JP2008217170A - Information processor and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To flexibly perform setting of a conversion rule for converting a text into an XML. <P>SOLUTION: An analysis definition preparation part 21 generates a setting screen for performing the designation of conditions for determining a block as the unit of a range as an analysis object in at least a text and the designation of conditions for extracting individual data in the block as a setting screen for setting a conversion rule for analyzing a text, and for converting it into the XML(extensible markup language), and on the basis of individual conditions designated through the setting screen, prepares analysis definition data including a character string where information showing the individual conditions with tags added thereto as the conversion rule. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an information processing apparatus and program capable of setting conversion rules used for data conversion.

  2. Description of the Related Art In recent years, the work of sequentially changing the format of an article distributed as a message from a news agency to a newspaper company or the like from text to XML (extensible markup language) (including XML-based languages such as NewsML) has been progressing. Each newspaper company is also working on creating an environment for managing articles in XML format. On the other hand, articles that newspaper companies send themselves are still often created in text format. In the future, it is desired that such texts are also converted to XML and managed as XML files, and the XML files can be used secondarily for other purposes.

There are various techniques for analyzing text and converting to XML. For example, Patent Document 1 discloses a method of cutting out a part suitable for reading through processing such as dividing text into sentence units. Patent Document 2 discloses a technique for creating an XML document through processing such as segmenting a natural sentence text by a morphological analysis or the like.
JP 2002-334070 A JP 2003-288332 A

  However, the conventional methods including Patent Document 1 and Patent Document 2 described above are flexible when it is desired to change or set the target to be extracted from the text, or to change or set the form of the conversion result for the target. Can not cope with. In particular, techniques for performing the above-described change work and setting work on a setting screen that is visually understandable for the operator are very promising techniques in the future, and these techniques have not yet been proposed.

  The present invention has been made in view of the above circumstances, and an object thereof is to provide an information processing apparatus and program capable of flexibly setting a conversion rule for converting text into XML.

  The information processing apparatus according to the present invention, as a setting screen for setting a conversion rule for analyzing text and converting it into XML (extensible markup language), at least a block that is a unit of a range to be analyzed in the text. A means for generating a setting screen capable of specifying a condition to be determined and a condition for extracting individual data in the block, and based on the individual conditions specified through the setting screen, And means for creating analysis definition data including character strings each having a tag attached to information indicating individual conditions.

  The program according to the present invention determines at least a block that is a unit of a range to be analyzed in the text as a setting screen for setting a conversion rule for analyzing text and converting it into XML (extensible markup language). A function for generating a setting screen that can specify a condition and a condition for extracting individual data in the block, and the individual conditions specified through the setting screen, The computer is realized with a function of creating analysis definition data including character strings each having a tag attached to information indicating a condition.

  According to the present invention, it is possible to flexibly set a conversion rule for converting text into XML.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a diagram illustrating an example of a configuration of an information processing system according to an embodiment of the present invention.

  The information processing system is provided in, for example, a newspaper company, and includes a server system 1, a communication unit 2, and a plurality of terminals 11, 12, 13,. The communication unit 2 receives an article distributed as a message from a communication company, and passes the received article to the server system 1. The server system 1 includes various servers such as a material management server that manages article materials and an article processing server that processes articles using article materials. Each server executes a requested process by using a corresponding database (DB) in response to a request from the terminals 11, 12, 13. Each of the terminals 11, 12, 13... Can request the server system 1 to execute various processes.

  FIG. 2 is a diagram illustrating an example of the configuration of the information processing apparatus according to the present embodiment.

  The information processing apparatus (computer) 20 corresponds to, for example, a certain server (for example, an article processing server) included in the server system 1 described above. The information processing apparatus 20 includes an analysis definition creation unit 21 and an analysis unit 22 as computer programs executed by a processor (CPU (central processing unit) or the like). The storage device 23 stores an analysis definition storage unit for storing an analysis definition file created by the analysis definition creation unit 21. The storage device 24 stores a text file that is input to the analysis unit 22. The storage device 25 stores an XML file output from the analysis unit 22. Here, the storage device 24 and the storage device 25 are physically separated. However, the storage device 24 and the storage device 25 may be physically separated as an internal configuration as a physically integrated storage device. .

The input text file and the output XML file are described assuming the file format, but the file format is not limited to the file format and may be configured in a database table format.

  The analysis definition creation unit 21 sets a conversion rule (hereinafter also referred to as “analysis definition”) that analyzes the text of an article and converts it into XML in response to a request from a certain terminal. It has the function to generate. In this setting screen, it is possible to specify a condition for determining a block that is a unit of a range to be analyzed in at least the text and a condition for extracting individual data in the block. The generated setting screen information is sent to the requesting terminal and displayed on the display unit of the terminal.

  The analysis definition creating unit 21 also includes an analysis definition including character strings each tagged with information indicating the individual conditions as the conversion rule based on the individual conditions specified through the generated setting screen. Data (file) is created and stored in the analysis definition storage unit in the storage device 23. Although the analysis definition data is described assuming a file format, the analysis definition data is not limited to the file format, and may be configured in a database table format.

  The analysis unit 22 determines the analysis range for the text stored in the storage device 24 and extracts the individual data according to the analysis definition data stored in the analysis definition storage unit in the storage device 23, thereby Conversion processing for converting the text into XML is performed, and the XML data after the conversion processing is stored in the storage device 25 or transmitted to the requesting terminal.

  FIG. 3 is a diagram for explaining an example of the analysis definition.

  Consider a case where the text to be input is, for example, a data list indicating the name, age, gender, and residence of three employees of a Tokyo store. In the analysis definition data, a condition for determining “analysis target range (block)” and a condition for extracting “individual data” in the analysis target range are defined.

  For example, as shown in FIG. 3, a parent block 31 and a plurality of child blocks 41 are defined as conditions for determining the “range to be analyzed”. The parent block 31 corresponds to the 8th line from the beginning of the text, and it is defined that root is assigned as an XML tag. Each of the plurality of child blocks 41 corresponds to two rows in the parent block 31, and it is defined that employee is assigned as an XML tag.

  On the other hand, as a condition for extracting “individual data”, the name, age, gender, and residence existing in each of the plurality of child blocks 41 are extracted at the positions shown in the condition in FIG. Is defined. In addition, it is defined that name, age, sex, and place are assigned as XML tags to the name, age, sex, and place of residence, respectively.

  XML data, which is a result of analysis / conversion based on such analysis definition data, is shown on the right side of FIG. As shown in this result, a hierarchical structure corresponding to the relationship between the parent block 31 and the child block 41 is formed, and each individual piece of extracted data is provided with a tag. (For example, the process of calculating the average age of employees, the process of calculating the gender ratio, the process of calculating the distribution of residences by prefecture, etc.) Recognize. In addition, when using the XML data, data access using a tag becomes possible.

  As a result, even if a new data item to be extracted is added or conversely deleted, it is not necessary to make a significant change in the processing procedure in the system using the XML data.

  When there are child blocks in each of the plurality of parent blocks, it is possible to define different individual data extraction conditions for each parent block when defining individual data extraction conditions in each child block. For example, as shown in FIG. 5, consider the case where there are blocks 52, 52, and 53 in the text, and there are child blocks in each of these blocks. In this case, the individual data extraction condition in the child block in the block 51, the individual data extraction condition in the child block in the block 52, and the individual data extraction condition in the child block in the block 53 are made different. Is also possible.

  As shown in FIG. 6, four types of conditions that can be specified when creating an analysis definition are “repetition definition”, “block conversion”, “block extraction condition”, and “individual extraction condition”. Can be mentioned. Hereinafter, these four conditions will be described in detail.

1. Repeat definition The repeat definition is used for the purpose of defining the range where the block extraction condition and individual extraction condition are repeatedly applied. The result of applying the repetition definition to the analysis target is converted (formatted) into “line units” and applied to all the repetitions (rows) of the block extraction condition and the individual extraction conditions.

2. Block Conversion Block conversion is used for the purpose of shaping a block so that block extraction conditions and individual extraction conditions can be set. The result of the block transformation applied to the analysis target is shaped in “row units”, and the block extraction condition and the individual extraction condition are repeatedly applied to the shaped block as appropriate.

3. Block extraction condition The block extraction condition is used for the purpose of dividing the analysis target into units called blocks in order to determine the application range of the individual extraction conditions. The result of applying the block extraction condition (s) to the analysis target is “line unit”. Therefore, when a plurality of blocks are defined (nested definition) for one block, the original block is treated as a separate row (block). Finally, by applying a plurality of block conditions, the block conditions are designated so as to be “minimum block unit” that is the application range of the individual extraction conditions. When the upper block definition is repeatedly specified, the same block extraction condition is applied to all blocks to be repeated.

4). Individual extraction conditions The individual extraction conditions are used for the purpose of defining individual data items and tag names that are finally extracted from the input text. The application range of the individual extraction condition (s) is only one block. Therefore, in order to define the individual extraction conditions, the application range is always determined by defining a block at the upper level. When the upper block definition is repeatedly specified, the same individual extraction condition is applied to all the blocks to be repeated.

  FIG. 7 is a diagram illustrating an example of processing for converting text into XML data with ruby.

  For example, let us consider a case where the text is “? * Day * day?

  In order to realize the extraction of characters with ruby using this text as an analysis target, the following processing is executed.

  (1) Set the range for repeated analysis.

  (2) The contents of the repeated analysis in (1) above are set.

      (2-1) Extract only the part “? ▲ * ■ (Δ * □)” separately from other character strings.

(However, ▲, ■, △, and □ indicate any characters.)
(2-2) Individual data is extracted with a pair of “▲ and Δ” and “■ and □”.

  The following results are obtained by performing a conversion process for attaching a tag by performing such extraction.

<root>
<Rb><R1> Book </ R1><R2> Hon </ R2></Rb><Rb><R1> Day </ R1><R2> That </ R2></Rb>
<Rb><R1> Sunny </ R1><R2> Sei </ R2></Rb><Rb><R1> Ten </ R1><R2> Ten </ R2></Rb>
</ root>
FIG. 8 is a diagram illustrating an example of an analysis definition setting screen. Here, the example of ruby analysis described in FIG. 7 is applied.

  This setting screen is divided into an area 70A and an area 70B. In the area 70A, the file name of the analysis definition to be created, the file name of the test message (text used for the test), the display method, and the like can be designated.

  In the area 70B, items for which various conditions described in FIG. 6 can be specified are provided. Here, an item 71 for specifying “repetition definition”, an item 72 for specifying “block conversion”, and an item 73 for specifying “block extraction condition” are provided. By looking at the hierarchical structure information on the left side of these items, the block or item to be set can be identified. In addition, by clicking on a desired hierarchy in the hierarchical structure information, it is possible to specify conditions for a desired block or item. In addition, an area 74 for displaying test data as a target in the test message and an area 75 for displaying in real time execution results obtained by converting the test data using the analysis definition being created are provided.

  Here, various items shown in the region 70B will be described in more detail.

  “Repetition” selects the content from the following when it is desired to repeatedly apply the same analysis condition to the currently selected block.

      “Not specified”: Do nothing.

      “Row direction”: Take out as child blocks with the number of rows specified in “Repetition value”. Repeat until the current block is empty.

“Row direction (automatic)”: Repeat block extraction until the current block is empty. Specify when the number of repeated lines is not exactly the same. (For example, when "Specified character string +" is specified for "End position")
“Column direction”: Extract as child blocks with the number of columns (number of characters) specified in “Repetition value”. Repeat until the current block is empty.

      “Column direction (CSV column)”: A character string designated by “Repetition value” is regarded as a delimiter of CSV and is divided into blocks in the column direction.

  In “Repetition value”, a value corresponding to the selection content of “Repetition” is input.

  ・ "Block conversion" selects the conversion contents to be performed on the block from the following.

      “Vertical / Horizontal conversion”: Convert blocks by swapping rows and columns.

      “Designated character padding”: The block is converted so that the number of character padding columns becomes the number specified by “conversion parameter”.

      “Specified character delimiter”: The block is converted by regarding the character string specified in “Conversion parameter” as a line feed.

      “Designated character delimiter (previous SP)”: The block is converted by regarding the character string designated by the “conversion parameter” as a line feed. If the number of columns in each row is different, a full-width space is added to the beginning of the row so that the number of columns is the same.

  In “Conversion parameter”, a value corresponding to the selected content of “Block conversion” is input.

  “Add” adds a child block or child item within the current block. Only one type of block or item can be added in the block. Multiple blocks and items can be added.

  -“Delete” deletes a child block or child item in the current block.

  • For “Tag name”, enter the tag name of the block / item.

  “Start position” selects the designation method of the start position of the block / item from the following.

      “Relative position”: The position corresponding to the “number of start characters” from the previous item in the block / the end position of the block is set as the start position.

      “Absolute position”: The position corresponding to the “number of start characters” from the beginning in the block is set as the start position.

      “Designated character string”: A row / column in which the character string designated by “start character string” exists is searched, and the row / column is moved from there by the number designated by “number of start characters”.

  -“Start Character Count”: Enter a value according to the selected content of “Start Position”.

  "End position": Select a block / item end position designation method from the following.

      “Relative position”: The position corresponding to the “number of end characters” from the start position is set as the end position.

      “Absolute position”: The position corresponding to the “number of end characters” from the beginning in the block is set as the end position.

      “Specified character string”: A row / column in which the character string specified by “End character string” exists is searched, and the number of lines / columns specified by “Number of end characters” is moved from there.

      “To the end”: Select the last line of the current block.

      “Designated character string +”: The next line is inspected from the current line, and if the character string designated by the “end character string” exists, the next line is also examined in the same manner. If the specified string does not exist, the check ends at that line.

  -“End Character Count”: Enter a value according to the selection at “End Position”.

  "End character string": Enter a value according to the selection at "End position".

  FIG. 9 is a diagram illustrating another example of the analysis definition setting screen. This setting screen is a screen for editing an item existing in the block that is the editing target in the setting screen of FIG. In the setting screen of FIG. 8, as in the setting screen of FIG. 9, an item 81 for designating “repetition definition” and an item 82 for designating “block conversion” are provided. In particular, unlike the setting screen of FIG. 8, an item 83 for designating “item extraction condition” is provided. In addition, an area 84 for displaying the test data that is the target in the test message and an area 85 for displaying the execution result obtained by converting the test data with the analysis definition being created in real time are provided.

  As an execution result display method, "XML display" and "table display" are prepared as shown in the area 70A, and a desired display method can be selected. In FIG. 8 and FIG. 9, the execution result is displayed by the method of “XML display”. The “table display” is displayed, for example, in a form as shown in an area 75 in FIG. Here, an example of baseball inning is shown. It can be seen that the information classified into three is displayed in a table format in the horizontal direction.

  FIG. 11 shows an example of the analysis definition file created through the setting screen described with reference to FIGS. As can be seen from FIG. 11, since the block or item to be extracted is defined by the XML method in combination with the tag, the conversion process to the XML data can be surely executed.

  Next, an example of the operation performed by the analysis definition creation unit 21 will be described with reference to FIG.

  The analysis definition creation unit 21 inputs a test message (step S11), sets block conversion condition settings for block extraction (step S13), block extraction condition settings (step S14) for each block, The first loop processing (steps S12 to S21) including individual data extraction condition setting (step S20) is executed.

  In the first loop process, after the block extraction condition is set (step S14), if there is a child block in the block, the child block extraction condition is set for each child block (step S17). The second loop processing (steps S16 to S19) including the individual data extraction condition setting (step S18) is executed. On the other hand, if there is no child block, the second loop process (steps S16 to S19) is not executed.

  After the first loop processing (steps S12 to S21), when the operator performs an operation such as pressing the execution button shown in FIG. 8 or FIG. 9, a test message conversion test is executed ( Step S22). When there is no problem in the test result and the operator performs an operation such as pressing the save button shown in FIG. 8 or FIG. 9, an analysis definition file is created (step S23), and the analysis definition creating unit 21 The operation ends.

  Next, an example of the operation performed by the analysis unit 22 will be described with reference to FIGS. 13 and 14.

  As shown in FIG. 13, the analysis unit 22 reads the analysis definition created by the analysis definition creation unit 21 (step S31), and executes a conversion process for converting the target text into XML according to the analysis definition (step S31). Step S32). When this conversion process is completed, the operation by the analysis unit 22 is completed. Next, details of the conversion processing in step S32 described above are shown in FIG.

  As shown in FIG. 14, the analysis unit 22 first extracts the processing range to be converted from the given text (step S41), and outputs a tag indicating the start (step S42). Then, loop processing (steps S43 to S47) including conversion processing for each block (determination of analysis range and extraction of individual data) is executed from the beginning to the end of the processing range to be converted.

  In this loop process, it is determined whether or not a condition for the child conversion process is defined in the block that is the object of the current conversion process (step S44). The child conversion process means a conversion process (determination of analysis range and extraction of individual data) performed for each child block when there are one or more child blocks in the parent block. If the condition for the child conversion process is not defined, the individual data in the block that is the object of the current conversion process is extracted, and the extracted result is output together with the tag (step S45). On the other hand, if the conditions for the child conversion process are defined, the child conversion process is executed (step S46).

  After completion of the loop processing (steps S43 to S47), a tag indicating the end is output, and the operation by the analysis unit 22 is ended.

  Thus, according to the above embodiment, in the creation of analysis definition data, in the setting screen that can specify conditions such as “repetition definition”, “block conversion”, “block extraction condition”, “individual extraction condition”, Since the setting operation can be performed while viewing the execution result using the test data, a conversion rule for converting the text into XML can be easily created. Also, the analysis definition can be easily changed when the object to be extracted from the text is to be changed or when the conversion result is to be changed. In addition, by creating the conversion result based on the created analysis definition as an XML file, it is possible to handle the meaning, relations, and format of the data in addition to the element values in a unified manner. Can be improved. Specifically, the following effects can be mentioned.

  The means for extracting elements from XML can be used universally because several standard methods are implemented on various platforms.

  -If element extraction from text and processing of the extracted results are performed separately, when changing the input text format, it is only necessary to change the analysis settings for the changed part, and change the analysis process The location can be localized.

  ・ Even if there is a change (addition or deletion) in the extraction result of an element, the value is obtained based on the XML tag name or structure in the subsequent processing that analyzes the input text and uses the analysis result secondarily. , The effect is only on the part where the changed tag is handled, and the changed part of the subsequent processing can be localized.

  -Even when different processing is performed on the same text format in subsequent processing, the text analysis part can be made common, so that versatility is enhanced.

(Modification 1)
As an additional function of the analysis definition creation unit 21 in the above embodiment, among the items for inputting conditions of various analysis definitions (block definition, element definition) displayed in the setting screen, a predetermined item is selected. If a condition is not specified, a function for displaying information such as a message indicating that the condition is inappropriate or illegal may be provided. In this case, for example, attribute information indicating “input required” or the like is given to a predetermined item in advance, so that it can be distinguished from other items. This technique facilitates preformat verification when parsing input text.

  In addition, as an additional function of the analysis definition creation unit 21 in the above embodiment, among predetermined items for inputting conditions of various analysis definitions (block definition, element definition) displayed in the setting screen, a predetermined item When a condition is not designated for the item, a function of performing a complementing process for designating a predetermined condition for the item may be provided. In this case, for example, attribute information indicating that automatic input of an “implicit value” that is a provisional value for supplementing a value when a value is not input may be given in advance to a predetermined item. Good. By this method, it is possible to prevent the occurrence of inconveniences such as malfunction when the input text is analyzed.

(Modification 2)
As an additional function of the analysis definition creating unit 21 in the above embodiment, a function of automatically generating a DTD (document type definition) from an analysis definition (block definition, element definition) used at the time of input text analysis may be provided. In this way, it is possible to perform format verification on the side using the extraction result (output XML file).

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

The figure which shows an example of a structure of the information processing system which concerns on one Embodiment of this invention. The figure which shows an example of a structure of the information processing apparatus which concerns on the embodiment. The figure for demonstrating an example of an analysis definition. The figure for demonstrating the conditions of an extraction item. The figure for demonstrating that the separate data extraction conditions in the child block in each block can be varied. The figure for demonstrating four types of conditions which can be specified when creating an analysis definition. The figure which shows an example of the process which converts a text into XML data with a ruby. The figure which shows an example of the setting screen of an analysis definition. The figure which shows another example of the setting screen of an analysis definition. The figure which shows an example of the setting screen containing the display of a table format. The figure which shows an example of the analysis definition file produced through the setting screen. The flowchart which shows an example of the operation | movement by an analysis definition preparation part. The flowchart which shows an example of operation | movement by an analysis part. 14 is a flowchart showing an example of details of conversion processing in FIG. 13.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Server system, 2 ... Communication part, 11, 12, 13 ... Terminal, 20 ... Information processing apparatus, 21 ... Analysis definition preparation part, 22 ... Analysis part, 23, 24, 25 ... Memory | storage device.

Claims (6)

As a setting screen for setting a conversion rule for analyzing text and converting it into XML (extensible markup language), specification of a condition for determining at least a block as a unit of a range to be analyzed in the text, and the block A means for generating a setting screen capable of specifying conditions for extracting individual data in
Based on the individual conditions specified through the setting screen, the conversion rule includes means for creating analysis definition data including character strings each tagged with information indicating the individual conditions. Information processing apparatus.   2. The information processing apparatus according to claim 1, further comprising means for converting the text into XML according to the analysis definition data. In the setting screen,
A first item for designating a condition for extracting the block;
A second item for specifying a condition for extracting individual data included in the block;
A third item for designating a range for repeatedly applying the condition specified by the first item or a range for repeatedly applying the condition specified by the second item;
The information processing apparatus according to claim 1, wherein a fourth item for designating a condition for shaping the extraction target block is displayed.   2. The information processing apparatus according to claim 1, further comprising: means for displaying information indicating that the predetermined item displayed in the setting screen is inappropriate when a condition is not specified for the predetermined item. 4. The information processing apparatus according to any one of 3.   And a means for performing a complementing process for designating a predetermined special condition for the item when the condition is not designated for the predetermined item displayed in the setting screen. The information processing apparatus according to claim 1, wherein the information processing apparatus is an information processing apparatus. As a setting screen for setting a conversion rule for analyzing text and converting it into XML (extensible markup language), specification of a condition for determining at least a block as a unit of a range to be analyzed in the text, and the block A function to generate a setting screen that can specify the conditions for extracting individual data in
Based on the individual conditions specified through the setting screen, the computer realizes, as the conversion rule, a function of creating analysis definition data including character strings each tagged with information indicating the individual conditions. A program characterized by

Images