Information processing device, information processing method, and information processing program

The information processing device uses transfer learning and fine-tuning to generate high-quality advertisement texts, addressing the issue of inappropriate text selection in conventional advertising systems by enhancing click-through rates and language suitability.

JP7580430B2Active Publication Date: 2024-11-11LY CORP
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
JP2022095840
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2022-06-14
Publication Date
2024-11-11
Estimated Expiration
2042-06-14

AI Technical Summary

Technical Problem

Conventional advertising technologies often fail to select appropriate advertisement texts, leading to ineffective advertising.

Method used

An information processing device that learns a model to generate appropriate advertisement texts by using transfer learning with pre-trained models like GPT-2, fine-tuning on high-performing advertisements, and generating new texts based on input advertisements.

Benefits of technology

The system effectively generates advertisement texts with high predicted click-through rates and language likeness, providing advertisers with optimized advertising content.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007580430000001
    Figure 0007580430000001
  • Figure 0007580430000002
    Figure 0007580430000002
  • Figure 0007580430000003
    Figure 0007580430000003
Patent Text Reader

Abstract

To provide an information processing apparatus, an information processing method, and an information processing program, capable of generating an appropriate advertising sentence.SOLUTION: In an information processing system in which a terminal device, a user terminal, an information providing apparatus, and an information processing apparatus are connected by wire or wirelessly so as to communicate with each other over a network, the information processing apparatus 100 includes an acquisition unit 131 and a learning unit 132. The acquisition unit 131 acquires information on a first advertising sentence which is an advertising sentence that satisfies a predetermined condition. The learning unit 132 trains, on the basis of the information on the first advertising sentence acquired by the acquisition unit 131, a learning model that receives, as input information, character information and outputs a sentence corresponding to the input information, as a second advertising sentence.SELECTED DRAWING: Figure 4
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to an information processing device, an information processing method, and an information processing program. [Background technology]

[0002] Conventionally, various techniques related to advertising have been proposed, such as a technique for selecting advertising copy for the same advertising target in a personalized manner. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2017-111479 A [Non-patent literature]

[0004] [Non-Patent Document 1] “Improving Language Understanding by Generative Pre-Training”, Alec Radford, Karthik Narasimhan, Tim Salimans, and Ilya Sutskever<Internet>https: / / s3-us-west-2.amazonaws.com / openai-assets / research-covers / language-unsupervised / language_understanding_paper.pdf(Retrieved June 1, 2022) Summary of the Invention [Problem to be solved by the invention]

[0005] However, there is room for improvement in the above conventional technology. For example, the above conventional technology merely selects an advertisement copy for an individual from a plurality of advertisement copies for an advertisement target, and does not necessarily provide an appropriate advertisement copy, for example, when the advertisement copies to be selected do not include an appropriate advertisement copy. Therefore, it is desired to be able to generate an appropriate advertisement copy.

[0006] The present application has been made in consideration of the above, and aims to provide an information processing device, an information processing method, and an information processing program that enable generation of appropriate advertising copy. [Means for solving the problem]

[0007] The information processing device according to the present application is characterized by comprising: an acquisition unit that acquires information related to a first advertising copy, which is an advertising copy that satisfies a predetermined condition; and a learning unit that learns a learning model that receives character information as input information and outputs a sentence corresponding to the input information as a second advertising copy, based on the information related to the first advertising copy acquired by the acquisition unit. Effect of the Invention

[0008] According to one aspect of the embodiment, it is possible to provide an effect of enabling appropriate advertising copy to be generated. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram illustrating an example of information processing according to an embodiment. [Diagram 2] FIG. 2 is a diagram showing a specific example of information processing. [Diagram 3] FIG. 3 is a diagram illustrating an example of the configuration of an information processing system according to the embodiment. [Figure 4] FIG. 4 is a diagram illustrating an example of the configuration of the information processing device according to the embodiment. [Diagram 5] FIG. 5 is a diagram illustrating an example of an advertisement information storage unit according to the embodiment. [Figure 6]FIG. 6 is a diagram illustrating an example of a model information storage unit according to the embodiment. [Figure 7] FIG. 7 is a diagram illustrating an example of a generation information storage unit according to the embodiment. [Figure 8] FIG. 8 is a flowchart showing an example of the flow of information processing executed by the information processing device. [Figure 9] FIG. 9 is a hardware configuration diagram illustrating an example of a computer that realizes the functions of the information processing device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, the information processing device, the information processing method, and the information processing program according to the present application will be described in detail with reference to the drawings. Note that the information processing device, the information processing method, and the information processing program according to the present application are not limited to the embodiments. In addition, the same parts in the following embodiments are denoted by the same reference numerals, and duplicated descriptions are omitted.

[0011] (Embodiment) [1. Information Processing] First, an example of information processing executed by the information processing device 100 will be described with reference to Fig. 1. Fig. 1 is a diagram showing an example of information processing according to an embodiment. In the following, a case will be described as an example in which the information processing device 100 generates an advertising copy (also called a "second advertising copy") using information on an advertising copy (also called a "first advertising copy") that satisfies a predetermined condition and a learning model (also called a "model" or a "generative model") that outputs a sentence (character string).

[0012] [1-1. Configuration of information processing system] Prior to describing the process shown in FIG. 1, the configuration of the information processing system 1 will be described with reference to FIG. 3. FIG. 3 is a diagram showing a configuration example of the information processing system according to the embodiment. As shown in FIG. 3, the information processing system 1 includes a terminal device 10, a user terminal 20, an information providing device 50, and an information processing device 100. The terminal device 10, the user terminal 20, the information providing device 50, and the information processing device 100 are connected to each other via a predetermined network N so as to be communicable with each other by wire or wirelessly. Note that the information processing system 1 shown in FIG. 3 may include a plurality of terminal devices 10, a plurality of user terminals 20, a plurality of information providing devices 50, and a plurality of information processing devices 100.

[0013] The information processing device 100 is a computer that learns a model (generative model) that receives character information as input information and outputs a sentence corresponding to the input information as a second advertising copy, based on information related to the first advertising copy. For example, the information processing device 100 learns a model that receives the first advertising copy as input information and outputs a sentence corresponding to the input information as a second advertising copy. Note that the model may be a model that uses any information related to the first advertising copy as input information, as long as the information processing device 100 is capable of learning. For example, the model may be a model that uses a part of the information in the first advertising copy. For example, the model may be a model that uses a characteristic keyword (character string) included in the first advertising copy as input information.

[0014] The information processing device 100 generates a second advertising copy using the trained model. For example, the information processing device 100 generates a second advertising copy corresponding to a first advertising copy (specified advertising copy) specified by an advertiser using the trained model. For example, the information processing device 100 inputs the first advertising copy specified by an advertiser into the model, and determines an advertising copy (also called a "candidate advertising copy") that is a candidate to be used for an advertisement, based on the second advertising copy output by the model.

[0015] The terminal device 10 is a computer (advertiser device) used by an advertiser. The advertiser may be a corporation or the like, and the user of the terminal device 10 may be, for example, a publicity officer of the advertiser. For example, the terminal device 10 is realized by a smartphone, a tablet terminal, a notebook PC, a desktop PC, a mobile phone, a PDA, or the like. The example in FIG. 1 shows a case where the terminal device 10 is a notebook PC as an example. The advertiser uses the terminal device 10 to transmit and receive information to and from the information processing device 100.

[0016] The terminal device 10 receives information provided from the information processing device 100 and transmits information to the information processing device 100. For example, the terminal device 10 receives content for accepting a designation by an advertiser (also referred to as "designation content") from the information processing device 100. For example, the terminal device 10 receives designation content for accepting a designation of a first advertising copy by an advertiser from the information processing device 100. For example, the terminal device 10 transmits information indicating a first advertising copy (designated advertising copy) designated by the advertiser by the designation content to the information processing device 100. The terminal device 10 transmits information indicating the first advertising copy designated by the advertiser to the information processing device 100. For example, the terminal device 10 transmits information indicating the first advertising copy designated by the advertiser to the information processing device 100.

[0017] In addition, an advertiser may use the terminal device 10 to request an agency to submit advertisement content (hereinafter also simply referred to as "advertisement") without submitting the advertisement to the information processing device 100. In this case, it is the agency that submits the advertisement to the information processing device 100. The term "advertiser" is a concept that includes not only advertisers but also agencies, and the term "advertiser device" is a concept that includes not only advertiser devices but also agency devices used by agencies.

[0018] The user terminal 20 is a computer used by a user. The user terminal 20 is a device (terminal device) that can be carried by a user. The user terminal 20 is realized, for example, by a smartphone, a tablet terminal, a notebook PC (Personal Computer), a mobile phone, a PDA (Personal Digital Assistant), or the like. FIG. 1 shows a case where the user terminal 20 is a smartphone. Note that, hereinafter, the user terminal 20 may be referred to as a user. That is, hereinafter, a user can also be read as a user terminal 20. For example, the user terminal 20 is a computer that displays advertisements provided by the information processing device 100. For example, the user terminal 20 displays various information such as advertisements by a display application.

[0019] The information providing device 50 is a computer in which information for providing various information related to a service is stored. For example, the information providing device 50 collects information from various external devices such as a web server, and provides the collected information to the information processing device 100. For example, the information providing device 50 is realized by, for example, a server device or a cloud system. For example, the information providing device 50 provides, as various types of information, information related to the service (service information) and information related to a user who uses the service (user information).

[0020] For example, the information providing device 50 may provide a service of providing data used for learning (also referred to as "learning data") to the information processing device 100. In this case, the information providing device 50 provides the information processing device 100 with data (learning data) used by the information processing device 100 to learn a model. For example, the information providing device 50 provides the information processing device 100 with data (learning data) used by the information processing device 100 to learn a generative model. The information providing device 50 may also provide a service of providing the information processing device 100 with a trained model. For example, the information providing device 50 provides the information processing device 100 with a model (for example, a prediction model M2) used to predict an advertising index and a model (for example, an estimation model M3) that estimates language-likeness (sentence-likeness) to the information processing device 100.

[0021] [1-2. Overview of processing in information processing systems] From here, an example of information processing performed by the information processing system 1 will be described with reference to FIG. 1. Note that detailed description of the information processing performed in the information processing system 1 that is similar to conventional processing will be omitted as appropriate. In FIG. 1, CTR (click-through rate) will be described as an example of an advertising index, but the advertising index is not limited to CTR and any index can be adopted. For example, the advertising index can be any index related to clicks, such as the number of clicks (number of clicks) other than CTR. Also, the advertising index can be any index related to conversion, such as the conversion rate and the number of times conversion was achieved, other than the index related to clicks. Note that the conversion here may be any behavior of the user. For example, the conversion may be a user behavior that leads to the purchase of a product (use of a service) or the benefit of member registration for the target of the advertisement. Note that purchase (use), member registration, etc. are only examples, and the conversion may be a request for information materials, viewing (display) of a site corresponding to the advertisement, visiting a physical store, etc. In this way, any behavior of the user that is considered as a conversion can be set arbitrarily. For example, what kind of user behavior is regarded as a conversion for each advertisement may be individually set by an administrator of the information processing device 100, an advertiser, or the like. Note that the above advertising index is merely an example, and any index can be adopted as the advertising index.

[0022] In FIG. 1, the information processing device 100 acquires information used to execute a learning process of a model M1 that generates advertising copy from the information providing device 50 (step S1). The information processing device 100 acquires data (learning data) used in the learning process of the model M1 from the information providing device 50. For example, the information processing device 100 acquires a first advertising copy that is an advertising copy that satisfies a predetermined condition from the information providing device 50. For example, the information processing device 100 acquires an advertising copy whose CTR is equal to or greater than a predetermined threshold as the first advertising copy from the information providing device 50. Note that, if the information processing device 100 has already acquired the learning data, it is not necessary to perform the process of step S1.

[0023] The information processing device 100 executes a learning process to learn a model (step S2). The information processing device 100 stores the information received from the information providing device 50 in the storage unit 120. The information processing device 100 stores the learning data used in the learning process of the model M1 acquired from the information providing device 50 in the model information storage unit 122 (see FIG. 4) in association with the model M1. The information processing device 100 learns the model M1 using the learning data. The information processing device 100 learns the model M1 that receives, as input information, a first advertising copy with a CTR equal to or greater than a predetermined threshold, and outputs, as a second advertising copy corresponding to the input information.

[0024] The information processing device 100 learns the model M1 using a technique related to transfer learning. For example, the information processing device 100 learns the model M1 based on a model that has been learned in advance (also referred to as a "pre-learning model"). For example, the information processing device 100 learns the model M1 using an open source model as the pre-learning model. For example, the information processing device 100 learns the model M1 using a model that outputs a character string corresponding to an input character string as the pre-learning model.

[0025] For example, the information processing device 100 learns a model M1 based on a pre-learning model trained based on any method such as the method disclosed in Non-Patent Document 1 (GPT (Generative Pre-trained Transformer)). In FIG. 1, the information processing device 100 learns a model M1 based on a pre-learning model trained by GPT-2 (Generative Pre-trained Transformer-2). Note that GPT-2 is merely an example, and the information processing device 100 may use a pre-learning model trained by any method related to natural language processing. For example, the information processing device 100 may use a pre-learning model trained by a method such as LSTM (Long short-term memory), Transformer, BERT (Bidirectional Encoder Representations from Transformers), GPT-3 (Generative Pre-trained Transformer-3), etc.

[0026] The information processing device 100 trains a model M1 by fine-tuning a pre-trained pre-learning model. The information processing device 100 re-trains the weights of the entire model by fine-tuning using the advertising copy, using the weights of the pre-learning model (network) as initial values, which is a model that outputs a character string corresponding to an input character string. This allows the information processing device 100 to generate a model M1 that has trained the features of the advertising copy.

[0027] 1, the information processing device 100 fine-tunes a pre-learning model that has been trained in advance, using a first advertising copy whose CTR is equal to or greater than a predetermined threshold, to train a model M1. For example, the information processing device 100 fine-tunes a model M1, using a first advertising copy whose CTR is equal to or greater than a predetermined threshold.

[0028] That is, the information processing device 100 performs fine tuning using a first advertising copy having a CTR equal to or greater than a predetermined threshold, thereby learning a model M1 that outputs a second advertising copy corresponding to the first advertising copy. The information processing device 100 generates a model M1 that has learned the features of the first advertising copy. In this manner, the information processing device 100 performs fine tuning using a first advertising copy having a CTR equal to or greater than a predetermined threshold, thereby learning the features of the first advertising copy having a high CTR, and learning a model M1 that outputs a second advertising copy that is expected to have a high CTR like the first advertising copy.

[0029] The above-described learning process is merely an example, and the information processing device 100 may learn the model M1 by any method as long as the model M1 is able to learn the model M1 that receives the first advertising copy as input information and outputs a copy corresponding to the first advertising copy as the second advertising copy. For example, the information processing device 100 may generate the model M1 without using a pre-learning model.

[0030] The information processing device 100 generates a new advertisement copy using the trained model M1. This will be described below. In FIG. 1, the information processing device 100 receives a request for an advertisement copy proposal from an advertiser (step S3). For example, the information processing device 100 receives information indicating an advertisement copy (specified advertisement copy) specified by the advertiser from the terminal device 10 used by the advertisement copy, thereby accepting a request for an advertisement copy proposal from the terminal device 10. In FIG. 1, the information processing device 100 receives information indicating a first advertisement copy FA, ​​which is a character string "XXXXXX", from the terminal device 10 used by the advertisement copy. Note that, although the character strings "XXXXXX", "YYYYYY", and "ZZZZZZ" are shown as abstract character strings in FIG. 1, each character string is assumed to be a character string indicating a specific content such as "carefully selected popular nursing care worker job sites".

[0031] Then, the information processing device 100 executes a generation process for generating a new advertising copy (step S4). The information processing device 100 executes a generation process for generating a second advertising copy corresponding to the first advertising copy FA. In FIG. 1, the information processing device 100 generates second advertising copies SA1, SA2, etc. corresponding to the first advertising copy FA. Note that when the second advertising copies SA1, SA2, etc. are described without distinction, they are referred to as "second advertising copy SA."

[0032] First, the information processing device 100 inputs input information including the first advertising copy FA to the model M1 (step S4-1). In FIG. 1, the information processing device 100 inputs the first advertising copy FA as input information to the model M1.

[0033] The model M1 to which the input information is input outputs the second advertising copy (step S4-2). In FIG. 1, the model M1 to which the first advertising copy FA is input outputs advertising copies as shown in the second advertising copies SA1, SA2, etc. In this way, the information processing device 100 generates the second advertising copies SA1, SA2, etc. corresponding to the first advertising copy FA by using the model M1.

[0034] Then, the information processing device 100 provides the advertiser with information based on the advertisement copy generated by the generation process (step S5). For example, the information processing device 100 transmits a proposed advertisement copy based on the second advertisement copies SA1, SA2, etc. generated by the generation process to the terminal device 10 used by the advertiser. An example of the proposed advertisement copy, etc. will be described with reference to FIG. 2.

[0035] In this way, the information processing device 100 can generate appropriate advertising copy by learning a learning model that inputs information on a first advertising copy that satisfies a predetermined condition as input information and outputs a copy corresponding to the input information as a second advertising copy. Moreover, the information processing device 100 can generate appropriate advertising copy by using the learned model. Moreover, the information processing device 100 can propose an appropriate advertising copy to an advertiser by proposing a proposed advertising copy based on information generated by the generation process to the advertiser.

[0036] [1-2-1. Processing example] On the premise of the information processing performed by the information processing system 1 described above, an example of an advertisement copy generated by the information processing device 100 will be described with reference to Fig. 2. Fig. 2 is a diagram showing a specific example of information processing. The generation of the second advertisement copy shown in Fig. 2 corresponds to the processing of step S4 in Fig. 1. Note that the description of the same points as those described above will be omitted as appropriate.

[0037] In Fig. 2, the information processing device 100 generates a second advertising copy corresponding to the first advertising copy AC1, which is "carefully selected popular nursing care worker job sites" (step S11). The information processing device 100 generates a second advertising copy as shown in the advertising information group INF1 by inputting "carefully selected popular nursing care worker job sites" into the model M1. The information processing device 100 generates second advertising copies such as "carefully selected recommended nursing care worker job sites," "carefully selected day service nursing care worker job sites," and "carefully selected job sites for inexperienced nursing care workers," by inputting "carefully selected popular nursing care worker job sites" into the model M1.

[0038] Furthermore, the information processing device 100 predicts the predicted CTR of each second advertising copy using the prediction model M2. For example, in response to an input of a character string, the prediction model M2 outputs a predicted CTR value (0 to 100 (%)) when the character string is used as an advertisement. The information processing device 100 predicts the predicted CTR of the second advertising copy by inputting the second advertising copy into the prediction model M2.

[0039] As shown in the advertisement information group INF1, the information processing device 100 predicts the predicted CTR of second advertisement copy such as "carefully selected recommended job sites for care workers," "carefully selected job sites for day service care workers," and "carefully selected job sites for inexperienced care workers." For example, the information processing device 100 predicts the predicted CTR of "carefully selected recommended job sites for care workers" to be "2.46(%)" by inputting "carefully selected recommended job sites for care workers" into the prediction model M2.

[0040] Furthermore, the information processing device 100 estimates the language-likeness of each second advertising copy using the estimation model M3. For example, the estimation model M3 outputs a value (0 to 1) of the language-likeness of a character string in response to an input of the character string. The information processing device 100 estimates the language-likeness of the second advertising copy by inputting the second advertising copy to the estimation model M3.

[0041] As shown in the advertisement information group INF1, the information processing device 100 estimates the language-likeness of the second advertisement copy such as "carefully selected recommended job sites for care workers," "carefully selected job sites for day service care workers," and "carefully selected job sites for inexperienced care workers." For example, the information processing device 100 inputs "carefully selected recommended job sites for care workers" into the estimation model M3, thereby estimating the language-likeness of "carefully selected recommended job sites for care workers" to be "0.907."

[0042] The information processing device 100 may provide information to the advertiser based on the advertising information group INF1. For example, the information processing device 100 provides a proposed advertising copy based on the second advertising copy generated by the model M1 to the terminal device 10 used by the advertiser (target advertiser) to which the information is to be provided. The information processing device 100 provides the second advertising copy generated by the model M1 as a proposed advertising copy to the terminal device 10 used by the target advertiser. In FIG. 2, the information processing device 100 transmits a second advertising copy determined from a plurality of second advertising copies such as "carefully selected recommended care worker job sites," "carefully selected day service care worker job sites," and "carefully selected inexperienced care worker job sites" as a proposed advertising copy to the terminal device 10 used by the target advertiser.

[0043] The information processing device 100 determines an advertisement candidate copy to be used for an advertisement from the second advertisement copy generated by the model M1. The information processing device 100 provides a proposed advertisement copy based on the determined advertisement candidate copy to the terminal device 10 used by the target advertiser. For example, the information processing device 100 transmits the advertisement candidate copy determined from the second advertisement copy generated by the model M1 to the terminal device 10 as a proposed advertisement copy.

[0044] In FIG. 2, the information processing device 100 determines a candidate advertisement copy from among a number of second advertisement copies, such as “carefully selected recommended job sites for care workers,” “carefully selected job sites for day service care workers,” and “carefully selected job sites for care workers with no experience,” based on a predetermined criterion.

[0045] For example, the information processing device 100 determines the second advertisement copy whose advertisement index satisfies a predetermined condition as the advertisement candidate copy. The information processing device 100 determines the second advertisement copy whose predicted CTR is equal to or greater than a first threshold as the advertisement candidate copy. In FIG. 2, the information processing device 100 determines the second advertisement copy whose predicted CTR is equal to or greater than a first threshold as the advertisement candidate copy among a plurality of second advertisement copies such as "carefully selected recommended care worker job sites", "carefully selected day service care worker job sites", and "carefully selected inexperienced care worker job sites". When the first threshold is "3", the information processing device 100 determines the second advertisement copy whose predicted CTR is equal to or greater than "3", such as "carefully selected inexperienced care worker job sites", as the advertisement candidate copy. In this case, the information processing device 100 transmits the second advertisement copy such as "carefully selected inexperienced care worker job sites" to the terminal device 10 as the proposed advertisement copy.

[0046] For example, the information processing device 100 determines the second advertisement copy that satisfies the condition of language-likeness (sentence-likeness) as the advertisement candidate copy. The information processing device 100 determines the second advertisement copy whose value indicating language-likeness is equal to or greater than the second threshold as the advertisement candidate copy. In FIG. 2, the information processing device 100 determines the second advertisement copy whose value indicating language-likeness is equal to or greater than the second threshold as the advertisement candidate copy among a plurality of second advertisement copies such as "carefully selected recommended care worker job sites", "carefully selected day service care worker job sites", and "carefully selected inexperienced care worker job sites". When the second threshold is "0.906", the information processing device 100 determines the second advertisement copy whose value indicating language-likeness is equal to or greater than "0.906", such as "carefully selected recommended care worker job sites", as the advertisement candidate copy. In this case, the information processing device 100 transmits the second advertisement copy whose value indicating language-likeness is equal to or greater than the second threshold to the terminal device 10 as the proposed advertisement copy.

[0047] For example, the information processing device 100 determines, as an advertisement candidate text, a second advertisement copy whose advertisement index satisfies a predetermined condition and whose language-likeness (sentence-likeness) condition is satisfied. The information processing device 100 determines, as an advertisement candidate text, a second advertisement copy whose predicted CTR is equal to or greater than a first threshold and whose value indicating language-likeness is equal to or greater than a second threshold.

[0048] In FIG. 2, the information processing device 100 determines, as an advertisement candidate copy, a second advertisement copy having a predicted CTR of a first threshold or more and a value indicating a language-likeness of a second threshold or more among a plurality of second advertisement copies such as "Carefully selected recommended care worker job sites", "Carefully selected day service care worker job sites", and "Carefully selected inexperienced care worker job sites". When the first threshold is "2.5" and the second threshold is "0.905", the information processing device 100 determines, as an advertisement candidate copy, a second advertisement copy such as "Carefully selected day service care worker job sites" having a predicted CTR of "2.5" or more and a value indicating a language-likeness of "0.905" or more. In this case, the information processing device 100 transmits the second advertisement copy such as "Carefully selected day service care worker job sites" to the terminal device 10 as a suggested advertisement copy.

[0049] As described above, the information processing device 100 automatically generates a sentence that sounds like an advertisement copy. The information processing device 100 learns a model using a technique related to transfer learning. For example, the information processing device 100 is a learning device that acquires an advertisement that satisfies a predetermined condition (high click-through rate, high conversion rate, etc.) and makes a model that generates a new sentence from a character string such as an input sentence learn the features of the acquired advertisement. For example, the information processing device 100 acquires an advertisement that satisfies a predetermined condition. For example, the information processing device 100 makes a model (GPT, etc.) that generates a new sentence from an input sentence learn a sentence included in the acquired advertisement as learning data. For example, the information processing device 100 may output a sentence included in an advertisement belonging to a certain product category as learning data. In addition, the information processing device 100 may receive a simple sentence that is about a product belonging to the same category and indicates a content to be emphasized or a product as an input sentence. The information processing device 100 learns a model so as to output each output sentence from the input sentence.

[0050] The information processing device 100 may use a model generated for each product category. The information processing device 100 receives input text and product category information from an advertiser. The information processing device 100 may input the input text to a model corresponding to the received product category and generate a plurality of output texts. The information processing device 100 obtains a score for each output text using, for example, a prediction model M2, an estimation model M3, and a trained advertising text evaluation model (evaluation model). The information processing device 100 arranges the output texts in order of score and displays them together with the scores. For example, the evaluation model may be a model trained to output a higher score when the advertising text of an actually delivered advertisement is input, as the click-through rate or the conversion rate is higher. The above-mentioned process is merely an example, and the information processing device 100 may appropriately use various information to determine an advertising copy to be proposed to an advertiser and make a proposal to the advertiser.

[0051] 2. Configuration of Information Processing Device Next, the configuration of the information processing device 100 according to the embodiment will be described with reference to Fig. 4. Fig. 4 is a diagram showing an example of the configuration of the information processing device 100 according to the embodiment. As shown in Fig. 4, the information processing device 100 has a communication unit 110, a storage unit 120, and a control unit 130. Note that the information processing device 100 may have an input unit (e.g., a keyboard, a mouse, etc.) that accepts various operations from an administrator of the information processing device 100, and a display unit (e.g., a liquid crystal display, etc.) that displays various information.

[0052] (Communication unit 110) The communication unit 110 is realized by, for example, a network interface card (NIC) etc. The communication unit 110 is connected to a network by wire or wirelessly, and transmits and receives information between the terminal device 10, the user terminal 20, and the information providing device 50.

[0053] (Storage unit 120) The storage unit 120 is realized by, for example, a semiconductor memory element such as a random access memory (RAM) or a flash memory, or a storage device such as a hard disk or an optical disk. As shown in Fig. 4, the storage unit 120 according to the embodiment has an advertisement information storage unit 121, a model information storage unit 122, and a generation information storage unit 123. The storage unit 120 may store various kinds of information, not limited to the above.

[0054] (Advertisement information storage unit 121) The advertisement information storage unit 121 according to the embodiment stores various information related to advertisements. For example, the advertisement information storage unit 121 stores various information related to advertisements whose delivery results have been collected. FIG. 5 is a diagram illustrating an example of the advertisement information storage unit according to the embodiment. The advertisement information storage unit 121 illustrated in FIG. 5 has items such as "advertising ID", "advertisement", "advertiser", "advertising target", and "advertising index".

[0055] "Advertisement ID" indicates identification information for identifying an advertisement. "Advertisement" indicates an advertisement submitted by an advertiser. While FIG. 5 shows an example in which conceptual information such as "AC1" is stored in "Advertisement," in reality, text used as the advertisement or a file path name indicating the storage location of such text is stored. Note that "Advertisement" may include images, or may be data such as video and audio, so long as it includes advertising text.

[0056] "Advertiser" refers to the advertiser corresponding to the advertisement. For example, "advertiser" refers to identification information (advertiser ID) for identifying the advertiser. Note that an advertiser may request an agency to perform tasks such as submitting advertisements. The term "advertiser" is intended to be a concept that includes not only advertisers but also agencies. "Advertising target" refers to the advertising target or service name that is the subject of the advertisement. Figure 5 shows an example in which conceptual information such as "GD1" is stored in "advertising target", but in reality, the specific name of the advertising target is stored. Note that the information for identifying each product is not limited to "advertising target", and information such as "advertising target ID" for identifying the advertising target may also be stored.

[0057] "Advertising index" indicates an advertising index corresponding to an advertisement. In FIG. 5, an example is shown in which conceptual information such as "IND1" is stored in "advertising index", but in reality, advertising indexes based on the delivery performance collected for the advertisement, or file path names indicating the storage locations of these, etc. are stored. For example, "advertising index" indicates the performance values ​​of various advertising indexes such as CTR and CVR corresponding to the advertisement.

[0058] 5, the advertisement (advertisement AC1) identified by the advertisement ID "A11" is an advertisement obtained from an advertiser identified by the advertiser "AP1", and the advertisement target is the advertisement target GD1. Also, the advertisement indicator of the advertisement AC1 is the advertisement indicator IND1.

[0059] The advertisement information storage unit 121 may store various information according to the purpose, not limited to the above. For example, the advertisement information storage unit 121 may store information on bidding prices, etc. For example, the advertisement information storage unit 121 may store information on the attributes of users targeted by each advertisement, etc.

[0060] (Model information storage unit 122) The model information storage unit 122 according to the embodiment stores information about the model. For example, the model information storage unit 122 stores information (model data) of a trained model (model) trained (generated) by a learning process. The model information storage unit 122 shown in FIG. 6 stores data used in learning (trained data) in association with the trained model (model). FIG. 6 is a diagram showing an example of a model information storage unit according to the embodiment. In the example shown in FIG. 6, the model information storage unit 122 includes items such as "model ID", "purpose", "model data", and "trained data". In the example of FIG. 6, the model information storage unit 122 stores data used in learning (trained data) in association with the trained model (model).

[0061] "Model ID" indicates identification information for identifying a model. "Use" indicates the use of the corresponding model. "Model data" indicates the data of the model. Figure 6 etc. shows an example in which conceptual information such as "MDT1" is stored in "model data", but in reality, various information constituting the model is included, such as information on the model configuration (network configuration) and information on parameters. For example, "model data" includes information including the nodes in each layer of the network, the functions employed by each node, the connection relationships of the nodes, and the connection coefficients set for the connections between the nodes.

[0062] "Learning data" refers to data used to train a trained model (model). In the "learning data", information indicating a data set used to train the corresponding model is stored. For example, the "learning data" stores data (input information) and answer information (output information) corresponding to the data as learning data (also called "data for learning"). FIG. 6 shows an example in which conceptual information such as "LDT1" is stored in the "learning data", but in reality, various information related to the data used to train the corresponding model, such as data (input information) and answer information (output information) corresponding to the data, is included.

[0063] In FIG. 6, the model (model M1) identified by the model ID "M1" indicates that its use is "advertising copy generation." In other words, model M1 indicates that it is a model that outputs (generates) a new character string (advertising copy) based on an input character string (advertising copy). Also, it indicates that the model data of model M1 is model data MDT1. Also, it indicates that the learning data used to learn model M1 is learning data LDT1.

[0064] Also, the model (prediction model M2) identified by the model ID "M2" indicates that its use is "CTR prediction." In other words, the prediction model M2 indicates that it is a model that outputs information (score) indicating a predicted CTR for input advertising data. For example, the prediction model M2 outputs a value from 0 to 100 (%). For example, the prediction model M2 outputs a predicted value of CTR corresponding to an input advertising copy. It indicates that the model data of the prediction model M2 is model data MDT2. It also indicates that the learning data used to train the prediction model M2 is learning data LDT2. The learning data LDT2 includes a plurality of training data in which the CTR value (actual value, etc.) of the advertising copy is associated with the advertising copy as correct answer information.

[0065] The model (estimation model M3) identified by the model ID "M3" indicates that its use is "language-likeness." In other words, the estimation model M3 indicates that it is a model that outputs information (score) indicating the degree to which an input character string (sentence) is appropriate as a natural language. For example, the estimation model M3 outputs a value from 0 to 1. For example, the estimation model M3 outputs a higher score the more appropriate the input advertising copy is as a natural language. Also, it indicates that the model data of the estimation model M3 is model data MDT3. Also, it indicates that the learning data used to train the estimation model M3 is learning data LDT3. The learning data LDT3 includes a plurality of training data in which a value indicating the language-likeness of the advertising copy is associated with the advertising copy as correct answer information.

[0066] The model information storage unit 122 may store various information according to the purpose, not limited to the above. For example, the model information storage unit 122 may store a model predicting other advertising indicators, not limited to CTR. For example, the model information storage unit 122 may store a model M4 predicting CVR. In this case, the model information storage unit 122 stores the model M4 predicting CVR in association with the data used for learning the model (a combination of the CVR, which is correct information, and the advertising copy from which the actual CVR was obtained).

[0067] (Generated information storage unit 123) The generated information storage unit 123 according to the embodiment stores various information related to the generated advertising copy. Fig. 7 is a diagram showing an example of the generated information storage unit according to the embodiment. The generated information storage unit 123 shown in Fig. 7 includes items such as "processing result ID", "input information", and "generated information".

[0068] "Processing result ID" indicates identification information for identifying the processing result. "Input information" indicates the input information used to generate the information. "Input information" includes items such as "First advertising copy".

[0069] "Generation information" indicates various results (information) obtained by the generation process. "Generation information" includes items such as "second ad copy," "estimated CTR," and "language-likeness." "Second ad copy" indicates ad copy generated based on the first ad copy. For example, "second ad copy" indicates ad copy generated using model M1.

[0070] "Predicted CTR" indicates a CTR predicted for the corresponding second advertising copy. For example, "predicted CTR" indicates a CTR predicted using the prediction model M2. For example, "predicted CTR" indicates a predicted CTR output by the prediction model M2 to which the second advertising copy is input.

[0071] "Language-likeness" indicates a predicted CTR for the corresponding second advertising copy. For example, "language-likeness" indicates a language-likeness estimated using the estimation model M3. For example, "predicted CTR" indicates a degree of language-likeness of the second advertising copy output by the estimation model M3 to which the second advertising copy is input.

[0072] For example, in the process identified by the process result ID "RS1", the first ad copy is "Carefully selected popular care worker job sites". It also shows that second ad copies generated from the first ad copy are "Carefully selected recommended care worker job sites", "Carefully selected day care worker job sites", etc. The second ad copy "Carefully selected recommended care worker job sites" has a predicted CTR of "2.46" and a language-likeness of "0.907".

[0073] Note that the generated information storage unit 123 is not limited to the above, and may store various kinds of information depending on the purpose.

[0074] (Control unit 130) Returning to the explanation of Fig. 4, the control unit 130 is a controller, and is realized, for example, by a CPU (Central Processing Unit), MPU (Micro Processing Unit), etc., executing various programs (corresponding to an example of an information processing program) stored in a storage device inside the information processing device 100 using a RAM as a working area. The control unit 130 is also a controller, and is realized, for example, by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

[0075] 4, the control unit 130 has an acquisition unit 131, a learning unit 132, a determination unit 133, a generation unit 134, and a provision unit 135, and realizes or executes the functions and actions of information processing described below. Note that the internal configuration of the control unit 130 is not limited to the configuration shown in FIG. 4, and may be other configurations as long as they perform the information processing described below.

[0076] (Acquisition part 131) The acquisition unit 131 acquires various information. The acquisition unit 131 acquires various information from external devices such as the terminal device 10, the user terminal 20, and the information providing device 50. The acquisition unit 131 also acquires various information from the storage unit 120. The acquisition unit 131 also acquires various information from the advertisement information storage unit 121, the model information storage unit 122, the generation information storage unit 123, and the like.

[0077] The acquisition unit 131 receives various information from an external information processing device via the communication unit 110. The acquisition unit 131 stores the received information in the storage unit 120. The acquisition unit 131 receives various information from the terminal device 10, the user terminal 20, or the information providing device 50. The acquisition unit 131 acquires information from the terminal device 10 used by the user. For example, the acquisition unit 131 acquires various information from the information providing device 50. For example, the acquisition unit 131 receives learning data used for learning a model from the information providing device 50. For example, the acquisition unit 131 receives a model from the information providing device 50.

[0078] The acquisition unit 131 acquires information about a first advertising copy that is an advertising copy that satisfies a predetermined condition. The acquisition unit 131 acquires information about a first advertising copy whose advertising indicator satisfies a predetermined condition. The acquisition unit 131 acquires information about a first advertising copy whose advertising indicator related to clicks satisfies a predetermined condition. The acquisition unit 131 acquires information about a first advertising copy whose advertising indicator related to the number of times the first advertising copy is clicked by a user satisfies a predetermined condition. The acquisition unit 131 acquires information about a first advertising copy whose click rate is equal to or greater than a predetermined threshold.

[0079] The acquisition unit 131 acquires information about a first advertising copy in which an advertising index related to conversion satisfies a predetermined condition. The acquisition unit 131 acquires information about a first advertising copy in which an advertising index related to the number of times the first advertising copy has led to conversion satisfies a predetermined condition. The acquisition unit 131 acquires information about a first advertising copy in which a conversion rate is equal to or greater than a predetermined threshold.

[0080] (Learning Section 132) The learning unit 132 executes a learning process to learn a learning model (model). The learning unit 132 learns a learning model that receives character information as input information and outputs a sentence corresponding to the input information as a second advertising copy, based on information related to the first advertising copy acquired by the acquisition unit 131.

[0081] The learning unit 132 learns the learning model based on information about a first advertising copy whose advertising indicator satisfies a predetermined condition. The learning unit 132 learns the learning model based on information about a first advertising copy whose advertising indicator related to clicks satisfies a predetermined condition. The learning unit 132 learns the learning model based on information about a first advertising copy whose advertising indicator related to the number of times the first advertising copy is clicked by a user satisfies a predetermined condition. The learning unit 132 learns the learning model based on information about a first advertising copy whose click rate is equal to or greater than a predetermined threshold.

[0082] The learning unit 132 learns the learning model based on information about a first advertising copy whose advertising indicator related to conversion satisfies a predetermined condition. The learning unit 132 learns the learning model based on information about a first advertising copy whose advertising indicator related to the number of times the first advertising copy led to conversion satisfies a predetermined condition. The learning unit 132 learns the learning model based on information about a first advertising copy whose conversion rate is equal to or greater than a predetermined threshold.

[0083] The learning unit 132 receives input information, which is a character string included in the first advertising copy, and learns a learning model that outputs a second advertising copy corresponding to the input information. The learning unit 132 receives input information, which is the first advertising copy, and learns a learning model that outputs a second advertising copy corresponding to the input information. The learning unit 132 receives input information, which is a character string that is a part of the first advertising copy, and learns a learning model that outputs a second advertising copy corresponding to the input information. The learning unit 132 generates a learning model that has learned the characteristics of the first advertising copy. The learning unit 132 generates a learning model that has learned the characteristics of the association between the input information and the second advertising copy.

[0084] For example, the learning unit 132 executes the learning process based on various information acquired by the acquisition unit 131. The learning unit 132 executes the learning process based on information from an external information processing device and information stored in the storage unit 120. The learning unit 132 executes the learning process based on information stored in the model information storage unit 122. The learning unit 132 stores a model generated by learning in the model information storage unit 122.

[0085] The learning unit 132 performs a learning process. The learning unit 132 performs various learning. The learning unit 132 learns various pieces of information based on the information acquired by the acquisition unit 131. The learning unit 132 learns (generates) a model. The learning unit 132 learns various pieces of information such as a model. The learning unit 132 generates a model by learning. The learning unit 132 learns the model using various machine learning techniques. For example, the learning unit 132 learns parameters of the model (network). The learning unit 132 learns the model using various machine learning techniques.

[0086] The learning unit 132 generates various learning models such as the model M1. The learning unit 132 learns network parameters. For example, the learning unit 132 learns network parameters of various learning models such as the model M1. The learning unit 132 performs a learning process using the learning data stored in the model information storage unit 122, thereby generating various learning models such as the model M1. For example, the learning unit 132 generates a model used for generating advertising copy. The learning unit 132 generates various learning models such as the model M1 by learning network parameters of various learning models such as the model M1.

[0087] The learning unit 132 performs a learning process based on the learning data (teacher data) stored in the model information storage unit 122. The learning unit 132 performs a learning process using the learning data stored in the model information storage unit 122, thereby generating various learning models such as model M1.

[0088] For example, the learning unit 132 performs a learning process using a method such as backpropagation so that the advertising copy output by the model M1 approaches correct answer information (information indicating a desired generation result) associated with the input information input to the model M1. For example, the learning unit 132 performs a learning process so that a synthetic image output by the model M1 to which input information including the first advertising copy is input approaches correct answer information associated with the input information.

[0089] For example, the learning unit 132 adjusts the value of the weight (i.e., the connection coefficient) that is taken into consideration when a value is transmitted between nodes by the learning process. In this way, the learning unit 132 learns the model M1 by a process such as backpropagation that corrects the parameters (connection coefficients) so as to reduce the error between the output in the model M1 and the correct answer information corresponding to the input. For example, the learning unit 132 generates the model M1 by a process such as backpropagation that minimizes a predetermined loss function. This allows the learning unit 132 to perform a learning process that learns the parameters of the model M1.

[0090] The learning method of the model is not limited to the above-mentioned method, and any known technology can be applied. The generation of each model may be performed by appropriately using various conventional technologies related to machine learning. For example, the generation of the model may be performed by using a technology related to machine learning of supervised learning, such as SVM (Support Vector Machine). Also, for example, the generation of the model may be performed by using a technology related to machine learning of unsupervised learning. For example, the generation of the model may be performed by using a deep learning technology. For example, the generation of the model may be performed by appropriately using various deep learning technologies such as DNN (Deep Neural Network), RNN (Recurrent Neural Network), and CNN (Convolutional Neural Network). Note that the description regarding the generation of the above model is an example, and the generation of the model may be performed by a learning method appropriately selected according to the obtainable information, etc. That is, the learning unit 132 may generate the model M1 by any method as long as the model M1 can be learned so as to output a score corresponding to the correct answer information when the input information included in the learning data is input.

[0091] As described above, the method of learning by the learning unit 132 is not particularly limited, but for example, learning data in which data (input information) and its correct answer information (output information) are linked may be prepared, and the learning data may be input to a computation model based on a multi-layer neural network for learning. Also, a method based on DNN such as CNN or 3D-CNN may be used. When targeting time-series data such as voice, the learning unit 132 may use a method based on a recurrent neural network (RNN) or LSTM (Long Short-Term Memory units), which is an extension of RNN.

[0092] Furthermore, the learning unit 132 may learn a prediction model M2 that is a model used to predict an advertising index. For example, the learning unit 132 may learn a prediction model M2 that is a model used to predict a CTR. For example, the learning unit 132 learns the prediction model M2 using learning data LDT2 including a plurality of teacher data in which a CTR value (such as an actual value of 0 to 100 (%)) of an advertising copy is associated with the advertising copy as correct answer information. For example, the learning unit 132 learns the prediction model M2 so that correct answer information (value) associated with an advertising copy is output when the advertising copy is input. For example, the learning unit 132 learns a prediction model M2 that outputs a predicted CTR value (0 to 100 (%)) when the character string is used as an advertisement in response to the input of the character string. Note that when the prediction model M2 is acquired from the information providing device 50, the learning unit 132 does not need to learn the prediction model M2.

[0093] Furthermore, the learning unit 132 may learn an estimation model M3 that is a model for estimating language-likeness (sentence-likeness). For example, the learning unit 132 may learn an estimation model M3 that is a model used for estimating the appropriateness of a character string as a natural language. For example, the learning unit 132 learns the estimation model M3 using learning data LDT3 including a plurality of teacher data in which a value of the language-likeness of an advertising copy (such as a value of 0 to 1) of the advertising copy is associated with the advertising copy as correct answer information. For example, the learning unit 132 learns the estimation model M3 so that correct answer information (value) associated with the advertising copy is output when the advertising copy is input. For example, the learning unit 132 learns an estimation model M3 that outputs a value of 0 to 1 in response to an input of a character string. Note that when the estimation model M3 is acquired from the information providing device 50, the learning unit 132 does not need to learn the estimation model M3.

[0094] (Decision unit 133) The determination unit 133 executes a determination process to determine various pieces of information. The determination unit 133 stores the information determined by the determination process in the storage unit 120. For example, the determination unit 133 executes the determination process based on various pieces of information acquired by the acquisition unit 131. The determination unit 133 executes the determination process based on various pieces of information stored in the storage unit 120. For example, the determination unit 133 executes the determination process based on various pieces of information received from an external information processing device.

[0095] The determination unit 133 executes the determination process using the model learned by the learning unit 132. The determination unit 133 executes the determination process using the model stored in the model information storage unit 122.

[0096] The determination unit 133 determines, from the second advertising copy generated by the learning model, an advertising candidate copy that is a candidate for use in an advertisement. The determination unit 133 determines, as the advertising candidate copy, the second advertising copy whose advertising indicator satisfies a predetermined condition. The determination unit 133 determines, as the advertising candidate copy, the second advertising copy whose language-likeness condition is satisfied.

[0097] (Generation unit 134) The generating unit 134 executes a generating process for generating various information. For example, the generating unit 134 executes the generating process based on various information acquired by the acquiring unit 131. For example, the generating unit 134 executes the generating process based on information stored in the storage unit 120.

[0098] The generating unit 134 executes a generating process based on the information determined by the determining unit 133. For example, the generating unit 134 generates content that proposes to an advertiser the advertisement candidate sentence determined by the determining unit 133. The generating unit 134 generates designation content.

[0099] The generating unit 134 generates content. For example, the generating unit 134 generates a screen (content) to be provided to the terminal device 10 by appropriately using various technologies such as Java (registered trademark). The generating unit 134 may generate the screen (content) to be provided to the terminal device 10 based on the format of CSS, JavaScript (registered trademark), or HTML. Also, for example, the generating unit 134 may generate the screen (content) in various formats such as JPEG (Joint Photographic Experts Group), GIF (Graphics Interchange Format), and PNG (Portable Network Graphics). The generating unit 134 generates the evaluation receiving content.

[0100] (Provider 135) The providing unit 135 provides various information. The providing unit 135 transmits various information to an external information processing device via the communication unit 110. The providing unit 135 transmits various information to the terminal device 10, the user terminal 20, and the information providing device 50. The providing unit 135 transmits a result of the determination process to the terminal device 10.

[0101] The providing unit 135 provides the information determined by the determining unit 133. The providing unit 135 transmits the determined advertisement candidate text to the terminal device 10. The providing unit 135 provides the information generated by the generating unit 134. The providing unit 135 transmits the determined advertisement to the user terminal 20.

[0102] The providing unit 135 provides the proposed advertising copy based on the second advertising copy generated by the learning model to the terminal device 10 used by the target advertiser, who is the advertiser to which information is to be provided. The providing unit 135 provides the second advertising copy generated by the learning model as a proposed advertising copy to the terminal device 10 used by the target advertiser. The providing unit 135 provides the proposed advertising copy based on the advertising candidate copy determined by the determining unit 133 to the terminal device 10 used by the target advertiser.

[0103] 3. Processing Procedure Next, the procedure of information processing executed by the information processing device 100 according to the embodiment will be described with reference to Fig. 8. Fig. 8 is a flowchart showing an example of the flow of information processing executed by the information processing device.

[0104] 8, the information processing device 100 acquires information on a first advertising copy that satisfies a predetermined condition (step S101). For example, the information processing device 100 acquires a first advertising copy that has a CTR equal to or greater than a predetermined value.

[0105] Furthermore, the information processing device 100 learns a learning model that receives character information as input information and outputs a sentence corresponding to the input information as a second advertising copy, based on information related to the first advertising copy (step S102). For example, the information processing device 100 learns a model M1 that receives the first advertising copy as input information and outputs a sentence corresponding to the first advertising copy as a second advertising copy, based on the first advertising copy, which is an advertising copy with a CTR equal to or greater than a predetermined value.

[0106] 4. Effects As described above, the information processing device 100 according to the embodiment includes the acquisition unit 131 and the learning unit 132. The acquisition unit 131 acquires information on a first advertising copy that is an advertising copy that satisfies a predetermined condition. The learning unit 132 learns a learning model that receives character information as input information and outputs a sentence corresponding to the input information as a second advertising copy, based on the information on the first advertising copy acquired by the acquisition unit 131.

[0107] In this way, the information processing device 100 according to the embodiment can enable the generation of appropriate advertising copy by learning a learning model that inputs character information as input information based on information about a first advertising copy that is an advertising copy that satisfies a predetermined condition and outputs a sentence corresponding to the input information as a second advertising copy.

[0108] Moreover, in the information processing device 100 according to the embodiment, the acquisition unit 131 acquires information related to a first advertising copy whose advertising index satisfies a predetermined condition.

[0109] In this manner, the information processing device 100 according to the embodiment can generate appropriate advertising copy by learning based on information on the first advertising copy whose advertising index satisfies a predetermined condition.

[0110] Moreover, in the information processing device 100 according to the embodiment, the acquisition unit 131 acquires information on a first advertising copy in which an advertising index related to a click satisfies a predetermined condition.

[0111] In this manner, the information processing device 100 according to the embodiment can generate appropriate advertising copy by learning based on information on the first advertising copy in which the advertising index related to clicks satisfies a predetermined condition.

[0112] Moreover, in the information processing device 100 according to the embodiment, the acquisition unit 131 acquires information on a first advertising copy in which an advertising indicator related to the number of times the first advertising copy has been clicked by users satisfies a predetermined condition.

[0113] In this way, the information processing device 100 according to the embodiment can generate appropriate advertising copy by learning based on information about a first advertising copy whose advertising indicator related to the number of times the first advertising copy has been clicked by a user satisfies a predetermined condition.

[0114] Moreover, in the information processing device 100 according to the embodiment, the acquisition unit 131 acquires information on a primary advertising copy having a click rate equal to or greater than a predetermined threshold.

[0115] In this manner, the information processing device 100 according to the embodiment can generate appropriate advertising copy by learning based on information on the first advertising copy having a click rate equal to or higher than a predetermined threshold.

[0116] Moreover, in the information processing device 100 according to the embodiment, the acquisition unit 131 acquires information on a first advertising copy in which an advertising index related to a conversion satisfies a predetermined condition.

[0117] In this manner, the information processing device 100 according to the embodiment can generate appropriate advertising copy by learning based on information on the first advertising copy in which the advertising index related to conversion satisfies a predetermined condition.

[0118] Moreover, in the information processing device 100 according to the embodiment, the acquisition unit 131 acquires information on a first advertising copy in which an advertising indicator related to the number of times the first advertising copy led to a conversion satisfies a predetermined condition.

[0119] In this way, the information processing device 100 according to the embodiment can generate appropriate advertising copy by learning based on information about a first advertising copy in which an advertising indicator related to the number of times the first advertising copy led to conversion satisfies a predetermined condition.

[0120] Moreover, in the information processing device 100 according to the embodiment, the acquisition unit 131 acquires information on a first advertising copy whose conversion rate is equal to or greater than a predetermined threshold.

[0121] In this manner, the information processing device 100 according to the embodiment can generate appropriate advertising copy by performing learning based on information on the first advertising copy having a conversion rate equal to or higher than a predetermined threshold.

[0122] Furthermore, in the information processing device 100 according to the embodiment, the learning unit 132 learns a learning model that receives input information, which is a character string included in a first advertising copy, and outputs a second advertising copy corresponding to the input information.

[0123] In this way, the information processing device 100 according to the embodiment can generate appropriate advertising copy by learning a learning model that takes input information, which is a character string included in a first advertising copy, as input and outputs a second advertising copy corresponding to the input information.

[0124] Moreover, in the information processing device 100 according to the embodiment, the learning unit 132 learns a learning model that receives input information that is a first advertising copy and outputs a second advertising copy that corresponds to the input information.

[0125] In this way, the information processing device 100 according to the embodiment can generate appropriate advertising copy by learning a learning model that takes input information, which is a first advertising copy, as input and outputs a second advertising copy corresponding to the input information.

[0126] Furthermore, in the information processing device 100 according to the embodiment, the learning unit 132 learns a learning model that receives input information, which is a character string that is a part of a first advertising copy, and outputs a second advertising copy that corresponds to the input information.

[0127] In this way, the information processing device 100 according to the embodiment can generate appropriate advertising copy by learning a learning model that takes input information, which is a character string of a first advertising copy, as input and outputs a second advertising copy corresponding to the input information.

[0128] Moreover, in the information processing device 100 according to the embodiment, the learning unit 132 generates a learning model that has learned the features of the first advertising copy.

[0129] In this manner, the information processing device 100 according to the embodiment can generate an appropriate advertising copy by generating a learning model that has learned the features of the first advertising copy.

[0130] Moreover, in the information processing device 100 according to the embodiment, the learning unit 132 generates a learning model that learns the characteristics of the association between the input information and the second advertising copy.

[0131] In this manner, the information processing device 100 according to the embodiment can generate an appropriate advertising copy by generating a learning model that has learned the characteristics of the relevance between the input information and the second advertising copy.

[0132] Moreover, the information processing device 100 according to the embodiment includes a providing unit 135. The providing unit 135 provides a proposed advertising copy based on the second advertising copy generated by the learning model to the terminal device 10 used by a target advertiser who is an advertiser to which information is to be provided.

[0133] In this way, the information processing device 100 according to the embodiment can provide the terminal device 10 used by the target advertiser, who is the advertiser to which the information is to be provided, with the proposed advertising copy based on the second advertising copy generated by the learning model, thereby providing the advertiser with appropriately learned composite information.

[0134] Furthermore, in the information processing device 100 according to the embodiment, the providing unit 135 provides the second advertising copy generated by the learning model as a suggested advertising copy to the terminal device 10 used by the target advertiser.

[0135] In this way, the information processing device 100 according to the embodiment can propose an appropriate advertising copy to the advertiser by providing the second advertising copy generated by the learning model as a suggested advertising copy to the terminal device 10 used by the target advertiser.

[0136] Moreover, the information processing device 100 according to the embodiment includes a determination unit 133. The determination unit 133 determines an advertisement candidate copy that is a candidate to be used for an advertisement from the second advertisement copy generated by the learning model. The provision unit 135 provides a proposed advertisement copy based on the advertisement candidate copy determined by the determination unit 133 to the terminal device 10 used by the target advertiser.

[0137] In this way, the information processing device 100 according to the embodiment determines candidate advertisement texts to be used in advertisements from the second advertisement texts generated by the learning model, and provides the terminal device 10 used by the target advertiser with suggested advertisement texts based on the determined candidate advertisement texts, thereby being able to propose appropriate advertisement texts to the advertiser.

[0138] Moreover, in the information processing device 100 according to the embodiment, the determination unit 133 determines the second advertising copy, the advertising indicator of which satisfies a predetermined condition, as the advertising candidate copy.

[0139] In this manner, the information processing device 100 according to the embodiment can determine an appropriate advertisement copy as an advertisement candidate copy by determining the second advertisement copy whose advertisement index satisfies a predetermined condition as the advertisement candidate copy.

[0140] Moreover, in the information processing device 100 according to the embodiment, the determination unit 133 determines the second advertising copy that satisfies the condition of language likeness as the advertisement candidate copy.

[0141] In this manner, the information processing device 100 according to the embodiment can determine an appropriate advertisement copy as the advertisement candidate copy by determining the second advertisement copy that satisfies the condition of language likeness as the advertisement candidate copy.

[0142] [5. Hardware Configuration] The information processing device 100 according to the embodiment described above is realized by a computer 1000 having a configuration as shown in Fig. 9. Fig. 9 is a hardware configuration diagram showing an example of a computer that realizes the functions of the information processing device. The computer 1000 has a CPU 1100, a RAM 1200, a ROM 1300, a HDD (Hard Disk Drive) 1400, a communication interface (I / F) 1500, an input / output interface (I / F) 1600, and a media interface (I / F) 1700.

[0143] The CPU 1100 operates and controls each unit based on a program stored in the ROM 1300 or the HDD 1400. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 is started up, programs that depend on the hardware of the computer 1000, and the like.

[0144] The HDD 1400 stores programs executed by the CPU 1100, data used by such programs, etc. The communication interface 1500 receives data from other devices via the network N and sends the data to the CPU 1100, and transmits data generated by the CPU 1100 to other devices via the network N.

[0145] The CPU 1100 controls output devices such as a display and a printer, and input devices such as a keyboard and a mouse, via the input / output interface 1600. The CPU 1100 acquires data from the input devices via the input / output interface 1600. The CPU 1100 also outputs generated data to the output devices via the input / output interface 1600.

[0146] The media interface 1700 reads a program or data stored in the recording medium 1800 and provides it to the CPU 1100 via the RAM 1200. The CPU 1100 loads the program from the recording medium 1800 onto the RAM 1200 via the media interface 1700 and executes the loaded program. The recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or a PD (Phase change rewritable Disc), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory.

[0147] For example, when the computer 1000 functions as the information processing device 100 according to the embodiment, the CPU 1100 of the computer 1000 executes programs loaded onto the RAM 1200 to realize the functions of the control unit 130. The CPU 1100 of the computer 1000 reads and executes these programs from the recording medium 1800, but as another example, the CPU 1100 may obtain these programs from another device via the network N.

[0148] Although some of the embodiments and variants of the present application have been described in detail above with reference to the drawings, these are merely examples, and the present invention can be implemented in other forms that incorporate various modifications and improvements based on the knowledge of those skilled in the art, including the aspects described in the lines of the disclosure of the invention.

[0149] [6. Other] Furthermore, among the processes described in the above embodiments and modifications, all or part of the processes described as being performed automatically can be performed manually, or all or part of the processes described as being performed manually can be performed automatically by a known method. In addition, the information including the processing procedures, specific names, various data and parameters shown in the above documents and drawings can be changed arbitrarily unless otherwise specified. For example, the various information shown in each drawing is not limited to the illustrated information.

[0150] In addition, each component of each device shown in the figure is a functional concept, and does not necessarily have to be physically configured as shown in the figure. In other words, the specific form of distribution and integration of each device is not limited to that shown in the figure, and all or part of them can be functionally or physically distributed and integrated in any unit according to various loads, usage conditions, etc.

[0151] Furthermore, the above-described embodiments and modifications can be appropriately combined as long as the processing contents are not contradictory.

[0152] Furthermore, the above-mentioned "section, module, unit" can be read as "means" or "circuit", etc. For example, an acquisition section can be read as an acquisition means or an acquisition circuit. [Explanation of symbols]

[0153] 1. Information Processing Systems 100 Information processing device 121 Advertising information storage unit 122 Model information storage unit 123 Generation information storage unit 130 Control section 131 Acquisition Department 132 Learning Department 133 Decision Section 134 Generation part 135 Provision Department 10 Terminal Equipment 20 User terminal 50 Information provision device N Network

Claims

1. An acquisition unit that acquires a first advertising copy that satisfies a predetermined condition; a learning unit that learns a learning model by a learning process based on machine learning, the learning model inputting character information including at least a part of the first advertising copy acquired by the acquisition unit as input information and outputting a sentence corresponding to the input information as a second advertising copy; a generation unit that generates a second advertising copy by inputting a first character information to the learning model trained by the learning unit, and having the learning model output a second advertising copy, and that generates an estimated value of the language-likeness of the second advertising copy by inputting the generated second advertising copy to an estimation model that estimates a language-likeness of a character string, and having the estimation model output a value indicating the language-likeness of the second advertising copy; An information processing device comprising:

2. The acquisition unit is The first advertising copy, the advertising indicator of which satisfies the predetermined condition, is acquired.

2. The information processing apparatus according to claim 1,

3. The acquisition unit is The first advertising copy in which the advertising indicator regarding a click satisfies the predetermined condition is obtained.

3. The information processing apparatus according to claim 2.

4. The acquisition unit is The first advertising copy, in which the advertising indicator related to the number of times the first advertising copy is clicked by users, satisfies the predetermined condition, is acquired.

4. The information processing apparatus according to claim 3.

5. The acquisition unit is The first advertising copy having a click rate equal to or greater than a predetermined threshold is acquired.

5. The information processing apparatus according to claim 4.

6. The acquisition unit is The first advertising copy in which the advertising indicator regarding the conversion satisfies the predetermined condition is obtained.

3. The information processing apparatus according to claim 2.

7. The acquisition unit is The first advertising copy is acquired, the advertising index relating to the number of times the first advertising copy has led to a conversion satisfying the predetermined condition.

7. The information processing apparatus according to claim 6,

8. The acquisition unit is The first advertisement copy having a conversion rate equal to or higher than a predetermined threshold is obtained.

8. The information processing apparatus according to claim 7,

9. The learning unit is The learning model is trained to receive input information, which is a character string included in the first advertising copy, and to output the second advertising copy corresponding to the input information.

2. The information processing apparatus according to claim 1,

10. The learning unit is The learning model is trained to receive the input information, which is the first advertising copy, and to output the second advertising copy corresponding to the input information.

10. The information processing apparatus according to claim 9,

11. The learning unit is The learning model is trained to receive input information, which is a character string that is a part of the first advertising copy, and to output the second advertising copy that corresponds to the input information.

10. The information processing apparatus according to claim 9,

12. The learning unit is The learning model that has learned the characteristics of the first advertising copy is generated.

2. The information processing apparatus according to claim 1,

13. The learning unit is The learning model that learns the characteristics of the relevance between the input information and the second advertisement copy is generated.

13. The information processing apparatus according to claim 12.

14. A provision unit that provides a proposed advertising copy based on the second advertising copy generated by the learning model to a terminal device used by a target advertiser who is an advertiser to which information is to be provided; The information processing apparatus according to claim 1 , further comprising:

15. The providing unit is The second advertising copy generated by the learning model is provided as the proposed advertising copy to the terminal device used by the target advertiser.

15. The information processing apparatus according to claim 14,

16. A determination unit that determines, from the second advertising copy generated by the learning model, an advertisement candidate copy to be used for an advertisement; Further equipped with The providing unit is The proposed advertisement copy based on the advertisement candidate copy determined by the determination unit is provided to the terminal device used by the target advertiser.

15. The information processing apparatus according to claim 14,

17. The determination unit is The second advertising copy, the advertising indicator of which satisfies the predetermined condition, is determined as the advertising candidate copy.

17. The information processing apparatus according to claim 16,

18. The determination unit is The second advertisement copy that satisfies the condition of language likeness is determined as the advertisement candidate copy.

17. The information processing apparatus according to claim 16,

19. 1. A computer-implemented information processing method, comprising: An acquisition step of acquiring a first advertising copy which is an advertising copy that satisfies a predetermined condition; a learning process for learning a learning model that uses character information including at least a part of the first advertising copy acquired by the acquisition process as input information and outputs an advertising copy corresponding to the input information as a second advertising copy by using a learning process based on machine learning; a generation process of generating a second advertising copy by inputting a piece of character information to the learning model trained by the learning process, and having the learning model output a second advertising copy; and inputting the generated second advertising copy to an estimation model that estimates a language-likeness of a character string, and having the estimation model output a value indicating the language-likeness of the second advertising copy, thereby generating an estimated value of the language-likeness of the second advertising copy; 13. An information processing method comprising:

20. An acquisition step of acquiring a first advertising copy that satisfies a predetermined condition; a learning step of learning a learning model that uses character information including at least a part of the first advertising copy acquired by the acquisition step as input information and outputs an advertising copy corresponding to the input information as a second advertising copy by a learning process based on machine learning; a generation step of generating a second advertising copy by inputting a piece of character information to the learning model trained by the learning step, and causing the learning model to output a second advertising copy, and inputting the generated second advertising copy to an estimation model that estimates a language-likeness of a character string, and causing the estimation model to output a value indicating the language-likeness of the second advertising copy; and An information processing program characterized by causing a computer to execute the above.

Citation Information

Patent Citations

  • Advertisement copywriting generation method and device, equipment and storage medium

    CN113129051A

  • Information extraction device, information extraction method, and information extraction program

    JP2015079449A

  • Advertisement text selection device and program

    JP2017111479A

  • Generation apparatus, generation method, and generation program

    JP2018128805A

  • Generating device and generating program

    JP2021068053A