JP2022113612A - Question answering system by associating content - Google Patents

Abstract

To solve the problem in which a plurality of users cannot develop questions and answers for content in an easy, intuitive, and highly reusable manner when the content such as books is browsed.SOLUTION: Other users' content is displayed by narrowing down to other users' content with similar associated places or content rated highly by a relevant user, at the same time content with overlapping ranges is organized with justification based on content range specification, and thereby questions and answers for content are developed in an easy, intuitive, and highly reusable manner.SELECTED DRAWING: Figure 2

Description

TECHNICAL FIELD The present invention relates to a system suitable for improving the comprehension of viewers and improving content for difficult parts of content such as electronic books.

When browsing content such as a book, there is a method of uniquely adding annotations to parts that are difficult for the reader to understand. The drawback of this method is that if the difficulty level is too high for the viewer, even annotations cannot be given, and the user's grasp of the content is delayed.

In order to understand difficult passages that are difficult for readers to understand, it is effective to obtain explanations from others, and it is also possible to solve them by asking questions to people close to you or at study sessions. However, this method has the drawback that the range of other people from whom answers can be obtained is limited, and the reusability of answers by third parties who are in trouble with the same question is lacking. Questions on online question-and-answer boards improve respondent coverage and reusability, but because questions and answers are deployed in a different context than the original content, the quality of responders and the quality of responses to content viewers are limited. It cannot be said that the reusability is good. Furthermore, when asking a question, it is necessary to provide an explanation that includes the context of the original content, which poses a problem that places a heavy burden on the questioner.

If reusability is emphasized, there is a means of linking content using hyperlinks in HTML on the WWW. , the question and answer cannot be expanded freely by specifying the location in the content. In contrast to this, hyperlinks in the opposite direction, the backlink structure of Non-Patent Document 1, can be associated with multiple contents by multiple subjects, but the location of the linked contents cannot be freely selected, and the range within the contents Because it is difficult to dynamically generate question answers from the backlink structure by specifying

The invention of Patent Document 1, which concludes a question-and-answer session at a conference, can be applied to a question-and-answer session for general contents to some extent, but there are some problems. One is that it is difficult to function unless the associated original content is as small as one opinion. Combined with hyperlinks, simple association over a range within the source content does not efficiently handle a large number of subtly varying range of link structures. In addition, since it is premised on simultaneous participation in a conference, it is not possible to efficiently process asynchronous participation after a certain period of time. In this way, Patent Document 1 functions in a place such as a meeting where the time, number of people, and content are limited, but when the quality and quantity of those events become large-scale, a large number of content with low relevance makes it practical. is a problem.

Patent 5648773 8 backlink check tools https://cluster-seo. com/blog/backlink-tools. html

There is a method of adding annotations to aid understanding of parts that are difficult for a reader to view when reading content such as a book. A first object of the present invention is to solve the situation in which the viewer alone cannot provide appropriate annotations and the understanding of the content does not progress.

It is difficult to ask appropriate questions when obtaining answers to questions offline or online from others, and the reusability of question answers is poor. Solution is the second object of the invention.

A third object of the present invention is to solve the lack of flexibility in associating multiple contents when trying to express a question and answer with hyperlinks such as HTML.

In the invention of Patent Document 1, when answering questions on large-scale content such as e-books, the lack of flexibility in the position in the content to be associated and the decrease in convenience due to a large number of content with weak relationships are solved by the present invention. This is the fourth purpose of

For the first purpose, as shown in FIG. 1, a plurality of viewers can individually associate other contents, and when viewing, as in claims 1, 2, and 9, another viewer's Resolved by making associated content available.

For the second purpose, by associating questions and answers directly with original content and places in it as shown in FIG.・The problem is solved by making the associated content available to the viewer after the association is made as shown in Figs. 2 and 9.

For the third purpose, it is possible to dynamically present a bulletin board format based on a plurality of associated contents that are associated by the viewer himself/herself or by other viewers as shown in claims 1, 2 and 9. to resolve.

For the fourth purpose, it is possible to dynamically present a bulletin board format based on a plurality of associated contents that are associated by the viewer himself/herself or other viewers as shown in claims 1, 2, and 9. , as in claims 2, 5 and 10, contents of other viewers who are less relevant to the viewer are not displayed; I will solve it by putting it out to.

As explained in the solutions for the first to third objectives, various effects are generated because the question answering and annotating can be done in a state directly related to the content. As a first effect, it becomes possible for the entire group of viewers to discuss and understand the parts of the linking source content that are difficult to understand, going beyond individual viewers. This eliminates restrictions on time and place, compared to study sessions where people gather directly at the same time and place. It means accumulated. The first effect is that multiple subjects can set associations in ranges that allow overlap, and not only simple range matching but also information attached to overlapping ranges can be displayed at the same time. This effect is caused mainly by claim 1.

As a second effect, the author of the content itself can quickly add and modify the content after grasping in detail how the content is received by the viewer. It can be said that the part where the viewer has added many annotations and questions and answers to his/her own content/copyright is an indicator that the explanation of the content is insufficient. Of course, this metric can also take into account how far the content has been viewed, what part of the content it took time to view, etc. Based on these metrics, more detailed improvement of own content can be taken into account. It can be a source of information for In particular, it is a clear advantage over conventional indicators that it is possible to identify points that need to be improved in related content such as specific question sentences, as well as in simple locations. Once you've identified areas for improvement, you can, of course, add annotations with new content to help resolve associations with the intent of the question. This effect is caused mainly by claim 1.

As explained in the solution for the fourth purpose, as a third effect, even if the first effect causes too many associations with other viewers, exclude associations that are unnecessary or less relevant for the viewer This makes it possible to efficiently obtain an answer that matches one's own knowledge level. This makes it possible to exert the first effect with little noise on all content users beyond a mere meeting or organization of a small number of people. Moreover, on a large-scale network, malicious users often commit fraud or other criminal acts by guiding other legitimate users to unauthorized content. In conventional SNSs on the Internet, these malicious users have committed criminal acts by recklessly associating hyperlinks to content to which they are guided. With this third idea, it can be judged that associations that are made indiscriminately are very different from those of legitimate users. can be done. This effect is mainly caused by claims 2 and 3.

As a fourth effect, it is possible to utilize the usage characteristics of each viewer. For example, a viewer who has read a lot of content in a certain field and has many answers that solve other viewers' questions can be considered to have high knowledge and skill in that field. By delivering targeted recruitment advertisements to such users, it becomes easier to secure human resources with the necessary skills. Similarly, it is possible to target human resources who have similar usage characteristics to those who are active in the company. Of course, it goes without saying that privacy and other considerations must be taken into consideration. Moreover, the above first to fourth effects can be achieved with a realistic amount of calculation based on the latest data precisely because the association data are aggregated as shown in FIG.

Screen example immediately before the system implements claim 1 In the "content display area", the content that is the source of the association that is currently being generated is displayed. By specifying the range of a part of it with the mouse or keyboard, a "marker indicating the association range" and a "content editing area being created" are generated. The "editing area for content being created" is a place to enter the content that will be the destination of the association that is currently being generated. After focusing with the mouse or keyboard, content can be entered using the keyboard or the like. After the association is generated, the content of the association source and the content of the destination are associated by "two-way link". In other words, if you select the "marker indicating the range of association" of the source content with a mouse, the content of the association destination will be displayed in a separate window or popup, and the "marker indicating that the entire content is the scope of association" of the content of the association destination. is selected with a mouse, the area around the "linking range marker" of the linking source content will be displayed in a separate window or popup. Although an example of a "bidirectional link" has been given here, the present invention can of course also be implemented with a unidirectional link used in HTML or the like. In addition, questions and comments can be written in the "content editing area being created". In the present invention, when content intended to be a question is added, there is a function of requesting answers from other viewers as in claim 3, and it is important to determine whether or not the content is a question. This judgment is for example added content? There are both automatic judgment such as mechanical judgment whether it has a mark, automatic judgment such as judging whether it is an interrogative sentence by natural language processing, and manual judgment which allows the content adder to select whether it is a question or not. In FIG. 1, automatic determination is assumed. It is also a system design choice whether or not to allow additional empty content, ie, content with no input. If it is allowed, if the additional content remains empty even after a certain period of time, it is possible to automatically determine that the user wanted to add some annotation but could not, and that the user intended to ask a question. The "Related content expansion button for the specified range" is related to the currently generated related range with the description of Related 3. It shows three things. By pressing this button, the screen shifts to FIG.

An example of a screen on which the system implements claim 1 Descriptions of common parts with FIG. 1 are omitted. Here, related content created by other viewers who are closely related to the currently generated link in FIG. 1 is displayed. Specifically, there are three "related contents in the specified range", and the left end of all of them is left from the left end of the "content editing area being created". It indicates that all the association range start positions within are before the "content editing area being created". The right ends of the 1st content and the 2nd and 3rd contents, respectively, on the left and right of the right end of the "content editing area being created" similarly indicate the relative positions of the end of the association range in the association source content. With this ingenuity, it is possible to display contents whose association ranges do not exactly match in the original document as contents that are related to some extent, while also making it possible to easily find direct answers whose ranges strictly match. In this figure, related content already created by other viewers that overlaps with the range of the currently generated relation is displayed, but even if there is no overlap, it may be close to a certain level or have a deep relationship. It is also possible to display the content to be displayed at the same time. For example, the number of characters, the number of pixels, and the number of seconds in the video are each less than a certain amount, or the content to be displayed can be increased or decreased by the similarity between the creator of the content and the viewer. is. As for the order of arrangement of the "related contents in the specified range", the order of creation time, the order of matching range of association, and the order of relevance of claim 2 can be considered. By selecting one of the "Specified range of related content", you can create further related content of the related content starting from the related content's related range instead of the "Content editing area being created" in the same format. may be displayed. In addition, a high evaluation may be given to the "related content within the specified range" as the viewer's own question has been resolved, or the viewer's own comment may be captured. These functions eliminate the need for additional responses from the viewers themselves.・It has the effect of reducing the risk that it may be deleted and become useless. For other viewers, there is an effect that the reply request as in claim 3 is not required more than necessary. The system of the present invention has the effect of making it easier to identify other viewers' content that is really useful to the viewer. Since the "related content expansion button in the specified range" in FIG. 1 has already been expanded in FIG. 2, it functions as a minimization function to return to FIG.

Screen example where the system implemented claim 1 and did not immediately produce an effect In the same situation as Figure 1, there is no duplicate or related related content created by other viewers in the "marker indicating the range of association" This is an assumed screen when In this case, it is possible to request other viewers to answer according to claim 3 by associating with "newly created question content" that intends to ask questions.

Derived screen example of FIG. 3 Instead of asking questions in FIGS. 1 and 3, it is possible to create a memo for the viewer himself, such as "newly created information content". This information is displayed when you view it yourself, and can also be displayed to other viewers under the conditions of claims 1 and 2 to help their understanding. At the time of content creation, as shown in Figs. 1 and 3, the entire content is the association range of mutual links. Then, you can move the marker that indicates that the entire content is the range of association and make it a marker that limits the range. By similarly moving the marker that indicates a mutual link to another range or another content and specifying the range, it is also possible to add a "link to further content", and associate and combine multiple contents. It can also be used as a notebook.

Screen example where the system implements claim 2 When displaying the "contents display area" as in Fig. 1, not only "markers associated in the past by the viewer and associated content" are displayed as annotations, but also " By simultaneously displaying a marker that indicates the association of another person with a high degree of similarity and the associated content, it is possible to add annotations in advance to areas where the viewer is likely to stumble. In addition to associations and content created by others with high similarity, content that has been highly rated by others with high similarity and reported that the question has been solved may be displayed with priority. . The degree of similarity with others can be calculated, for example, by using the Jaccard coefficient or the like for the similarity of the location of association with the content of the association source. By summing up the association locations used for calculation by intention, that is, the similarity of the association location intended for the question + the similarity of the association location intended for information + ..., the similarity including the intention is calculated. Also good. It is possible to provide only important information for the viewer by displaying only the other's associated content whose calculated similarity is above a certain value or whose similarity is higher than the predetermined ranking. If the link destination content is too long for display, you can also cut it to a certain number of characters centered on the link destination location (the middle point between the start and end positions of the reciprocal link in the link destination content) and use it as an annotation. .

If the creator of the content being browsed is the same as the creator of the link itself or the link destination content, it is a so-called creator's official link, so by always displaying it, it is possible to provide a stable supplementary explanation by the official. . If the creator and the viewer of the content being viewed are the same, that is, when viewing content created by oneself, by making it possible to view all the associations, it is possible to check all the reputations of the self-made content. If the creator of the association itself or the content to be associated is the same as the viewer, it is so-called self-annotation. Of course, display/non-display according to the relationship between the above-mentioned various subjects may be changed by setting as the viewer likes. Both of these display availability and content display of others with high similarity to the viewer may be adopted. You may use the similarity of a subject.

Example of a screen where the system implements claim 3 When there is an association intended to ask a question, the system of claim 3 can selectively issue an answer request to a person who is likely to be able to answer the question. Based on this, it is an example of a screen presented to a person who has issued an answer request. The name of the content for which the question was asked in the ``name of the content for which an answer was requested'', and the range of association with the intended question and its surroundings in the ``where the answer was requested and its surrounding context'' is the ``associated question'' Since the content of the question is displayed in "Content", it is possible to collectively judge whether the question is likely to be asked here. The display order of the answer requests in this list can be the order of creation time of the question content, the order of the time the answer request was made, or the order of content comprehension level indicating the degree of possibility of answering, which will be explained later with reference to FIG. 11 . When it is determined that an answer is possible, by pressing the "Answer button", the location of the linked content that intends to ask a question can be browsed, and the content that intends to provide information can be added there as shown in Fig. 1. By moving to , it is possible to respond smoothly.

An example of a database that manages the content itself "Name" is the title of the content, "Type" is the type of content, "Subject" is the creator of the content, and "Content" is the content itself and the details of the content. Identifiers such as file paths and URLs are recorded. When using in a wide range such as the whole world, it is conceivable that the "name" may be duplicated, so an ID column or the like that can be uniquely identified may be added separately. According to the nature of the present invention, the "type" can be anything that can specify a range within the content, such as text, image, and moving image. "Creator" basically identifies the creator, but it can be any entity that can maintain continuity, such as a creation group or brand. In addition to this, there may be additional information such as the content creation time and the identifier of the content of the previous version. In particular, if you have the identifier of the previous version of the content, you can manage the version of the content.

An example of a database that manages associations One of the content identifiers associated with "source content" is associated with "original content location" The location in the original content is associated with "authoring subject" The intent of the association of the destination content with the source content is recorded in , the content identifier of the other side of the association is recorded in 'related destination content', and the location in the destination content associated with 'destination content location' is recorded. It is the core data of the present invention, and in claim 1, information will be displayed when the association source or destination content and the source or destination content location overlap or are related to the association being created. In claim 2, the association source or destination content and the location of the source or destination content are aggregated for each creator, and the associations that are similar between the viewer and the association creator are prioritized and automatically displayed as annotations in the content. do. In claim 3, it is used for extracting a subject who has made many associations with intentions other than questions, that is, who has a high degree of understanding of the content.

Flowchart example for implementing claim 1 Triggered by "manually specifying a range on the content" for the "content display state", "contents including other people associated with a range overlapping the specified range Search for. For example, if the 5th to 10th characters are selected, the 4th to 6th characters, the 0th to 10th characters, the 6th to 8th characters, and the 8th to 12th characters are related. , means that the associated content will be included in the search results. Conversely, associations in the range of 2nd to 5th characters and in the range of 11th to 20th characters are not included in the search results. After that, "the display of a plurality of associated contents with position adjustment based on the relative relationship between the start and end positions of the specified range and the retrieved associated range" is performed. Here, as shown in FIG. 2, the position is adjusted by shifting the left and right positions of the "related content in the specified range" and the "content editing area being created" with respect to the specified range. In other words, if the associated position on the original content of the "content editing area being created" is, for example, the 5th to 10th characters, the first "related content in the specified range" is associated with the 4th to 8th characters on the original content. The 2nd and 3rd characters are associated with the 4th to 12th characters on the original content. By expressing the relative position information in this natural form, it is possible to simultaneously grasp information with a deep relationship in which the range of association completely matches, and information with a loose relationship in which the range of association is close at the same time.

An example of a flow chart for implementing claim 2 When there is a "related content display request", "extract information relating to the content being displayed". That is, if the content being displayed is the 100th to 200th characters of the content A, the association information having a range overlapping even one character is extracted. This "processing for each extracted association information" is performed as follows. First, a determination is made as to whether the association subject is the viewing subject or the content subject being viewed. This is a measure for preferentially displaying annotations given by the viewer himself or the author of the content being viewed. In the case of Yes in this conditional branch, "display the corresponding association on the content". As for the method of display, for example, it is possible to overlay display on the originally displayed content. In the case of No in this conditional branch, "calculate the degree of similarity between the browsing subject and the corresponding associated subject from the similarity of the location and intention of the association". A similarity measure between sets is available for this computation. For example, when the Jaccard coefficient is used, the association locations of subject A are the 4th to 6th characters of content C and the 8th to 14th characters of content D, and the association locations of subject B are the 5th to 8th characters of content C and content D. , the degree of similarity between subjects A and B = (association location of subject A∩association location of subject B)/(association location of subject A∪association location of subject B) = (content C 5th to 6th characters + 8th to 14th characters of content D) / (4th to 8th characters of content C + 6th to 20th characters of content D) = (2 characters + 7 characters) / (5 characters + 15 characters) = 0.45. It is also possible to calculate a more precise degree of similarity by taking into account the intention as well, such as the degree of similarity only for the linking points intended for questions + the degree of similarity only for the linking points intended for information. The degree of similarity of browsed content may be additionally used. This can be done by calculating the similarity of the browsed contents by the similarity between sets and adding them. Reading time may be taken into consideration. This can be done by limiting the associations used for the calculation to the 100 recent creations or created in the last month. It is also possible to assign a higher weight to the most recent rather than restrictive. Consideration of this browsing time makes it possible to reduce the amount of calculation and to perform similarity calculation in consideration of time characteristics. You may use the relationship with the content of an association source. This can be done by limiting the association used for similarity calculation to only the contents of the association source, only the contents of the same creator (author), or only the same field (mathematics, etc.). By taking into consideration the relationship with the content being displayed, only the content that is closely related can be used for calculation, making it possible to reduce the amount of calculation and remove noise derived from content that is less relevant. After calculating the degree of similarity, it is determined whether the degree of similarity between the browsing subject and the related subject is equal to or greater than a predetermined value. If the degree of similarity is equal to or higher than a predetermined value, for example, 0.3 or higher, or if it is within the top 10 ranks, "the relevant association is displayed on the content" as indicated by the conditional branch Yes. In the case of No, the display is skipped and it is determined whether or not "the association information still remains". If it remains, it repeats from "processing for each extracted association information", and if it does not remain, the process ends.

Flowchart example for implementing claim 3 When there is "occurrence of an association with the intention of asking a new question", first Only people who recently browsed the question part are extracted as answer request candidates. Next, "calculate the content comprehension level of the viewing subject by the number of intentions other than questions in association with the content". This is a process of judging whether or not the content is likely to have knowledge capable of answering the question, that is, the degree of understanding of the content, rather than simply viewing the content. Here, it was considered that the more content that had intentions other than questions, that is, content that was a supplement to the content on its own, was associated with the content, the higher the degree of content comprehension. Of course, other scales that correlate with content comprehension may be used. For example, instead of using the number of content other than questions, it is possible to use the number of content that has been evaluated by other viewers, and use the amount of content that is useful to others as an indicator of content comprehension. . In addition, the fact that the last viewing time of the linking part is late, the amount of viewing of the linking source content, and the like can also be indicators of understanding of the content. Next, "process the viewers in descending order of content comprehension". First, determine whether the question is "New question solved?" "Resolved" means that the subject who made the association with the intention of creating a new question adds content that will be the answer by himself or presses a button that the question has been solved with other content, including those created by others. can be determined by Receiving responses to an already understood part over and over can be a nuisance for readers, so it is convenient for readers to not ask for answers and stop receiving new answers if the problem is marked as solved. , a good answer can also be a means to get more information about any one. If "solved" is Yes, the process is terminated, and if "no", "request the corresponding browsing subject to answer the new question". Through the extraction and classification up to this point, it is possible to request a response only when the question is unsolved, from the viewer who has recently viewed the content and who has a high degree of understanding of the content and is likely to be able to respond. Next, by "waiting for a specified time", instead of requesting a large number of responses all at once, people who are likely to request responses will respond and buy time until the question is resolved. After that, if "Are there still viewing subjects remaining?" do.

Software Configuration Example for Implementing the Present Invention In order to implement claim 1, functional blocks like the upper half of FIG. 12 are required. See FIG. 7 for the content DB and FIG. 8 for the association DB. The viewing history DB manages a triplet of content ID, viewer ID, and last viewing time. When a range is specified for the content being browsed, the 'contents/association creating means' adds new associations and contents to the 'association DB' and 'contents DB' from the data input in FIG. In addition, the 'associated multiple content display means' searches the 'association DB' and the 'content DB' for related contents, and displays the contents related to the specified range as shown in FIG. Unread content may be highlighted by using the “browsing history DB”.

In order to carry out claim 2, functional blocks such as those shown in the lower left of FIG. 12 are required. Based on conditions such as only the contents being browsed, or the same creator, or the same field, etc., a set of contents to be calculated is extracted from the "content DB", the association with them is extracted from the "association DB", and the "association similarity calculation means , the ``subject similarity calculation means'' is configured, and the contents associated with the subjects that have obtained a high degree of similarity are presented to the user by the ``content display means''. To update the contents of a browsing history DB when displaying contents. See description of FIG. 10 for these details.

In order to implement claim 3, functional blocks such as those shown in the lower right of FIG. 12 are required. When an association intended for a new question is added to the 'association DB', a set of contents to be calculated is extracted from the 'content DB' under conditions such as only the content of the association source, or the same creator, or the same field. The "association DB" is searched for the number of associations with intentions other than questions to these, and the number is collected for each subject to form a "content comprehension degree calculation means." It is possible to issue an answer request to a person who is likely to be able to answer efficiently by executing the "answer request means" such as e-mail or FIG. See description of FIG. 11 for these details.

Example of hardware configuration diagram for implementing the present invention Client machines such as PCs and mobile terminals that users directly access are connected to the "network", and data for the core processing of the claims of the present invention are stored. A "server" that connects the "databases" connected to each other is also connected. A LAN, the Internet, or the like can be used as the “network”. The client machine only needs to be able to create and browse content and associations. Specifically, it is sufficient if it has a display device such as a liquid crystal display, a pointing device such as a mouse, and a text input device such as a keyboard. In addition, a CPU, a RAM, a secondary storage device, etc. are held, and each of the "contents/association creation means", "associated multiple contents display means", "contents display means", and "response request means" in FIG. It functions as an interface for inputting/outputting display/input information to the user. "Database" holds various DBs in FIG. The "server" implements the other processes in FIG.

A client-server configuration as shown in FIG. 13 is effective for carrying out the present invention. Of course, it is also possible to distribute and manage the "database" and "server", or to change the embodiment to P2P or the like. An example in a client-server configuration is summarized below.

If you want to add a question or an annotation to the e-book content you are browsing, you can specify the range of the content you are browsing with a pointing device such as a mouse, and then go through the process of the flowchart in FIG. The viewer can be presented with the creation screen and the associated content created by others with high relevance as shown in FIG. By implementing claim 1, multiple subjects can freely set questions, comments, etc. within the range where overlap is allowed, and not only simple range matching, but also other's information given to the overlapping range. can be used flexibly.

If claim 1 is straightforwardly implemented, a large number of annoying contents of malicious users and contents that are not interested in viewers whose knowledge and tastes differ greatly from those of viewers will be displayed. In order to reduce the noise for the viewer, claim 2 can be implemented according to the flow chart of FIG. The invention of claim 2 functions effectively as shown in FIG. 5 not only when a range is specified for contents, but also when selecting contents to be displayed when nothing is specified. In this case, since there is no action that requests information such as specifying a range, the negative impression of noise tends to be stronger, so set a strict default threshold, and use a smaller number and a higher degree of confidence that it will be useful. You can also narrow down to content.

If the content is already associated with a large number of items, there is a high possibility that information that already answers the question can be found according to claims 1 and 2. However, if it is new content that cannot be found or is almost unrelated in the first place, there will be no valid answer even if the existing content is searched with claims 1 and 2. It is the third aspect of the present invention to request others to answer such unsolved questions in the order in which they are likely to be able to answer them. FIG. 6 shows the request collection screen. Although the response is requested on the screen here, other means such as e-mail may also be used. If you use email together, you will be asked to register an email address for your user account, and you will be notified there. In the flow chart of Fig. 11, the processing is to list all the candidate respondents when asking a new question. become. This is to always add the content creator to the answer candidate, and after "waiting for the specified time" in FIG. If there is a subject other than the viewing subject who has already requested a reply, it can be handled by modifying it to continue the process. This latter modification allows new viewers to appear during the "waiting for a specified time" period, giving later viewers a chance to answer the first viewer's question.

INDUSTRIAL APPLICABILITY The present invention can be implemented in the publishing industry, communication industry, education industry, etc., and has clear industrial applicability.

The reference numerals themselves in the drawings are simply explained as meaningful text, and since they have already been described in detail in the brief explanation of the drawings, they will be omitted.

Claims (4)

A system that simultaneously visualizes multiple linked contents according to the degree of matching with each linked range based on the specified range for a structure that specifies a content range and associates different contents. For a structure in which another content is associated by specifying the range of content, using the intention of association and the relationship between the four parties of the content management entity of the originating content, the management entity of the association management entity, the content management entity of the destination destination, and the viewing entity, A system for determining whether to visualize associations and associated content Triggered by the addition of a new association with the intention of an unsolved question to the structure that specifies the scope of the content and associates another content, it has an intention other than a question to the associated content. A system that prioritizes the number of associations or the number of high evaluations from them and requests responses to the associations that triggered them A computer program functioning as each means of the system according to claims 1 to 3

Images