information should include the identified expressions "common expression 1" to "common expression 3", and a constraint 703 that specifies that the explanatory information should not include the identified expressions "unique expression 1" to "unique expression 3".
[0063] The generation unit 105 provides the generated prompt to a large-scale language model to generate explanatory information to be attached to the target item.
[0064] For example, the generation unit 105 inputs a prompt, "Item P1-1, Color C1, Size M1" and "Image 1" to the external server 400 having a large-scale language model, and causes the external server 400 to generate explanatory information to be attached to "Item P1-1." The generation unit 105 then receives the generated explanatory information from the external server 400, and displays the received explanatory information in the column 605 of the creation page.
[0065] An example of a creation page on which the generated explanatory information is displayed is shown in Fig. 6. In a column 605 of the creation page 600-2 in Fig. 6, explanatory information including "common expression 1" to "common expression 3" and not including "unique expression 1" to "unique expression 3" is displayed.
[0066] The editing unit 106 allows the shop that is going to put up the target item to edit the explanatory information generated for the target item.
[0067] For example, the editing unit 106 accepts an edit from "Shop S1" which is trying to sell "Item P1-1" to change the explanatory information displayed in the column 605 of the creation page 600-2 in Fig. 6 from "The effect is obtained" to "It brings about an effect." Then, when the button 606 on the creation page 600-2 in Fig. 6 is selected, the explanatory information with the change is copied to the clipboard.
[0068] The storage unit 107 stores difference information indicating the difference between the description information generated for the target item and the description information published for the target item in association with the shop where the target item is to be put up for sale.
[0069] The explanatory information generated for the target item is explanatory information generated by the generation unit 105. Moreover, the explanatory information published for the target item is explanatory information generated by the generation unit 105 or explanatory information edited by a shop and published in the online market.
[0070] The difference information is information indicating the difference between the explanatory information generated by the generation unit 105 for the target item and the explanatory information published in the online market. If the explanatory information generated by the generation unit 105 is used as is, there is no difference between the explanatory information generated for the target item and the explanatory information published for the target item, and if the explanatory information generated by the generation unit 105 is edited, there is a difference between the explanatory information generated for the target item and the explanatory information published for the target item. In other words, the difference information is information indicating the corrections made by the shop.
[0071] For example, when a sales page for "item P1-1" is published on the online market, the storage unit 107 compares the explanatory information generated for "item P1-1" with the explanatory information included in the sales page for "item P1-1." Then, when the storage unit 107 detects a difference between the generated explanatory information and the published explanatory information, it stores the difference in the recording medium 14 as difference information.
[0072] An example of a difference information table including difference information is shown in Fig. 7. In the difference information table 800 in Fig. 7, a shop ID for identifying a shop selling an item in the online market, an item ID for identifying an item being sold by the shop in the online market, and difference information indicating the content of the difference between the generated description information and the published description information are registered in association with each other. The record in the first row of the difference information table 800 indicates that "shop S1" (shop ID "S1") is selling "item P1-1" (item ID "P1-1") in the online market, and the difference between the generated description information and the published description information is "the effect is obtained" in the generated description information and "it brings about an effect" in the published description information.
[0073] Then, the creation unit 104 creates a prompt for generating explanatory information based on the difference information accumulated in the past in association with the shop that is going to put up the target item.
[0074] For example, when "Shop S1" newly puts "Item P10-1" up for sale on the online market, the creation unit 104 refers to the difference information table 800 in Fig. 7 and creates a prompt for generating explanatory information based on the received specification information of "Item P10-1", the identified common expression, and the difference information associated with "Shop S1", the explanatory information not including idiosyncratic expressions. Specifically, the creation unit 104 adds to the constraint conditions included in the prompt a condition for changing the explanatory information to the published explanatory information indicated in the difference information if the initial draft of the explanatory information includes the generated explanatory information indicated in the difference information.
[0075] (4. Generation process by the generation device of the embodiment) The generation process executed by the generation device 100 of this embodiment will be described with reference to Fig. 8. When the generation device 100 receives an instruction to generate explanation information from the shop terminal 200, the generation device 100 starts the generation process shown in Fig. 8. For example, when the button 604 is selected on the creation page 600-1 in Fig. 4, the generation device 100 starts the generation process in Fig. 8.
[0076] The reception unit 101 receives, from a shop, input of specification information to be attached to a target item to be put up for sale on an online market (step S101).
[0077] For example, when “Shop S1” selects button 604 on creation page 600-1 of FIG. 4, the reception unit 101 receives the item name “Item P1-1, Color C1, Size M1” displayed in field 601 and “Image 1” selected in area 602 as specification information.
[0078] The selection unit 102 selects a plurality of similar items that are being sold in the online market by disclosing specification information similar to the specification information related to the received input (step S102).
[0079] For example, the selection unit 102 selects as similar items "Item P1-3" to "Item P1-M" from "Item P1-2" to "Item P1-N" sold on the online market that have specification information similar to the specification information related to the received input and whose number of views, number of sales, or sales amount on the online market are equal to or greater than a predetermined threshold, excluding "Item P1-2" sold by "Shop S1."
[0080] The identification unit 103 identifies common expressions that appear in common in the multiple pieces of explanatory information attached to the selected multiple similar items, and unique expressions that appear in any one of the multiple pieces of explanatory information attached to the selected multiple similar items (step S103).
[0081] For example, the identifying unit 103 accesses the online market server 300, refers to the description information included in the sales pages of "Item P1-3" to "Item P1-M," and identifies "Common Expression 1" to "Common Expression 3" that are commonly used in the description information of "Item P1-3" to "Item P1-M." The identifying unit 103 also identifies "Unique Expression 1" to "Unique Expression 3" that is used in any one of the description information of "Item P1-3" to "Item P1-M."
[0082] The creation unit 104 creates a prompt for generating explanatory information based on the specification information related to the received input and the identified common expression, the explanatory information not including the identified specific expression (step S104).
[0083] For example, the creation unit 104 generates a prompt containing instructions to generate explanatory information based on "Item P1-1, Color C1, Size M1" and "Image 1", which includes "Common Expression 1" to "Common Expression 3" identified by the identification unit 103, but does not include "Peculiar Expression 1" to "Peculiar Expression 3" identified by the identification unit 103.
[0084] The generation unit 105 provides the generated prompt to the large-scale language model to generate explanatory information to be attached to the target item (step S105).
[0085] For example, the generation unit 105 inputs a prompt, “Item P1-1, Color C1, Size M1” and “Image 1” to an external server 400 having a large-scale language model, causes the external server 400 to generate explanatory information to be attached to “Item P1-1”, receives the generated explanatory information and displays it in column 605 of the creation page.
[0086] According to this embodiment, description information for an item to be put up for sale on the online market is generated based on specification information for the item to be put up for sale on the online market and common expressions commonly used in description information for similar items sold on the online market. This makes it possible to automatically generate a rough draft of description information that conveys the appeal of the item and does not miss any important points.
[0087] Furthermore, according to this embodiment, difference information indicating the difference between explanatory information generated for an item and explanatory information published for the item in the online market is accumulated, and a prompt for generating explanatory information is created based on the past difference information. As a result, if a shop has previously modified explanatory information generated by the generation device 100, a similar modification can be automatically performed in advance when newly generating explanatory information for an item. This eliminates the need for the shop to make the same modification.
[0088] Furthermore, according to this embodiment, similar items with explanatory information to be referenced when identifying common expressions are selected based on the number of views, number of sales, or sales amount in the online market. This allows the explanatory information of items with high numbers of views or sales to be used as a reference, and allows the creation of a rough draft of appropriate explanatory information.
[0089] Furthermore, according to this embodiment, the multiple similar items to which the explanatory information to be referred to when identifying the common expression is added are selected so that the multiple similar items are sold in different shops. This allows common expressions adopted by different shops to be used as common expressions, and allows the generation of explanatory information including such expressions.
[0090] Furthermore, according to this embodiment, a plurality of similar items to which description information that can be referenced to identify common expressions is selected, excluding items sold by the shop that is trying to sell the item. This allows the user to refer to description information from other shops, rather than the description information from the shop that is trying to sell the item.
[0091] In addition, according to the present embodiment, a specific expression that appears in one of the description information of similar items sold on the online market is identified, and a prompt is created for generating description information that does not include the specific expression. This makes it possible to generate description information that excludes expressions unique to shops selling similar items, and automatically generate description information that does not plagiarize expressions from other shops.
[0092] (5. Modifications) Although the embodiment of the present invention has been described above, various modifications and applications are possible in carrying out the present invention.
[0093] In the above embodiment, the functional units of the generating device 100 are shown in Fig. 3, but are not limited thereto. The generating device 100 may functionally include only the receiving unit 101, the creating unit 104, the generating unit 105, the editing unit 106, and the storing unit 107.
[0094] Specifically, the receiving unit 101 receives an input of specification information to be attached to a target item to be put up on the electronic market from a shop. The creating unit 104 creates a prompt for generating explanatory information based on the specification information related to the received input. The generating unit 105 generates explanatory information to be attached to the target item by providing the created prompt to a large-scale language model. The editing unit 106 causes the shop that is going to put up the target item to edit the explanatory information generated for the target item. The storage unit 107 stores difference information representing the difference between the explanatory information generated for the target item and the explanatory information published for the target item in association with the shop that is going to put up the target item. Then, the creating unit 104 creates a prompt for generating explanatory information further based on the difference information previously stored in association with the shop that is going to put up the target item. That is, the generating device 100 generates explanatory information based on the specification information of the target item and the content previously edited by the shop that is going to put up the target item, without referring to explanatory information published by other shops in the electronic market.
[0095] In the above embodiment, the generation device 100 generates the description information of the item that the shop is going to put up for sale by referring to the description information attached to similar items sold by other shops, but this is not limited to this. The generation device 100 may generate the description information by referring to the description information attached to similar items sold by the shop that is going to put up the target item.
[0096] Specifically, the selection unit 102 selects multiple similar items so that they are items that are already being sold on the online market by the shop that is listing the target item and that have the highest number of views, sales volume, or sales amount on the online market.
[0097] For example, among items sold in the online market that have similar specification information to the specification information of "Item P11-1" that "Shop S1" is putting up for sale, the items sold by "Shop S1" are designated as "Item P11-2" to "Item P11-K (K: a positive number equal to or greater than 2)." In this case, the selection unit 102 selects, from among "Item P11-2" to "Item P11-K," items whose number of views, number of sales, or sales amount in the online market is equal to or greater than a predetermined threshold value as multiple similar items.
[0098] This allows the shop to refer to description information of items that are being sold by the shop itself and that have been viewed a lot or are selling well, making it possible to generate a rough draft of appropriate description information.
[0099] Alternatively, the selection unit 102 selects multiple similar items so as to include top items, which are items already being sold on the online market by the shop that is planning to list the target item, and which are items with the highest number of views, sales volume, or sales amount on the online market, and lower items, which are lower ranking items.
[0100] For example, the selection unit 102 selects multiple similar items from among "item P11-2" to "item P11-K," with the items whose number of views, number of sales, or sales amount in the online marketplace is equal to or greater than a first threshold as higher-level items, and the items whose number of views, number of sales, or sales amount in the online marketplace is equal to or less than a second threshold as lower-level items. The first and second thresholds for the number of views, number of sales, and sales amount are arbitrarily set by the user of the generation device 100. Hereinafter, the higher-level items included in the selected multiple similar items are referred to as "item P11-2" to "item P11-5," and the lower-level items are referred to as "item P11-12" to "item P11-15."
[0101] Next, the identification unit 103 identifies a higher-level expression that appears in explanatory information attached to a higher-level item included in the selected multiple similar items but does not appear in explanatory information attached to a lower-level item included in the selected multiple similar items, and a lower-level expression that appears in explanatory information attached to a lower-level item included in the selected multiple similar items but does not appear in explanatory information attached to a higher-level item included in the selected multiple similar items.
[0102] For example, the identification unit 103 identifies higher-level expressions that appear in the explanation information attached to the higher-level items "Item P11-2" to "Item P11-5" but do not appear in the explanation information attached to the lower-level items "Item P11-12" to "Item P11-15". The identification unit 103 also identifies lower-level expressions that appear in the explanation information attached to the lower-level items "Item P11-12" to "Item P11-15" but do not appear in the explanation information attached to the higher-level items "Item P11-2" to "Item P11-5". Hereinafter, the identified higher-level expressions are referred to as "higher-level expression 1" to "higher-level expression 3", and the identified lower-level expressions are referred to as "lower-level expression 1", to "lower-level expression 3".
[0103] Then, the generator 104 generates a prompt for generating explanatory information that includes the identified higher-level expression and does not include the identified lower-level expression.
[0104] For example, the creation unit 104 creates a prompt containing instructions to generate explanatory information based on the item name and image accepted as specification information, which includes ``super-representation 1'' to ``super-representation 3'' identified by the identification unit 103, but does not include ``sub-representation 1'' to ``sub-representation 3'' identified by the identification unit 103.
[0105] This allows the shop to refer to description information of items sold by the shop itself that have high numbers of views and sales, as well as description information of items that have low numbers of views and sales, and to generate a rough draft of appropriate description information.
[0106] In the above embodiment, the identification unit 103 determines an expression that appears in any one of the multiple pieces of explanatory information attached to the multiple selected similar items as a unique expression, but this is not limited thereto. The identification unit 103 may determine an expression that appears only in any one of the multiple pieces of explanatory information attached to the multiple selected similar items as a unique expression. Alternatively, the identification unit 103 may determine a small threshold H according to the number L of multiple pieces of explanatory information, and determine an expression that appears in a maximum of H pieces of explanatory information as a unique expression. For example, if the number L is 2 to 100, the threshold H is set to 1, and if the number L is 100 to 1000, the threshold H is set to 2. The threshold H can be arbitrarily adjusted by the user of the generation device 100.
[0107] In addition, by applying an operating program that specifies the operation of the generating device 100 according to the above embodiment to an existing personal computer or information terminal device, it is also possible to cause the personal computer or information terminal device to function as the generating device 100 according to the embodiment.
[0108] The present invention allows various embodiments and modifications without departing from the broad spirit and scope of the present invention. The above-described embodiments are for the purpose of explaining the present invention and do not limit the scope of the present invention. That is, the scope of the present invention is indicated by the claims, not the embodiments. Various modifications made within the scope of the claims and the scope of the invention equivalent thereto are considered to be within the scope of the present invention. [Industrial Applicability]
[0109] According to the present invention, it is possible to provide a generating device, a generating method, and a program capable of supporting the creation of explanatory information that plays an advertising role for shops in an online market. [Explanation of symbols]
[0110] 11 CPU 12 ROM 13 RAM 14 Recording media 15 Output Devices 16 Communication Devices 17 Input Devices 18 Bus 100 generator 101 Reception 102 Selection section 103 Specific part 104 Creation Department 105 Generation part 106 Editorial Department 107 Storage Unit 200 Shop terminals 300 Electronic Market Server 400 External Server 500 Computer Network 600-1,600-2 Created page Columns 601, 603, and 605 602 areas 604,606 Buttons 607 Radio Button 700 prompt 701 Instructions 702,703 Constraints 800 Difference Information Table
Claims
1. a reception unit that receives input of specification information to be attached to a target item to be put up for sale on an electronic marketplace that allows a shop to sell the item by disclosing specification information for identifying the item and explanatory information for promoting the purchase of the item from the shop; a selection unit that selects a plurality of similar items being sold in the online market by disclosing specification information similar to the specification information related to the received input; an identification unit that identifies a common expression that appears in a plurality of pieces of explanatory information attached to the plurality of selected similar items; a generating unit that generates a prompt for generating explanatory information based on the specification information related to the received input and the identified common expression; a generating unit that generates explanatory information to be attached to the target item by providing the generated prompt to a large-scale language model; A generating device comprising:
2. an editing unit that allows a shop that intends to sell the target item to edit the generated explanatory information about the target item; a storage unit that stores difference information representing a difference between description information generated for the target item and description information published for the target item in association with a shop that is going to put up the target item; Further equipped with The creation unit creates a prompt for generating explanatory information based on difference information accumulated in the past in association with the shop where the target item is to be put up for sale. The generating device according to claim 1 .
3. The selection unit selects the plurality of similar items based on the number of views, the number of sales, or the sales amount in the electronic marketplace.
3. The generating device according to claim 1 or 2.
4. The selection unit selects the plurality of similar items such that the plurality of similar items are sold at different shops.
3. The generating device according to claim 1 or 2.
5. The selection unit selects the plurality of similar items from the shops selling the plurality of similar items, excluding a shop that is going to sell the target item.
5. The generating device according to claim 4.
6. The identification unit further identifies a unique expression appearing in any one of a plurality of pieces of explanatory information attached to the selected plurality of similar items, The creating unit creates a prompt for generating the explanatory information that does not include the identified specific expression.
3. The generating device according to claim 1 or 2.
7. The selection unit selects the plurality of similar items as items that are already being sold in the online market by the shop that is going to put up the target item and that have a high number of views, sales volume, or sales amount in the online market.
3. The generating device according to claim 1 or 2.
8. the selection unit selects the plurality of similar items to include higher-ranked items that are items already being sold on the online market by the shop that is going to put up the target item, and that are items with higher views, sales numbers, or sales amounts on the online market, and lower-ranked items that are lower-ranked items; the identification unit identifies a higher-level expression that appears in explanatory information attached to a higher-level item included in the selected plurality of similar items but does not appear in explanatory information attached to a lower-level item included in the selected plurality of similar items, and a lower-level expression that appears in explanatory information attached to a lower-level item included in the selected plurality of similar items but does not appear in explanatory information attached to a higher-level item included in the selected plurality of similar items, The generating unit generates a prompt for generating the explanatory information, the prompt including the identified higher-level expression and not including the identified lower-level expression.
3. The generating device according to claim 1 or 2.
9. a reception unit that receives input of specification information to be attached to a target item to be put up for sale on an electronic marketplace that allows a shop to sell the item by disclosing specification information for identifying the item and explanatory information for promoting the purchase of the item from the shop; a generating unit that generates a prompt for generating explanatory information based on the specification information related to the received input; a generating unit that generates explanatory information to be attached to the target item by providing the generated prompt to a large-scale language model; an editing unit that allows a shop that intends to sell the target item to edit the generated explanatory information about the target item; a storage unit that stores difference information representing a difference between description information generated for the target item and description information published for the target item in association with a shop that is going to put up the target item; Equipped with The creation unit creates a prompt for generating explanatory information based on difference information accumulated in the past in association with the shop where the target item is to be put up for sale. A generating device characterized by:
10. A generation method executed by a generation device, comprising: a receiving step of receiving, from a shop, specification information for identifying an item and explanatory information for promoting purchase of the item, to be attached to a target item to be put up for sale on an electronic marketplace that allows the shop to sell the item by disclosing the specification information for identifying the item and explanatory information for promoting purchase of the item; a selection step of selecting a plurality of similar items sold in the electronic market by disclosing specification information similar to the specification information related to the received input; A step of identifying a common expression that appears in a plurality of pieces of explanatory information attached to the plurality of selected similar items; a creating step of creating a prompt for generating explanatory information based on the specification information related to the received input and the identified common expression; a generating step of generating explanatory information to be attached to the target item by applying the generated prompt to a large-scale language model; A generating method comprising:
11. A generation method executed by a generation device, comprising: a receiving step of receiving, from a shop, specification information for identifying an item and explanatory information for promoting purchase of the item, to be attached to a target item to be put up for sale on an electronic marketplace that allows the shop to sell the item by disclosing the specification information for identifying the item and explanatory information for promoting purchase of the item; a generating step of generating a prompt for generating explanatory information based on the specification information related to the received input; a generating step of generating explanatory information to be attached to the target item by applying the generated prompt to a large-scale language model; an editing step of having a shop that intends to sell the target item edit the generated explanatory information about the target item; a storing step of storing difference information representing a difference between description information generated for the target item and description information published for the target item in association with a shop which is going to put up the target item; Equipped with In the creating step, a prompt is created for generating explanatory information based on difference information accumulated in the past in association with the shop where the target item is to be put up for sale. A generating method comprising:
12. Computer, a reception unit that receives input of specification information to be attached to a target item to be put up for sale on an electronic marketplace that allows a shop to sell the item by disclosing specification information for identifying the item and explanatory information for promoting the purchase of the item from the shop; a selection unit that selects a plurality of similar items being sold in the online market by disclosing specification information similar to the specification information related to the received input; an identification unit that identifies a common expression that commonly appears in a plurality of pieces of explanatory information attached to the plurality of selected similar items; a generating unit that generates a prompt for generating explanatory information based on the specification information related to the received input and the identified common expression; a generating unit that generates explanatory information to be attached to the target item by applying the generated prompt to a large-scale language model; A program characterized by causing the program to function as a
13. Computer, a reception unit that receives input of specification information to be attached to a target item to be put up for sale on an electronic marketplace that allows a shop to sell the item by disclosing specification information for identifying the item and explanatory information for promoting the purchase of the item from the shop; a generating unit for generating a prompt for generating explanatory information based on the specification information related to the received input; a generating unit that generates explanatory information to be attached to the target item by applying the generated prompt to a large-scale language model; an editing unit that allows a shop that is going to sell the target item to edit the generated explanatory information about the target item; a storage unit that stores difference information representing a difference between explanatory information generated for the target item and explanatory information published for the target item in association with a shop that is going to put up the target item; Function as a The creation unit creates a prompt for generating explanatory information based on difference information accumulated in the past in association with the shop where the target item is to be put up for sale. A program characterized by:
Citation Information
Patent Citations
Device and method for predicting share of every merchandise, and program for the same
JP2005284470A
Information processing apparatus, information processing method, and information processing program
JP2018106284A
Generation apparatus, generation method, and generation program
JP2018128805A
Information processing device, information processing method, and information processing program
JP2022167665A
JPP7411137B
