JP2012230540A - Document generation support method, document generation support device and document generation support program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To support generation of document data 11a with high comprehension easiness.SOLUTION: A document generation support device 1 includes: browsing order acquisition means 31 for acquiring browsing order data showing flow of browsing of a reader, and calculating assumed browsing coordinates for every component to output assumed browsing coordinate data 14a in the document data 11a; object word extraction means 32 for extracting a word which matches to an object word in components and easiness of the word, calculating distance between the assumed browsing coordinates and the extracted word in the components to output word information data 15a in which the word, the easiness of the word and the distance are associated with one another; and word evaluation means 33 for calculating correction priority of the word so that correction priority becomes high to output weighted word information data 17a for a word with low easiness and at close distance for the respective words of the word information data 15a.

Description

  The present invention relates to a document generation support method, a document generation support apparatus, and a document generation support program that support generation of document data composed of a plurality of components.

  With the recent increase in the types and opportunities of devices used by users, device manufacturers have to assume that each device is used by a wide range of users. Accordingly, it is preferable that the instruction manual of the device is described so that any user can easily understand it.

  In order to create an instruction manual that is easy for everyone to understand, it is preferable to avoid technical terms used in the equipment and explain in plain words when explaining how to use the equipment in the instruction manual. However, it is generally difficult for the creator to determine what words are technical terms and are not common, and in what circumstances they should not be used.

  The word ease is used in determining what words are technical terms and not common. For example, the word familiarity described in Non-Patent Document 1 corresponds to the word ease. The word familiarity is scored on the basis of data obtained as a result of a questionnaire survey on whether or not the word is familiar with the technical term used for the device. A word having a high word familiarity is easy to be familiar and has a high degree of ease, while a word having a low word familiarity is difficult to become familiar and has a low degree of ease. There is also a method of outputting an analysis result that is easy for humans to understand by using the familiarity of words in the instruction manual (see Patent Document 1).

JP 2007-11774 A

Sato, Kasahara, Kanesugi, Amano, “Selection of basic vocabulary based on word familiarity”, Journal of the Japanese Society for Artificial Intelligence, Vol. 19, no. 6, pp. 502-510, 2004

  However, even in the same technical terms, the evaluation of the understandability of the instruction manual may differ depending on the location where the instruction manual is used. For example, a word in a place where the reader is easily noticed has a greater influence on the evaluation of comprehension than a place where the reader is difficult to see.

  Therefore, in the instruction manual, the ease of understanding of the instruction manual should be evaluated in consideration of the position where technical terms are used and the configuration of the paper surface. However, with the methods described in the above-mentioned patent documents and non-patent documents, it is difficult to evaluate the ease of understanding in consideration of the position of technical terms and the configuration of paper.

  In addition, there is a problem that it is difficult for the creator of the instruction manual to objectively evaluate because it is familiar with technical terms. Therefore, usability experts may analyze and evaluate the instruction manual, or an assumed reader may conduct a manual evaluation experiment. However, there is a problem that such a method is very expensive in terms of time and cost.

  Therefore, it is expected that the computer will evaluate the ease of understanding of the document data such as the instruction manual and develop the technology that supports the generation of the document data in consideration of the reader's browsing flow and the page structure assumed by the creator. Has been.

  Accordingly, an object of the present invention is to provide a document generation support method, a document generation support apparatus, and a document generation support program that support generation of document data with high comprehension.

  In order to solve the above-mentioned problem, a first feature of the present invention relates to a document generation support method for supporting generation of document data composed of a plurality of components. In other words, the document generation support method according to the first feature of the present invention stores the target word ease data in which the target word to be evaluated for document data is associated with the ease of the target word. In the document data, the browsing order data indicating the reading flow of the reader is acquired, and for each component, the assumed reading coordinates that are easily caught by the reader in the reading flow are calculated, and the assumed reading coordinate data , A word that matches the target word in the component and the ease of the word are extracted, the assumed browsing coordinates in the component and the distance between the extracted words are calculated, and the word, A step of outputting word information data in which the degree of ease and the distance are associated, and for each word of the word information data, a word having a low degree of ease and a short distance So that the modified priority have increased, comprising the step of calculating the modified priority of the word, and outputs the weighted word information data.

  Here, in the step of outputting the weighted word information data, the correction priority may be calculated in consideration of the display attribute of the word.

  Further, the method may further include a step of calculating a correction priority for each component based on the word correction priority and outputting weighted component information data.

  A second feature of the present invention relates to a document generation support apparatus that supports generation of document data composed of a plurality of components. That is, the document generation support apparatus according to the second feature of the present invention stores a target word that stores target word ease data in which a target word to be evaluated for document data is associated with the ease of the target word. In the ease data storage unit and document data, browsing order data indicating the reader's browsing flow is acquired, and for each component, assumed browsing coordinates that are easily caught by the reader's eyes in the browsing flow are calculated. The browsing order acquisition means for outputting the assumed browsing coordinate data, the word that matches the target word in the component and the ease of the word are extracted, and the assumed browsing coordinate in the component and the distance between the extracted words Target word extracting means for calculating word and outputting word information data in which the word, the ease and distance of the word are associated, and each word of the word information data. Te, so that the distance Simplicity is low modification priority for the word is higher closer comprises calculates the modification priority of the word, the word evaluation means for outputting the weighted word information data.

  Here, the word evaluation means may calculate the correction priority in consideration of the display attribute of the word.

  Further, it may further comprise a component evaluation means for calculating a correction priority for each component based on the word correction priority and outputting weighted component information data.

  A third feature of the present invention relates to a document generation support program for causing a computer to execute the steps described in the first feature of the present invention.

  According to the present invention, it is possible to provide a document generation support method, a document generation support apparatus, and a document generation support program that support generation of document data with high comprehension.

It is a functional block diagram of the document generation assistance apparatus which concerns on embodiment of this invention. It is a figure explaining an example which displayed the preview of document data concerning an embodiment of the invention. It is a figure explaining an example of the source of document data concerning an embodiment of the invention. It is a figure explaining a component in the preview display of document data concerning an embodiment of the invention. It is a figure explaining an example of data structure and data of object word ease data concerning an embodiment of the invention. It is a figure explaining an example of a data structure and data of browsing order data concerning an embodiment of the invention. It is a figure explaining an example of data structure and data of assumption inspection coordinate data concerning an embodiment of the invention. In embodiment of this invention, it is a figure explaining an example of the screen which the creator input the flow of browsing. In an embodiment of the present invention, it is a figure explaining correspondence with a component of document data, a flow of browsing, and an assumed browsing coordinate. It is a flowchart explaining the browsing order acquisition process by the browsing order acquisition means which concerns on embodiment of this invention. It is a figure explaining an example of a data structure and data of word information data concerning an embodiment of the invention. It is a figure explaining an example of a data structure and data of component element information data concerning an embodiment of the invention. In embodiment of this invention, it is a figure explaining the distance with the flow of word information data. It is a flowchart explaining the target word extraction process by the target word extraction means which concerns on embodiment of this invention. It is a figure explaining an example of a data structure and data of weighted word information data concerning an embodiment of the invention. It is a flowchart explaining the word evaluation process by the word evaluation means which concerns on embodiment of this invention. It is a figure explaining an example of the data structure and data of weighted component information data concerning an embodiment of the invention. It is a flowchart explaining the component evaluation process by the component evaluation means which concerns on embodiment of this invention. It is a figure explaining an example of the output result screen by the output means which concerns on embodiment of this invention. (Part 1) It is a figure explaining an example of the output result screen by the output means which concerns on embodiment of this invention. (Part 2) It is a flowchart explaining the output process by the output means which concerns on embodiment of this invention. In the browsing order acquisition means which concerns on the modification of this invention, it is an example of the browsing order which a creator inputs. In the modification of this invention, it is a figure explaining a response | compatibility with the component of document data, and assumption browsing coordinates.

  Next, embodiments of the present invention will be described with reference to the drawings. In the following description of the drawings, the same or similar parts are denoted by the same or similar reference numerals.

(Embodiment)
The document generation support apparatus 1 according to the embodiment of the present invention evaluates the document data 11a created by the creator in consideration of the reader's browsing flow and the page layout assumed by the creator, and is highly understandable. Supports generation of document data. For example, the document generation support device 1 is configured so that a difficult word close to the reader's browsing flow, a word that is conspicuous in size or color, or a component that includes many such words is preferentially corrected. Can be proposed. The creator can intensively correct the correction target according to the proposal. As described above, the document generation support apparatus 1 according to the embodiment of the present invention can support the generation of the document data 11a that can be easily understood by the reader.

  Here, the document data 11a is, for example, data of a document distributed by an expert to a general user. Specifically, the document data 11a is an instruction manual for equipment, a maintenance manual, or the like.

  As shown in FIG. 1, the document generation support apparatus 1 according to the embodiment of the present invention is a general computer including a storage device 10, a central processing control device 30, an input device 40, and an output device 50.

  In the example shown in FIG. 1, the first storage device 10a and the second storage device 10b and two storage devices are provided for convenience, but only one storage device or a plurality of storage devices may be provided. Also good. Further, the storage devices 10 a and 10 b may not be built in the document generation support device 1. The storage devices 10 a and 10 b may be storage devices that can be read by the document generation support device 1.

  The first storage device 10a shown in FIG. 1 stores data stored in advance prior to processing. On the other hand, data output by the processing is stored in the second storage device 10b. In the embodiment, when the first storage device 10a and the second storage device 10b are not particularly distinguished, they may be simply referred to as the storage device 10.

  The input device 40 is a mouse, a keyboard, or the like. The creator can give an instruction to the central processing control device 30 via the input device 40. The output device 50 is a liquid crystal display or the like. The output device 50 can display data on a screen in accordance with an instruction from the central processing control device 30 and allow a creator to view the data. The document generation support apparatus 1 is realized by installing and executing a program for executing predetermined processing in such a general computer.

  The storage device 10 includes a document data storage unit 11, a target word ease data storage unit 12, a browsing order data storage unit 13, an assumed browsing coordinate data storage unit 14, a word information data storage unit 15, a component information data storage unit 16, A weighted word information data storage unit 17, a weighted component information data storage unit 18, and an evaluation result data storage unit 19 are provided. Here, the storage device 10 is, for example, a hard disk or RAM, and is a nonvolatile or volatile storage device.

  The document data storage unit 11 is a storage area in the storage device 10 in which the document data 11a is stored. The document data 11a is composed of a plurality of components. Here, the component is a unit of processing in the embodiment of the present invention. The component may be a paragraph delimited by a line feed or a sentence delimited by “.”. Further, when the document data 11a is a structured document such as HTML, the component may be in a range delimited by a tag such as div. The div tag is a tag related to layout, character size, color, and the like. When the constituent element is a sentence unit, the document data 11a is divided into sentences by, for example, character string processing using a regular expression.

  The document data 11a is data that allows the document generation support apparatus 1 to read characters included in the document data 11a. The document data 11a is electronic data such as PDF and HTML, for example, and is data in which word information such as character size and color can be set. The document data 11a may be data obtained by digitizing characters using an OCR (Optical Character Reader) from data obtained by reading a paper surface with a scanner or the like.

  In the embodiment of the present invention, the document data 11a is data expressed in the HTML format, and a description will be given of a case where the constituent elements are defined based on a unit delimited by a “div” tag.

  An example of the document data 11a according to the embodiment of the present invention will be described with reference to FIG. The document data 11a shown in FIG. 2 is a preview display of the HTML format source data shown in FIG. The document data 11a is a horizontal writing format in which characters flow from left to right, and is a two-column format. The document data 11 a has “instruction manual” as a title, and includes items of “overview”, “procedure 1”, “procedure 2”, “procedure 3”, and “notes”. The “Notes” item is provided small in the upper right and is referred to as an annotation.

  As shown in FIGS. 3 and 4, in the embodiment of the present invention, the document data 11a includes five components E1 to E5. FIG. 4 shows the preview display shown in FIG. 2 in association with the components E1 to E5 of the document data 11a.

  The first component E1 is a character string enclosed in </ div> immediately after <div class = “introduction”> in the HTML source shown in FIG. As shown in FIG. 4, the first component E <b> 1 includes a subtitle character string “Summary” and a character string following the character string. The second component E2 is a character string enclosed in </ div> immediately after <div class = “step1”> in the HTML source shown in FIG. As shown in FIG. 4, the second component E <b> 2 includes a subtitle character string “procedure 1” and a character string following the character string. The third component E3 is a character string surrounded by </ div> immediately after <div class = "step2"> in the HTML source shown in FIG. As shown in FIG. 4, the third component E3 includes a character string of the subtitle “Procedure 2” and a subsequent character string. The fourth component E4 is a character string surrounded by </ div> immediately after <div class = “step3”> in the HTML source shown in FIG. As shown in FIG. 4, the fourth component E4 includes a character string of the subtitle “Procedure 3” and a character string subsequent thereto. The fifth component E5 is a character string enclosed in </ div> immediately after <div class = “note”> in the HTML source shown in FIG. 3. As shown in FIG. 4, the fifth component E5 includes a subtitle character string “notes” and a subsequent character string.

  Here, in the embodiment of the present invention, the case where the component is determined based on the <div> tag will be described, but it may be determined under any condition. Further, the document data 11a shown in FIGS. 2 to 4 is composed only of characters, but may include figures.

  The target word ease data storage unit 12 is a storage area in the storage device 10 in which the target word ease data 12a is stored. As shown in FIG. 5, the target word ease data 12a is data in which a target word to be evaluated for document data is associated with the ease of the target word. Here, the ease of the target word is a numerical value of the difficulty of the target word for the reader. The ease of the target word is set such that, for example, a word familiar to the reader has a high degree of ease, and a word unfamiliar with the degree of ease is set low. In the example illustrated in FIG. 5, the ease is a value from 0 to 1. Words with a degree of ease closer to “1” are easier to understand, and words with a degree of ease closer to “0” are more difficult to understand.

  As the ease of the target word ease data 12a, for example, word familiarity can be considered (Patent Literature 1 and Non-Patent Literature). The word familiarity is calculated by analyzing data obtained by conducting a questionnaire survey in advance as to whether or not a word is familiar with a technical term used in a device or the like.

  The browsing order data storage unit 13 is a storage area in which the browsing order data 13 a and the browsing order data 13 b are stored in the storage device 10. The assumed browsing coordinate data storage unit 14 is a storage area in which the assumed browsing coordinate data 14 a of the storage device 10 is stored. The word information data storage unit 15 is a storage area in which the word information data 15 a of the storage device 10 is stored. The component element information data storage unit 16 is a storage area in which the component element information data 16a of the storage device 10 is stored. The weighted word information data storage unit 17 is a storage area in which the weighted word information data 17a of the storage device 10 is stored. The weighted component information data storage unit 18 is a storage area in which the weighted component information data 18a of the storage device 10 is stored. The evaluation result data storage unit 19 is a storage area in which the evaluation result data 19a of the storage device 10 is stored.

  Here, the browsing order data 13a, the browsing order data 13b, the assumed browsing coordinate data 14a, the word information data 15a, the component element information data 16a, the weighted word information data 17a, the weighted component information data 18a, and the evaluation result data 19a are: Since it is data output by each means of the central processing control device 30 described later, it will be described later.

  The central processing control device 30 includes a browsing order acquisition unit 31, a target word extraction unit 32, a word evaluation unit 33, a component evaluation unit 34, and an output unit 35.

  The browsing order acquisition unit 31 acquires browsing order data 13a indicating a reader's browsing flow F in the document data 11a. Furthermore, the browsing order acquisition means 31 calculates the assumed browsing coordinates which are coordinates that are easily caught by the reader in the browsing flow F for each component, and outputs the assumed browsing coordinate data 14a.

  The browsing order data 13a is data indicating a reader's browsing flow F in the document data 11a in units of components. The browsing order data 13a is also a range that the creator wants the reader to read in the document data 11a. The browsing order data 13a is data indicating the browsing flow F by listing the identifiers of the constituent elements as shown in FIG. 6A, for example. In the data shown in FIG. 6A, the browsing flow F is shown in the order of the first component E1, the second component E2, the third component E3, and the fourth component E4. Yes. In FIG. 6A, the browsing order data 13a does not include information on the fifth component E5. This indicates that the fifth component E5 does not exist on the viewing flow F of the creator.

  In the case where the creator inputs the browsing flow F with the input device 40, the browsing order data storage unit 13 further displays the coordinate data of the browsing flow F as the browsing order data 13b as shown in FIG. You may remember. The browsing order data 13b is data in which the x coordinate and the y coordinate of the trajectory of the browsing flow F indicated by the creator are associated with each other.

  The assumed browsing coordinate data 14a is assumed browsing coordinate data for each component. Assumed browsing coordinate data 14a is data in which components and assumed browsing coordinates are associated with each other, for example, as shown in FIG. In FIG. 7, the assumed browsing coordinates are not associated with the fifth component E5. This indicates that the fifth component E5 does not exist on the viewing flow F of the creator.

  Here, the assumed browsing coordinates are coordinates that are easily caught by the reader in consideration of the browsing flow F, and are calculated for each component. In the embodiment of the present invention, it is assumed that the assumed browsing coordinates are coordinates on the browsing flow F and first overlap each component. Assumed viewing coordinates may be any coordinates as long as they are easily caught by the reader. For example, the method of calculating the assumed browsing coordinates may differ depending on whether the document data 11a is written vertically or horizontally.

  As shown in FIG. 8, the browsing order acquisition unit 31 provides the creator with an input device such as a mouse or a pen tablet for the flow of reading that the creator assumes in the preview display of the document data 11 a shown in FIG. 2. 40, the creator inputs the browsing flow F. In the example shown in FIG. 6, the browsing flow F is input in the flow from the top to the bottom of the left column of the screen and further from the top to the bottom of the right column. Note that the precautions displayed in the upper right of the screen are out of the browsing flow F.

  The browsing order acquisition unit 31 displays the document data 11a with the browsing flow F superimposed on the preview display. The browsing order acquisition means 31 calculates the x coordinate and y coordinate of the trajectory of the browsing flow F in the preview display of the document data 11a, and stores it in the browsing order data storage unit 13 as the browsing order data 13b shown in FIG. Remember.

  Further, the browsing order acquisition unit 31 compares the input coordinates of the browsing flow F with the coordinates of the display area of the component in the preview of the document data 11a, and calculates the browsing flow in units of components. The browsing order acquisition unit 31 stores the browsing flow in units of components calculated here as browsing order data 13 a in the browsing order data storage unit 13. In the example shown in FIG. 6, the flow of browsing in units of constituent elements is the order of the first constituent element E1, the second constituent element E2, the third constituent element E3, and the fourth constituent element E4. Here, the fifth component E5 deviates from the browsing flow.

  The browsing order acquisition unit 31 calculates assumed browsing coordinates based on the browsing order data 13a and the browsing order data 13b. Assumed browsing coordinates are calculated for each component. In the embodiment of the present invention, the browsing order acquisition means 31 calculates the coordinates on the browsing flow F for each component, which first overlap each component, the identifier of the component, and its configuration The assumed browsing coordinate data 14a is calculated by associating the assumed browsing coordinates of the element.

  As shown in FIG. 8, the assumed viewing coordinates of the first component E1 are coordinates on the viewing flow F and are a point P1 that overlaps the upper side of the first component E1. The assumed viewing coordinates of the second component E2 are coordinates on the viewing flow F, and are a point P2 that overlaps with the upper side of the second component E2. The assumed viewing coordinates of the third component E3 are coordinates on the viewing flow F and are a point P3 that overlaps the left side of the third component E3. The assumed browsing coordinates of the fourth component E4 are coordinates on the browsing flow F, and are a point P4 overlapping the upper side of the fourth component E4. Here, since the fifth component E5 and the browsing flow F do not overlap, there is no assumed browsing coordinate for the fifth component E5. Therefore, in the assumed browsing coordinate data 14a shown in FIG. 7, the assumed browsing coordinates are not associated with the fifth component E5.

  With reference to FIG. 10, the browsing order acquisition process by the browsing order acquisition means 31 which concerns on embodiment of this invention is demonstrated.

  First, the browsing order acquisition unit 31 reads the document data 11a from the storage device 10 in step S101, and acquires the browsing flow F in step S102. As described above, the creator uses the input device 40 to input the browsing flow F on the preview display of the document data 11a. The browsing order acquisition means 31 can acquire the coordinates of the browsing flow F input by the creator and output the browsing order data 13b.

  In step S103, the browsing order acquisition unit 31 determines the browsing order of the components of the document data 11a according to the browsing flow F acquired in step S102. The browsing order acquisition unit 31 compares the display area of the component and the coordinates of the browsing flow F, determines the browsing flow in units of the component, and outputs the browsing order data 13a. When the flow of browsing in units of constituent elements is determined, the processes in steps S105 and S106 are repeated for each constituent element.

  First, in step S105, the browsing order acquisition unit 31 determines whether the component and the browsing flow F overlap each other. When the browsing order data 13 includes the identifier of the constituent element, the browsing order acquisition unit 31 determines that the constituent element and the browsing flow F overlap. When the component and the browsing flow F overlap, in step S106, the browsing order acquisition unit 31 calculates assumed browsing coordinates in the component. On the other hand, when not overlapping, the browsing order acquisition means 31 repeats the process of step S105 and step S106 about the following component.

  When the processing of step S105 and step S106 is completed for all the constituent elements, the browsing order acquisition unit 31 assumes that the identifiers of the constituent elements are associated with the assumed browsing coordinates when the assumed browsing coordinates can be calculated. The browsing coordinate data 14a is output.

  Next, the target word extracting unit 32 will be described. The target word extraction unit 32 extracts, for each component, a word that matches the target word in the component and the ease of the word. Further, the target word extraction unit 32 calculates the assumed browsing coordinates in the component and the distance between the extracted words, and outputs the word information data 15a that associates the word, the ease of the word, and the distance. Further, the target word extraction means 32 outputs the component element information data 16a.

  The word information data 15a is data that associates a word, the ease of the word, and the distance. The word information data 15a may further be associated with an appearance position in the flow and a display attribute of the word. The display attributes are, for example, word decoration information such as word size and color.

  The word information data 15a includes, for example, an example of the data structure and data shown in FIG. The word information data 15a is data in which a word identifier, a word, an appearance position, a distance from a flow, a character size, a character color, and an ease included in the document data 11a are associated with each other. Here, the target word extraction means 32 numbers the word identifier. The word, character size, and character color are acquired from the document data 11a. When the document data 11a is in the HTML format, the target word extraction unit 32 can easily obtain the target word extraction unit 32 from the tag of the source file. The easiness is an easiness associated with the word in the target word easiness data 12a. Here, when a word is displayed as a part of the explanatory text in the figure, the larger the figure, the larger the character size of the word, and the higher the importance of the word in the explanatory text. May be set.

  The appearance position is a word position in the browsing flow F. Whether the word appears at the beginning of the browsing flow F, appears backward, or is outside the range of the browsing flow F is indicated. In calculating the appearance position, for example, browsing order data 13a shown in FIG. The target word extraction unit 32 associates the first half, the middle, and the second half of the flow of browsing for each component, and associates out-of-range with the components not included in the browsing order data 13a. The example shown in FIG. 11 shows the first half of the appearance position of the word included in the first component E1, the middle part of the appearance position of the word included in the second component E2 and the third component E3, and the fourth configuration. The appearance position of the word included in the element E4 is associated with the latter half, and the appearance position of the word included in the fifth component E5 is associated with out of range. In the embodiment of the present invention, the case where the appearance position is determined based on the components will be described, but the present invention is not limited to this.

  The distance to the flow is the distance from the word and the assumed viewing coordinates for each component. In the embodiment of the present invention, the distance to the flow is a distance considering a single browsing flow. In the embodiment of the present invention, the distance from the flow is generally different from the physical distance between the word and the assumed viewing coordinates. In the embodiment of the present invention, the distance from the flow is calculated for each component in consideration of the flow that the reader browses based on the assumed viewing coordinates of each component.

  As shown in FIG. 12, the constituent element information data 16a is data in which constituent element identifiers are associated with identifiers of words included in the constituent elements.

  The target word extraction unit 32 extracts a target word from the document data 11a. The target word extracting unit 32 extracts nouns by performing morphological analysis on the document data 11a. The target word extraction means 32 extracts the target word contained in the target word ease data 12a among the extracted nouns as a word used in the document data 11a. The word extracted here becomes an evaluation target of the document generation support apparatus 1 according to the embodiment of the present invention. Further, the target word extraction means 32 acquires the character size and character color of the word, and acquires the ease of the word with reference to the target word ease data 12a. The target word extraction unit 32 assigns a word identifier to the acquired word, associates the word, the character size of the word, the color of the character, and the ease, and generates word information data 15a. Store in the storage unit 15.

  The target word extracting unit 32 generates the component information data 16a by associating the acquired word with the component included in the word. In addition, the target word extraction unit 32 calculates the appearance positions of the constituent elements in consideration of the browsing flow F, such as the first half, the middle board, the second half, and out of range. In the embodiment of the present invention, the target word extraction unit 32 stores the information on the appearance position of the component in the word information data 15a as the information on the appearance position of the word.

  Further, the target word extraction unit 32 calculates a distance from each word flow. In the embodiment of the present invention, the distance from the flow is calculated based on the browsing flow F in the entire document data 11a and the flow that the reader browses based on the assumed browsing coordinates in each component.

  With reference to FIG. 13, the calculation method of the distance with the flow which concerns on embodiment of this invention is demonstrated. FIG. 13 is an enlarged view of the first component E1 of the document data 11a. A part of the browsing flow F overlaps the first component E1. As shown in FIG. 8, this browsing flow F flows from the top to the bottom of the diagram shown in FIG. In the embodiment of the present invention, the assumed browsing coordinates are coordinates on the browsing flow F, and are the coordinates that first overlap each component. Accordingly, the assumed viewing coordinate of the first component E1 is P1. In the embodiment of the present invention, the distance from the flow is calculated for each component in consideration of the flow that the reader browses based on the assumed viewing coordinates of each component. Therefore, the distance L from the flow is expressed by the following formula 1.

Distance to flow L = basic distance + number of characters k × coefficient n
= SQRT ((x1-x0) ^ 2 + (y1-y0) ^ 2) + k * n
... (Formula 1)
Here, the basic distance is a distance between the assumed viewing coordinate P1 and the top left vertex P0 of the first component E1. The number of characters is the number of characters from the beginning of the character string of the first component E1 to the word. The coefficient n is a length corresponding to the character width of each word.

  The basic distance is an evaluation target from the beginning of the character string of the first component E1 when the reader browses the document data 11a according to the browsing flow F and browses the character string of the first component E1. The distance that the reader moves his eyes to the word. The target word extraction unit 32 adds the number k of characters from the head of the character string of the first component E1 to the word to be evaluated multiplied by the coefficient n to the basic distance, thereby adding a distance L to the flow. As described above, the distance to the word can be calculated from the assumed browsing coordinates P1.

  Here, in the browsing order data 13a, the distance L to the flow is calculated even for components that are outside the range of the browsing flow F. For a word included in such a component, the target word extraction unit 32 calculates the distance L between this word and the flow as the distance from the closest assumed viewing coordinate. For example, in the example illustrated in FIG. 9, the fifth component E <b> 5 that starts with “Caution” is out of the range of the browsing flow F. Therefore, the target word extraction means 32 specifies the assumed browsing coordinates P3 of the third component E3 as the assumed browsing coordinates closest to the fifth component E5, and determines the distance between the assumed browsing coordinates P3 and the word, Calculated as a distance L to the flow.

  With reference to FIG. 14, the target word extraction process by the target word extraction means 32 which concerns on embodiment of this invention is demonstrated.

  The target word extraction unit 32 reads the document data 11a from the storage device 10 in step S201, and reads the target word ease data 12a in step S202. Next, the target word extracting unit 32 repeats the processing from step S203 to step S209 for each component of the document data 11a.

  In step S203, the target word extraction unit 32 extracts the target word registered in the target word ease data 12a from the component. In step S204, the target word extracting unit 32 associates the word extracted in step S203 with the constituent element and outputs the same to the constituent element information data 16a. Further, the target word extraction unit 32 calculates the appearance position of the component in the browsing flow F, such as the first half, the middle, the second half, and the out of range for the component.

  Next, the processing from step S206 to step S209 is repeated for each word extracted in step S203.

  The target word extraction unit 32 extracts the ease of the word from the target word ease data 12a in step S206, and calculates the distance from the assumed viewing coordinates in step S207. In step S208, the target word extracting unit 32 acquires display attributes such as the size and color of the characters of the word.

  In step S209, the target word extraction unit 32 determines the word, the appearance position calculated in step S205, the ease acquired in step S206, the distance from the assumed viewing coordinates calculated in step S207, and the word acquired in step S208. A record that associates the display attribute with is generated. The target word extraction unit 32 outputs the generated record to the word information data 15a.

  When the processing from step S203 to step S209 is completed for all the constituent elements and words, the target word extraction processing by the target word extraction unit 32 is completed.

  Next, the word evaluation means 33 will be described. For each word in the word information data 15a, the word evaluation means 33 calculates a correction priority of the word so that the correction priority is high for a word with a low degree of ease and a short distance, and the weighted word information data 17a is obtained. Output. Here, the word correction priority is a degree of necessity of correcting the word in the document data 11a. A word having a high correction priority is preferably corrected by an operator in order to improve the understanding level of the document data 11a. Here, the word evaluation means 33 may calculate the correction priority in consideration of not only the ease of the word and the distance from the flow but also the display attribute of the word.

  With reference to FIG. 15, the weighted word information data 17a according to the embodiment of the present invention will be described. The weighted word information data 17a shown in FIG. 15 is different from the word information data 15a shown in FIG. 11 in that it has correction priority and score items. This score is calculated from each item of the word information data 15a. In the example shown in FIG. 15, the word correction priority is the order in which the scores are sorted in descending order.

  In the embodiment of the present invention, the word correction priority is set higher for difficult words that are easily caught by the reader based on the browsing flow F. Specifically, the easier the word is, the shorter the distance from the assumed viewing coordinate, the first half of the appearance position, the larger the character size, and the higher the color of the character, the more the character should be corrected. . In the example shown in FIG. 15, the score is set so that the word to be corrected has a higher score. Further, the correction priority value is set to be smaller for the word to be corrected.

  Here, the word evaluation means 33 determines each element point of (1) word ease, (2) distance from assumed viewing coordinates, (3) appearance position, (4) character size, and (5) character color. A case will be described in which overall evaluation is performed to determine the word correction priority.

  The word evaluation means 33 sets (1) X1 = 1-easyness for (1) word ease. The word evaluation unit 33 sets (2) a distance point with respect to the distance from the assumed viewing coordinates (2). The word evaluation means 33 (3) sets X3 = position point for the appearance position. The word evaluation means 33 sets X4 = size point for (4) character size. The word evaluation means 33 (6) sets X5 = attraction level for the character color. Here, each element point of the distance point, the position point, the size point, and the attractiveness point is converted and set within a range between 0 and 1 based on each data of the word information data 15a. Each element point is set to be closer to 1 as the distance from the assumed viewing coordinate is closer, the appearance position is in the first half, the character size is larger, and the character color is higher in degree of attraction.

  The word evaluation means 33 can calculate the score Y of the word by Equation 2 using each of the element point stores X1 to X5.

Y = 20 × ΣXi (I = 1..5) (Formula 2)
Here, in calculating the score Y for the word “LAN” identified by the word “W1” shown in FIG. 15, each element point is set in advance. Here, each element point has (1) X1 = 1−0.7 = 0.3 for the word ease (= 0.7), and (2) X2 for the distance (= close) from the viewing order. = 1 point, (3) For the appearance position (= first half), X3 = 1 point, (4) For the character size (= normal), X4 = 0.9 point, (5) For the character color (= black) It is assumed that X5 = 0.9 point is set.

  In this case, the score Y is calculated as 20 × ((1−0.7) + 1 + 1 + 0.9 + 0.9) = 82 points. Although this word is a word with high ease, it is highly likely that it will be caught by the reader's eyes, so the score Y is calculated to be high. Here, the score Y is calculated in consideration of all of the word ease, the distance point, the position point, the size point, and the attractiveness point, but the present invention is not limited to this example. For example, even if the score Y is calculated using only the word ease and distance points, the word can be evaluated in consideration of the flow of reading by the reader. Further, the score Y may be calculated by weighting the word ease, distance point, position point, size point, and attraction degree point.

  The word evaluation means 33 outputs the score thus calculated to the weighted word information data 17a as the score of the word identifier “W1”. When the word evaluation means 33 calculates the score Y for all the words in the word information data 15a, the word evaluation means 33 sorts in descending order based on the score Y. The word evaluation means 33 gives a correction priority based on the sorted order. The word evaluation means 33 associates the word with the score calculated for this word and the modification priority, and outputs it to the weighted word information data 17a.

  With reference to FIG. 16, the word evaluation process by the word evaluation means 33 which concerns on embodiment of this invention is demonstrated.

  The word evaluation means 33 reads the word information data 15a from the storage device 10 in step S301, and processes step S302 for each word in the word information data 15a. The word evaluation means 33 calculates the score of the word of the word information data 15a about step S302.

  When the scores are calculated for all the words in the word information data 15a, in step S303, the word evaluation means 33 gives a correction priority based on the scores calculated in step S302. In step S303, the word evaluation unit 33 associates the word, the score calculated in step S302, and the modification priority given in step S303, and outputs the result to the weighted word information data 17a. Here, the word evaluation means 33 may also associate each item of the word information data 15a and output it to the weighted word information data 17a.

  Next, the component element evaluation means 34 will be described. The component evaluation means 34 calculates a correction priority for each component based on the word correction priority, and outputs weighted component information data 18a. The component evaluation means 34 refers to the weighted word information data 17a and acquires the score of each target word in the document data 11a. The component evaluation means 34 refers to the component information data 16a, and calculates the score for each component based on the score corresponding to the word included in the component for each component.

  As shown in FIG. 17, the weighted component information data 18a is data in which a component identifier, a word identifier included in the component, a modification priority, and a score are associated with each other. The weighted component information data 18a shown in FIG. 17 is different from the component information data 16a shown in FIG. 12 in that it includes items of correction priority and score. This score is calculated from the score of the weighted word information data 17a. In the example shown in FIG. 17, the correction priority of words is the order in which the scores are sorted in descending order.

  As a method for calculating the score for each component, there is a method in which the score obtained by adding the scores of all target words included in the component is used as the score for each component. Using the score calculated for each component, the component evaluation means 34, for example, the modification priority for each component so that the modification priority of this component increases as the score for each component increases. Is calculated.

  With reference to FIG. 18, the component evaluation process by the component evaluation means 34 is demonstrated. The component evaluation means 34 reads the component element information data 16a from the storage device 10 in step S401, and reads the weighted word information data 17a in step S402. Next, the constituent element evaluation unit 34 repeats the processing of step S403 and step S404 for each constituent element of the document data 11a.

  In step S403, the constituent element evaluation means 34 acquires identifiers of all words included in the constituent element from the constituent element information data 16a, and acquires the scores of these words from the weighted word information data 17a. In step S404, the component evaluation means 34 sets the total score of the words acquired in step S403 as the score of the component.

  When the scores have been calculated for all the constituent elements, the constituent element evaluation means 34 assigns a correction priority in step S405 based on the scores calculated in step S404. In step S406, the component evaluation means 34 associates the identifier of the component, the score calculated in step S404, and the modification priority given in step S405, and outputs the result to the weighted component information data 18a.

  Next, the output means 35 will be described. The output means 35 is a means for outputting an evaluation result based on the weighted word information data 17a and the weighted component information data 18a to the output device 50 and allowing the creator to view it. Various methods can be considered as a method for outputting the evaluation result by the output means 35. For example, a method of displaying a list as tabular data, a method of superimposing evaluation results on a preview screen, or the like can be considered.

  When the evaluation results are displayed as a list of table data, the output unit 35 displays the weighted word information data 17a shown in FIG. 15 and the weighted component information data 18a shown in FIG. 18 on the output device 50 in a table format. To do.

  An example of the display screen when the evaluation result is displayed on the preview screen will be described with reference to FIGS. 19 and 20. In this case, the output unit 35 generates evaluation result data 19a in which the evaluation result is superimposed on the document data 11a, and outputs the evaluation result data 19a to the output device 50.

  FIG. 19 is an example of a result display screen P101 in which the output unit 35 expresses the correction priority with the background color. In the example shown in FIG. 19, the output means 35 outputs evaluation result data 19a with a background according to the correction priority for the words and components to be evaluated. When the document data 11a is in the HTML format, the output unit 35 outputs the evaluation result data 19a in which the background color is set using a tag based on the document data 11a. The output means 35 sets a background with a high degree of attractiveness for words and components having a high correction priority, and sets a background with a low degree of attractiveness as the correction priority decreases. The output unit 35 does not superimpose the background on words and components that do not need to be particularly corrected.

  The output means 35 sets and displays these background colors on the preview screen of the document data 11a according to the evaluation result data 19a. Therefore, the creator can grasp the correction priority based on the background color, and thus can specify an item to be corrected in consideration of the flow of the entire preview screen.

  FIG. 20 is an example of a result display screen P102 in which the output unit 35 expresses the modification priority of the component by a symbol. In the example illustrated in FIG. 20, the output unit 35 outputs evaluation result data 19 a with symbols attached to the evaluation target components according to the correction priority. The output means 35 sets a symbol “x” for a component with a high modification priority, and sets “◯” for a component with a low modification priority.

  The output means 35 attaches and displays these symbols on the preview screen of the document data 11a according to the evaluation result data 19a. By visualizing the correction dominance in this way, the creator can grasp the correction priority. As a result, the creator can specify items to be corrected or can improve the document data 11a while considering the flow of the entire preview screen. Further, in the example shown in FIG. 20, the score calculated from the correction priority of the entire document data 11a is displayed. For example, when all the constituent elements are “◯”, the output means 35 may display 100 points. The result display screen described with reference to FIGS. 19 and 20 is an example, and various result display screens are conceivable.

  The document generation support apparatus 1 according to the embodiment of the present invention designates a range that the creator intends to read, by designating a reader's browsing flow assumed by the creator. In consideration of the reader's browsing flow assumed by the creator specified in this way, the document generation support device 1 individually determines the words and components included in the document data as to whether or not the reader can easily understand. To evaluate. As a result, the document generation support device 1 can point out to the creator a word or a constituent element that hinders the reader's understanding. The creator can generate document data that is easy for the reader to understand by correcting the words and components pointed out by the document generation support apparatus 1.

  In addition, the document generation support apparatus 1 according to the embodiment of the present invention can determine the modification priority in units of components in the document data 11a by considering the flow of reading by the reader. Therefore, the document generation support apparatus 1 can preferentially point out an improved portion even for document data expressed on a limited page.

  The document generation support apparatus 1 according to the embodiment of the present invention can automate the evaluation of document data regarding whether or not a reader can easily understand by computer processing. Thereby, since it is not necessary for an expert or a test subject to confirm document data, the production cost of document data can be reduced significantly.

  As described above, the document generation support apparatus 1 according to the embodiment of the present invention can support generation of document data with high understanding.

(Modification)
Here, the browsing order acquisition means 31a which concerns on the modification of the browsing order acquisition means 31 is demonstrated. In the embodiment of the present invention, the browsing order acquisition unit 31 causes the creator to input the trajectory of the browsing flow F using the input device 40, but the browsing order acquisition unit 31a allows the creator to A case of inputting the browsing order of will be described.

  As shown in FIG. 22, the creator inputs the assumed browsing flow shown for each component of the document data 11a. In the example shown in FIG. 22, first read the “Instruction Manual” of the title, and then assume browsing in the order of “Summary”, “Procedure 1”, “Procedure 2”, and “Procedure 3”. Indicates that In the parentheses following “Summary”, “Procedure 1”, “Procedure 2”, and “Procedure 3”, a character string delimited by <p> to </ p> of the source data shown in FIG. 3 is input. Here, the browsing order input method is not limited to the example shown in FIG. The browsing order acquisition unit 31a may be input in any way as long as the reading order of the reader assumed by the creator is known in units of components.

  When the browsing flow is input in this way, the browsing order acquisition unit 31a generates browsing order data 13b. As shown in FIG. 6A, the browsing order data 13b is data in which identifiers of constituent elements are listed according to an assumed browsing order.

  The document data 11a is horizontal writing in which characters flow from left to right. Therefore, in the modified example, the assumed browsing coordinates are the coordinates of the upper left corner of each component as the coordinates that are easily caught by the reader in consideration of the browsing flow F.

  Therefore, as shown in FIG. 23, the browsing order acquisition unit 31a sets the top left vertex of the first component E1 as the assumed browsing coordinate P1 of the first component E1. Similarly, the browsing order acquisition unit 31a calculates an assumed browsing coordinate P2 of the second component E2, an assumed browsing coordinate P3 of the third component E3, and an assumed browsing coordinate P4 of the fourth component E4. The browsing order acquisition unit 31a stores the assumed browsing coordinates P1 to P4 calculated in this way as the assumed browsing coordinate data 14a described with reference to FIG.

  Since the browsing flow differs depending on the format of the document data 11a, the assumed browsing coordinates may be different. For example, when the document data 11a is in the vertical writing format, the vertex at the upper right end of the component may be the assumed coordinate of the component.

  As described above, various methods are conceivable as a method for determining the flow of assumed browsing and a method for determining assumed viewing coordinates.

(Other embodiments)
As described above, the embodiments and modifications of the present invention have been described. However, it should not be understood that the descriptions and drawings constituting a part of this disclosure limit the present invention. From this disclosure, various alternative embodiments, examples, and operational techniques will be apparent to those skilled in the art.

  For example, the document generation support apparatus described in the best embodiment of the present invention may be configured on a single piece of hardware as shown in FIG. 1, or a plurality of pieces of hardware depending on the functions and the number of processes. It may be configured above. Moreover, you may implement | achieve on the existing information processing system.

  It goes without saying that the present invention includes various embodiments not described herein. Therefore, the technical scope of the present invention is defined only by the invention specifying matters according to the scope of claims reasonable from the above description.

DESCRIPTION OF SYMBOLS 1 Document production | generation assistance apparatus 10 Memory | storage device 11 Document data memory | storage part 12 Target word ease data storage part 13 Browsing order data memory | storage part 14 Assumed browsing coordinate data memory | storage part 15 Word information data memory | storage part 16 Component element information data memory | storage part 17 Weighted Word information data storage unit 18 Weighted component information data storage unit 19 Evaluation result data storage unit 30 Central processing control device 31 Browsing order acquisition unit 32 Target word extraction unit 33 Word evaluation unit 34 Component evaluation unit 35 Output unit 40 Input device 50 output device

Claims (7)

A document generation support method for supporting generation of document data composed of a plurality of components,
Storing target word ease data in which a target word to be evaluated for document data is associated with the ease of the target word;
In the document data, browsing order data indicating a reader's browsing flow is acquired, and for each component, an assumed browsing coordinate which is a coordinate that is easily perceived by the reader in the browsing flow is calculated. Outputting data, and
The word that matches the target word in the component and the ease of the word are extracted, the assumed browsing coordinates in the component and the distance between the extracted words are calculated, and the word and the ease of the word And outputting word information data associated with the distance;
For each word of the word information data, the step of calculating the correction priority of the word and outputting the weighted word information data so that the correction priority is high for a word with low ease and a short distance. A document generation support method characterized by the above. The document generation support method according to claim 1, wherein the step of outputting the weighted word information data calculates the modification priority in consideration of a display attribute of the word. The document generation according to claim 1, further comprising: calculating a correction priority for each component based on the correction priority of the word and outputting weighted component information data. Support method. A document generation support device that supports generation of document data composed of a plurality of components,
A target word ease data storage unit for storing target word ease data in which target words to be evaluated for document data are associated with the ease of the target words;
In the document data, browsing order data indicating a reader's browsing flow is acquired, and for each component, an assumed browsing coordinate which is a coordinate that is easily perceived by the reader in the browsing flow is calculated. Browsing order acquisition means for outputting data;
The word that matches the target word in the component and the ease of the word are extracted, the assumed browsing coordinates in the component and the distance between the extracted words are calculated, and the word and the ease of the word And target word extraction means for outputting word information data associated with the distance,
For each word of the word information data, a word evaluation means for calculating the correction priority of the word and outputting the weighted word information data so that the correction priority is high for a word with low ease and a short distance A document generation support apparatus comprising: The document generation support apparatus according to claim 4, wherein the word evaluation unit calculates the modification priority in consideration of a display attribute of the word. The component evaluation means which calculates the correction priority for every said component based on the said correction priority of the said word, and outputs weighted component information data, The further characterized by the above-mentioned. Document generation support device.   A content collection program for causing a computer to execute the steps according to any one of claims 1 to 3.