JP2004118741A - Curtailed word formation supporting device and curtailed word formation supporting program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a curtailed word formation supporting device and a program therefor suitable to be used by a number of unspecified people as dictionary information and capable of forming a curtailed word easily which can get consensus comprehensively. <P>SOLUTION: The curtailed word formation supporting device and the program therefor include a basic abbreviation substitution processing part 14, an abbreviated name formation processing part 15, and the other definition selection processing part 16 whereby the formation of a candidate for an abbreviated name is performed for one word using a different method. A curtailed name forming pattern determination part 12 selects at least one of the processing parts 14, 15 and 16 on the basis of data shown concerning the word etc. to be processed in dictionary data 21 read by a dictionary data reading part 11 and a forming pattern determination rule 22. A formation processing part 13 conducts the formation of an abbreviated name candidate using the selected processing part, and the abbreviated name candidate formed by the processing part(s) is used as the candidate for the curtailed word. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an abbreviated word creation support device and an abbreviated word creation support program for supporting creation of abbreviated words related to phrases included in the electronic catalog dictionary when creating an electronic catalog dictionary.
[0002]
[Prior art]
An electronic catalog system is a system for providing product information using a computer network. "ISO 13548 (Parts Library)" has been established as an international standard for constructing an electronic catalog system for electronically providing product information on the Internet. In this "ISO 13548", an electronic catalog is composed of a dictionary and contents, and a uniform data structure is provided for sharing and reuse of product information.
[0003]
In the dictionary defined by “ISO13548”, the main components are a product category represented on a tree as a concept hierarchy and a technical attribute (specification item) to which the product category belongs. Short names are associated with names for defining product categories and specification items.
[0004]
Since short names do not include words that have gained widespread consensus on all of the names used to define product categories and specifications, it is important that the workers have appropriate I had to consider short names.
[0005]
As a method of generating a short name of a product classification, the method of Patent Document 1 is known.
[0006]
In addition, as a general method of shortening character data, a method of converting a character using an abbreviation dictionary to keep the number of characters within a limit (for example, see Patent Literature 2), a generation method using an abbreviation generation rule (for example, see Patent Literature 3) It has been known.
[0007]
[Patent Document 1]
JP 2000-11178 A
[0008]
[Patent Document 2]
JP-A-2002-7104
[0009]
[Patent Document 3]
JP-A-11-25117
[0010]
[Problems to be solved by the invention]
As described above, in the related art, a dictionary creation operator has performed the creation of abbreviated names, which required a large number of man-hours.
[0011]
By using the technique of Patent Document 1, it is possible to reduce the burden on the dictionary creation worker. However, in the technique of Patent Document 1, since the creation of the abbreviated name is based on the information of the concept hierarchy in the product classification, it is not possible to generate the abbreviated name of the specification item having no such information.
[0012]
Further, the technique disclosed in Patent Document 2 is a method in which a character string is replaced with previously registered abbreviation data. If there is no corresponding abbreviation data, it is necessary to input the abbreviation data each time. It is difficult to achieve a sufficient effect in the development of a dictionary.
[0013]
Further, the technique of Patent Document 3 is a method of generating an abbreviation by dividing a character string, combining characters, and combining the extracted characters. However, this is a method for an individual to efficiently search a sentence. Lack of flexibility in the expressiveness of the rules. For this reason, it is difficult to generate a short name that can imagine the meaning of the original name as a standard short name used by an unspecified majority. This is not a suitable technique.
[0014]
Therefore, the present invention has been made in view of such circumstances, and the purpose of the present invention is to provide abbreviations that are suitable for use by an unspecified number of people as information in a dictionary and that can obtain a broad consensus. It is to enable easy creation.
[0015]
[Means for Solving the Problems]
In order to achieve the above object, the present invention provides a method for generating a plurality of candidate words, each of which generates a candidate character string as a shortened word by a different method based on information indicated in an electronic catalog dictionary with respect to a phrase to be abbreviated. Means for selecting at least one of the plurality of candidate generating means by a selecting means based on information indicated in the electronic catalog dictionary and a predetermined rule with respect to a phrase to be abbreviated for word creation. The candidate selecting unit selects the candidate character string generated by the selected candidate generating unit as the shortened word candidate.
[0016]
By adopting such means, at least one of the plurality of candidate generating means for generating the candidate character string as the shortened word in a different manner is different from the electronic catalog dictionary with respect to the phrase for which the shortened word is to be created. And the candidate character string generated by the validated candidate generating means is selected as a candidate for a shortened word. Therefore, a plurality of methods are appropriately used to generate abbreviated word candidate character strings.
[0017]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
[0018]
FIG. 1 is a block diagram of an electronic catalog dictionary creation device according to the present embodiment. The electronic catalog dictionary creation device supports creation of a short name (SHORTNAME), which is a requirement of the IS01584 standard, in the definition of dictionary data of the IS01584 standard.
[0019]
As shown in FIG. 1, the electronic catalog dictionary creation device has a processor 1, a main memory 2, a hard disk device 3, a CD-ROM drive 4, a keyboard / mouse 5, and a display 6. These components are connected to each other via a bus 7.
[0020]
The processor 1 performs various control processes by executing software processing according to a program stored in the hard disk device 3.
[0021]
The main memory 2 temporarily stores software actually used by the processor 1 and other data.
[0022]
The hard disk device 3 stores a program for an operating system used by the processor 1. The hard disk device 3 stores an application program, a driver program, or any other data in addition to an operating system program. The hard disk drive 3 stores a short name creation support program 3a. The abbreviated name creation support program 3a causes the processor 1 to execute processing to be described later for creating abbreviated names for registration in the electronic catalog dictionary.
[0023]
The CD-ROM drive 4 reads data from the CD-ROM medium M.
[0024]
The keyboard / mouse 5 receives various instruction inputs from the user.
[0025]
The display 6 displays images for notifying various information to the user.
[0026]
A general-purpose personal computer, for example, is used as the electronic catalog dictionary creation device. Therefore, initially, the short name creation support program 3a is not installed. Therefore, as a part of the introduction work of the information processing system, the short name creation support program 3a stored in the CD-ROM medium M is installed in the hard disk device 3.
[0027]
Here, the short name creation support program 3a causes the processor 1 to operate as each processing unit as shown in FIG.
[0028]
FIG. 2 is a block diagram showing a relationship between processing units realized by the processor 1 based on the short name creation support program 3a. FIG. 2 also shows various data stored in the main memory 2 and the hard disk device 3 for reference by each processing unit.
[0029]
As shown in FIG. 2, the processor 1 includes a dictionary data reading unit 11, a short name generation pattern determining unit 12, a generation processing control unit 13, a basic abbreviation substitution processing unit 14, an abbreviation generation processing unit 15, and another definition selection processing unit 16. And it operates as each of the short name registration unit 17. Also, dictionary data 21, generation pattern determination rules 22, constraint condition setting data 23, abbreviation candidate generation condition data 24, basic abbreviation data 25, morphological analysis dictionary 26, abbreviation generation rule setting data 27, and abbreviation candidate list 28 are prepared. You.
[0030]
The dictionary data reading unit 11 reads a dictionary element for which a short name is to be created from the dictionary data 21.
[0031]
The short name generation pattern determination unit 12 determines a list of processing contents to be executed based on the information included in the read dictionary element and the generation pattern determination rule 22. The processing content list determined here indicates one of the basic abbreviation replacement processing unit 14, the abbreviation generation processing unit 15, and the other definition selection processing unit 16 that should be used for generating abbreviation candidates.
[0032]
The generation processing control unit 13 causes the processing unit shown in the processing content list among the basic abbreviation substitution processing unit 14, the abbreviation generation processing unit 15, and the other definition selection processing unit 16 to execute the abbreviation candidate creation processing. The generation processing control unit 13 controls the basic abbreviation substitution processing unit 14, the abbreviation generation processing unit 15, and the other definition selection processing unit 16 so as to generate an abbreviation candidate that matches the condition indicated in the constraint condition setting data 23.
[0033]
The basic abbreviation replacement processing unit 14 generates abbreviation candidates by replacing the character string that is the basis of the abbreviation with the basic abbreviation data 25 under the conditions indicated in the abbreviation candidate generation condition data 24. The basic abbreviation substitution processing unit 14 adds the generated abbreviation candidate to the abbreviation candidate list 28.
[0034]
The abbreviation generation processing unit 15 analyzes the syntax of the character string that is the source of the abbreviation with reference to the morphological analysis dictionary 26, and identifies abbreviation candidates based on the analysis result and the rules indicated in the abbreviation generation rule setting data. Generate. The abbreviation generation processing unit 15 adds the generated abbreviation candidate to the abbreviation candidate list 28.
[0035]
The other definition selection processing unit 16 selects, as abbreviation candidates, an abbreviation determined for a phrase indicating the same object in a language different from the character string that is the base of the abbreviation. The other definition selection processing unit 16 adds the generated abbreviation candidate to the abbreviation candidate list 28.
[0036]
The short name registration unit 17 determines an abbreviation from the abbreviation candidates stored in the abbreviation candidate list 28, and registers the abbreviation in the dictionary data 21.
[0037]
Next, the operation of the electronic catalog dictionary creating apparatus configured as described above will be described.
FIG. 3 is a flowchart showing an overall processing algorithm by the processor 1 in the present embodiment.
In step ST1-1, the processor 1 first reads a dictionary element E for which a short name is to be created from the dictionary data 21 as a function of the dictionary data reading unit 11.
[0038]
FIG. 4 is a diagram showing the data structure of the dictionary data 21.
In the present embodiment, the dictionary data 21 is expressed as tabular data, and one dictionary element is expressed as one row. The dictionary element is a CODE field 21a, PREFNAME. EN field 21b, PREFNAME. JA field 21c, SHORTNAME. EN field 21d, SHORTNAME. Each field includes a JA field 21e and a Data Type field 21f. Note that each field of the dictionary element is composed of the requirements specified in IS01584, but FIG. 4 shows a portion necessary for the description of the present invention. The CODE field 21a indicates CODE, which is an identifier of a dictionary element. PREFNAME. In the EN field 21b, a PREFNAME. EN is indicated. PREFNAME. In the JA field 21c, a PREFNAME. JA is indicated. SHORTNAME. In the EN field 21d, SHORTNAME. EN is indicated. SHORTNAME. In the JA field 21e, SHORTNAME. JA is indicated. The Data Type field 21f indicates a Data Type that defines the format of the dictionary element.
[0039]
For example, in the row R1 in FIG. 4, the dictionary element PREFNAME. EN is “Maximum Operational Rated Voltage”, and PREFNAME. Indicates that EN is "maximum rated working voltage". Also, SHORTNAME. EN and SHORTNAME. Indicates that no JA has been registered. Furthermore, it indicates that Data Type is “Real”, that is, a real number type.
[0040]
Subsequently, in step ST1-2, the processor 1 determines the short name generation pattern based on the information included in the dictionary element E and the generation pattern determination rule 22 as a function of the short name generation pattern determination unit 12, and further executes the short name generation pattern. Determine the list of processing contents to be performed.
[0041]
FIG. 5 is a diagram showing a data structure of the generation pattern determination rule 22.
The generation pattern determination rule 22 includes a determination condition field 22a and a generation processing field 22b. The judgment condition field 22a describes a judgment condition regarding a dictionary element to be generated. Here, the determination condition is SHORTNAME. Of the dictionary element to be processed. The presence or absence of data in the EN field 21d, SHORTNAME. It consists of the presence / absence of data in the JA field 21e and the value of the Data Type field 21f. The generation processing field 22b indicates the contents of processing applied to dictionary elements that match each determination condition.
[0042]
For example, in row R2 in FIG.
(SHORTNAME.JA = none) ∧ (SHORTNAME.EN = none) ∧ (DataType = (Real∨Integer))
Regarding the processing of the dictionary element satisfying the condition of “SHORTNAME. The abbreviation substitution processing (PreferredName.EN) and the abbreviation generation processing (PreferredName.EN) are applied to the generation of EN. A rule is described in which another definition selection processing (SHORTNAME.EN) is applied to the generation of JA.
[0043]
After that, the processor 1 starts the first loop processing in step ST1-4 as a function of the generation processing control unit 13. When the first loop process is started, in step ST1-5, the processor 1 reads a constraint condition from the constraint condition setting data 23 according to the process content listed above, and determines the process content.
[0044]
FIG. 6 is a diagram showing the data structure of the constraint condition setting data 23. The constraint condition setting data includes a generation target field 23a, a maximum character string length field 23b, and an available character field 23c. The generation target field 23a indicates a feed to be generated. The maximum character string length field 23b indicates the maximum character string length of the generated short name. The usable character field 23c indicates a type of usable character. In the usable character field 23c, flags relating to a plurality of character types are prepared in advance, and indicate that a character type whose corresponding flag is "Y" is available. In the present embodiment, a constraint defined in IEC61360 is described as a constraint.
[0045]
Thereafter, in step ST1-6, the processor 1 checks whether or not the abbreviation replacement processing is included in the above-listed processing contents, and if it is included, in step ST1-7, the basic abbreviation replacement processing is performed. Execute
[0046]
FIG. 7 is a flowchart showing a processing algorithm in the basic abbreviation replacement processing. The basic abbreviation replacement processing is executed by the processor 1 as a function of the basic abbreviation replacement processing unit 14.
[0047]
In the basic abbreviation replacement process, the processor 1 first reads a character string X corresponding to a field described in a processing target included in the processing content from a dictionary element E to be processed in step ST2-1.
[0048]
Next, in step ST2-2, the processor 1 reads the abbreviation candidate generation condition data 24. Next, in step ST2-3, the processor 1 searches the partial character strings of the character string X for a word that matches the word described in the basic abbreviation data 25 based on the read abbreviation candidate generation condition data 24. I do.
[0049]
FIG. 8 is a diagram showing the data structure of the abbreviation candidate generation condition data 24. In the present embodiment, the abbreviation candidate generation condition data 24 includes a matching method field 24a and a partition character field 24b. The matching method field 24a indicates designation information of a matching method between the character string X and the word stored in the basic abbreviation data 25. Specifically, a list of available matching methods and a flag indicating whether or not the matching methods can be used are prepared, and the matching method with the flag “Y” should be used. . In the example of FIG. 8, it is indicated that the matching method “longest match on the left” is specified as the matching method to be used. Therefore, in step ST2-3, the processor 1 performs a search using the matching method specified as described above.
[0050]
FIG. 9 is a diagram showing the data structure of the basic abbreviation data 25. The basic abbreviation data 25 includes a word field 25a and an abbreviation field 25b. The word field 25a indicates a word serving as a headline for a search. The abbreviation field 25b indicates an abbreviation corresponding to the word indicated in the corresponding word field 25a. As the basic abbreviation data 25, an existing standard or the like may be used, and IEC60027, IEC60747, IEC60748 or the like, which is the standard of the character symbol of the technical attribute, may be stored in the abbreviation field. 25 can broadly obtain consensus.
[0051]
Then, in step ST2-4, the processor 1 checks whether a matching word is found.
[0052]
If a matching word is found, in step ST2-5, the processor 1 converts the partial character string searched from the character string X into an abbreviation for the matching word as a partitioning character specified in the abbreviation candidate generation condition data 24. To generate a character string X 'replaced with a character string replaced with the character string X'.
[0053]
As shown in FIG. 8, the abbreviation candidate generation condition data 24 includes a list of usable partition characters and a flag indicating whether or not the partition characters can be used. The flag is set to “Y”. Is the one that should be used. The example in FIG. 8 indicates that the character “-” is specified as a partition character.
[0054]
Subsequently, in step ST2-6, the processor 1 checks whether or not the character string X 'satisfies the constraint indicated by the constraint condition setting data 23. When it is confirmed that the constraint is satisfied, the processor 1 adds the character string X ′ to the abbreviation candidate list 28 in step ST2-7. When the abbreviated name candidate is added to the abbreviated name candidate list 28 or when it is confirmed in step ST2-6 that the constraint is not satisfied, the processor 1 converts the character string X 'into the character string X in step ST2-8. Then, the processes after step ST2-3 are repeated.
[0055]
When it is confirmed in step ST2-4 that no matching word is found, the processor 1 determines in step ST2-9 whether the character string X satisfies the constraint indicated by the constraint condition setting data 23. Confirm. Then, when it is confirmed that the constraint is satisfied, the processor 1 adds the character string X to the abbreviation candidate list 28 as an abbreviation candidate in step ST2-10. When the abbreviation candidate is added to the abbreviation candidate list 28 as described above, or when it is confirmed in step ST2-9 that the constraint is not satisfied, the processor 1 ends the basic abbreviation substitution process.
[0056]
When the basic abbreviation replacement processing is completed in this way, the processor 1 checks in step ST1-8 in FIG. 3 whether the processing contents listed in step ST1-3 include the abbreviation generation processing. I do. If it is confirmed in step ST1-6 that the listed processing contents do not include the abbreviation replacement processing, the processor 1 proceeds to step ST1-7 without performing the basic abbreviation replacement processing in step ST1-7. Check at -8. Then, if it is confirmed that the abbreviation generation process is included, the processor 1 executes the abbreviation generation process in step ST1-9.
[0057]
FIG. 10 is a flowchart showing a processing algorithm in the abbreviation generation processing. Note that this abbreviation generation processing is executed by the processor 1 as a function of the abbreviation generation processing unit 15.
[0058]
First, in step ST3-1, the processor 1 sets the abbreviation generation rule according to the input from the user, and stores it as the abbreviation generation rule setting data 27.
[0059]
FIG. 11 is a diagram showing the data structure of the abbreviation generation rule setting data 27. The abbreviation generation rule setting data 27 includes setting information 27a of a method of omitting a word and setting information 27b of a partition character when performing character concatenation.
[0060]
Omission methods can be set separately for the subject and the qualifier, and a list of applicable processing methods and the availability of these processing methods can be set for each. And a flag indicating “Y” is to be used.
[0061]
In this embodiment, the following four omission methods are provided.
[0062]
(1) Extract n characters from the beginning of each morpheme.
[0063]
{Circle around (2)} Extracting the last n characters connected to the last n characters of each morpheme.
[0064]
(3) Extract the vowels (excluding the head) of each morpheme that are omitted.
[0065]
(4) Extract the entire morpheme (= no processing).
[0066]
Separate characters can be set separately for the case between the qualifier and the subject and between the qualifiers. A list of applicable partition characters for each of them and the availability of those And a flag indicating the flag is set to "Y".
[0067]
Therefore, in step ST3-1, the processor 1 accepts a user specification of which omission method and partitioning character to use, and sets the corresponding flag to "Y". Regarding the omission method, it is allowed to specify a plurality of processes. For example, the processing of (1) and (4) can be selected for the morpheme corresponding to the subject, and the processing of (2) and (3) can be selected for the morpheme corresponding to the modifier. As for the partitioning character, the user can specify the partitioning character between the subject and the modifier and between the modifiers. Thus, more flexible abbreviation generation can be performed.
[0068]
Next, in step ST3-2, the processor 1 reads, from the dictionary element E, a character string Y corresponding to a field described as a processing target included in the processing content. Then, in step ST3-3, the processor 1 analyzes the syntax of the character string Y based on the morphological analysis dictionary 26, and generates a character string list L1 divided for each morpheme. For morphological analysis, we extract words and analyze the relationship between words, and if we can divide them into a set of nouns that are the subject of a sentence and a set of modifiers that modify the nouns, we can achieve this by using existing parsing technology Is possible.
[0069]
FIG. 12 is a diagram showing a data structure of the morphological analysis dictionary 26. The morphological analysis dictionary includes a character string field 26a describing a word, a word part of speech field 26b, and a semantic feature field 26c.
[0070]
Subsequently, the processor 1 starts a first loop process in step ST3-4. When the first loop process is started, in step ST3-5, the processor 1 executes the abbreviation generation rule indicated by the abbreviation generation rule setting data 27 for each of the elements E1 of the character string list L1 and sets a set of abbreviations of E1. (E2, E3,... En) are generated. Then, the processor 1 generates a list L2 having a set (E1, E2, E3,... En) obtained by adding E1 to the set (E2, E3,... En). After that, in step ST3-6, the processor 1 checks whether or not the generation of the list L2 has been completed for all the elements of the character list L1. Until the end can be confirmed, the processor 1 checks in steps ST3-4 to ST3-3. 7 is repeatedly executed.
[0071]
When it is confirmed in step ST3-6 that the generation of all the lists L2 has been completed, the processor 1 determines in step ST3-8 the elements of the list L2, ie, a list of each morpheme and one or more generated abbreviations. A set u is generated, which includes elements that are possible combinations of each other.
[0072]
Next, the processor 1 starts a second loop process in step ST3-9. When the second loop process is started, in step ST3-10, the processor 1 generates a concatenated character string Y 'via the partition character specified by the abbreviation generation rule setting data 27 for each element of the set u. Further, in step ST3-11, the processor 1 confirms whether or not the concatenated character Y 'satisfies the constraint of the constraint condition setting data 23. If the constraint is satisfied, the concatenated character Y' is regarded as an abbreviated candidate list. Add to 28. When the abbreviated name candidate is added to the abbreviated name candidate list 28 or when it is confirmed in step ST3-11 that the constraint is not satisfied, the processor 1 determines in step ST3-13 about all the elements of the set U. It is determined whether or not the generation of the concatenated character string Y 'has been completed, and the second loop of steps ST3-9 to ST3-14 is repeatedly executed until the completion is confirmed.
[0073]
If it is confirmed in step ST3-13 that the generation of the concatenated character string Y 'has been completed for all the elements of the set U, the processor 1 ends the abbreviation generation processing.
[0074]
When the abbreviation generation processing is completed in this way, the processor 1 checks in step ST1-10 in FIG. 3 whether or not the processing content listed in step ST1-3 includes another definition selection processing. I do. If it is confirmed in step ST1-8 that the listed processing contents do not include the abbreviation generation processing, the processor 1 proceeds to step ST1-9 without performing the abbreviation generation processing in step ST1-9. Confirmation at 10 is performed. Then, if it is confirmed that the other definition selection process is included, the processor 1 executes another definition selection process in step ST1-11.
[0075]
FIG. 13 is a flowchart showing a processing algorithm in the other definition selection processing. The other definition selection process is executed by the processor 1 as a function of the other definition selection processing unit 16.
[0076]
First, in step ST4-1, the processor 1 reads a character string S corresponding to a field described as a processing target included in the processing content from the dictionary element E. Next, in step ST4-2, the processor 1 checks whether or not the character string S satisfies the constraint of the constraint condition setting data 23. If the constraint is satisfied, the processor 1 sets the character string S as an abbreviated candidate. Add to list 28. This makes it possible to use a short name defined for a certain language as a candidate for a short name in another language, such as adding an English short name to a candidate as a Japanese short name.
[0077]
When the abbreviation candidate is added to the abbreviation candidate list 28 in this way, or when it is confirmed in step ST4-2 that the constraint is not satisfied, the processor 1 ends the other definition selection process.
[0078]
After completing the other definition selection process, the processor 1 executes the short name registration process in step ST1-12 in FIG. When it is confirmed in step ST1-10 that the other process is not included in the listed processing contents, the processor 1 proceeds to step ST1-11 without performing the other definition selecting process. The process proceeds to ST1-12 to execute abbreviated name registration processing.
[0079]
FIG. 14 is a flowchart showing the processing algorithm of the short name registration processing. This short name registration process is executed by the processor 1 as a function of the short name registration unit 17.
[0080]
The processor 1 first reads the abbreviation candidate list 28 in step ST5-1. Subsequently, in step ST5-2, the processor 1 accepts that the setting of the display form of the abbreviation candidate list 28 is input by the user. Here, the setting for accepting the input is, for example, the display order, the maximum display number, and the like. The display order is assumed to be alphabetical order, eyewear order, character string length order, or the like. The maximum display number is an upper limit of the display number when the number of candidates is large.
[0081]
Next, in step ST5-3, the processor 1 displays the information of the abbreviated name candidate list 28 in a display form according to the input setting. Then, in step ST5-4, the processor 1 accepts the user's designation as to whether or not to select an abbreviation from the displayed list. If the selection from the displayed list is designated, the processor 1 proceeds to step ST5-5. In, one abbreviation is selected from the abbreviation candidate list based on the user's designation.
[0082]
However, if the selection from the displayed list is rejected by the user, in step ST5-6, the processor 1 receives a user instruction as to whether or not to manually input the abbreviation. In ST5-7, the input of the abbreviation by the user is accepted.
[0083]
After receiving the input of the abbreviation by the user or receiving the designation of the abbreviation from the list in step ST5-5, the processor 1 determines the specified abbreviation as the short name in step ST5-8. Then, the short name is written to the corresponding output destination of the dictionary element E specified as the target of generation. Then, the processor 1 ends the short name registration processing.
[0084]
If manual input is rejected by the user in step ST5-6, the processor 1 returns the process to step ST5-2, accepts resetting of the display mode, and repeats the subsequent processing.
[0085]
Thereafter, in step ST1-13, the processor 1 checks whether or not the processing regarding all the elements of the processing content list generated in step ST1-3 has been completed. The first loop from step 4 to step ST1-14 is repeatedly executed.
If it is confirmed in step ST1-13 that the processing for all elements has been completed, the processor 1 ends the processing in FIG.
[0086]
Hereinafter, the manner in which abbreviations are registered by the above processing will be described with reference to specific examples.
Here, a case is considered where the element in row R1 in FIG. 4 is to be processed. The element to be processed is SHORTNAME. EN and SHORTNAME. Since no JP has been set and Data Type is “Real”, the row R2 matches the determination condition in the generation pattern determination rule 22 in FIG. Therefore, for this element, SHORTNAME. For EN, abbreviation substitution processing (PreferredName.EN) and abbreviation generation processing (PreferredName.EN) are performed, and SHORTNAME.EN is also executed. For JA, the other definition selection processing (SHORTNAME.EN) is applied. The original data used in the abbreviation replacement process (PreferredName.EN) and the abbreviation generation process (PreferredName.EN) is “Maximum Operational Rated Voltage”.
[0087]
FIG. 15 is a diagram illustrating a state in which abbreviation candidates are generated by the abbreviation substitution process.
As shown in FIG. 15, the character string “Maximum Operational Rated Voltage” is partially and gradually replaced by a basic abbreviation. Here, the character string “Max-Op-V_r” is included in the abbreviation candidate list 28 as the abbreviation candidate. Will be added.
[0088]
FIG. 16 is a diagram illustrating a manner in which abbreviation candidates are generated by the abbreviation generation process.
As shown in FIG. 16, a character list L1, a list L2, and a set U are respectively generated, and five character strings such as “Man-Opl-Rated Vol” and “Man-Opl-Ratd Vol” are newly generated from these. An abbreviation candidate is generated and added to the abbreviation candidate list 28.
[0089]
FIG. 17 shows a short name candidate list 28 showing SHORTNAME. It is a figure showing signs that EN is determined.
Among the total of six abbreviation candidates as shown in FIG. 17A generated by the abbreviation substitution process and the abbreviation generation process, the candidate “Max-Op-V_r” as shown in FIG. Is SHORTNAME. It is determined as EN.
[0090]
FIG. 18 is a diagram showing a state of the other definition selecting process.
As shown in FIG. 18A, SHORTNAME. The other definition selection process for JA is SHORTNAME. It is defined that EN is directly added to the abbreviation candidate list 28 as an abbreviation candidate. Therefore, as shown in FIG. 18B, SHORTNAME. The phrase “Max-Op-V_r” determined as EN is added to the abbreviation candidate list 28 as an abbreviation candidate. After selection by the user, as shown in FIG. 18C, a candidate “Max-Op-V_r” is displayed as SHORTNAME. It is determined as JA.
[0091]
The SHORTNAME. EN and SHORTNAME. JA is added to the dictionary data 21 as shown in FIG. 19, and the dictionary data 21 is updated.
[0092]
As described above, according to the present embodiment, the basic abbreviation replacement processing unit 14, the abbreviation generation processing unit 15, and the other definition selection processing unit 16 that generate the abbreviation candidates by different methods are provided. SHORTNAME. EN and SHORTNAME. JA, and an appropriate processing unit is selected based on the data type of the name from which the short name is created, and abbreviated name candidates are generated by this processing unit. This makes it possible to provide standard short name candidates that are suitable for use by an unspecified number of dictionaries and that can be widely consented.
[0093]
Further, according to the present embodiment, the abbreviation candidate specified by the user is determined as the abbreviated name, so that the created abbreviated name is updated in the dictionary data 21 as in the present embodiment, for example. It can be easily used for example. By automatically performing the update process of the dictionary data 21 as in the present embodiment, it is possible to perform the process of creating the dictionary data 21 very efficiently.
[0094]
Further, according to the present embodiment, the provision of the other definition selection processing unit 16 makes it possible to employ abbreviated names based on other languages for a certain language. Correspondingly, it is possible to create an appropriate abbreviated name.
[0095]
Note that the present invention is not limited to the above embodiment. For example, the morpheme analysis used by the abbreviation generation processing unit 15 is to use word segmentation information from the English definition corresponding to the case where it is difficult to accurately analyze morphemes when generating abbreviated Japanese names. It is also possible to use a new morphological analysis technique utilizing other language information, such as performing more accurate morphological analysis.
[0096]
Further, it is not always necessary to include all of the basic abbreviation replacement processing unit 14, the abbreviation generation processing unit 15, and the other definition selection processing unit 16, and a processing unit that generates abbreviation candidates by a method different from these processing units may be provided. You may do it.
[0097]
Further, the abbreviation name may be automatically determined in accordance with a predetermined condition, for example, by setting the number of characters to be the smallest among the abbreviation candidates.
[0098]
Further, the present invention can be widely applied to the creation of abbreviated words other than the abbreviated names in an electronic catalog, such as encoding a character string or generating a function name in software development.
[0099]
Further, the short name creation support program 3a is not limited to be installed from the CD-ROM medium M, but may be installed from a floppy disk (R) or another type of storage medium, or may be installed on the Internet or the like. You may make it install via a network.
[0100]
In addition, various modifications can be made without departing from the spirit of the present invention.
[0101]
【The invention's effect】
According to the present invention, at least one of the plurality of candidate generating means for generating the candidate character string as the shortened word by a different method is indicated in the electronic catalog dictionary with respect to the phrase for which the shortened word is to be created. Of the candidate character string generated by the valid candidate generating means as a candidate for a shortened word, a plurality of methods are appropriately used. Generation of abbreviated word candidate character strings will be performed, and it is possible to easily create abbreviated words that are suitable for use by unspecified majority as dictionary information and that can obtain a wide consensus. it can.
[Brief description of the drawings]
FIG. 1 is a block diagram of an electronic catalog dictionary creation device according to an embodiment of the present invention.
FIG. 2 is a block diagram showing a configuration of a processing system realized by a processor 1 by software processing based on a short name creation support program 3a in FIG.
FIG. 3 is a flowchart showing an overall processing algorithm by a processor 1 in FIG. 1;
FIG. 4 is a view showing a data structure of dictionary data 21 in FIG. 2;
FIG. 5 is a diagram showing a data structure of a generation pattern determination rule 22 in FIG. 2;
FIG. 6 is a diagram showing a data structure of constraint condition setting data 23 in FIG. 2;
FIG. 7 is a flowchart illustrating a processing algorithm in a basic abbreviation replacement process.
FIG. 8 is a diagram showing a data structure of abbreviation candidate generation condition data 24 in FIG. 2;
FIG. 9 is a diagram showing a data structure of basic abbreviation data 25 in FIG. 2;
FIG. 10 is a flowchart illustrating a processing algorithm in abbreviated name generation processing.
FIG. 11 is a diagram showing a data structure of abbreviation generation rule setting data 27;
FIG. 12 is a diagram showing a data structure of a morphological analysis dictionary 26;
FIG. 13 is a flowchart showing a processing algorithm in another definition selection processing.
FIG. 14 is a flowchart illustrating a processing algorithm of a short name registration process.
FIG. 15 is a diagram showing a state in which abbreviation candidates are generated by the abbreviation substitution process.
FIG. 16 is a diagram illustrating a manner in which abbreviation candidates are generated by abbreviation generation processing.
FIG. 17 shows an example of a short name candidate list. The figure which shows a mode that EN is determined.
FIG. 18 is a diagram showing a state of another definition selecting process.
FIG. 19: Confirmed SHORTNAME. EN and SHORTNAME. The figure which shows an example of the dictionary data updated by JA.
[Explanation of symbols]
1. Processor
2 ... Main memory
3 ... Hard disk drive
3a: Short name creation support program
4: CD-ROM drive
5. Keyboard / mouse
6. Display
7 ... Bus
11: Dictionary data reading unit
12: Short name generation pattern judgment unit
13: generation processing control unit
14 Basic abbreviation replacement processing unit
15: Abbreviation generation processing unit
16 Other definition selection processing unit
17… Short name registration section
21 ... Dictionary data
22 ... Generation pattern judgment rule
23… Constraint condition setting data
24 ... Abbreviated name candidate generation condition data
25 Basic abbreviation data
26 ... Dictionary for morphological analysis
27 ... Abbreviation generation rule setting data
28 ... Abbreviated name candidate list
M: CD-ROM media

Claims (10)

In a shortened word creation support device that supports creation of shortened words related to phrases included in an electronic catalog dictionary,
A plurality of candidate generating means for generating a candidate character string as the shortened word in a different manner based on information indicated in the electronic catalog dictionary with respect to a phrase to be shortened,
Selecting means for selecting at least one of the plurality of candidate generating means based on information indicated in the electronic catalog dictionary and a predetermined rule with respect to a phrase to be abbreviated,
An abbreviation word creation support device, comprising: candidate selection means for selecting a candidate character string generated by the candidate generation means selected by the selection means as a candidate for the shortened word. 2. The abbreviated word creation support apparatus according to claim 1, further comprising: abbreviated word determination means for determining the abbreviated word according to a predetermined determination condition from among the abbreviated word candidates selected by the candidate selection means. . 3. The abbreviated word creation method according to claim 2, wherein the abbreviated word determination means receives a user designation of abbreviated words from among the abbreviated word candidates and determines abbreviated words according to the user designation. Support device. 4. The abbreviated word creation support device according to claim 1, further comprising a registration unit that registers the abbreviation word determined by the abbreviation word determination unit in the electronic catalog dictionary. 2. The candidate generating unit according to claim 1, wherein the candidate generating unit generates the candidate character string by replacing a character string included in a phrase to be abbreviated with a basic abbreviation prepared in advance. A shortened word creation support device according to any one of claims 4 to 4. The candidate generating unit includes a unit that analyzes the syntax of a phrase for which a shortened word is to be created, and processes the analysis result based on a predetermined shortened word generating rule to generate the candidate character string. The abbreviation creation support apparatus according to any one of claims 1 to 5, wherein: The electronic catalog dictionary is multilingualized,
4. The method according to claim 1, wherein the candidate generating unit includes a unit that uses a shortened word defined for another language corresponding to a phrase to be abbreviated as the candidate character string. 6. The shortened word creation support device according to any one of 6. What causes the computer to perform processing for supporting the creation of abbreviated words related to phrases included in the electronic catalog dictionary, the computer,
A plurality of candidate generating means for generating a candidate character string as the shortened word in a different manner based on information indicated in the electronic catalog dictionary with respect to a phrase to be shortened,
Selecting means for selecting at least one of the plurality of candidate generating means based on information indicated in the electronic catalog dictionary and a predetermined rule with respect to a phrase to be abbreviated,
A shortened word creation support program that functions as candidate selecting means for selecting all of the candidate character strings generated by the candidate generating means selected by the selecting means as candidates for the shortened word. 9. The computer according to claim 8, further comprising causing the computer to function as shortened word determining means for determining the shortened word from predetermined candidates of the shortened word selected by the candidate selecting means in accordance with a predetermined determining condition. Abbreviated word creation support program. The computer-readable storage medium according to claim 9, further comprising causing the computer to function as a registration unit that registers the abbreviation word determined by the abbreviation word determination unit in the electronic catalog dictionary.

Images