JPH10320388A - Text preparation device and method therefor and dictionary registration method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a text preparing device for easily and quickly preparing a text being a target by re-using an example text. SOLUTION: An input keyword being the keyword of a text being a target is inputted through a keyword inputting part 221. A keyword synthetic name indicating the concept or category of the input keyword is obtained from a keyword dictionary 31 by a synthetic name obtaining part 211, and an example text related with the obtained keyword synthetic name is obtained from an example text dictionary 32 by a text obtaining part 212. The prescribed part of the obtained example text is replaced with the input keyword by a replacing part 213 so that a text being a target can be prepared. Thus, a proper text can be prepared only by the input of the input keyword, and simple and quick text preparation can be attained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sentence creating apparatus for creating sentences described in various pamphlets and catalogs, a sentence creating method, a method of registering information in a dictionary used for creating sentences, and a method of sending sentences to a computer. It relates to a recording medium to be created.

[0002]

2. Description of the Related Art Brochures such as guides for various accommodation facilities and travel courses, and catalogs of products, etc., contain texts describing accommodation facilities, travel courses, products, and the like.
FIG. 13 shows an example of publication of accommodation facility information as an example of such a pamphlet. In the area shown by the oblique lines in FIG. 13, standard sentences such as the characteristics of accommodation facilities, meals, and traffic guidance are described. Such texts have conventionally been created by copywriters in line with proposals and the like.

[0003] Therefore, in order to perform such a text creation operation for an enormous amount of catalogs and the like, a technique has been adopted in which a large number of workers spend a large amount of time and perform human naval tactics.

[0004]

However, in printed materials such as catalogs, most of the sentences described are fixed sentences, and the work actually performed by an operator is to create many similar sentences according to a plan. Work. In other words, the same work is repeatedly performed every time there is an order to create a text. In addition, if multiple workers are in charge of writing a printed matter with many sentences such as a general pamphlet, after writing the sentence, check for typographical errors, terms,
Unification of wording, elimination of duplicate sentences, etc. are required, and more time and effort are required.

Accordingly, the present invention has been made in view of the above problems, and a sentence that has already been created is registered as a case sentence, and the case sentence is reused to reduce the labor of the sentence creation work and to quickly perform the work. It is an object of the present invention to provide a text creation device that can realize text creation, a text creation method, a method of registering information in a dictionary used for text creation, and a recording medium that allows a computer to create text.

[0006]

According to the first aspect of the present invention, there is provided a sentence creating apparatus for creating a target sentence by reusing a case sentence, including (a) a replacement word associated with a replacement word identifier. A case sentence storage means for storing a plurality of case sentences, (b) a keyword storage means for storing a plurality of dictionary keywords in association with a keyword identifier, and (c) a keyword input means for receiving an input of an input keyword which is a keyword of the target sentence (D) by searching the keyword storage means to find a specific dictionary keyword corresponding to the input keyword and to acquire a keyword identifier associated with the specific dictionary keyword as a target identifier; (E) a specific replacement corresponding to the target identifier by searching the case sentence storage means. A target case in which a case identifier including a replacement word associated with the specific replacement word identifier is obtained from a plurality of case sentences stored in the case sentence storage means as a target case sentence. A sentence obtaining unit, and (f) a replacing unit that creates the target sentence by replacing a replacement word included in the target case sentence with the input keyword.

According to a second aspect of the present invention, there is provided the text creating apparatus according to the first aspect, wherein priority is assigned to each case sentence stored in the case sentence storage means, and the target case sentence acquisition means is provided. However, (e-1) means for acquiring one most appropriate case sentence according to the priority order as a target case sentence is provided.

According to a third aspect of the present invention, there is provided the text creating apparatus according to the first aspect, wherein (g) means for displaying a plurality of target sentences in a list display among the target sentences created by the replacing means. (H) means for receiving selection of a target sentence from the list display.

According to a fourth aspect of the present invention, there is provided the text creating apparatus according to any one of the first to third aspects, wherein: (i) a case text is stored in the case text storage means, and a dictionary keyword is stored in the keyword storage means. The dictionary registration means further comprises: (i-1) a sentence input means for receiving an input of a case sentence, and (i-2) morphologically analyzing the case sentence to generate a plurality of morphemes. And (i-3)
Designation means for receiving designation of a replacement word from among the plurality of morphemes and receiving input of a replacement word identifier corresponding to the replacement word, and (i-4) associating the specified replacement word with the replacement word identifier. And (i-5) storing the replacement word identifier as a keyword identifier, and storing the replacement word as a dictionary keyword associated with the keyword identifier. Means for storing the keyword in the means.

According to a fifth aspect of the present invention, there is provided the text creating apparatus according to any one of the first to fourth aspects, further comprising: (j) means for receiving an input of a field of the target sentence.

The invention according to claim 6 is a sentence creation method for creating a target sentence by reusing a case sentence, wherein (a) a keyword input step of inputting an input keyword which is a keyword of the target sentence; b) a target identifier obtaining step in which the target identifier obtaining means searches the keyword storage means to find a specific dictionary keyword corresponding to the input keyword, and obtains a keyword identifier associated with the specific dictionary keyword as a target identifier. And (c) the target case sentence acquisition means finds a specific replacement word identifier corresponding to the target identifier by searching the case sentence storage means, and a plurality of case sentences stored in the case sentence storage means. Target sentences that include the replacement word associated with the specific replacement word identifier Has a goal case sentence acquisition step of acquiring a chapter and a replacement step of generating the target sentence by replacing the input keyword replacement word is (d) substituted means included in the target case sentence.

According to a seventh aspect of the present invention, in the sentence creating method according to the sixth aspect, priority is assigned to each case sentence stored in the case sentence storage means, and the target case sentence obtaining step is performed. (C-1) a step of acquiring one of the most appropriate case sentences according to the priority order as a target case sentence.

The invention according to claim 8 is the sentence creating method according to claim 6, wherein (e) displaying a plurality of target sentences in a list from among the target sentences created in the replacing step. (F) selecting a target sentence from the list display.

According to a ninth aspect of the present invention, there is provided the text creating method according to any one of the sixth to eighth aspects, further comprising: (g) inputting a field of the target sentence prior to the keyword inputting step. .

According to a tenth aspect of the present invention, there is provided a dictionary for storing a dictionary keyword in the keyword storage means used in the text creation method according to any one of the sixth to ninth aspects, and for storing a case sentence in the case sentence storage means. (H) a text input process of inputting case text,
(i) a morphological analysis of the case sentence to generate a plurality of morphemes, and (j) designation of specifying a replacement word from among the plurality of morphemes and inputting a replacement word identifier corresponding to the replacement word. And (k) a case sentence registration step of storing the designated replacement word in the case sentence storage means together with the case sentence in association with the replacement word identifier, and (l) using the replacement word identifier as a keyword identifier, A keyword registration step of storing a replacement word in the keyword storage means as a dictionary keyword associated with the keyword identifier.

According to an eleventh aspect of the present invention, there is provided a computer-readable recording medium having recorded thereon a program for causing a computer to function as the text creating apparatus according to any one of the first to fifth aspects.

[0017]

BEST MODE FOR CARRYING OUT THE INVENTION

<1. Outline of Device Configuration> FIG. 1 is a diagram showing a configuration of a text creating device 1 according to the present invention. This sentence creation device 1
Is to perform text creation work on behalf of a text writer such as a copywriter, etc. Even if the worker does not have the text creation ability as much as the text creator, it should automatically or semi-automatically create the text It is a device that helps you to do. It also has a function of accumulating past sentences created by a sentence creator to realize more efficient sentence creation.

The sentence creating apparatus 1 includes a program section 2, a dictionary section 3 in which information used in a sentence creating operation by the program section 2 is stored, a display 4 for displaying instructions to the worker, and the like. A keyboard 51 and a mouse 52 for inputting the above-mentioned instructions, and a control unit 6 for controlling these elements are connected to the bus line 7.

The program section 2 actually performs a text creation process in accordance with an instruction of a worker. The management section 21 performs processing in the text creation apparatus 1, a display to the worker via the display 4, a keyboard 51, and the like. It has a text creation / dictionary registration instructing unit 22 for acquiring the content of an instruction from the operator via the mouse 52, and an editing unit 23 for the operator to edit the text. Note that the program section 2 is provided in the text creating apparatus 1 in a state stored in a readable storage device such as a fixed disk or a semiconductor storage device.

It is to be noted that the editing unit 23 normally handles texts created for pamphlets and the like, and that texts are automatically or semi-automatically created by an operator.
When a text creator having a writing ability wants a text to be stored in the text creation device 1 while creating a text, the text creation / dictionary registration instructing unit 22 is activated and the management unit 21 performs processing. It has become. In other words,
The dictionary registration instructing unit 22 is called from the editing unit 23.

The dictionary section 3 stores the contents of work performed by the text creator as information, and uses the information when automatically or semi-automatically preparing a text. The dictionary unit 3 includes a keyword dictionary 31 for registering a keyword of a sentence to be created, a case sentence dictionary 32 for registering a sentence created in the past as a case sentence, and a morpheme used for sentence analysis performed at the time of creating a sentence. It has a dictionary 33, and the contents of these dictionaries are stored in writable and readable storage means such as a fixed disk.

The control unit 6 controls operations other than the text creation of the text creation device 1, and mainly according to instructions from the program unit 2, the program unit 2, the dictionary unit 3, the display 4, the keyboard 51, and the mouse 52. The communication between the control unit 6 and the bus line 7 and various arithmetic processing are performed.

The configuration of the text creating apparatus 1 shown in FIG. 1 can be constructed by using an ordinary computer.
That is, a disk reader 9 is further connected to the bus line 7, and the optical disk,
By reading the contents of the program unit 2 and the dictionary unit 3 recorded on a recording medium 91 such as a magnetic disk or a magneto-optical disk into a semiconductor storage device or a fixed disk device, a computer can be used as the document creation device 1 It is. Also, the program unit 2 is used to improve the operation speed.
Only the hardware may be configured as dedicated hardware.

FIG. 2 is a diagram showing how information is exchanged between the program unit 2 and the dictionary unit 3 when writing a sentence and registering a case sentence.

As described above, the text creation / dictionary registration instructing unit 2
Numeral 2 serves as an interface with the operator, and includes a field input section 220 for inputting in advance the field of a sentence to be created, and a keyword input section 221 and an output section 222 which operate when creating the sentence. doing. Further, a sentence input unit 22 that operates at the time of dictionary registration work.
3 and a designation unit 224. In response to this, the management unit 21 that operates in response to a signal from the text creation / dictionary registration instructing unit 22 has an integrated name acquisition unit 211, a text acquisition unit 212, and a replacement unit 213 that operate at the time of text creation. Morphological analysis unit 2 that operates during dictionary registration work
14 and a registration unit 215.

The function of each of these components will be briefly described. When a sentence is prepared, after performing predetermined initial settings, a worker inputs a key word (keyword) of the sentence to be prepared. . This is performed by the operator via the keyword input unit 221. The input keyword (hereinafter, referred to as “input keyword”) is transferred to the integrated name obtaining unit 211, and the integrated name obtaining unit 211 obtains an integrated name (described later) corresponding to the input keyword from the keyword dictionary 31. The integrated name obtained by the integrated name obtaining unit 211 is transferred to the sentence obtaining unit 212, and the sentence obtaining unit 212 obtains a case sentence corresponding to the integrated name from the case sentence dictionary 32. Thereafter, the case sentence acquired by the sentence acquisition unit 212 is transferred to the replacement unit 213, and a target sentence is created. The created text is transferred to the output unit 222 and becomes a text being edited by the editing unit 23.

In the operation of registering a case sentence in the sentence creating apparatus 1, first, a sentence creator having the ability to create a sentence inputs a sentence as an example. This is sentence input unit 2
23. The input case sentence is transferred to the morphological analysis unit 214 and morphological analysis using the morphological dictionary 33 is performed. When the input example sentence is decomposed into morphemes, it is displayed on the display 4, and the creator specifies a morpheme to be changed when the example sentence is reused from these morphemes. This operation is performed via the specifying unit 224. The contents specified by the text creator are registered in the registration unit 21
5 and then registered in the keyword dictionary 31 and the case sentence dictionary 32 in a reusable state.

As described above, in the sentence preparation apparatus 1, the sentence input unit 223, the designation unit 224, the morphological analysis unit 214, the morphological dictionary 33, and the registration unit 215 function only for the registration of the case sentences. These components constitute the dictionary registration unit 28.

Therefore, when the present invention is used only for creating a sentence, the present invention may have a configuration without the dictionary registration unit 28, and the sentence creating apparatus 1 shown in FIGS. 1 and 2 creates a more appropriate sentence. A dictionary registration unit 28 for constructing a keyword dictionary 31 and a case sentence dictionary 32 so as to be able to perform this is further added.

The above is the outline of the configuration of the text creating apparatus 1. Subsequently, the keyword dictionary 31 and the case text dictionary 32
Will be described, and then the operation of the text creating apparatus 1 in text creation and dictionary registration will be described.

<2. Structure of Information in Keyword Dictionary and Case Sentence Dictionary> Before describing the text creation operation, the structure of information stored in the keyword dictionary 31 and the case sentence dictionary 32 in advance will be described.

FIG. 3 is a diagram showing the structure of information in the keyword dictionary 31. As shown, the keyword dictionary 31
Internally, the keyword 312 is stored so as to be associated with the keyword integrated name 311, and the keyword integrated name 311 is classified and stored according to field. In addition, as viewed from the keyword integrated name 311, a plurality of keywords 312 are associated with one keyword integrated name 311. The classification is used for convenience of searching for the keyword 312 or the like.
"Meat" when writing text only in the field of
The lower layer (keyword 312 side) plays a role as the keyword dictionary 31. Therefore, when writing text for "fresh food" in general, all fields under "fresh food"("meat","vegetables,
“Fish” or the like plays a role as the keyword dictionary 31.

FIG. 4 is a diagram showing a specific example of the keyword integrated name 311 and the keyword 312 in the keyword dictionary 31. This example shows the keyword 312 in the field of “meat”. In this example, the keywords “cow”, “pig” and “bird” are associated with the keyword integrated name 311 of “type”. As described above, the keyword integrated name 311 is a name indicating the concept or category of the keyword 312 associated therewith.
Specifically, in addition to “type”, there are “use”, “production place”, “part”, “processing information”, and the like. The keyword 312 associated with the keyword integrated name 311 is “meat”
It is a specific word belonging to the concept or the category indicated by the keyword integrated name 311 in the field of "."

FIG. 5 is a diagram showing an example of the classification of the case sentences 321 stored in the case sentence dictionary 32. Case sentence 3
As shown in FIG. 5, reference numeral 21 is stored by being classified according to the field in the same manner as the keyword dictionary 31. For example, when a text creation work is performed only in the field of "meat", a layer lower than "meat" (case sentence 321 side) The part () plays a role as the case sentence dictionary 32. Also, if you are writing text about "fresh food" in general,
All of the following functions as the case sentence dictionary 32. Note that the classification in the keyword dictionary 31 and the case sentence dictionary 32 is provided for using the inside of the dictionary according to the field, and when it is not assumed that the dictionary is used for each field, it is not necessary to classify according to the field. Absent.

FIG. 6 is a diagram showing the structure of information relating to each case sentence 321 in FIG. FIG. 7 is a diagram showing a specific example of the structure of the information shown in FIG.

As shown in FIG. 6, a specific morpheme (a division unit when a sentence is divided for each part of speech) which is a word included in each case sentence 321 is associated with the case sentence 321 as a replacement word 322. Further, each replacement word 322 is associated with a replacement word integrated name 323 which is a name indicating the concept and the category of the replacement word 322. Note that the name used as the replacement word integrated name 323 is of the same type as that used as the keyword integrated name 311 in the keyword dictionary 31.

The case sentence 321 is further associated with a priority 325 and a sentence length 326 if necessary.
The priority order 324 is also associated with the replacement word integrated name 323. These are the case sentences 3 used to create the sentences.
This is used when it is not possible to specify 21 as one, and its use mode will be described later.

The case sentence shown in FIG. 7 "Kobe beef, which has been fattened and fattened, is sliced for sukiyaki. Please use it for various dishes besides sukiyaki." , "Shoulder loin", "sukiyaki", "slice", and "sukiyaki" are associated as the replacement word 322 of the case sentence 321, and the replacement word "Kobe" is replaced with the replacement word integrated name 323 of "production area". Associated. Similarly, for the other replacement words “cow”, “shoulder loin”, “sukiyaki”, and “slice”, the replacement word integrated name “type”,
It is associated with “part”, “use”, and “processing information”. As a method for associating the replacement word 322 with the case sentence 321, the word itself of the replacement word 322 may be associated. However, even if a similar word exists in another position of the case sentence 321, the position of the replacement word 322 is also determined. It is preferable to register the word itself and the position of the word as information of the replacement word 322 in association with the case sentence 321 so that the case sentence can be specified.

<3. Contents of Work and Operation of Apparatus in Sentence Creation> In the sentence creation work using the sentence creation apparatus 1, an automatic mode in which an operator only inputs an input keyword which is a keyword of a sentence intended to be created, There is a semi-automatic mode for selecting the most appropriate one from a plurality of prepared sentences.

<3.1 Sentence Creation Process in Automatic Mode> FIG. 8 is a flowchart showing the work of the worker and the operation of the sentence creation device 1 in the automatic mode.
At this time, the flow of information in the program section 2 and the dictionary section 3 is shown by a solid line. Hereinafter, the operation when the example sentence shown in FIG. 7 is reused in the automatic mode to create the sentence and the operation of the sentence creating apparatus 1 will be described with reference to FIGS.
This will be described with reference to FIG.

In the text creation work in the automatic mode, a worker inputs a target text field in advance as a premise of the work (step S11). This work is sentence creation / dictionary registration instruction part 2
2 is performed through the operation of the field input unit 220. Specifically, the operator operates the keyboard 5 at the input position displayed on the display 4 by the operation of the field input unit 220.
The user inputs a field using the mouse 1 or the mouse 52, or selects and inputs a desired field from the displayed fields. As a specific example, it is assumed that the operator intends to create a sentence relating to “meat” and has selected and designated the field “meat”.

By this field selecting operation, the use range of the keyword dictionary 31 and the case sentence dictionary 32 is determined according to the selected field. Specifically, the keyword dictionary 3
The content shown in FIG. 4 is selected as 1 and the case sentence dictionary 3
For example, a case sentence below the field "meat" shown in FIG.

Next, the keyword of the target sentence, which is the sentence to be created by the operator, is input as an input keyword (step S12). This operation is performed through the operation of the keyword input unit 221. That is, the keyword input unit 221 performs display for keyword input on the display 4, and the operator uses the keyboard 51 and the mouse 52 to input as many input keywords as necessary. Specifically, it is assumed that the operator has input “shabu-shabu”, “Tottori”, “pig”, “peach”, and “1.5 mm thick slice” according to the instruction handed in advance.

When the input keywords are input, these input keywords are sent to the integrated name acquisition unit 211 of the management unit 21. The integrated name obtaining unit 211 searches the keyword dictionary 31 for a keyword 312 that matches the input keyword, and obtains a keyword integrated name 311 (target identifier) associated with the searched keyword 312 (step S13). Specifically, the keyword integrated name 31 associated with each of the keywords “shabu-shabu”, “Tottori”, “pig”, “peach”, and “1.5 mm thick slice”
As shown in FIG. 4, “use”, “production area”, “type”,
“Part” and “processing information” are acquired.

When the keyword integrated names 311 are acquired, they are sent to the sentence acquisition unit 212, and the most appropriate case sentence 321 (target case sentence) is searched and acquired from the case sentence dictionary 32 (step S14). That is, in the data structure of the case sentence 321 shown in FIG.
21 has a replacement word 322
3, a case sentence in which the names of these replacement word integrated names 323 match the names of the acquired keyword integrated names 311 is searched. In the above example, the case sentence 321 shown in FIG. 7 having “use”, “place of origin”, “type”, “part”, and “processing information” as the replacement word integrated name 323 is searched for and acquired.

When a plurality of case sentences 321 are searched, the highest priority is given based on the priority 325 and the sentence length 326 (see FIG. 6) which are previously associated with the case sentences 321 and the case The correct length is searched. If there is no case sentence 321 in which the number of types of the replacement word integrated name 323 matches the number of types of the keyword integrated name 311, the case where the degree of overlap between the keyword integrated name 311 and the replacement word integrated name 323 is the highest is determined. The text 321 is selected.

For example, case sentences 321 and 4 having three replacement word integrated names “usage”, “place of origin”, and “kind”.
Unification words, “use”, “locality”, “type”,
When there is a case sentence 321 having a “part”,
The latter case sentence 321 in which the four replacement word integrated names 323 match the keyword integrated name 311 corresponding to the five input keywords is selected. Also, all of the case sentence 321 has the same replacement word integrated name 323 as the keyword integrated name 311 corresponding to the five input keywords, but one example sentence 321 further has another alternative word integrated name 323 and the other example sentence 321. Has two other replacement word integration names 323, five of the six replacement word integration names 323 are the same as the keyword integration name 311 (ie,
The case where the number of types of the keyword integrated name 311 is higher than the number of types of the replacement word integrated name 323 is higher) The former case sentence 321 is determined as the most appropriate case sentence 321. Further, a replacement word 32 of a plurality of case sentences 321
If the degree of duplication of the two is equal, the replacement word integrated name 3
Priority 324 and case sentence 321 associated with H.23
One case sentence 321 is determined based on the priority order 325 associated with.

When the most appropriate case sentence 321 is acquired by the sentence acquisition unit 212, the case sentence 321 is sent to the replacement unit 213, and the replacement unit 213 replaces the replacement word 322.
Is replaced with the corresponding input keyword to create a target sentence (step S15). That is, in the example of the example sentence 321 shown in FIG.
"Cow", "shoulder loin", "sukiyaki", "slice",
Each of "Sukiyaki" is an input keyword "Tottori",
"Pig", "Peach", "Shabu-shabu", "1.5mm thick slice" and "Shabu-shabu" have been replaced with "The delicate and fattened Tottori pork peach is sliced 1.5mm thick for shabu-shabu. Shabu-shabu Go for the various dishes. "

When the target text is created, it is sent to the output unit 222, and is output from the output unit 222 to the editing unit 23, and then becomes a text to be edited.

Next, another example of text creation in the automatic mode will be briefly described. "Sukiyaki" as another example,
When four input keywords of “Kagoshima”, “cow”, and “red peach” are input, the integrated name acquisition unit 211 inputs the four keyword integrated names “use”, “production area”, “type”, and “part”. Is obtained.

After that, the sentence acquisition unit 212 sends the case sentence 3
21 is obtained, but it is assumed that there is no replacement word integrated name 323 having the same type and number of these keyword integrated names 311, and the example sentence 321 shown in FIG. 7 has been obtained.

In this case, since all the replacement words 322 cannot be replaced, only the replacement word 322 corresponding to the input keyword is replaced. In other words, the replacement word "slice" associated with the replacement word unified name "processing information" remains without being replaced, and the target sentence created is `` for red peach of savory and fattened Kagoshima beef for sukiyaki. Sliced. For sukiyaki and various dishes. "

If the number of types of the replacement word unification name 323 of the determined case sentence 321 is smaller than the number of types of the obtained keyword integration name 311, all the replacement words 322 are replaced with the input keywords. However, in this case, there is an input keyword that is not included in the target sentence.

As described above, in the text preparation apparatus 1 in the automatic mode, a text is automatically generated only by inputting a field or an input keyword at the time of text generation. The burden on the person is greatly reduced. In addition, it is not always necessary for a professional copy writer or the like having the ability to create sentences to perform the work, and the cost can be further reduced.

Furthermore, since the text is created using a machine, the frequency of typographical errors and mismatches in terms and phrases can be greatly reduced.

The input of the field is performed to limit the search range of the keyword dictionary 31 and the case sentence dictionary 32. If the search range is not limited, the input operation of the field is unnecessary.

The text creation operation in the automatic mode and the operation of the text creation device 1 have been described above. Next, the text creation operation and the operation of the text creation device 1 in the semi-automatic mode will be described.

<3.2 Sentence Creation Step in Semi-Automatic Mode> FIG. 9 is a diagram showing the transfer of information of each component of the program section 2 and the dictionary section 3 (see FIG. 1) in the semi-automatic mode by solid arrows. FIG. 3 is a diagram corresponding to FIG. 2 in an automatic mode. FIG. 10 is a diagram showing a work content of the worker in the semi-automatic mode and a flow of the operation of the text creating apparatus 1. The difference between FIG. 9 and FIG. 2 is that the sentence creation / dictionary registration instructing unit 22 shows a candidate sentence selection unit 225, which is not shown in FIG. 2 for convenience of illustration. When the text creation device 1 is used only in the automatic mode, the candidate text selection unit 225 is unnecessary.

In the text creation process in the semi-automatic mode, first, the operator inputs the field of the target text, and the use range of the keyword dictionary 31 and the case text dictionary 32 is determined (step S21). The input keyword is input by the operator (step S22). These input keywords are sent to the integrated name acquisition unit 211, and the keyword integrated name 311 (target identifier) associated with the same keyword 312 as the input keyword is acquired from the keyword dictionary 31 (step S23). These steps are similar to the operations and operations in the automatic mode. For the following explanation, the field “meat” is selected and input in this work, and the input keywords “cow”, “domestic”, “sukiyaki”,
It is assumed that “loin”, “1.6 mm thickness”, and “slice” have been input. Then, it is assumed that “type”, “production place”, “use”, “part”, “thickness information”, and “processing information” have been acquired as the keyword integrated name 311.

Next, the keyword integrated name 311 obtained by the integrated name obtaining unit 211 is sent to the sentence obtaining unit 212, and a predetermined number of case sentences 321 (target case sentences) are searched and obtained from the case sentence dictionary 32. You. The search for the case sentence 321 at this time is almost the same as the search in the automatic mode except that a plurality of case sentences are obtained, and the type and number of the keyword integrated name 311 and the replacement word integrated name 323 of the case sentence 321 are determined. Degree of overlap, priority 324,
325 and a predetermined number of appropriate case sentences are acquired from the most appropriate case sentences determined by the sentence length 326 (step S24).

As specific examples, the following three case sentences 321 are shown.
It is assumed that has been selected. Note that <> indicates the replacement word 3
22 and these replacement word integrated names 323 are
It is assumed that it is the same as any one of the three keyword integrated names 311.

[<Domestic><Cow><Inside peach and outside peach with a lot of redness> used. <1.6mm thick> Thin <slice>
You can make it softer. "<Slices of <shoulder loin> containing moderate fat into <about 1.5mm thickness><slice>.""<Slice> for <1.6mm thickness><sukiyaki>." These case sentences 321 The sentence is sent to the replacement unit 213, and the replacement word 322 of each case sentence 321 is replaced with the input keyword, thereby creating three candidate sentences (a predetermined number of target sentences) (step S25). These candidate sentences are sent to the candidate sentence selection unit 225 and displayed as a list on the display 4 (step S26). The above 3 displayed
Candidate sentences obtained from the two case sentences are as follows: "Since <Loose> of <Domestic><Cow> has been <sliced> to <1.6mm thickness>, it can be softer."<Slice> for <1.6mm thick><Slice> for <Sukiyaki> of <1.6mm thick>.

When the list of candidate sentences is displayed, the operator selects the most appropriate one from the list of candidate sentences using the keyboard 51 and the mouse 52, and selects the final target sentence from the above three candidate sentences. Determined from sentences (step S2
7).

The determined target sentence is sent to the output unit 222 and output as an editable sentence in the editing unit 23 (step S28).

As described above, in the text preparation apparatus 1 in the semi-automatic mode, the burden of the text preparation work can be greatly reduced, and an appropriate text can be prepared by the operator selecting a candidate text. it can. In addition, the worker is not always required to have the ability to create a sentence, so that the work of creating a sentence can be facilitated and cost can be reduced.

Further, in the case where a large amount of text has to be prepared, for example, as in a general pamphlet, there has been a problem that similar texts are scattered because a plurality of people worked at the same time. By using the semi-automatic mode of the apparatus 1, it is possible for one worker to quickly create a sentence, and it is also possible to prevent a problem that a similar case sentence is repeatedly reused.

<4. Work Contents and Device Operation in Dictionary Registration> Next, an appropriate keyword dictionary 31 and an example sentence dictionary 32 (hereinafter referred to collectively as “text creation A dictionary registration operation performed to construct a dictionary will be described.

The existence of a proper sentence creation dictionary is a prerequisite for the sentence creation work. Unless a proper sentence creation dictionary is given in advance, the sentence creation work is actually performed in the automatic mode or the semi-automatic mode. Is after the dictionary registration operation described below is repeatedly performed. In addition, the dictionary registration work is not performed only before the text creation work.For example, if an appropriate candidate text is not displayed in the semi-automatic mode, the worker himself / herself inputs an appropriate target text, and enters this into the text. It may be registered in the creation dictionary. In the dictionary registration work, it is assumed that a target sentence, which is a sentence to be created, is input as a case sentence, and the worker (sentence creator) is assumed to have a sentence creation ability.

The dashed arrows in FIG. 2 indicate the flow of information in the program section 2 and the dictionary section 3 in the dictionary registration work. In the dictionary registration work, as described above, the dictionary registration unit 28
Is mainly performed, and the processing result is registered in the keyword dictionary 31 and the case sentence dictionary 32.

The dictionary registration section 28 has a sentence input section 223 and a designation section 22 which serve as an interface with an operator.
4. It has a morphological analysis unit 214 and a registration unit 215 that perform processing inside the text creation device 1, and a morphological dictionary 33 used for morphological analysis. The processing performed by these components will be described below.

FIG. 11 is a flowchart showing the work of a worker having the ability to create a sentence in the dictionary registration work and the steps of the operation of the sentence creating apparatus 1.

In the dictionary registration work, the operator first inputs the field of the case sentence to be registered (step S31). Thus, the range registered in the text creation dictionary is determined.
This input operation is performed in accordance with the display on the display 4 through the operation of the field input unit 220 in the same manner as the text creation operation.

When the field is determined, the operator next inputs a case sentence to be registered using the keyboard 51 (step S32). This input operation is performed through the operation of the text input unit 223.

The input case sentence is converted to a morphological analysis unit 214.
The morphological analysis is performed here with reference to the morphological dictionary 33. That is, the case sentence is decomposed into morphemes that are words for each part of speech, and a series of morphemes is obtained (step S33). FIG. 12 shows an example of the result of the morphological analysis of the first sentence of the example of the example sentence shown in FIG.

When the morpheme analysis unit 214 acquires morphemes, these morphemes are sent to the designation unit 224 and displayed in a list. Here, the operator designates a morpheme to be replaced when the example sentence is reused from the displayed morpheme as a replacement word. In the example shown in FIG. 12, "Kobe", "cow",
“Shoulder roast”, “Sukiyaki”, and “Slice” are specified. Note that "Sukiyaki" is similarly designated for the second sentence of the example of the example sentence shown in FIG. Further, for each of the specified replacement words, the operator inputs a replacement word integrated name indicating the concept and the category of the replacement word. The input of the replacement word integrated name is selected from a predetermined list and input. These operations are performed through the operation of the designation unit 224 of the text creation / dictionary registration instruction unit 22 (step S34). Specifically, for the replacement word specified earlier, the "Origin", "Type", "Part",
“Use”, “processing information”, and “use” are selected and input.

By designating a replacement word and selecting and inputting a replacement word integrated name, a case sentence 321, a replacement word 322, and a replacement word integrated name 323 to be registered in the case sentence dictionary 32 are determined, and these are sent to the registration unit 215. Is registered in the case sentence dictionary 32 with the structure shown in FIG. 7 (step S35). In addition, the priorities 324, 325
May be input by the operator, the priority 324 may be determined according to the number of registered replacement word integrated names, and the priority 325 is determined in the order of registration. You may make it. As for the sentence length 326, the length of the registered case sentence is automatically registered.

The registration unit 215 also registers the specified replacement word and the selectively input replacement word integrated name in the keyword dictionary 31 (step S36). That is, if the same keyword 312 as the replacement word specified in the keyword integration name 311 that is the same as the name of the selected and input replacement word integration name is not registered, the replacement word is used as the keyword 312 in the keyword integration name 311. Register in association. As a result, the search of the keyword dictionary 31 at the time of writing a sentence allows more keywords 312 to be searched.
For the input keyword (FIG. 8:
Step S12, FIG. 10: Step S22) allows input of a wider variety of keywords. As a result, various sentences can be created as target sentences.

As described above, the registered contents of the dictionary for creating a sentence are additionally updated by the dictionary registration work, and it becomes possible to create more various kinds of sentences in the following sentence creation work.

Note that this dictionary registration work can be performed, for example, when a copy writer creates a sentence, and does not have the problem of taking time only for the construction of a dictionary.

Further, by using a structure in which the dictionary for creating sentences is classified according to fields, it is possible to quickly create more appropriate sentences.

<5 Modification> Although the text creating apparatus 1 according to the embodiment of the present invention has been described above, the present invention is not limited to the above embodiment.

For example, in the semi-automatic mode in the above-described embodiment, a target sentence is acquired after a candidate sentence is created. A list of acquired example sentences (sentences before replacement) is displayed. The operator may select and specify the most appropriate sentence from the list display, and then obtain the target sentence through the replacement process.

In the above embodiment, the case sentence dictionary 3
Although the priority order is registered inside the keyword dictionary 31, the priority order may be registered inside the keyword dictionary 31. In this case, for example, the priority is associated with the acquired keyword integrated name 311 and the most appropriate case sentence 321 is determined from the priority of the keyword integrated name 311 corresponding to the replacement word integrated name 323 associated with the case sentence 321. Can be determined.

Further, in the text creation operation in the above embodiment, the keyword integrated name 311 associated with the same keyword 312 as the input keyword is obtained. The keyword integrated name 311 associated with the keyword 312 having some relation may be acquired. As a result, keywords that are not registered in the keyword dictionary 31 can be used as input keywords.

In the description of the above-described embodiment, it has been described that the expressions and expressions of sentences can be unified. However, conversely, the degree of change of expressions and expressions is set as an initial setting, and the expressions are not biased. You may do so.

Further, in the above-described embodiment, the sentence creation apparatus for Japanese is used, but it is needless to say that other languages such as English can be similarly applied.

[0087]

As described above, according to the first to ninth and eleventh aspects of the present invention, it is possible to easily and quickly create an appropriate target sentence for an input keyword. A worker who does not can create a large amount of sentences in a short time, and can save time and costs required for creating the sentences. In addition, typographical errors, terminology and
Inconsistent wording is also prevented.

According to the second or seventh aspect of the present invention, since the target text can be created only by the input of the input keyword by the operator, the target text can be created more quickly, and the text creation is required. Time and expenses can be significantly reduced.

According to the third or eighth aspect of the present invention, a worker can obtain a target sentence only by inputting an input keyword and selecting a target sentence, so that a more appropriate target sentence can be quickly created. Can be. Further, even when a large number of sentences are created, it is possible to prevent the same sentence from being duplicated.

In the invention according to claim 4 or 10, since the dictionary keyword can be stored in the keyword storage means and the case sentences can be stored in the case sentence storage means, the contents stored in the keyword storage means and the case sentence storage means can be stored. It can be more appropriate, and a more appropriate target sentence can be created.

According to the fifth or ninth aspect of the invention, since the field of the target sentence is input, the range of use of the keyword storage means and the case sentence storage means can be limited, and the target sentence can be created more appropriately and quickly. Becomes

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a configuration of a text creation device (computer).

FIG. 2 is a diagram showing a flow of information in a program section and a dictionary section in an automatic mode and dictionary registration work.

FIG. 3 is a diagram showing a structure of information of a keyword dictionary.

FIG. 4 is a diagram showing an example of registered contents of a keyword dictionary.

FIG. 5 is a diagram showing an example of classifying case sentences.

FIG. 6 is a diagram showing a structure of information of a case sentence dictionary.

FIG. 7 is a diagram showing an example of a case sentence and its structure.

FIG. 8 is a flowchart showing a text creation process in the automatic mode.

FIG. 9 is a diagram showing a flow of information in a program section and a dictionary section in a semi-automatic mode.

FIG. 10 is a flowchart showing a text creation process in a semi-automatic mode.

FIG. 11 is a flowchart showing steps in a dictionary registration operation.

FIG. 12 is a diagram illustrating an example of a morphological analysis result of a case sentence.

FIG. 13 is a diagram showing an example of the contents of a pamphlet.

[Explanation of symbols]

Reference Signs List 1 sentence creation device (computer) 4 display 28 dictionary registration unit 31 keyword dictionary 32 case sentence dictionary 91 recording medium 211 integrated name acquisition unit 212 sentence acquisition unit 213 substitution unit 214 morphological analysis unit 215 registration unit 220 field input unit 221 keyword input unit 223 sentence input unit 224 designation unit 225 candidate sentence selection unit 311 keyword integrated name 312 keyword 321 case sentence 322 replacement word 323 replacement word integrated name 324, 325 Priority S11-S16, S21-S28, S31-S36 Step

Claims (11)

[Claims] 1. A sentence creating apparatus for creating a target sentence by reusing a case sentence, comprising: (a) case sentence storage means for storing a plurality of case sentences including a replacement word associated with a replacement word identifier; (B) a keyword storage unit that stores a plurality of dictionary keywords in association with a keyword identifier; (c) a keyword input unit that receives an input of an input keyword that is a keyword of the target sentence; and (d) a search for the keyword storage unit. By,
A target identifier obtaining means for finding a specific dictionary keyword corresponding to the input keyword and obtaining a keyword identifier associated with the specific dictionary keyword as a target identifier; (e) searching for the case sentence storage means, A specific replacement word identifier corresponding to the target identifier is found, and a case sentence including a replacement word associated with the specific replacement word identifier is selected from a plurality of case sentences stored in the case sentence storage means. Target case sentence acquisition means for acquiring as a case sentence; and (f) replacement means for creating the target sentence by replacing a replacement word included in the target case sentence with the input keyword. Text preparation device. 2. The sentence creating apparatus according to claim 1, wherein priority is assigned to each case sentence stored in said case sentence storage means, and said target case sentence acquisition means comprises: (e- 1) A text creating apparatus, comprising: means for acquiring, as a target case text, one most appropriate case text according to the priority order. 3. The sentence creating apparatus according to claim 1, wherein: (g) means for displaying a plurality of target sentences in a list from among the target sentences created by the replacing means; Means for receiving selection of a target sentence from the list display. 4. The sentence creating apparatus according to claim 1, wherein: (i) a case sentence is stored in said case sentence storage means, and a dictionary keyword is stored in said keyword storage means. Means, and the dictionary registration means, (i-1) sentence input means for receiving the input of the case sentence, and (i-2) analysis means for morphologically analyzing the case sentence to generate a plurality of morphemes, (I-3) designation means for receiving designation of a replacement word from among the plurality of morphemes and receiving input of a replacement word identifier corresponding to the replacement word, and (i-4) A case sentence registration unit that is stored in the case sentence storage unit together with the case sentence in association with the replacement word identifier, (i-5) the replacement word identifier is a keyword identifier,
A keyword registering unit for storing the replacement word in the keyword storage unit as a dictionary keyword associated with the keyword identifier. 5. The text creating apparatus according to claim 1, further comprising: (j) means for receiving an input of a field of the target sentence. 6. A sentence creating method for creating a target sentence by reusing a case sentence, comprising: (a) a keyword inputting step of inputting an input keyword that is a keyword of the target sentence; and (b) obtaining a target identifier. Means for searching a keyword storage means for finding a specific dictionary keyword corresponding to the input keyword, and for obtaining a keyword identifier associated with the specific dictionary keyword as a target identifier; (c) The target case sentence obtaining means searches the case sentence storage means to find a specific replacement word identifier corresponding to the target identifier, and the specific case word out of a plurality of case sentences stored in the case sentence storage means. Obtain a case sentence containing the replacement word associated with the replacement word identifier as the target case sentence A target case sentence acquiring step, and (d) a replacing step of replacing the replacement word included in the target case sentence with the input keyword to create the target sentence. How to make. 7. The sentence creating method according to claim 6, wherein priority is assigned to each case sentence stored in said case sentence storage means, and said target case sentence obtaining step comprises: (c- 1) a step of acquiring one of the most appropriate case sentences according to the priority order as a target case sentence. 8. The sentence creating method according to claim 6, wherein (e) the target sentence created in the replacing step includes:
A text creation method, further comprising: displaying a plurality of target texts in a list display; and (f) selecting a target text from the list display. 9. The sentence creating method according to claim 6, further comprising: (g) inputting a field of the target sentence prior to the keyword inputting step. How to write sentences. 10. A dictionary registration method in which a dictionary keyword is stored in the keyword storage means used in the text creation method according to claim 6, and a case text is stored in the case text storage means. (H) a sentence input step of inputting a case sentence; (i) an analysis step of morphologically analyzing the case sentence to generate a plurality of morphemes; and (j) specifying a replacement word from among the plurality of morphemes. And (k) a case sentence registration step of storing the designated replacement word in the case sentence storage means together with the case sentence in association with the replacement word identifier. (L) the replacement word identifier is a keyword identifier, and the replacement word is a dictionary keyword associated with the keyword identifier. Dictionary registration method characterized by having a keyword registration step of storing in the word storage means. 11. A computer-readable recording medium on which a program for causing a computer to function as the text creating apparatus according to claim 1 is recorded.