JP7528697B2 - Information processing program, information processing device, and information processing method - Google Patents

Description

The present invention relates to information processing.

In order to improve user satisfaction with responses to questions, an information processing device is known that extracts basic response information, which is the answer to a question input by a user, and additional response information that has a relationship with the answer, from multiple pieces of knowledge information (see, for example, Patent Document 1).

There is also a known question and answer information providing system that displays question and answer information suited to the actual usage of a site by a user in response to a request from the user to display question and answer information (see, for example, Patent Document 2). There is also a known dialogue device that improves the accuracy of estimating the knowledge of a dialogue partner user (see, for example, Patent Document 3). Distributed representations of sentences and documents are also known (see, for example, Non-Patent Document 1).

JP 2015-133047 A JP 2017-27208 A JP 2019-28604 A

Q. Le et al., “Distributed Representations of Sentences and Documents”, Cornell University, arXiv:1405.4053v2, 2014.

According to the technology of Patent Document 1, by utilizing knowledge information about the user, it is possible to respond to questions from the user by providing not only an answer to the question but also additional information. Furthermore, according to the technology of Patent Document 2, it is possible to display question and answer information that matches the actual usage behavior of the user using the website.

However, it is difficult to respond to a user's question by taking into account the user's situation, such as the user's level of understanding of the information related to the question or the background of the question.

This problem is not limited to responses to questions from users, but also occurs when responding to various pieces of information input by users.

In one aspect, the present invention aims to estimate a user's level of understanding or interest in words contained in information entered by a user using a website.

In one proposal, the information processing program causes the computer to perform the following processes:

Based on input information entered by a user who viewed the website, the computer identifies content on the website that is related to words included in the input information. The computer calculates an index showing the viewing status of content related to the word based on the user's viewing history of content on the website.

According to one aspect, it is possible to estimate a user's level of understanding or interest in words contained in information entered by a user using a website.

FIG. 2 is a functional configuration diagram of an information processing device. 13 is a flowchart of information processing. FIG. 1 is a configuration diagram of an information processing system. FIG. 1 is a first functional configuration diagram of a server. FIG. 1 illustrates a website. 13 is a flowchart of an index calculation process. FIG. 2 is a second functional configuration diagram of the server. 13 is a flowchart of a first response method determination process. FIG. 3 is a third functional configuration diagram of the server. FIG. 11 is a diagram showing a method for changing a threshold value TH1. FIG. 13 is a diagram showing a method for changing a threshold value TH2. 13 is a flowchart of a second response method determination process. FIG. 4 is a fourth functional configuration diagram of the server. FIG. 5 is a fifth functional configuration diagram of the server. 13 is a flowchart of a third response method decision process. FIG. 2 is a hardware configuration diagram of an information processing device.

The following describes the embodiment in detail with reference to the drawings.

Figure 1 shows an example of the functional configuration of an information processing device (computer) according to an embodiment. The information processing device 101 in Figure 1 includes a storage unit 111, an identification unit 112, and a calculation unit 113. The storage unit 111 stores a browsing history 121 of the content on a website of a user who has viewed the website.

Figure 2 is a flowchart showing an example of information processing performed by the information processing device 101 of Figure 1. The identification unit 112 identifies, based on the input information entered by the user, among the contents within the website, contents related to the words included in the input information (step 201). The calculation unit 113 calculates an index indicating the browsing status of the contents related to the words based on the browsing history 121 (step 202).

The information processing device 101 in FIG. 1 can estimate a user's level of understanding or interest in words contained in information entered by a user using a website.

FIG. 3 shows an example of the configuration of an information processing system including the information processing device 101 of FIG. 1. The information processing system of FIG. 3 includes a server 301, a terminal device 302, a terminal device 303, and terminal devices 304-1 to 304-M (M is an integer equal to or greater than 1).

Terminal device 302 is the terminal device of user A. Terminal device 303 is the terminal device of an operator who belongs to the organization of the website owner, and terminal devices 304-i (i = 1 to M) are terminal devices of personnel who belong to the same organization. Among the personnel are experienced personnel. An experienced personnel is a person who has specialized knowledge or a lot of experience.

The server 301 communicates with the terminal devices 302, 303, and 304-i via the communication network 305, and the terminal device 304-i communicates with the server 301 and the terminal device 302 via the communication network 305. The server 301 corresponds to the information processing device 101 in FIG. 1.

Figure 4 shows a first example of the functional configuration of server 301 in Figure 3. Server 301 in Figure 4 includes a memory unit 411, a communication unit 412, an identification unit 413, and a calculation unit 414. Memory unit 411, identification unit 413, and calculation unit 414 correspond to memory unit 111, identification unit 112, and calculation unit 113, respectively, in Figure 1. Communication unit 412 communicates with terminal device 302, terminal device 303, and terminal device 304-i via communication network 305.

The storage unit 411 stores multiple pieces of content 421 in a website and the browsing history 422 of multiple users who have viewed the website. User A accesses the server 301 using the terminal device 302 to view the content 421 in the website.

The browsing history 422 corresponds to the browsing history 121 in FIG. 1 and includes browsing histories associated with the user IDs of multiple users. The browsing history of each user includes identification information of the content 421 viewed by the user and the viewing date and time. The identification information of the content 421 is, for example, the title displayed on the page of the content 421.

The browsing history of each user may include an operation history for the content 421. The operation history indicates a move operation for moving from one page of the content 421 to another page of the content 421, a scroll operation within the page of the content 421, etc.

The website has an input field for accepting questions, and user A inputs question 423 into the input field using terminal device 302, which then sends an email including user A's user ID and question 423 to server 301. Communication unit 412 receives the email from terminal device 302, and storage unit 411 stores question 423 included in the received email. Question 423 is an example of input information entered by a user.

Even if user A viewed a large number of pieces of content 421 before entering question 423, if the user has not viewed enough of the pieces of content 421 related to a word included in question 423, it is unlikely that the user will understand that word. On the other hand, the greater the number of pieces of content 421 viewed by user A, the greater the possibility that user A has a strong interest in the website. In this case, user A can be said to be an important user to the website owner.

Therefore, by using the browsing history 422 to understand the situation of user A who entered question 423, it becomes possible to select an appropriate response method to question 423.

The identification unit 413 extracts one or more words from the text of the question 423, extracts one or more words from the text contained in each of the multiple pieces of content 421, and calculates the similarity between the question 423 and each piece of content 421 using the extracted words. Then, the identification unit 413 identifies, from among the multiple pieces of content 421, a piece of content 421 related to the word contained in the question 423, based on the similarity.

For example, the identification unit 413 regards the text of the question 423 as a document and converts the question 423 into a document vector. The identification unit 413 also regards the text of each piece of content 421 as a document and converts each piece of content 421 into a document vector.

For example, a paragraph vector based on Bag of Words described in Non-Patent Document 1 can be used as a document vector. Instead of a paragraph vector based on Bag of Words, a document vector based on TF-IDF (Term Frequency-Inverse Document Frequency) can be used.

Next, the identification unit 413 calculates the cosine similarity between the document vector of the question 423 and the document vector of each content 421, and selects N pieces of content (N is an integer equal to or greater than 1) from among the multiple pieces of content 421 in descending order of similarity. This identifies N pieces of content 421 related to the words included in the question 423.

Instead of the cosine similarity between document vectors, the number or percentage of matching words between the question 423 and each piece of content 421 may be used as the similarity between the question 423 and each piece of content 421. By calculating the similarity between the question 423 and each piece of content 421, it is possible to accurately identify the piece of content 421 related to the words included in the question 423.

The calculation unit 414 extracts the browsing history corresponding to the user ID of user A from the browsing history 422. Then, the calculation unit 414 uses the extracted browsing history of user A to calculate the level of understanding 424 indicating the browsing status of the N pieces of content 421 identified by the identification unit 413, and stores it in the memory unit 411. The level of understanding 424 is an example of an index indicating the browsing status of content related to a word.

The calculation unit 414 calculates the level of understanding 424 using the number of views, viewing time, or reading completion rate of user A for N pieces of content 421. The number of views indicates the number of pieces of content 421 viewed by user A. The number of views may be page views. The viewing time indicates the time from the time viewing of a page of content 421 begins to the time viewing ends. The reading completion rate indicates the proportion of the area viewed by user A out of the entire area of the page of content 421. The area viewed by user A is identified, for example, from the operation history of a scroll operation within the page.

For example, when calculating the level of understanding 424 using the number of views, the proportion of the content 421 that was viewed by user A out of the N pieces of content 421 can be used as the level of understanding 424.

When calculating the level of understanding 424 using the viewing time, the proportion of the N pieces of content 421 that have been viewed by user A for a predetermined period of time or more can be used as the level of understanding 424. When calculating the level of understanding 424 using the completion rate, the proportion of the N pieces of content 421 that have been viewed by user A for a predetermined period of time or more can be used as the level of understanding 424.

By calculating the level of understanding 424 using the number of views, viewing time, or completion rate of user A, the viewing status of content 421 related to the words included in question 423 can be reflected in level of understanding 424. The higher the level of understanding 424, the higher the possibility that user A understands the words included in question 423.

Next, the calculation unit 414 calculates an interest level 425 indicating the browsing status of the website using the number of views or the viewing time of the multiple contents 421 on the website by user A, and stores the calculated interest level 425 in the memory unit 411. The interest level 425 is an example of an index indicating the browsing status of the website.

For example, when calculating interest level 425 using the number of views, the number of pieces of content 421 that have been viewed by user A within a specified period of time among the multiple pieces of content 421 in the website can be used as interest level 425. The specified period may be one day, one week, or one month, and may be one session or a specified number of sessions. The proportion of pieces of content 421 that have been viewed by user A within a specified period of time among the multiple pieces of content 421 in the website may be used as interest level 425.

When calculating interest level 425 using viewing time, the sum of viewing times of content 421 viewed by user A within a specified period of time among multiple content 421 on a website can be used as interest level 425.

By calculating interest level 425 using the number of views or viewing time of user A, the website viewing status can be reflected in interest level 425. Interest level 425 corresponds to the amount of work from when user A accesses the website to when he or she inputs question 423, and the higher interest level 425 is, the higher the possibility that user A is interested in the website.

The communication unit 412 outputs the level of understanding 424 and the level of interest 425 to the terminal device 303 by transmitting the question 423, the level of understanding 424, and the level of interest 425 together with the user ID of user A to the terminal device 303. The terminal device 303 displays the question 423, the level of understanding 424, and the level of interest 425 on the screen together with the user ID of user A.

The operator estimates the degree of user A's understanding of the words included in question 423 and the degree of user A's interest in the website from the displayed level of understanding 424 and level of interest 425. Then, the operator determines the appropriate response method for question 423 based on the estimation result.

Figure 5 shows an example of a website. Website 501 in Figure 5 includes content C1 to content C10. The titles of the pages of each content are as follows:

C1 Service - Procedure C2 Service - Fee C3 Service - Suspension C4 Design - Implementation Procedure C5 Design - Procedure C6 Design - Fee C7 Design - Planning C8 Product - Purchase Procedure C9 Product - Performance C10 Product - Fee

User A's browsing history 502 includes the titles and browsing times of content C2, content C6, content C10, content C9, content C8, and content C1. For example, the title of content C2 is "Service - Fees" and the browsing time is "13:10." User A's question 503 includes the text "I would like to request a design. What should I do?"

For example, the identification unit 413 extracts the word "design" from the text of the question 503, and uses "design" to calculate the similarity between the question 503 and each piece of content in the website 501. The identification unit 413 then extracts content C4 to content C7 related to "design" based on the similarity, and generates a content set 504 that includes the extracted content.

In this case, N=4, and of the four pieces of content included in the content set 504, the number of pieces of content included in the browsing history 502 is one. Therefore, when calculating the level of understanding 424 using the number of views, the level of understanding 424 becomes 1/4=25%, which shows that user A's level of understanding of "design" is low.

On the other hand, since the total number of contents included in browsing history 502 is six, if the number of contents viewed is used as interest level 425, interest level 425 will be 6. If the proportion of contents included in browsing history 502 out of the 10 contents in website 501 is used as interest level 425, interest level 425 will be 6/10 = 60%. Therefore, it can be seen that user A has a high degree of interest in the website.

Even if the owner of the website is a company that supports new business ventures for its customers and does not propose new businesses, but rather posts a description of the business on the website, user A may request a proposal for a new business. In this case, for example, a question 423 such as "Could you propose a new business venture in the context of XX?" is sent.

Now, let's imagine that, without understanding User A's situation, the company sends a response stating, "We are a company that assists with proposals. We ask that you make any proposals yourself." If User A has already carefully examined the website and understands the company's business, the redundant explanation contained in this response may make User A uncomfortable.

Next, let's assume that the reply is, "Understood. I would like to discuss the specific requirements." If User A has not thoroughly examined the website and does not understand the company's business, this reply may result in the meeting being a waste of time for both parties.

Therefore, by presenting the level of understanding 424 and the level of interest 425 to the operator, the operator can estimate the situation of user A and determine an appropriate response method. The operator determines the response method from the level of understanding 424 and the level of interest 425, for example, using the following judgment conditions:

(A1) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A's understanding of the words contained in question 423 is low, and User A's interest in the website is also low. Therefore, the operator determines that User A is unlikely to become a customer and that User A may simply not have found the information he or she is looking for, and selects an answer using a beginner's template as the response method.

In this case, the terminal device 303 automatically sends to the terminal device 302 an e-mail containing an explanatory text such as "We are a company that will assist you with your proposal. Here's how we can assist you," which is registered in the template for beginners. The terminal device 303 may also automatically send a response using chat instead of e-mail.

(A2) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A has a high level of understanding of the words contained in question 423, and his/her level of interest in the website is low. Therefore, the operator determines that although User A may become a customer, User A's level of understanding 424 is high and therefore a response by email is sufficient, and selects a response using a template for those who understand as the response method.

In this case, the terminal device 303 automatically sends to the terminal device 302 an e-mail containing a question such as "Understood. But first, let me confirm a few points about the requirements," which is registered in the template for those who understand. The terminal device 303 may also automatically send a response using chat instead of e-mail.

(A3) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a low level of understanding of the words contained in question 423, and is very interested in the website. The operator therefore determines that User A is likely to become a customer and that polite response from the staff member will increase his or her favorable impression of the website, and selects a response method in which the staff member will check and answer question 423.

In this case, the operator notifies one of the staff members of the response method. The staff member uses terminal device 304-i to send a response such as, "We ask that you make the proposal. We are a company that supports our customers' proposals. Is that okay?" to terminal device 302.

(A4) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a high degree of understanding of the words contained in question 423, and User A is also highly interested in the website. Therefore, the operator determines that since User A is likely to become a customer and has a high degree of understanding 424, an appropriate answer from a person with knowledge not contained in the website is desirable. The operator then selects an answer from a skilled person with more detailed knowledge as the response method.

In this case, the operator notifies one of the experienced staff of the method of response, and the experienced staff uses terminal device 304-i to transmit a response such as, "Understood. Please discuss the specific requirements with us now." to terminal device 302.

For example, in the case of website 501 in FIG. 5, threshold TH1 of level of understanding 424 may be a number in the range of 50% to 90%. If the number of views of content in website 501 is used as level of interest 425, threshold TH2 of level of interest 425 may be a number in the range of 6 to 9. If the proportion of content included in browsing history 502 is used as level of interest 425, threshold TH2 may be a number in the range of 50% to 90%.

The operator can also select the following response methods using the same judgment conditions as (A1) to (A4).

(B1) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A has little understanding of the words contained in question 423, and User A is also little interested in the website. Therefore, the operator determines that User A does not understand the basic points and that it is desirable to start with a simple explanation, and selects a simple explanation by a person in charge as the response method.

In this case, the operator notifies one of the staff members of the method of response, and the staff member uses terminal device 304-i to send a response such as, "We are a company that will support your proposal. Here is how we can help you..." to terminal device 302.

(B2) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A has a high level of understanding of the words contained in question 423, but is not very interested in the website. The operator decides that although he does not know whether User A intends to enter into a contract, since User A is capable of understanding, it would be better to start by confirming the details, and selects the response method to have the person in charge confirm the details.

In this case, the operator notifies one of the staff members of how to respond, and the staff member uses terminal device 304-i to send a response such as, "Understood. First, let me confirm a few points about the requirements." to terminal device 302.

(B3) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a low level of understanding of the words contained in question 423, but is very interested in the website. Therefore, the operator determines that although User A is a potential customer, it is possible that he or she may not understand the question, and therefore it is desirable to respond carefully to avoid any misunderstandings, and selects a careful response by a person in charge as the response method.

In this case, the operator notifies one of the staff members of the response method. The staff member uses terminal device 304-i to send a response such as, "We ask that you make the proposal. We are a company that supports our customers' proposals. Is that okay?" to terminal device 302.

(B4) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a high degree of understanding of the words contained in question 423, and is also very interested in the website. Therefore, the operator determines that since User A is likely to become a good customer, a positive response that assumes a contract or purchase is desirable, and selects a positive response from the person in charge as the response method.

In this case, the operator notifies one of the staff members of the response method, and the staff member uses terminal device 304-i to send a response such as, "Understood. Please discuss the specific requirements with us now." to terminal device 302.

According to the information processing system of FIG. 3, by presenting the level of understanding 424 and the level of interest 425 to the operator, the operator can estimate the level of user A's understanding of the words included in question 423 and the level of user A's interest in the website. This allows the operator to determine an appropriate response method, and can avoid an insufficient or excessive response to question 423, thereby improving the user's satisfaction with the answer.

Figure 6 is a flowchart showing an example of the index calculation process performed by the server 301 in Figure 4. First, the communication unit 412 receives an email from the terminal device 302 (step 601). Then, the identification unit 413 calculates the similarity between the question 423 included in the email and each piece of content 421 (step 602), and identifies the content 421 related to the words included in the question 423 based on the similarity (step 603).

Next, the calculation unit 414 extracts the browsing history of user A corresponding to the user ID included in the email from the browsing history 422 (step 604), and calculates the understanding level 424 and the interest level 425 using the browsing history of user A (step 605). The communication unit 412 then transmits the question 423, the understanding level 424, and the interest level 425 together with the user ID of user A to the terminal device 303 (step 606).

The operator may also determine the response method based only on the level of understanding 424. In this case, the operator may select either response method (A1) or (A2), or either response method (A3) or (A4), depending only on the magnitude relationship between the level of understanding 424 and the threshold TH1. The operator may select either response method (B1) or (B2), or either response method (B3) or (B4), depending only on the magnitude relationship between the level of understanding 424 and the threshold TH1.

However, by determining the response method using not only the level of understanding 424 but also the level of interest 425, the variety of response methods can be increased compared to response methods determined using only the level of understanding 424.

Figure 7 shows a second example of the functional configuration of the server 301 in Figure 3. The server 301 in Figure 7 has a configuration in which a determination unit 415 is added to the server 301 in Figure 4. The determination unit 415 determines the response method using the understanding level 424 and the interest level 425. The determination unit 415 performs the following process, for example, using the same judgment conditions as (A1) to (A4).

(D1) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A has little understanding of the words included in question 423, and is also little interested in the website. Therefore, the determination unit 415 selects a response using a template for beginners as the response method.

In this case, the determination unit 415 automatically sends an e-mail including an explanatory text such as "We are a company that will support your proposal. How we can support you...", which is registered in the template for beginners, to the terminal device 302 via the communication unit 412. The determination unit 415 may automatically send a response using chat instead of e-mail.

(D2) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A has a high level of understanding of the words contained in question 423, and is not very interested in the website. Therefore, the determination unit 415 selects a response using a template for those who understand the question as the response method.

In this case, the determination unit 415 automatically sends an e-mail including a question such as "Understood. But first, let me confirm a few points about the requirements," which is registered in the template for those who understand, to the terminal device 302 via the communication unit 412. The determination unit 415 may automatically send a response using chat instead of e-mail.

(D3) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a low level of understanding of the words contained in question 423, and is highly interested in the website. Therefore, the determination unit 415 selects a response method in which the person in charge checks and answers question 423.

In this case, the decision unit 415 notifies the terminal device 304-i of one of the staff members of the response method via the communication unit 412. The staff member uses the terminal device 304-i to send a response such as, "We ask that you make the proposal yourself. We are a company that supports our customers' proposals. Is that okay?" to the terminal device 302.

(D4) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a high level of understanding of the words contained in question 423, and is also highly interested in the website. Therefore, the decision unit 415 determines that an appropriate response from a person with knowledge not contained in the website is desirable, and selects an answer from a more experienced person with more detailed knowledge as the response method.

In this case, the decision unit 415 notifies the terminal device 304-i of one of the experienced staff of the response method via the communication unit 412. The experienced staff uses the terminal device 304-i to send a response such as "Understood. Please discuss the specific requirements with us." to the terminal device 302.

The decision unit 415 can also select the following response method using the same judgment conditions as (A1) to (A4).

(E1) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A has little understanding of the words contained in question 423, and is also little interested in the website. Therefore, the decision unit 415 selects a simple explanation by the person in charge as the response method.

In this case, the determination unit 415 notifies the terminal device 304-i of one of the staff members of the response method via the communication unit 412. The staff member uses the terminal device 304-i to send a response such as "We are a company that will support your proposal. Here is how we can support you..." to the terminal device 302.

(E2) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A has a high level of understanding of the words contained in question 423, and is not very interested in the website. Therefore, the decision unit 415 selects detailed confirmation by a person in charge as the response method.

In this case, the determination unit 415 notifies the terminal device 304-i of one of the staff members of the response method via the communication unit 412. The staff member uses the terminal device 304-i to send a response such as "Understood. First, let me confirm a few points about the requirements." to the terminal device 302.

(E3) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a low level of understanding of the words contained in question 423, and is highly interested in the website. Therefore, the decision unit 415 selects a careful response by a person in charge as the response method.

In this case, the decision unit 415 notifies the terminal device 304-i of one of the staff members of the response method. The staff member uses the terminal device 304-i to send a response such as, "We ask that you make the proposal. Our company supports our customers' proposals. Is that okay?" to the terminal device 302.

(E4) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a high level of understanding of the words contained in question 423, and is also highly interested in the website. Therefore, the decision unit 415 selects a positive response from the person in charge as the response method.

In this case, the decision unit 415 notifies the terminal device 304-i of one of the staff members of the response method via the communication unit 412. The staff member uses the terminal device 304-i to send a response such as "Understood. Please discuss the specific requirements with us now" to the terminal device 302.

According to the server 301 in FIG. 7, the response method is automatically determined using the understanding level 424 and the interest level 425, so the operator does not need to determine the response method, and the operator's workload is reduced. In this case, the terminal device 303 in FIG. 3 can be omitted.

Figure 8 is a flowchart showing an example of the first response method determination process performed by the server 301 in Figure 7. The processes in steps 801 to 805 are similar to the processes in steps 601 to 605 in Figure 6.

When the calculation of the level of understanding 424 and the level of interest 425 is completed, the determination unit 415 determines the response method using the level of understanding 424 and the level of interest 425 (step 806) and performs response control processing (step 807). The response control processing is processing for automatically sending a response to the terminal device 302, or processing for notifying the terminal device 304-i of the person in charge of the response method.

The determination unit 415 may also determine the response method based only on the level of understanding 424. In this case, the determination unit 415 may select either the response method (D1) or (D2), or either the response method (D3) or (D4), depending only on the magnitude relationship between the level of understanding 424 and the threshold TH1. The determination unit 415 may select either the response method (E1) or (E2), or either the response method (E3) or (E4), depending only on the magnitude relationship between the level of understanding 424 and the threshold TH1.

However, by determining the response method using not only the level of understanding 424 but also the level of interest 425, the variety of response methods can be increased compared to response methods determined using only the level of understanding 424.

Figure 9 shows a third example of the functional configuration of the server 301 in Figure 3. The server 301 in Figure 9 has the same configuration as the server 301 in Figure 7. The memory unit 411 further stores person in charge information 911. The person in charge information 911 indicates the status of the person in charge of responding to the question 423. The status of the person in charge indicates the number of customers in negotiation, the number of customers available, the number of people available, the number of experienced people available, etc. The decision unit 415 decides the response method using the level of understanding 424, the level of interest 425, and the person in charge information 911.

For example, when selecting either a reply with a wording similar to a refusal or a reply with a polite explanation as the response method, the decision unit 415 changes the threshold TH1 of the understanding level 424 according to the state of the person in charge indicated by the person in charge information 911. If the state of the person in charge is one in which they want to increase the number of customers, the threshold TH1 is changed in a direction that increases the likelihood that a reply with a wording similar to a refusal will be selected. On the other hand, if the state of the person in charge is one in which they do not want to increase the number of customers, the threshold TH1 is changed in a direction that increases the likelihood that a reply with a wording similar to a refusal will be selected.

The determination unit 415 performs the following process, for example, using the same judgment conditions as (A1) to (A4).

(F1) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is less than the threshold TH2

The determination unit 415 selects automatic transmission of a message similar to a refusal as the response method, and automatically transmits an email containing a standard message similar to a refusal to the terminal device 302 via the communication unit 412.

(F2) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is less than the threshold TH2

The determination unit 415 selects automatic transmission of a detailed explanatory text as the response method, and automatically transmits an email including a standard phrase of the detailed explanatory text to the terminal device 302 via the communication unit 412.

(F3) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

The determination unit 415 selects a reply written by the person in charge that resembles a refusal as the response method, and notifies the terminal device 304-i of one of the persons in charge of the response method via the communication unit 412. The person in charge uses the terminal device 304-i to transmit the reply written in a manner similar to a refusal to the terminal device 302.

(F4) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

The decision unit 415 determines that an appropriate response from a staff member with knowledge not included in the website is desirable, and selects a response in the form of a detailed explanation from a skilled staff member with more detailed knowledge as the response method. The decision unit 415 then notifies the terminal device 304-i of one of the skilled staff members of the response method via the communication unit 412. The skilled staff member uses the terminal device 304-i to transmit the response in the form of a detailed explanation to the terminal device 302.

Figure 10 shows an example of how to change the threshold value TH1. Figure 10(a) shows an example of how to change the threshold value TH1 based on the number of customers currently in negotiation. A state in which the number of customers currently in negotiation is 20 or less corresponds to the off-season, a state in which the number of customers currently in negotiation is between 21 and 100 corresponds to the normal season, and a state in which the number of customers currently in negotiation is 101 or more corresponds to the busy season.

The determination unit 415 sets the threshold value TH1 to 40% when the number of customers currently negotiating is 20 or less, sets the threshold value TH1 to 60% when the number of customers currently negotiating is between 21 and 100, and sets the threshold value TH1 to 80% when the number of customers currently negotiating is 101 or more. As a result, the more customers currently negotiating, the more likely it is that a response with a wording that is closer to a rejection will be selected.

Figure 10(b) shows an example of a change method based on the number of available customers. A state in which the number of available customers is 10 or less corresponds to a high load state, a state in which the number of available customers is between 11 and 50 corresponds to a normal state, and a state in which the number of available customers is 51 or more corresponds to a low load state.

The determination unit 415 sets the threshold value TH1 to 90% when the number of available customers is 10 or less, sets the threshold value TH1 to 60% when the number of available customers is between 11 and 50, and sets the threshold value TH1 to 30% when the number of available customers is 51 or more. As a result, the more available customers there are, the more likely an answer with a detailed explanatory text will be selected.

When selecting either a response by a general staff member or a response by a skilled staff member as the response method, the decision unit 415 changes the threshold value TH1 according to the number of available skilled staff members indicated by the staff member information 911. If the number of available skilled staff members is small, the threshold value TH1 is changed in a direction that increases the likelihood that a response by a general staff member will be selected. On the other hand, if the number of available skilled staff members is large, the threshold value TH1 is changed in a direction that increases the likelihood that a response by a skilled staff member will be selected.

The determination unit 415 can also change the person in charge information 911 based on the response method selection result. For example, the determination unit 415 determines the number K of questions 423 whose understanding level 424 is equal to or greater than the threshold value TH1 among the questions 423 of multiple users for which a response method was determined within the most recent specified period, and compares this number with a specified number P.

If K is equal to or greater than P, the determination unit 415 multiplies P by the order rate R to calculate the update amount Δ. On the other hand, if K is less than P, the determination unit 415 sets the update amount Δ to 0. The update amount Δ represents the number of customers who are estimated to start negotiations soon, and is updated every day.

When the person in charge information 911 indicates the number of customers in negotiation, the determination unit 415 adds the update amount Δ to the number of customers in negotiation, and when the person in charge information 911 indicates the number of customers available, the determination unit 415 subtracts the update amount Δ from the number of customers available. For example, when P = 10, R = 0.5, and K is greater than or equal to P, Δ = 5.

Then, the determination unit 415 determines the threshold TH1 using the changed person in charge information 911. This makes it possible to dynamically change the threshold TH1 based on the response method selection result.

When selecting either automatic response transmission or response by a representative as the response method, the determination unit 415 changes the threshold TH2 of the interest level 425 according to the number of available representatives indicated by the representative information 911. If the number of available representatives is small, the threshold TH2 is changed so that the automatic response transmission is more likely to be selected. On the other hand, if the number of available representatives is large, the threshold TH2 is changed so that the probability of selection of response by a representative is higher.

The determination unit 415 performs the following process, for example, using the same judgment conditions as (A1) to (A4).

(G1) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is less than the threshold TH2

The determination unit 415 selects automatic transmission of a simple answer as the response method, and automatically sends an email including a standard phrase for the simple answer to the terminal device 302 via the communication unit 412.

(G2) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is less than the threshold TH2

The determination unit 415 selects automatic transmission of a detailed response as the response method, and automatically sends an email including a standard text of the detailed response to the terminal device 302 via the communication unit 412.

(G3) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

The determination unit 415 selects the response by the person in charge as the response method, and notifies the terminal device 304-i of one of the persons in charge of the response method via the communication unit 412. The person in charge transmits the response to the terminal device 302 using the terminal device 304-i.

(G4) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

The decision unit 415 determines that an appropriate response from a staff member with knowledge not included in the website is desirable, and selects a response from a more knowledgeable and experienced staff member as the response method. The decision unit 415 then notifies the terminal device 304-i of one of the experienced staff members of the response method via the communication unit 412. The experienced staff member uses the terminal device 304-i to transmit the response to the terminal device 302.

Figure 11 shows an example of how to change the threshold value TH2. A state in which the number of available agents is 5 or less corresponds to a state in which there are few agents, a state in which the number of available agents is 6 to 10 corresponds to a normal state, and a state in which the number of available agents is 11 or more corresponds to a state in which there are many agents.

The determination unit 415 sets the threshold value TH2 to 10 when the number of available agents is 5 or less, sets the threshold value TH2 to 3 when the number of available agents is 6 to 10, and sets the threshold value TH2 to 0 when the number of available agents is 11 or more. As a result, the more available agents there are, the more likely an answer by an agent will be selected.

The determination unit 415 can also change the number of available agents indicated by the agent information 911 based on the selection of the response method. For example, when response by an agent is selected as the response method, the determination unit 415 decrements the number of available agents by 1, and when response by an agent is completed, the determination unit 415 increments the number of available agents by 1.

Then, the determination unit 415 determines the threshold value TH2 using the changed person in charge information 911. This makes it possible to dynamically change the threshold value TH2 based on the result of the selection of the response method.

The server 301 in FIG. 9 can change the response method depending on the status of the person in charge of responding, making it possible to appropriately adjust the workload of the person in charge.

Figure 12 is a flowchart showing an example of the second response method determination process performed by the server 301 in Figure 9. The processes in steps 1201 to 1205 are similar to the processes in steps 601 to 605 in Figure 6.

When the calculation of the level of understanding 424 and the level of interest 425 is completed, the determination unit 415 determines the response method using the level of understanding 424, the level of interest 425, and the person in charge information 911 (step 1206), and performs response control processing (step 1207). The response control processing is processing for automatically sending a response to the terminal device 302, or processing for notifying the terminal device 304-i of the person in charge of the response method. Then, the determination unit 415 changes the person in charge information 911 based on the result of the selection of the response method (step 1208).

Figure 13 shows a fourth example of the functional configuration of the server 301 in Figure 3. The server 301 in Figure 13 has a configuration similar to that of the server 301 in Figure 7. The storage unit 411 further stores user information 1311. The user information 1311 indicates user attributes. The user attributes are, for example, information indicating whether each user is an important customer or not. The determination unit 415 determines the response method using the understanding level 424, the interest level 425, and the user information 1311.

The determination unit 415 determines whether or not user A, who inputs question 423, is a valued customer based on whether or not the user A is registered in the customer database or purchase database of the website owner, and registers the user information 1311. Users who are not valued customers are registered as general customers.

If user A is a known user who is registered in the customer database or the purchasing database, user A is determined to be a valued customer. On the other hand, if user A is an unknown user who is not registered in the customer database or the purchasing database, user A is determined to be a general customer.

The determination unit 415 may determine whether or not user A is a valued customer based on the browsing history of user A included in the browsing history 422. For example, if user A has viewed a website within the most recent specified period, user A is determined to be a valued customer, and if user A has not viewed a website within the most recent specified period, user A is determined to be a general customer. The specified period may be one or several quarters, or may be one or several years.

The determination unit 415 may determine whether or not user A is an important customer based on whether or not user A is a decision maker within the company. If user A is a decision maker, user A is determined to be an important customer, and if user A is not a decision maker, user A is determined to be a general customer.

The determination unit 415 may determine whether or not user A is an important customer based on whether or not user A is an employee of an existing customer company. If user A is an employee of an existing customer company, user A is determined to be an important customer, and if user A is not an employee of an existing customer company, user A is determined to be a general customer.

If user A is a general customer, the determination unit 415 determines the response method in the same manner as (D1) to (D4), for example. If user A is a valuable customer, the determination unit 415 performs the following process using the same judgment conditions as (A1) to (A4), for example.

(H1) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A has little understanding of the words contained in question 423, and is also little interested in the website. In this case, user A is a potential customer, but may be mistaken about the words contained in question 423. Since user A has the potential to become a customer, it is advisable to carefully review question 423 and respond to it.

The decision unit 415 then selects a response method in which the person in charge checks the background of the question 423 and responds, and notifies the terminal device 304-i of one of the people in charge of the response method via the communication unit 412. The person in charge uses the terminal device 304-i to transmit a response that checks the background of the question 423 to the terminal device 302.

(H2) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is less than the threshold TH2

User A has a high level of understanding of the words contained in question 423, and is not very interested in the website. In this case, User A is a potential customer with a high level of understanding, and is behaving as usual. Since User A is an important customer with a high level of understanding, it is desirable to proceed with the business negotiations as quickly as possible.

The decision unit 415 then selects a response method that allows the person in charge to proceed to the negotiation stage, and notifies the terminal device 304-i of one of the people in charge of the response method via the communication unit 412. The person in charge uses the terminal device 304-i to send a response that allows the person in charge to proceed to the negotiation stage to the terminal device 302.

(H3) When the level of understanding 424 is less than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a low level of understanding of the words contained in question 423, and is very interested in the website. In this case, User A is a good prospect, but may be confused about the words contained in question 423. User A is very likely to become a customer, so it is desirable to resolve the points of doubt contained in question 423.

The decision unit 415 then selects a response method in which the person in charge responds quickly and listens carefully to the background of the question 423, and notifies the terminal device 304-i of one of the persons in charge of the response method via the communication unit 412. The person in charge uses the terminal device 304-i to transmit a response in which the person in charge listens carefully to the background of the question 423 to the terminal device 302.

(H4) When the level of understanding 424 is equal to or greater than the threshold TH1 and the level of interest 425 is equal to or greater than the threshold TH2

User A has a high level of understanding of the words contained in question 423, and is also very interested in the website. In this case, User A is a promising customer, but is behaving differently than usual. Although User A is an important customer, it is advisable to respond somewhat cautiously and prioritize User A's understanding.

Therefore, the decision unit 415 determines that an appropriate response from a staff member with knowledge not included in the website is desirable, and selects a response method in which a more experienced staff member with more detailed knowledge will respond after carefully listening to the background of the question 423. The decision unit 415 then notifies the terminal device 304-i of one of the experienced staff members of the response method via the communication unit 412. The experienced staff member uses the terminal device 304-i to transmit a response in which the experienced staff member carefully listens to the background of the question 423 to the terminal device 302.

The server 301 in FIG. 13 allows the response method to be changed depending on the attributes of the user, making it possible to prioritize responses to important customers over general customers, thereby making response operations more efficient.

The response method determination process performed by the server 301 in FIG. 13 is similar to the response method determination process in FIG. 8. In this case, in step 806, the determination unit 415 determines the response method using the understanding level 424, the interest level 425, and the user information 1311.

Figure 14 shows a fifth example of the functional configuration of the server 301 in Figure 3. The server 301 in Figure 14 has a configuration similar to that of the server 301 in Figure 7. The storage unit 411 further stores question and answer information 1411. The question and answer information 1411 is information on a Q&A database or FAQ (Frequently Asked Questions) related to the content 421 in the website, and includes multiple combinations of questions and answers.

The identification unit 413 uses the question-and-answer information 1411 to identify answer candidates for the question 423. For example, the identification unit 413 extracts questions similar to the question 423 from the question-and-answer information 1411 by text matching, and selects answers associated with the extracted questions as answer candidates.

The identification unit 413 may identify answer candidates using a prediction model that predicts answers from questions. In this case, the identification unit 413 generates a prediction model by having a learning model learn combinations of questions and answers included in the question and answer information 1411, and identifies answer candidates corresponding to the question 423 using the generated prediction model.

Next, the identification unit 413 extracts one or more words from the text of the answer candidate, extracts one or more words from the text contained in each of the multiple contents 421, and uses the extracted words to calculate the similarity between the answer candidate and each content 421. The identification unit 413 then uses the calculated similarity as the similarity between the question 423 and each content 421 to identify, from among the multiple contents 421, the content 421 that is related to the words contained in the question 423.

According to the server 301 in FIG. 14, the level of understanding 424 is calculated using content 421 similar to answers registered in the Q&A database or FAQ, so the level of understanding of the expected answer to the question 423 is reflected in the level of understanding 424. This improves the accuracy of the level of understanding 424, making it possible to select a more appropriate response method.

Figure 15 is a flowchart showing an example of the third response method determination process performed by the server 301 of Figure 14. First, the communication unit 412 receives an email from the terminal device 302 (step 1501). Next, the identification unit 413 identifies answer candidates to the question 423 contained in the email (step 1502), and calculates the similarity between the question 423 and each piece of content 421 using the identified answer candidates (step 1503). The processes of steps 1504 to 1508 are similar to the processes of steps 803 to 807 of Figure 8.

In the server 301 of FIG. 4, FIG. 9, or FIG. 13, the identification unit 413 can identify answer candidates using the question and answer information 1411, as in the server 301 of FIG. 14. In this case, the identification unit 413 identifies the content 421 related to the words included in the question 423, using the similarity between the answer candidates and each content 421.

The configuration of the information processing device 101 in FIG. 1 is merely an example, and some of the components may be omitted or changed depending on the use or conditions of the information processing device 101. The configurations of the information processing system in FIG. 3 and the server 301 in FIGS. 4, 7, 9, 13, and 14 are merely examples, and some of the components may be omitted or changed depending on the use or conditions of the information processing system. The server 301 may calculate the level of understanding 424 and the level of interest 425 using other input information entered by user A instead of the question 423.

The flowcharts in Figures 2, 6, 8, 12, and 15 are merely examples, and some processing may be omitted or changed depending on the configuration or conditions of the information processing device 101 or the information processing system.

The website shown in FIG. 5 is merely an example, and the content within the website varies depending on the website. The method of changing threshold TH1 shown in FIG. 10 and the method of changing threshold TH2 shown in FIG. 11 are merely examples, and threshold TH1 or threshold TH2 may be changed based on other criteria.

The response method determination methods shown in (A1) to (H4) are merely examples, and the operator or the determination unit 415 may determine the response method based on another determination method.

Figure 16 shows an example of the hardware configuration of an information processing device used as the information processing device 101 in Figure 1 and the server 301 in Figures 4, 7, 9, 13, and 14. The information processing device in Figure 16 includes a CPU (Central Processing Unit) 1601, a memory 1602, an input device 1603, an output device 1604, an auxiliary storage device 1605, a media drive device 1606, and a network connection device 1607. These components are hardware and are connected to each other by a bus 1608.

The memory 1602 is, for example, a semiconductor memory such as a read only memory (ROM), a random access memory (RAM), or a flash memory, and stores programs and data used in processing. The memory 1602 may operate as the memory unit 111 in FIG. 1 or the memory unit 411 in FIG. 4, FIG. 7, FIG. 9, FIG. 13, and FIG. 14.

The CPU 1601 (processor) operates as the identification unit 112 and calculation unit 113 in FIG. 1, for example, by executing a program using the memory 1602. The CPU 1601 also operates as the identification unit 413 and calculation unit 414 in FIG. 4, FIG. 7, FIG. 9, FIG. 13, and FIG. 14, by executing a program using the memory 1602. The CPU 1601 also operates as the determination unit 415 in FIG. 7, FIG. 9, FIG. 13, and FIG. 14, by executing a program using the memory 1602.

The input device 1603 is, for example, a keyboard, a pointing device, etc., and is used to input instructions or information from the user of the information processing device. The output device 1604 is, for example, a display device, a printer, etc., and is used to output inquiries or instructions to the user of the information processing device, and processing results. The processing results may be the level of understanding 424 and the level of interest 425.

The auxiliary storage device 1605 is, for example, a magnetic disk device, an optical disk device, a magneto-optical disk device, a tape device, or the like. The auxiliary storage device 1605 may be a hard disk drive or a flash memory. The information processing device can store programs and data in the auxiliary storage device 1605 and load them into the memory 1602 for use. The auxiliary storage device 1605 may operate as the storage unit 111 in FIG. 1 or the storage unit 411 in FIGS. 4, 7, 9, 13, and 14.

The medium drive device 1606 drives the portable recording medium 1609 and accesses the recorded contents. The portable recording medium 1609 is a memory device, a flexible disk, an optical disk, a magneto-optical disk, etc. The portable recording medium 1609 may be a CD-ROM (Compact Disk Read Only Memory), a DVD (Digital Versatile Disk), a USB (Universal Serial Bus) memory, etc. A user of the information processing device can store programs and data on this portable recording medium 1609 and load them into the memory 1602 for use.

In this way, the computer-readable recording medium that stores the programs and data used in the processing is a physical (non-transitory) recording medium such as memory 1602, auxiliary storage device 1605, or portable recording medium 1609.

The network connection device 1607 is a communication interface circuit that is connected to the communication network 305 and performs data conversion associated with communication. The information processing device receives programs and data from an external device via the network connection device 1607 and can load them into the memory 1602 for use. The network connection device 1607 may operate as the communication unit 412 in Figures 4, 7, 9, 13, and 14.

Note that the information processing device does not need to include all of the components in FIG. 16, and some components may be omitted depending on the purpose or conditions of the information processing device. For example, if an interface with the user is not required, the input device 1603 and the output device 1604 may be omitted. If the portable recording medium 1609 is not used, the medium drive device 1606 may be omitted.

The terminal device 302, the terminal device 303, and the terminal devices 304-1 to 304-M in FIG. 3 can be information processing devices similar to those in FIG. 16.

Although the disclosed embodiments and their advantages have been described in detail, it will be understood that those skilled in the art may make various modifications, additions, and omissions without departing from the scope of the present invention as expressly set forth in the claims.

REFERENCE SIGNS LIST 101 Information processing device 111, 411 Storage unit 112, 413 Identification unit 113, 414 Calculation unit 121, 422, 502 Browsing history 301 Server 302, 303, 304-1 to 304-M Terminal device 305 Communication network 412 Communication unit 415 Determination unit 421 Content 423, 503 Question 424 Understanding level 425 Interest level 501 Website 504 Content collection 911 Person in charge information 1311 User information 1411 Question and answer information 1601 CPU
1602 Memory 1603 Input device 1604 Output device 1605 Auxiliary storage device 1606 Media drive device 1607 Network connection device 1608 Bus 1609 Portable recording medium

Claims (12)

Identifying content within the website that is related to words included in the input information based on input information entered by a user who browsed the website;
Calculating an index indicating a browsing status of content related to the word based on the browsing history of the content in the website of the user;
An information processing program for causing a computer to execute processing. The information processing program according to claim 1, characterized in that the process of calculating an index indicating the browsing status of the content related to the word includes a process of calculating an index indicating the browsing status of the content related to the word using the number of views, browsing time, or reading completion rate of the content related to the word by the user. The information processing program according to claim 1 or 2, further comprising causing the computer to execute a process of calculating an index indicating the browsing status of the website based on the browsing history. The information processing program according to claim 3, characterized in that the process of calculating an index indicating the browsing status of the website includes a process of calculating an index indicating the browsing status of the website using the number of views or the viewing time of the content on the website by the user. The information processing program according to claim 3 or 4, further comprising causing the computer to execute a process of outputting an indicator showing the browsing status of content related to the word and an indicator showing the browsing status of the website. The information processing program according to claim 3 or 4, further comprising causing the computer to execute a process of determining a response method to the input information based on an index indicating the browsing status of content related to the word and an index indicating the browsing status of the website. The information processing program according to claim 6, characterized in that the process of determining a response method to the input information includes a process of determining the response method based on an indicator indicating the browsing status of content related to the word, an indicator indicating the browsing status of the website, and the status of a person in charge of executing the response. The information processing program according to claim 6, characterized in that the process of determining a response method to the input information includes a process of determining the response method based on an index indicating the browsing status of content related to the word, an index indicating the browsing status of the website, and attributes of the user. The information processing program according to any one of claims 1 to 8, characterized in that the process of identifying content related to the word includes a process of extracting the word from the input information, a process of calculating a similarity between the word and text included in the content in the website, and a process of identifying content related to the word from among the content in the website based on the similarity. the input information is information indicating a question of the user,
The information processing program according to any one of claims 1 to 8, characterized in that the process of identifying content related to the word includes a process of identifying answer candidates for the question, a process of calculating a similarity between the answer candidates and text contained in content within the website, and a process of identifying content within the website that is related to the word based on the similarity. a storage unit that stores a browsing history of a content on a website of a user who has browsed the website;
an identification unit that identifies, from among contents within the website, contents related to words included in the input information input by the user;
A calculation unit that calculates an index indicating a browsing status of content related to the word based on the browsing history;
An information processing device comprising: Identifying content within the website that is related to words included in the input information based on input information entered by a user who browsed the website;
Calculating an index indicating a browsing status of content related to the word based on the browsing history of the content in the website of the user;
An information processing method characterized in that the processing is executed by a computer.

Images