Keyword output system, keyword output method, and program

Abstract

This invention provides a keyword output system, keyword output method, and program for creating advertisements. [Solution] In a keyword output system which is an information processing system 1000 in which a server device 100, a client device 110, a user terminal device 120, and a server device 130 are communicated to each other via a network 150, the server device has at least one control unit. The control unit acquires telephone call data. The call data is the call data of the client's telephone. The control unit generates text information by transcribing the call data, acquires keywords that are targeted for the client's search advertising based on the text information, and outputs the keywords.

Description

Technical Field

[0001] The present invention relates to a keyword output system, a keyword output method, and a program.

Background Art

[0002] As one type of Internet advertisement, search-linked advertisements are known. A search-linked advertisement is an advertisement that displays an advertisement related to a keyword searched by a general user on a search engine on a search result screen.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] For an advertiser (client) who issues search-linked advertisements, it is a difficult problem to efficiently select appropriate keywords.

Means for Solving the Problems

[0005] According to one aspect of the present invention, a keyword output system is provided. This keyword output system has at least one or more control units. The control unit acquires call data of a telephone. The call data is the call data of the client's telephone. The control unit generates character information by writing down the call data. The control unit acquires keywords that are target objects of the client's search-linked advertisements based on the character information. The control unit outputs the keywords.

Brief Description of the Drawings

[0006] [Figure 1] FIG. 1 is a diagram showing an example of the system configuration of an information processing system. [Figure 2]Figure 2 shows an example of the hardware configuration of a server device. [Figure 3] Figure 3 is a flowchart illustrating an example of information processing in an information processing system. [Figure 4] Figure 4 is a flowchart showing an example of information processing in the information processing system of Modification Example 1. [Figure 5] Figure 5 is a flowchart showing an example of information processing in the information processing system of Modified Example 2. [Figure 6] Figure 6 is a flowchart showing an example of information processing in the information processing system of the modified example 3. [Modes for carrying out the invention]

[0007] Embodiments will be described below with reference to the drawings. The various features shown in the embodiments below (including the modified examples shown below; the same applies hereafter) can be combined with each other.

[0008] <Embodiment 1> 1. System Configuration Figure 1 shows an example of the system configuration of the information processing system 1000. As shown in Figure 1, the information processing system 1000 includes, as a system configuration, at least a server device 100, a client device 110, a user terminal device 120, and a server device 130. The server device 100, the client device 110, the user terminal device 120, and the server device 130 are connected to each other via a network 150. The network 150 includes a WAN (Wide Area Network) and / or a LAN (Local Area Network) built using wired and / or wireless connections. The network 150 also includes a telephone network. The information processing system 1000 is an example of a keyword output system.

[0009] The server device 100 is the device that performs the main processing described in the specification. The server device 100 also provides the functionality of an automated telephone answering service.

[0010] The client device 110 is a device belonging to a user company (hereinafter also simply referred to as "the company") that utilizes the functions of the automated telephone answering service. The user terminal device 120 is a terminal device used by users who make phone calls to the company.

[0011] Server device 130 is a device that provides services for large-scale language models. A large-scale language model is a computer language model composed of an artificial neural network with many parameters, and is a language model trained using a vast amount of unlabeled text through self-supervised or semi-supervised learning.

[0012] Although Figure 1 shows only one user terminal device 120 for simplification, the information processing system 1000 may include multiple user terminal devices 120 and user terminals 130.

[0013] Furthermore, while Figure 1 illustrates a smartphone as an example of the client device 110 and user terminal device 120, it is not limited to smartphones.

[0014] Here, the information processing system described in the claims may consist of multiple devices or of a single device. If the information processing system described in the claims consists of a single device, an example of such a device is a server device 100. If the information processing system described in the claims consists of multiple devices, an example of such multiple devices is a cloud system that provides the functions of the server device 100. The cloud system may consist of, for example, multiple server devices.

[0015] 2. Hardware Configuration FIG. 2 is a diagram showing an example of the hardware configuration of the server device 100. The server device 100 includes, as a hardware configuration, a control unit 210, a storage unit 220, a communication unit 230, and an internal bus 240. The control unit 210, the storage unit 220, and the communication unit 230 are electrically connected via the internal bus 240.

[0016] The control unit 210 is a CPU (Central Processing Unit) or the like, and controls the entire server device 100.

[0017] The storage unit 220 is any one of an HDD (Hard Disk Drive), a ROM (Read Only Memory), a RAM (Random Access Memory), an SSD (Solid State Drive), or any combination thereof, and stores programs and data (for example, call data, blacklist, etc. described later) used when the control unit 210 executes processing based on the programs.

[0018] When the control unit 210 reads out the programs stored in the storage unit 220 and executes processing based on the read programs, the processing of the flowcharts in FIGS. 3 to 5 described later and the functions of the server device 100 are realized. The storage unit 220 is an example of a storage medium. In the first embodiment, the data used when the control unit 210 executes processing based on the programs is described as being stored in the storage unit 220, but it may be stored in the storage unit of another device communicable with the server device 100. That is, the data may be stored in the storage unit of any device as long as the control unit 210 can refer to and / or acquire it.

[0019] The communication unit 230 connects the server device 100 to the network 150 and controls communication with other devices. Note that each hardware configuration constituting the server device 100 is not limited to one, and two or more may exist as necessary. Also, it may have hardware other than the configuration shown in FIG. 2.

[0020] 3. Information Processing The information processing according to Embodiment 1 will be described below.

[0021] (1) Automatic Telephone Answering Service The control unit 210 controls to display a setting screen related to an automatic telephone answering service on the enterprise device (client device 110) according to a request. The control unit 210 can create a plurality of call flows according to a setting operation via the setting screen. A call flow is a series of processes regarding how the server device 100 responds and processes when a customer makes a call. As a branch response after pushing, the control unit 210 can allow selection from functions such as a voice guidance function, an SMS (Short Message Service) transmission function, a telephone transfer function, a voice recording function, etc.

[0022] The voice guidance function is a function in which AI (Artificial Intelligence) automatically reads out text created according to the business content of the enterprise. The SMS transmission function is a function that automatically transmits preset text to the receiving telephone number by SMS. It is a function that can be used for various purposes such as the address of an enterprise (store), route guidance such as a route, event guidance, and guidance to an inquiry form. The telephone transfer function is a function that transfers a call to a person in charge according to requirements, date and time, etc. The voice recording function is a function that accepts a customer's requirements by voice recording (answering machine function). The enterprise can play back voice data via the setting screen and take corresponding actions such as making a callback according to the content. In the automatic telephone answering service, all call contents between the enterprise and the customer are recorded.

[0023] (2) Outline of Keyword Output Processing The control unit 210 acquires call data of a telephone for the automatic telephone answering service. The call data is call data of a telephone between a client (enterprise) and a customer. The control unit 210 generates character information by transcribing the call data. The control unit 210 acquires keywords that are target objects of the client's search-linked advertisement based on the character information. The control unit 210 outputs the keywords. This configuration allows advertisers (clients) running search engine advertising to efficiently select appropriate keywords.

[0024] (3) Details of keyword output processing Figure 3 is a flowchart showing an example of keyword output processing in the information processing system 1000. In step S310, the control unit 210 retrieves customer identification information and call data, which are stored in the storage unit 220 or the like in association with customer identification information that identifies the customer. The call data is the call data from the client's telephone. The customer identification information is, for example, the customer's telephone number.

[0025] In step S320, the control unit 210 generates transcribed text information based on the acquired call data. For example, the control unit 210 inputs the acquired call data into a trained model. The trained model is a model that has been trained using the call data as input data and the transcribed text information as output data. The control unit 210 receives the text information output from the trained model.

[0026] Furthermore, if character information for call data has already been generated in the processing related to the automated telephone answering service, and this character information is stored in the storage unit 220 or the like in association with customer identification information, the control unit 210 may also retrieve the character information for call data from the storage unit 220 or the like.

[0027] In step S330, the control unit 210 generates a prompt for the large-scale language model provided by the server device 130. The prompt includes the character information generated in step S320. That is, the control unit 210 generates a prompt containing character information. The prompt further includes instructions to extract keywords that the client's search advertising targets from the character information. More preferably, the prompt includes instructions to extract keywords and instructions to generate keywords similar to the extracted keywords (hereinafter also referred to as extended keywords).

[0028] The control unit 210 transmits the generated prompt to the service provider 130, requesting the acquisition of data corresponding to the prompt. Upon receiving the prompt, the large-scale language model of the service provider 130 outputs data based on the instructions contained in the prompt. The output data includes multiple keywords targeted by the client's search advertising and their respective extended keywords. The multiple keywords and their respective extended keywords will hereafter also be simply referred to as the keyword group.

[0029] In step S340, the control unit 210 determines whether or not it has received data (keyword group) output from the service provision device 130. If the control unit 210 determines that it has received data output from the service provision device 130, it proceeds to step S350. If the control unit 210 determines that it has not received data output from the service provision device 130, it repeats the process in step S340. If the control unit 210 fails to receive data output from the service provision device 130 even after repeating the process in step S340 for a predetermined time, it terminates the process shown in the flowchart in Figure 3. Step S330, or the processing in steps S330 and S340, is an example of the process by which the control unit 210 obtains candidate keywords to be targeted for the client's search advertising based on text information. The processing in steps S330, or the processing in steps S330 and S340, can also be described as an example of the process of obtaining data output from a large-scale language model as keywords based on a prompt.

[0030] In step S350, the control unit 210 temporarily stores the keyword group received from the service provider 130 in the storage unit 220 or the like. The keyword group temporarily stored in the storage unit 220 or the like can be called keyword candidates. The control unit 210 then deletes keywords from the keyword candidates that are included in the blacklist stored in a predetermined storage area of ​​the storage unit 220 or the like. Here, the blacklist is a list of keywords that are set as inappropriate as keywords to be targeted for the client's search advertising. Inappropriate keywords to be targeted for the client's search advertising include, for example, malicious words, obscene words, etc. The blacklist is an example of a predetermined list of keywords to be excluded. The process in step S350 can also be described as an example of the control unit 210 selecting a keyword from multiple keyword candidates based on the list.

[0031] In step S360, the control unit 210 removes a keyword from the keyword candidates based on the search volume of each of the multiple keywords included in the keyword candidates stored in the memory unit 220, etc. Search volume refers to the number of searches performed on a search engine within a predetermined period (for example, in the last month). A keyword searched 10,000 times in a month would have a search volume of "10,000". For example, the control unit 210 obtains the search volume of each of the multiple keywords from a predetermined system. Here, the predetermined system is a system that outputs the search volume for each keyword. An example of a predetermined system is a keyword planner. For example, the control unit 210 can obtain the search volume of each of the multiple keywords via an API (Application Programming Interface) provided by a keyword planner, etc.

[0032] The control unit 210 removes keywords from the keyword group whose search volume is above a first threshold. The first threshold is a threshold indicating that the search volume is too high (for example, 100,000). This is because keywords with a high keyword volume can display ads to many users, but the competition is also fierce. The control unit 210 also removes keywords from the keyword group whose search volume is below a second threshold. The second threshold is a threshold indicating that the search volume is too low (for example, 50). This is because there are few users who search for the relevant keyword in the first place, making it difficult to expect effective advertising. The first and second thresholds are assumed to be stored in advance in the storage unit 220, etc.

[0033] The process in step S360 can be considered an example of a process that retrieves search volume information for each of several keyword candidates. Furthermore, the process in step S360 can be considered an example of a process that selects a keyword from several keyword candidates based on the search volume information.

[0034] In step S370, the control unit 210 removes keywords with low relevance from the keyword candidates temporarily stored in the memory unit 220, etc., which are included in the character information of the call data. For example, the control unit 210 determines the relationships between keywords using a predetermined model. Examples of predetermined models include BERT (language model) and a pre-trained word2vec model.

[0035] The process in step S370 is an example of a process for selecting a keyword from multiple keyword candidates by excluding keywords with low relevance to keywords that appear frequently in the text information. Furthermore, the process in step S370 is an example of a process for excluding keywords with low relevance to keywords that appear frequently in the text information from the keyword candidates using a predetermined model. The predetermined model is a model that outputs the relationships between input keywords.

[0036] In step S380, the control unit 210 outputs keyword candidates temporarily stored in the storage unit 220 or the like as keywords to be targeted for the client's search advertising. The control unit 210 may output the keywords by transmitting the keyword candidates to the client device 110 that can communicate via the network 150, or it may output the keywords by writing the keyword candidates to a file or the like and storing them in a predetermined recording area of ​​the storage unit 220 or the like. Alternatively, the control unit 210 may output the keywords by setting the keyword candidates temporarily stored in the storage unit 220 or the like as keywords to be targeted for the client's search advertising in a predetermined system.

[0037] The process in step S380 is an example of a process that outputs keywords.

[0038] According to Embodiment 1, advertisers (clients) who run search-linked ads can efficiently select appropriate keywords.

[0039] (Variation 1) The following describes Modification 1 of Embodiment 1. Modification 1 is included in Embodiment 1 and is not a different embodiment from Embodiment 1. Configurations and processes not described in Modification 1 are the same as in Embodiment 1.

[0040] Figure 4 is a flowchart showing an example of information processing in the information processing system of Modification Example 1. In step S410, the control unit 210 of the modified example 1 (hereinafter simply referred to as the control unit 210) acquires the company's web data. Web data is data stored on a computer called a web server and transmitted to a client such as a web browser using the HTTP protocol. Examples of web data include text data, image data, video data, audio data, structured data, etc. The control unit 210 acquires the company's web data, for example, by using web scraping technology. The process in step S410 is an example of a process for retrieving data from a client's web page.

[0041] In step S420, the control unit 210 generates a prompt for the large-scale language model provided by the server device 130. The control unit 210 generates the prompt based on the character information generated in step S320 and the web data acquired in step S410. For example, the prompt may include instructions to take the company's web data into consideration and extract keywords that will be targeted for the client's search advertising from the character information.

[0042] Step S420, or the process in steps S420 and S340, is an example of a process that retrieves keywords based on textual information and web page data.

[0043] According to Modification 1, keywords can be obtained that also take into account data from a company's web page.

[0044] (Modification 2) The following describes a modified example 2 of Embodiment 1. Modified example 2 is included in Embodiment 1 and is not a different embodiment from Embodiment 1. Configurations and processes not described in Modified Example 2 are the same as in Embodiment 1.

[0045] Figure 5 is a flowchart showing an example of information processing in the information processing system of Modified Example 2. In step S510, the control unit 210 of the modified example 2 (hereinafter simply referred to as the control unit 210) acquires the company's business information. The company's business information is information that shows what kind of business the company is conducting, what kind of products and / or services it is providing, and how it is generating revenue. The control unit 210 may acquire the company's business information from the company's web data, or it may acquire business information based on the results of a survey of the company's representatives, etc. The process in step S510 is an example of a process that retrieves business information that shows the content of the client's business.

[0046] In step S520, the control unit 210 generates a prompt for the large-scale language model provided by the server device 130. The control unit 210 generates the prompt based on the character information generated in step S320 and the client's business information obtained in step S510. For example, the prompt may include instructions to take the company's business information into consideration and extract keywords that will be targeted for the client's search advertising from the character information.

[0047] Step S520, or the process in steps S520 and S340, is an example of a process that retrieves keywords based on textual information and business information.

[0048] According to variation 2, keywords can be obtained that also take into account the company's business information.

[0049] (Variation 3) The following describes a modification 3 of Embodiment 1. Modification 3 is included in Embodiment 1 and is not a different embodiment from Embodiment 1. Configurations and processes not described in Modification 3 are the same as in Embodiment 1.

[0050] Figure 6 is a flowchart showing an example of information processing in the information processing system of the modified example 3. The flowchart in Figure 6 has an additional step, S610, compared to the flowchart in Figure 3. In step S610, the control unit 210 generates text for a search-linked advertisement related to the keywords output in step S380. For example, the control unit 210 generates a prompt that includes a keyword and an instruction to generate text for a search-linked advertisement that will be displayed in conjunction with the keyword, and sends it to the large-scale language model. The control unit 210 then receives the text for the search-linked advertisement that will be displayed in conjunction with the keyword, which is output from the large-scale language model.

[0051] In step S620, the control unit 210 outputs the generated search-linked advertisement text. The control unit 210 may output the search-linked advertisement text by transmitting it to a client device 110 that can communicate via the network 150, or it may output the search-linked advertisement text by writing the search-linked advertisement text to a file or the like and storing it in a predetermined recording area such as the storage unit 220. Alternatively, the control unit 210 may output keywords by setting the search-linked advertisement text temporarily stored in the storage unit 220 or the like as the client's search-linked advertisement document in a predetermined system.

[0052] According to Modification 3, it is possible to generate search engine advertising text related to keywords.

[0053] <Note> This embodiment includes the following disclosures.

[0054] (Note 1) It is a keyword output system, Having at least one control unit, The control unit, Obtain telephone call data, The aforementioned call data is the call data from the client's phone. The aforementioned call data is transcribed to generate text information, Based on the aforementioned text information, the keywords to be targeted for the client's search advertising are obtained. Output the aforementioned keyword. Keyword output system.

[0055] (Note 2) The keyword output system described in Appendix 1, The control unit, We obtain business information that shows the nature of the client's business, Based on the aforementioned text information and the aforementioned business information, the keyword is obtained. Keyword output system.

[0056] (Note 3) The keyword output system described in Appendix 1, The control unit, Obtain the data from the aforementioned client's web page, Based on the aforementioned text information and the data of the web page, the keyword is obtained. Keyword output system.

[0057] (Note 4) The keyword output system described in Appendix 1, The control unit, A prompt containing the aforementioned character information is generated, The prompt further includes instructions to extract keywords that the client's search advertising will target from the text information. Based on the aforementioned prompt, the data output from the large-scale language model is obtained as the aforementioned keyword. Keyword output system.

[0058] (Note 5) The keyword output system described in Appendix 4, The prompt includes instructions to extract the keyword and instructions to generate keywords similar to the extracted keyword. Keyword output system.

[0059] (Note 6) The keyword output system described in Appendix 1, The control unit, Based on the aforementioned textual information, multiple keyword candidates to be targeted for the client's search advertising are obtained. Obtain search volume information for each of the aforementioned multiple keyword candidates. Based on the search volume information, select a keyword from the multiple keyword candidates. Keyword output system.

[0060] (Note 7) The keyword output system described in Appendix 1, The control unit, Based on the aforementioned textual information, multiple keyword candidates to be targeted for the client's search advertising are obtained. By excluding keywords with low relevance to keywords that appear frequently in the aforementioned textual information from the keyword candidates, a keyword is selected from the multiple keyword candidates. Keyword output system.

[0061] (Note 8) The keyword output system described in Appendix 7, The control unit, Using a predetermined model, keywords with low relevance to frequently occurring keywords in the text information are excluded from the keyword candidates. The aforementioned predetermined model is a model that outputs the relationships between input keywords, Keyword output system.

[0062] (Note 9) The keyword output system described in Appendix 1, The control unit, Based on the aforementioned textual information, multiple keyword candidates to be targeted for the client's search advertising are obtained. Obtain a predetermined list of keywords to be excluded, Based on the above list, select a keyword from the above multiple keyword candidates. Keyword output system.

[0063] (Note 10) The keyword output system described in Appendix 1, The control unit, To generate text related to the aforementioned search-linked advertisement, Keyword output system.

[0064] (Note 11) A keyword output method performed by a keyword output system, Obtain telephone call data, The aforementioned call data is the call data from the client's phone. The aforementioned call data is transcribed to generate text information, Based on the aforementioned text information, the keywords to be targeted for the client's search advertising are obtained. Output the aforementioned keyword. Keyword output method.

[0065] (Note 12) It is a program, Computers, A program to function as a keyword output system as described in any one of the appendices 1 through 10.

[0066] Although various embodiments of the present invention have been described, these are provided as examples only and are not intended to limit the scope of the invention. Novel embodiments can be implemented in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. Embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims of the invention and its equivalents.

[0067] The embodiments and modifications described above may be implemented in any combination. [Explanation of symbols]

[0068] 100: Server device 210: Control Unit 220: Storage section 230: Communications Department 240: Internal bus

Claims

1. It is a keyword output system, Having at least one control unit, The control unit, Obtain telephone call data, The aforementioned call data is the call data from the client's phone. The aforementioned call data is transcribed to generate text information, Based on the aforementioned text information, the keywords to be targeted for the client's search advertising are obtained. Output the aforementioned keyword. Keyword output system.

2. A keyword output system according to claim 1, The control unit, We obtain business information that shows the nature of the client's business, Based on the aforementioned text information and the aforementioned business information, the keyword is obtained. Keyword output system.

3. A keyword output system according to claim 1, The control unit, Obtain data from the aforementioned client's web page, Based on the aforementioned text information and the data of the web page, the keyword is obtained. Keyword output system.

4. A keyword output system according to claim 1, The control unit, A prompt containing the aforementioned character information is generated, The prompt further includes instructions to extract keywords that the client's search advertising will target from the text information. Based on the aforementioned prompt, the data output from the large-scale language model is obtained as the aforementioned keyword. Keyword output system.

5. A keyword output system according to claim 4, The prompt includes instructions to extract the keyword and instructions to generate keywords similar to the extracted keyword. Keyword output system.

6. A keyword output system according to claim 1, The control unit, Based on the aforementioned textual information, multiple keyword candidates to be targeted for the client's search advertising are obtained. Obtain search volume information for each of the aforementioned multiple keyword candidates. Based on the search volume information, select a keyword from the multiple keyword candidates. Keyword output system.

7. A keyword output system according to claim 1, The control unit, Based on the aforementioned textual information, multiple keyword candidates to be targeted for the client's search advertising are obtained. By excluding keywords with low relevance to keywords that appear frequently in the aforementioned textual information from the keyword candidates, a keyword is selected from the multiple keyword candidates. Keyword output system.

8. A keyword output system according to claim 7, The control unit, Using a predetermined model, keywords with low relevance to frequently occurring keywords in the text information are excluded from the keyword candidates. The aforementioned predetermined model is a model that outputs the relationships between input keywords, Keyword output system.

9. A keyword output system according to claim 1, The control unit, Based on the aforementioned textual information, multiple keyword candidates to be targeted for the client's search advertising are obtained. Obtain a predetermined list of keywords to be excluded, Based on the above list, select a keyword from the above multiple keyword candidates. Keyword output system.

10. A keyword output system according to claim 1, The control unit, To generate text related to the aforementioned search-linked advertisement, Keyword output system.

11. A keyword output method performed by a keyword output system, Obtain telephone call data, The aforementioned call data is the call data from the client's phone. The aforementioned call data is transcribed to generate text information, Based on the aforementioned text information, the keywords to be targeted for the client's search advertising are obtained. Output the aforementioned keyword. Keyword output method.

12. It is a program, Computers, A program for functioning as a keyword output system according to any one of claims 1 to 10.

Images