Essay support device, essay support program, and user terminal

Abstract

To provide an essay support device, an essay support program, and a user terminal capable of presenting a user with useful information regarding an essay structure comprising an argument and a plurality of supporting points.SOLUTION: An essay support device (14) instructs information output means (22) to display an evaluation result screen 100 containing an evaluation result regarding persuasive strength of an entire or a part of essay structure, or proposal content according to the evaluation result. The evaluation result screen 100 includes: a title column 102 indicating an essay's title; an essay information column 104 containing visual information on a plurality of essay elements; and an evaluation information column 106 containing evaluation result information (54) indicating the evaluation result or proposal information (56) indicating the proposal content.SELECTED DRAWING: Figure 9

Description

[Technical Field]

[0001] The present invention relates to an essay support device, an essay support program, and a user terminal. [Background technology]

[0002] Recently, with the development of artificial intelligence technology, natural language processing using machine learning has become widespread. For example, evaluation techniques for evaluating various texts through natural language processing have been proposed.

[0003] Patent document 1 discloses a method for estimating the discourse structure of text that includes tree structures and non-tree structures by obtaining a sequence of sentence vectors corresponding to each sentence included in the text, and estimating the discourse relationships between each sentence using a bilinear attention model for the sequence of sentence vectors. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2021-163134 Summary of the Invention [Problem to be solved by the invention]

[0005] Incidentally, in an argument structure consisting of a premise assertion and multiple pieces of evidence that affirm or deny the assertion, the persuasiveness of the assertion can be increased or decreased depending on the content, order, or combination of the evidence. For example, quantitatively evaluating the persuasiveness of an assertion and presenting the evaluation results can be a useful support service for users.

[0006] The present invention has been made in consideration of such problems, and its purpose is to provide an essay support device, essay support program, and essay support system that can accurately evaluate the persuasiveness of an argument in an essay structure consisting of an argument and multiple evidence. [Means for solving the problem]

[0007] The essay support device in a first aspect of the present invention comprises: a raw score assignment unit that assigns raw scores to one or more of the grounds of an essay, each of which indicates the persuasiveness of the argument, with respect to an essay structure consisting of an argument that is the premise of the essay and a plurality of grounds for either affirming or denying the argument; a score calculation unit that calculates a score corresponding to the first node or the second node by repeating a synthesis process that combines the raw scores assigned by the raw score assignment unit from the lower to the upper hierarchical levels of the tree structure identified from the tree structure data, with respect to tree structure data indicating a tree structure in which a first node corresponding to an argument element that indicates the content of the argument is at the upper level and a second node corresponding to an argument element that indicates the content of the grounds is at the lower level; and an evaluation processing unit that performs an evaluation process that evaluates the persuasiveness of the entire essay structure or a part of it based on the scores calculated by the score calculation unit.

[0008] In a second aspect of the present invention, the essay support program causes one or more computers to execute the following steps: an assignment step for assigning raw scores to one or more of the arguments indicating the persuasiveness of an argument, which is based on an argument that forms the premise of an argument and a plurality of arguments that affirm or deny the argument; a calculation step for calculating a score corresponding to the first node or the second node by repeating a synthesis process for combining the assigned raw scores from the lower to the upper levels of the tree structure identified from the tree structure data, which indicates a tree structure in which a first node corresponding to an argument element that indicates the content of the argument is at the upper level and a second node corresponding to an argument element that indicates the content of the arguments is at the lower level; and an evaluation step for performing an evaluation process to evaluate the persuasiveness of the entire essay structure or a portion of it based on the calculated scores.

[0009] The essay support system in a third aspect of the present invention comprises a user terminal owned by a user and an analysis server capable of communicating with the user terminal, wherein the analysis server executes the following steps: an acquisition step of acquiring, from the user terminal, data to be evaluated that indicates an essay structure consisting of an argument that forms the premise of the essay and a plurality of reasons for either affirming or denying the argument; an assignment step of assigning raw scores to one or more of the reasons for the acquired data to be evaluated, which raw scores indicate the degree of persuasiveness of the argument; a calculation step of calculating a score corresponding to the first node or the second node by repeating a synthesis process for tree structure data indicating a tree structure in which a first node corresponding to an argument element that indicates the content of the argument is at the top and a second node corresponding to an argument element that indicates the content of the reasons, from the lower to the upper hierarchical levels of the tree structure identified from the tree structure data; and an evaluation step of performing an evaluation process to evaluate the degree of persuasiveness of the essay structure as a whole or in part based on the calculated scores. [Effects of the Invention]

[0010] According to the present invention, it is possible to accurately evaluate the persuasiveness of an argument in a discourse structure consisting of an argument and a plurality of evidences. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a diagram illustrating the overall configuration of an essay support system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a block diagram of the essay support device shown in FIG. [Figure 3] 2 is a flowchart showing an example of an evaluation operation by the essay support system of FIG. 1. [Figure 4] FIG. 2 is a diagram showing a first example of an input screen displayed on the output device of FIG. [Figure 5] 1. FIG. 4 is a diagram showing a second example of the input screen displayed on the output device of FIG. [Figure 6] FIG. 10 is a diagram illustrating an example of a tree structure before a score is calculated. [Figure 7]FIG. 10 is a diagram illustrating an example of a tree structure after calculation of scores. [Figure 8] 3 is a diagram showing an example of a data structure of the tree structure data of FIG. 2. FIG. [Figure 9] FIG. 2 is a diagram showing a first example of an evaluation result screen displayed on the output device of FIG. [Figure 10] FIG. 10 is a diagram showing a second example of the evaluation result screen displayed on the output device of FIG. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings. To facilitate understanding of the description, the same components in each drawing will be denoted by the same reference numerals as much as possible, and duplicate explanations will be omitted. Furthermore, the term "part" may be replaced with other terms such as unit, module, device, or element.

[0013] [Configuration of Essay Support System 10] <Overall structure> FIG. 1 is a diagram illustrating the overall configuration of an essay support system 10 according to one embodiment of the present invention. The essay support system 10 is configured to provide an "essay support service" for supporting users in writing essays. This "essay" can be widely applied to presentations for various purposes, uses, or aspects. Examples of presentations include creating talk scripts, writing reports, creating content marketing articles, writing papers, or creating business plans.

[0014] Here, we assume that the argument structure consists of an "assertion" that is the premise of the argument, and "evidence" that affirms or denies the assertion. In the following, to clearly distinguish the attributes of the evidence, evidence that affirms the assertion will be referred to as "affirmative evidence," and evidence that denies the assertion will be referred to as "negative evidence."

[0015] Specifically, this essay support system 10 includes one or more user terminals 12 and an analysis server 14 (corresponding to an "essay support device"). Each user terminal 12 is configured to be able to communicate with the analysis server 14 via a network NT. This allows the analysis server 14 to acquire evaluation target data D1 from the user terminal 12 and provide output data D2 to the user terminal 12.

[0016] The user terminal 12 is a computer owned by a user of the writing support service. The user terminal 12 may be, for example, a portable device such as a tablet, laptop, smartphone, or wearable device, or a stationary device such as a personal computer. Specifically, the user terminal 12 includes a processor 16, memory 18, an input device 20, an output device 22, and a communication unit 24. A user interface (UI) is constructed by combining the input function of the input device 20 and the output function of the output device 22.

[0017] The input device 20 is configured to allow various pieces of information related to the discourse structure to be input. The input device 20 may be, for example, a mouse, a keyboard, a touch sensor, a microphone, or a combination thereof.

[0018] The output device 22 is configured to be able to output information in a manner that stimulates the five senses, such as sight, hearing, touch, etc. The output device 22 may be, for example, a display, a lamp, a speaker, a vibrator, or a combination thereof.

[0019] The analysis server 14 is a computer that performs overall control over the above-mentioned essay support service, and may be either a cloud-based or on-premise type. Here, the analysis server 14 is illustrated as a single computer, but the analysis server 14 may instead be a group of computers that form a distributed system.

[0020] <Configuration of analysis server 14> Fig. 2 is a block diagram of the analysis server 14 shown in Fig. 1. The analysis server 14 includes a communication unit 30, a control unit 32, and a storage unit 34.

[0021] The communication unit 30 is an interface that transmits and receives electrical signals to and from external devices. As a result, the analysis server 14 receives the evaluation target data D1 from the user terminal 12 and transmits the output data D2 that it has generated to the user terminal 12.

[0022] The control unit 32 is configured with a processor including a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit). The control unit 32 reads and executes programs and data stored in the storage unit 34, thereby functioning as a data acquisition unit 40, a raw score assignment unit 42, a data generation unit 44, a score calculation unit 46, an evaluation processing unit 48, and an output instruction unit 50.

[0023] The data acquisition unit 40 acquires various data related to the evaluation process of the narrative structure, for example, the evaluation target data D1. The data acquisition unit 40 may acquire the data via communication from an external device including the user terminal 12, or may acquire the data by reading it from the storage unit 34.

[0024] The raw score assigning unit 42 assigns raw scores to one or more of the multiple grounds for the argument structure identified from the evaluation target data D1 acquired by the data acquiring unit 40. This "raw score" refers to an index that indicates the persuasiveness of an argument. The raw score is defined so that the larger the value, the more persuasive the argument is, and the smaller the value, the less persuasive the argument is. Furthermore, the raw score may be assigned manually based on a subjective evaluation by a user, or may be assigned automatically based on an objective evaluation through regression calculations.

[0025] The raw score assigning unit 42 may assign raw scores to evidence (or argument elements) before the tree structure is constructed by the data generating unit 44, or may assign raw scores to evidence (or nodes) after the tree structure is constructed. Alternatively, the raw score assigning unit 42 may assign raw scores to evidence (or nodes) simultaneously with the construction of the tree structure by the data generating unit 44.

[0026] The data generation unit 44 generates tree structure data 52 indicating the narrative structure (here, a tree structure) of the evaluation target from the evaluation target data D1 acquired by the data acquisition unit 40. This tree structure has a data structure in which nodes (hereinafter also referred to as "first nodes") corresponding to narrative elements indicating the content of the argument are at the top and nodes (hereinafter also referred to as "second nodes") corresponding to narrative elements indicating the content of the evidence are at the bottom. This tree structure may be constructed manually based on the user's subjective opinion, or may be constructed automatically using various methods including multidimensional classification, structural analysis, and neural networks.

[0027] The data generating unit 44 may generate the tree-structured data 52 before the raw scores are assigned by the raw score assigning unit 42, or may generate the tree-structured data 52 after the raw scores are assigned. Alternatively, the data generating unit 44 may generate the tree-structured data 52 simultaneously with the assignment by the raw score assigning unit 42.

[0028] The score calculation unit 46 performs information processing (hereinafter referred to as "combining processing") to combine the raw scores assigned by the raw score assigning unit 42. Examples of combining processing include [1] "statistical processing" for obtaining statistical values ​​including maximum, minimum, mode, and median values, [2] "regression processing" for obtaining a composite value using a regression calculation model, and [3] "template matching processing" for selecting identical or similar patterns from multiple types of input / output patterns to obtain a composite value. Examples of regression calculation models include linear regression (simple regression or multiple regression), ridge regression, lasso regression, elastic net regression, logistic regression, random forest, gradient boosting tree, support vector machine, and neural network regression. Model parameters of the regression calculation model are determined by statistical methods or machine learning (or deep learning).

[0029] The score calculation unit 46 repeats the above-described synthesis process from the lower to the upper layers of the tree structure identified from the tree structure data 52 to calculate the score corresponding to the first node or the second node. The score calculation unit 46 may recursively perform the same synthesis process regardless of the layer of the tree structure, or may sequentially perform different synthesis processes according to the layer of the tree structure. Note that, when the raw score already assigned to the same node differs from the score obtained by the synthesis process, the score calculation unit 46 may [1] select either the raw score or the score, or [2] select another value obtained by combining the raw score and the score (i.e., a composite value).

[0030] The evaluation processing unit 48 performs information processing (hereinafter also referred to as "evaluation processing") to evaluate the persuasiveness of the entire or part of the essay structure based on the scores calculated by the score calculation unit 46. Examples of evaluation processing include [1] an "overall evaluation" calculation that evaluates the persuasiveness of the entire essay structure from the score of the first node in the tree structure, or [2] a "relative evaluation" calculation that evaluates the relative persuasiveness of parts of the essay structure from the score of the second node belonging to the same hierarchical level in the tree structure. Through this evaluation processing, evaluation result information 54 or proposal information 56, which will be described later, is generated.

[0031] The output instruction unit 50 instructs the information output means to output the evaluation results by the evaluation processing unit 48 or the contents of a proposal based on the evaluation results. This "information output means" may be the output device 22 provided in an external device such as the user terminal 12 (FIG. 1), or may be an output device (not shown) provided in the analysis server 14. The form of output may be, for example, one that stimulates the five senses, such as sight, hearing, and touch.

[0032] The storage unit 34 stores programs and data necessary for the control unit 32 to control each component. The storage unit 34 is a non-transitory, computer-readable recording medium. In the example of Fig. 2, the storage unit 34 stores tree-structure data 52, evaluation result information 54, and proposal information 56 in addition to the evaluation target data D1 and output data D2.

[0033] The evaluation target data D1 includes information about the essay structure to be evaluated (hereinafter referred to as structural information). Examples of structural information include the content of the essay element, the attributes of the essay element (assertion, affirmative evidence, negative evidence), the superior-subordinate relationship with other essay elements, and raw scores.

[0034] The tree structure data 52 includes various information for identifying the tree structure 90 (see FIGS. 6 and 7) indicated by the narrative structure. The narrative elements may be in the form of a string of words, a single or multiple sentences, a list, or a diagram. Parameters relating to the shape of the tree structure 90 (e.g., the number of nodes, the depth of the hierarchy, etc.) may be determined arbitrarily.

[0035] The evaluation result information 54 includes, for example, [1] "time information" such as date and time, [2] "structural information" regarding the essay structure, [3] "score information" regarding the score for all or part of the essay structure, or [4] "history information" regarding the past history of scores.

[0036] The proposal information 56 includes, for example, [1] "time information" such as date and time, [2] "essay element information" about essay elements that were evaluated as having relatively low scores, or [3] "measure information" about measures to improve the score. Examples of measures include adding, changing, or deleting evidence, changing the order of evidence, or changing the combination of evidence.

[0037] The output data D2 includes various information to be output to the user. Examples of the output data D2 include the evaluation result information 54 or the proposal information 56 described above.

[0038] [Operation of Essay Support System 10] The essay support system 10 in this embodiment is configured as described above. Next, the operation of the essay support system 10, particularly the evaluation operation by the analysis server 14, will be described with reference to FIGS.

[0039] Fig. 3 is a flowchart showing an example of the evaluation operation by the essay support system 10 of Fig. 1. Steps SP10 and SP28 of the flowchart are performed by the user terminal 12. Steps SP12 to SP26 of the flowchart are executed by the analysis server 14.

[0040] In step SP10, the user terminal 12, upon starting up the web application, causes the output device 22 to display a screen for inputting text information to be evaluated (hereinafter referred to as "input screens 60, 80").

[0041] Fig. 4 is a diagram showing a first example of an input screen 60 displayed on the output device 22 in Fig. 1. The input screen 60 is provided with an assertion input field 62 and two types of grounds input fields 64, 66.

[0042] The claim input field 62 includes a user control (here, a text box) that allows input of a statement element that indicates the content of the claim. In the example of Fig. 4, the sentence "Patents should be applied for proactively" is input as the claim.

[0043] The evidence input field 64 is configured to include user controls that allow input of argument elements that indicate the content of the negation evidence (here, "anticipated counterarguments"). In the example of FIG. 4, the evidence input field 64 is provided with auxiliary lines 71 that indicate the hierarchical level, a plus icon 72, and a button 73 labeled "Generated by AI." Selecting the plus icon 72 adds a new text box 74 for inputting the negation evidence. Selecting the button 73 allows a so-called generation AI to automatically generate typical negation evidence for the claim.

[0044] The rationale input field 66 is configured to include user controls that allow input of narrative elements that indicate the content of the affirmative rationale (simply "rationale" in this case). In the example of FIG. 4, in addition to the auxiliary line 71 and plus icon 72 described above, the rationale input field 66 is provided with a text box 74 and a button 75 labeled "Delete." By selecting the plus icon 72, a new text box 74 for inputting the affirmative rationale is added. By editing the text box 74, a narrative element that indicates the content of the affirmative rationale is input. By selecting the button 75, the corresponding text box 74 is deleted.

[0045] Fig. 5 is a diagram showing a second example of the input screen 80 displayed on the output device 22 in Fig. 1. On the input screen 80, an essay information field 82, a raw score information field 84, and two buttons 86 and 88 are provided.

[0046] The statement information field 82 includes visible information about the statement structure (here, a tree structure) of the assertion and evidence. In the example of FIG. 5, the visible information is a hierarchical representation of multiple statement elements along with their attributes. Specifically, the statement information field 82 lists, from top to bottom, the following statements: [1] A patent should ultimately be applied for (assertion: first level), [2] It will lead to increased corporate value (evidence: second level), [3] It is often argued that it is advantageous for fundraising and loans (evidence: third level), [4] It will increase the possibility of M&A to acquire the patent itself or a team with development capabilities (evidence: third level), [5] Competition will intensify if another company acquires the patent (evidence: second level), [6] Product development that avoids the patent will be necessary (evidence: second level), [7] Profits may decline due to brand power (evidence: third level), and [8] Obtaining a patent alone does not necessarily result in sales, so smaller companies should focus on activities that have a significant impact on business sales, such as sales (evidence: second level).

[0047] The raw score information field 84 includes a raw score indicating the level of persuasiveness. In the example of FIG. 5, raw scores are assigned to all grounds belonging to the lowest level in each branch. For example, the third content is assigned the highest score of 5, while the seventh content is assigned the lowest score of 1. This means that "it is advantageous for fundraising and financing" is a relatively persuasive ground, while "brand power may reduce profits" is a relatively less persuasive ground. This raw score may be [1] a value manually entered by the user via an input field (not shown), or [2] a value automatically calculated by the user terminal 12 or the analysis server 14 in response to a user request.

[0048] By selecting the [Evaluate] button 86, it is possible to start the evaluation operation for the narrative structure displayed in the narrative information column 82. By selecting the [Cancel] button 88, it is possible to cancel the display of the input screens 60 and 80.

[0049] 3, the analysis server 14 checks whether a predetermined operation for requesting evaluation of the narrative structure (hereinafter referred to as "evaluation request operation") has been received. If the evaluation request operation has not yet been received (step SP12: NO), the analysis server 14 remains in step SP12 until the operation is received.

[0050] 5, when the user terminal 12 receives a selection operation of the [Evaluation] button 86, it generates evaluation target data D1 including the contents of the essay information column 82 and the raw score information column 84, and transmits a communication signal including the evaluation target data D1 (hereinafter, evaluation request signal) to the analysis server 14. Thereafter, the analysis server 14 receives the evaluation request signal from the user terminal 12 and proceeds to the next step SP14.

[0051] In step SP14, the analysis server 14 (more specifically, the data acquisition unit 40) acquires the evaluation target data D1 through reception of the evaluation request signal transmitted in step SP14.

[0052] In step SP16, the analysis server 14 (more specifically, the data generation unit 44) uses the evaluation target data D1 acquired in step SP14 to construct a tree structure 90 indicating the narrative structure of the evaluation target. This step SP18 includes [1] an assignment step (SP16A) of assigning raw scores to one or more grounds, and [2] a generation step (SP16B) of generating tree structure data 52.

[0053] FIG. 6 is a diagram showing an example of a tree structure 90 before the score is calculated. The tree structure 90 in FIG. 6 is a symbolic representation of the essay structure shown in the essay information column 82 in FIG. 5. Specifically, this tree structure 90 is composed of one first node 92, multiple second nodes 94, and multiple links 96. The first node 92 indicates an essay element (ID=001) that indicates the content of the argument. The second nodes 94 indicate essay elements (ID=002 to 008) that indicate the content of the evidence. The numbers written in the upper right or upper left of the second node 94 correspond to the raw score assigned to each evidence.

[0054] In step SP18 of Figure 3, the analysis server 14 (more specifically, the score calculation unit 46) repeats the synthesis process of synthesizing raw scores for the tree structure 90 constructed in step SP16, and calculates the score corresponding to the first node 92 or the second node 94.

[0055] Fig. 7 is a diagram showing an example of a tree structure 90 after the scores have been calculated. In the example of Fig. 7, the synthesis process is performed by [1] a first operation that selects the maximum value from among the raw scores, and [2] a second operation that selects only the maximum and minimum values ​​from among the raw scores and adds them together. The numbers in boxes correspond to the scores for each node calculated through the synthesis process.

[0056] For example, the score of the second node 94 (ID=002) is calculated through the first calculation. During this first calculation, the raw scores of the two second nodes 94 (ID=003, 004) directly below that node are referenced. Specifically, "5", which is the maximum of the two raw scores (5 and 3), is calculated as the score for "ID=002". Also, "2", which is the maximum of the raw scores (2 and 1) of "ID=006" and "ID=007", is calculated as the score for "ID=002".

[0057] Furthermore, the score for the first node 92 (ID=001) is calculated through a second calculation. During this second calculation, the scores or raw scores of the three second nodes 94 (ID=002, 005, 008) directly below that node are referenced. Specifically, the score for "ID=001" is calculated as "7", which is the sum of the maximum value "5" and the minimum value "2" of the three scores / raw scores (5, 2, and 4). In this way, a score is obtained that matches the reality of increasing persuasiveness by "accumulating evidence".

[0058] 3, the analysis server 14 (more specifically, the data generation unit 44) associates the score calculated in step SP18 with the corresponding first node 92 or second node 94. As a result, the tree structure data 52 is updated.

[0059] Fig. 8 is a diagram showing an example of the data structure of the tree structure data 52 of Fig. 2. This tree structure data 52 is data in a table format showing the correspondence between [1] "node ID" indicating the identification information of a node, [2] "statement element" indicating the content of an assertion or evidence, [3] "link information" indicating the connection relationship between nodes, [4] "attributes" of the statement element, [5] assigned "raw score," and [6] calculated "score." This tree structure data 52 can visually reproduce the tree structure 90 shown in Fig. 7.

[0060] 3, the analysis server 14 (more specifically, the evaluation processing unit 48) performs evaluation processing on the scores calculated in step SP20. Through this evaluation processing, evaluation result information 54 or proposal information 56 regarding the evaluation target data D1 is generated.

[0061] In step SP24, the analysis server 14 (more specifically, the data generation unit 44) generates output data D2 including the evaluation result information 54 or the proposal information 56 obtained in the evaluation process in step SP22.

[0062] In step SP26, the analysis server 14 (more specifically, the output instruction unit 50) instructs the corresponding user terminal 12 to output the output data D2 generated in step SP24. Specifically, the output instruction unit 50 of the analysis server 14 transmits a communication signal (hereinafter referred to as an evaluation response signal) including the output data D2 to the user terminal 12. Thereafter, the user terminal 12 receives the evaluation response signal from the analysis server 14 and proceeds to the next step SP28.

[0063] In step SP28, the user terminal 12 accepts the instruction given in step SP28 and outputs to the output device 22 the evaluation result information 54 or the proposal information 56 included in the output data D2.

[0064] Fig. 9 is a diagram showing a first example of an evaluation result screen 100 displayed on the output device 22 in Fig. 1. The evaluation result screen 100 has a title field 102 showing the title of the essay, an essay information field 104, and an evaluation information field 106.

[0065] The comment information column 104 includes visual information about the comment structure (here, a tree structure) of the assertion and the evidence. In the example of Fig. 9, the visual information is formed by hierarchically notating a plurality of comment elements.

[0066] The evaluation information field 106 includes evaluation result information 54 and suggestion information 56 regarding the essay structure in the essay information field 104. In the example of FIG. 9, the evaluation information field 106 displays suggestions such as "Adding a comparison to the next argument will make it more persuasive" or "Please add persuasive evidence to the next argument." For example, by selecting the hyperlink 108, specific explanations regarding "adding a comparison" or "adding evidence" are displayed in separate tabs or windows, respectively.

[0067] Fig. 10 is a diagram showing a second example of the evaluation result screen 110 displayed on the output device 22 in Fig. 1. In the example of Fig. 10, an argument is made in response to the assertion that "the development of chatbots should be entrusted to Company X." The evaluation result screen 110 has a title field 112 showing the title of the argument, and an argument information field 114.

[0068] The essay information column 114 includes visible information about the essay structure (here, a tree structure) of the assertion and evidence. Marks 116 are attached to highlight four specific essay elements that received high scores. Specifically, the specific elements are [1] "numerous achievements," [2] "track record of operations at the level of a listed company," [3] "we promise surprisingly fast delivery," and [4] "we will deliver your first model in 60 days!"

[0069] By visually checking the evaluation result screens 100, 110, the user can grasp at a glance the quality of the essay structure he or she has created, as well as areas for improvement. In this way, the essay support system 10 completes the evaluation operation for evaluating the essay structure constructed by the user, in accordance with the flowchart shown in Figure 3.

[0070] [Summary of the embodiment] As described above, the essay support system 10 in this embodiment includes a user terminal 12 owned by a user and an essay support device (here, an analysis server 14) capable of communicating with the user terminal 12. The analysis server 14 includes a raw score assigning unit 42 that assigns raw scores indicating the persuasiveness of an argument to one or more arguments in an essay structure consisting of an argument and multiple arguments, a data generating unit 44 that generates tree structure data 52 representing a tree structure 90 in which a first node 92 corresponding to an argument element indicating the content of the argument is at the top and a second node 94 corresponding to an argument element indicating the content of the arguments is at the bottom, a score calculating unit 46 that calculates a score corresponding to the first node 92 or the second node 94 by repeating a synthesis process that synthesizes raw scores from the lower to the upper levels of the tree structure 90 identified from the tree structure data 52, and an evaluation processing unit 48 that performs an evaluation process to evaluate the persuasiveness of the essay structure as a whole or in part based on the scores.

[0071] According to the essay support method and essay support program of this embodiment, one or more computers (or processors) execute an assignment step (SP16A) of assigning raw scores indicating the persuasiveness of an argument to one or more grounds, a generation step (SP16B) of generating tree structure data 52 indicating a tree structure 90, a calculation step (SP18) of repeatedly combining raw scores from lower to higher levels of the tree structure 90 to calculate a score corresponding to the first node 92 or the second node 94, and an evaluation step (SP22) of performing evaluation processing to evaluate the persuasiveness of the entire essay structure or a part of it based on the scores.

[0072] In this way, by repeatedly performing a synthesis process that synthesizes raw scores from the lower to the higher levels of the tree structure 90 to calculate the score, a score can be obtained that matches the reality that persuasiveness increases through the accumulation of evidence, making it possible to accurately evaluate the persuasiveness of an argument.

[0073] The analysis server 14 may further include an output instruction unit 50 that instructs the information output means (here, the output device 22) to output the evaluation results by the evaluation processing unit 48 or suggestions based on the evaluation results, in addition to or separately from the display of the tree structure 90. By outputting the evaluation results or suggestions, the user can be encouraged to improve their essay structure.

[0074] The evaluation process may also include a calculation to evaluate the persuasiveness of the entire argument structure from the score of the first node 92 in the tree structure 90. Alternatively, the evaluation process may include a calculation to relatively evaluate the persuasiveness of parts of the argument structure from the score of the second node 94 that belongs to the same hierarchical level in the tree structure 90.

[0075] Furthermore, the output instruction unit 50 may instruct the display of an essay structure in which, among the multiple essay elements constituting the essay structure, essay elements corresponding to second nodes with relatively high or low scores are visually emphasized. This allows the user to easily grasp which essay elements are relatively persuasive or low by visually checking the essay structure.

[0076] The present invention is not limited to the above-described embodiment, and can be freely modified without departing from the spirit and scope of the present invention. Alternatively, the respective configurations may be arbitrarily combined within the scope of no technical contradiction. Alternatively, the execution or execution order of each step constituting the flowchart may be changed within the scope of no technical contradiction.

[0077] In the above embodiment, the analysis server 14 generates the tree-structured data 52, but the subject of the generation process is not limited to this. For example, the user terminal 12 may generate the tree-structured data 52 on behalf of the analysis server 14, and then transmit the evaluation target data D1 including the tree-structured data 52 to the analysis server 14. In this case, the generation process by the data generation unit 44 is omitted through the acquisition of the tree-structured data 52 by the data acquisition unit 40. [Explanation of symbols]

[0078] 10. Essay writing support system, 12. User terminal, 14. Analysis server (essay writing support device), 22. Output device (information output means), 40. Data acquisition unit, 42. Raw score assignment unit, 44. Data generation unit, 46. Score calculation unit, 48. Evaluation processing unit, 50. Output instruction unit, 52. Tree structure data, 54. Evaluation result information, 56. Proposal information, 60, 80. Input screen, 90. Tree structure, 92. First node, 94. Second node, 96. Link, 100, 110. Evaluation result screen, D1. Evaluation target data, D2. Output data

Claims

1. an output instruction unit that instructs an information output means to display an evaluation result screen containing an evaluation result regarding the persuasiveness of the entire or a part of an argument structure consisting of an argument that is the premise of the argument and a plurality of reasons for affirming or denying the argument, or a proposal content based on the evaluation result; The evaluation result screen includes: a title field indicating the title of said essay; an essay information column containing visible information about a plurality of essay elements that constitute the essay structure; an evaluation information field including evaluation result information indicating the evaluation result or proposal information indicating the proposal content,

2. the logical structure is a tree structure in which a first node corresponding to a statement element indicating the content of the claim is at a higher level and a second node corresponding to a statement element indicating the content of the grounds is at a lower level; The visible information is formed by highlighting, among a plurality of argument elements constituting the argument structure, argument elements having relatively high or low scores indicating the persuasiveness of the argument. The essay support device according to claim 1 .

3. a raw score assigning unit that assigns raw scores indicating the persuasiveness of the claim to one or more of the grounds; a score calculation unit that calculates the score corresponding to the first node or the second node by repeating a synthesis process of synthesizing the raw scores assigned by the raw score assignment unit from lower to higher hierarchical levels of the tree structure; and Further provided with 3. The essay support device according to claim 2.

4. the output instruction unit further instructs the information output means to display an input screen for inputting text information to be evaluated; The input screen includes: The essay information column; a raw score information field containing the raw scores corresponding to the essay elements; are provided, 4. The essay support device according to claim 3.

5. the input screen is provided with an input field in which the raw score value can be input through a user operation; 5. The essay support device according to claim 4.

6. With regard to an argument structure consisting of an argument that is the premise of an argument and a plurality of reasons for affirming or denying the argument, an instruction step is executed on one or more computers to instruct an information output means to display an evaluation result screen including an evaluation result regarding the persuasiveness of the entire or a part of the argument structure or a proposal based on the evaluation result; The evaluation result screen includes: a title field indicating the title of said essay; an essay information column containing visible information about a plurality of essay elements that constitute the essay structure; an evaluation information field including evaluation result information indicating the evaluation result or proposal information indicating the proposal content.

7. an output device for displaying an evaluation result screen including evaluation results regarding the persuasiveness of the entire or a part of an argument structure consisting of an argument that is the premise of the argument and a plurality of reasons for affirming or denying the argument, or proposals based on the evaluation results; The evaluation result screen includes: a title field indicating the title of said essay; an essay information column containing visible information about a plurality of essay elements that constitute the essay structure; and an evaluation information field including evaluation result information indicating the evaluation result or proposal information indicating the proposal content.

Images