JP2005141295A - Device, method and program for retrieving document - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To permit a retrieving person to precisely and comprehensively designate a position and a character string by an easy operation when the position in a document is designated and the desired character string is retrieved. <P>SOLUTION: A document group being an object of retrieval is previously analyzed and an attribute name list and a value list in a table, which appear in it, are created. Attribute names extracted at that time are classified into a group where the names are the same in terms of meaning or a group where the names are similar (To put it more concretely, the group of the names whose notation is the same or similar). All the attribute names appearing in the table in the document group being the object of retrieval are displayed on the list at every group in accordance with an instruction from the retrieving person. When the prescribed attribute is designated, a value appearing in the attribute is displayed in the list. Even if there are the tables which have delicate notation shakes in the attribute names, the values, etc. in large quantities, retrieval omission due to the shake is avoided and precise retrieval can be preformed. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a document search apparatus, a document search method, and a document search program that search for an electronic document including a specified character string at a specified position.

  In the conventional document search technique, although a character string in a document can be normally specified as a search condition, a position in the document of the character string cannot be specified. Therefore, in the case of a document having a table structure such as a spreadsheet, a document including a specified character string as an attribute (column item) X value and a document including an attribute (column item) Y value are the same. Has been searched for.

  Therefore, it is conceivable that the appearance position can be specified together with an arbitrary character string as a search condition. As a result, for example, only documents including a character string designated by a designated attribute can be searched.

  However, in the case described above, it is necessary for the searcher to accurately specify the desired attribute. This designation is not always easy when there are a large number of documents to be searched. For example, there are customer lists A, B,..., N, and when an attempt is made to search for a document including a customer residing in Tokushima City, information indicating the customer's address is the attribute “address” in document A, If the document B is stored in the attribute “current address”,..., And the document N is stored in the attribute “address / location”, the searcher determines that the character string “Tokushima City” should appear as “address” “ All of "Current Address" ... "Address / Location" must be specified.

  Moreover, even if the value is substantially the same, the format is often different for each table. For example, the telephone number is described as “088-666-****” in document A, “088 (666) *****” in document B,..., “088666 *** in document N”. If the search has been made, the searcher must designate all of the character strings “088-666-***”, “088 (666) *****”, and “088666 ***”.

  As described above, in the prior art, even the same information in terms of meaning may be stored under different attribute names for each table or in different formats. If you don't know exactly what format the name string appears in what name attribute, you will not be able to search for it. In addition, there is a problem that the pickup of the attribute and the format is based on the searcher's experience and manual work at present.

  In order to solve the above-described problems caused by the prior art, the present invention classifies and displays a plurality of attributes appearing in a search target document having a table structure by semantically the same or similar, An object of the present invention is to provide a document search apparatus, a document search method, and a document search program capable of displaying a list of actual values of attributes.

  In order to solve the above-described problems and achieve the object, a document search apparatus according to claim 1 is a document search apparatus for searching for an electronic document including a specified character string at a specified position. Attribute name extraction means for extracting the name of each attribute in the table included in the table, attribute name classification means for classifying the names extracted by the attribute name extraction means into groups of semantically identical or similar, Attribute name display means for displaying the name classified by the attribute name classification means for each of the groups.

  According to the first aspect of the present invention, it is possible to display a list of table attribute names appearing in the search target document group for each group.

  The document search apparatus according to the invention of claim 2 is the document search apparatus according to claim 1, wherein the attribute name classification means extracts the names extracted by the attribute name extraction means with the same or similar notation. It is characterized by classifying into groups.

  According to the invention of claim 2, the plurality of attributes are classified for each group based on the similarity of the names.

  According to a third aspect of the present invention, there is provided the document retrieval apparatus according to the first or second aspect, further comprising a value extracting means for extracting a value of each attribute whose name is extracted by the attribute name extracting means. And value display means for displaying the value extracted by the value extraction means.

  According to the invention of claim 3, it is possible to display a list of values in the table appearing in the search target document group.

  According to a fourth aspect of the present invention, there is provided a document retrieval method for retrieving an electronic document including a designated character string at a designated position, wherein the name of each attribute in a table included in the electronic document is obtained. The attribute name extraction step to extract, the attribute name classification step to classify the names extracted in the attribute name extraction step into groups of semantically the same or similar, and the names classified in the attribute name classification step And an attribute name display step for displaying each group.

  According to the invention of claim 4, it is possible to display a list of table attribute names appearing in the search target document group for each group.

  According to a fifth aspect of the present invention, there is provided the document search method according to the fourth aspect of the present invention, wherein in the attribute name classification step, the names extracted in the attribute name extraction step are identical or similar to each other. It is characterized by classifying into groups.

  According to the invention of claim 5, the plurality of attributes are classified for each group based on the similarity of the names.

  According to a sixth aspect of the present invention, there is provided the document retrieval method according to the fourth or fifth aspect, further comprising a value extracting step of extracting a value of each attribute whose name is extracted in the attribute name extracting step. And a value display step for displaying the value extracted in the value extraction step.

  According to the invention of claim 6, it is possible to display a list of values in the table appearing in the search target document group.

  Moreover, according to the document search program concerning invention of Claim 7, a computer can be made to perform the method described in any one of Claims 4-6.

  According to the document search device, the document search method, and the document search program according to the present invention, the document search device, the document search method, and the document search program capable of searching for a character string by designating a position in the document. In this case, the position and the character string can be accurately and comprehensively designated by the searcher with an easy operation.

  Exemplary embodiments of a document search apparatus, a document search method, and a document search program according to the present invention will be explained below in detail with reference to the accompanying drawings.

  FIG. 1 is an explanatory diagram showing an example of a hardware configuration of a document search apparatus according to an embodiment of the present invention. In the figure, 101 indicates a CPU that controls the entire apparatus, 102 indicates a ROM that stores basic input / output programs, and 103 indicates a RAM that is used as a work area of the CPU 101.

  Reference numeral 104 denotes an HDD (hard disk drive) that controls reading / writing of data with respect to the HD (hard disk) 105 according to the control of the CPU 101, and 105 denotes an HD that stores data written according to the control of the HDD 104. Yes.

  Reference numeral 106 denotes an FDD (flexible disk drive) that controls reading / writing of data with respect to the FD (flexible disk) 107 according to the control of the CPU 101, and 107 denotes a detachable FD that stores data written according to the control of the FDD 106. Respectively.

  Reference numeral 108 denotes a CD-RW drive that controls reading / writing of data with respect to the CD-RW 109 according to the control of the CPU 101, and reference numeral 109 denotes a removable CD-ROM that stores data written according to the control of the CD-RW drive 108. RW is shown respectively.

  Reference numeral 110 denotes a cursor, menu, window, or display for displaying various data such as characters and images, 111 denotes a keyboard having a plurality of keys for inputting characters, numerical values, various instructions, and the like, and 112 denotes various types. A mouse that performs selection and execution of an instruction, selection of a processing target, movement of a mouse pointer, and the like is shown.

  Reference numeral 113 denotes a network I / F that is connected to a network such as a LAN or a WAN via a communication cable 114, and functions as an interface between the network and the CPU 101. Reference numeral 100 denotes a bus for connecting the above-described units. ing.

  Next, FIG. 2 is an explanatory diagram functionally showing the configuration of the document search apparatus according to the embodiment of the present invention. In the figure, reference numeral 200 denotes a document storage unit, which holds a plurality of documents to be searched by a document search unit 204 described later. Specifically, these documents include documents created by so-called “spreadsheet software”, HTML documents including <table> tags, more generally XML documents including predetermined tags meaning tables, or It is assumed that image data (GIF file, PDF file, etc.) including a table image divided by ruled lines and the like includes at least one table in the document.

  A document analysis unit 201 analyzes each document held in the document storage unit 200 and first determines the direction (vertical or horizontal), and determines the table region and the relevant area from the positional relationship of the ruled lines in the document. In the function section that specifies the heading area in the area (in the case of a table with attributes arranged in the column direction, this usually corresponds to the first line of the table or several lines from the first line), the area of each cell, etc. is there.

  Reference numeral 202 denotes an attribute name extraction unit which extracts individual heading character strings, that is, names of individual attributes constituting the table (attribute names) from the heading area specified by the document analysis unit 201, and counts the number of appearances. It is a functional part to do. That is, the extracted attribute name, the document name from which the attribute name is extracted, and the number of appearances of the attribute name (initial value is 1) are associated and registered in the attribute name list. If an entry with the same attribute name already exists in the list, a new extraction source document name is added to the entry, and the number of appearances is increased by one. For example, if there are three tables having an attribute named “address” through all the documents in the document storage unit 200, the number of appearances finally associated with the attribute name “address” is three.

  Further, the attribute name extraction unit 202 includes an attribute name classification unit 202a that classifies the extracted attribute names into semantically identical or similar items.

  Some attribute names, such as “telephone number” and “TEL”, are semantically identical even if the notation is completely different. Conversely, even if the notation is the same, it may be used in a completely different meaning for each table. However, with the exception of the above, in general, attribute names that are semantically identical or similar are considered to have the same or similar notation.

Therefore, the attribute name classification unit 202a classifies the attribute names extracted by the attribute name extraction unit 202 into groups (categories) having the same or similar notation. For example, the attribute name extracted by the attribute name extraction unit 202 is as follows:-Request-Thickness-Other-None-Solder-Solder specifications-Solder diameter-Soldering-Gas pressure-Glass base-Glass center-Glass adhesion・ Lens ・ Lens type ・ Lens angle ・ Use ・ Use lens ・ Use device ・ Use model ・ If it seems to be a remark, these are: ・ Request ・ Thickness ・ Other ・ None ・ Solder Solder specifications Solder diameter Soldering ・ Gas pressure Glass base Glass center Glass bonding ・ Lens Lens type Lens angle ・ Used lens Used equipment Used model ・ Remarks

  Note that, as a grouping policy, it is considered that the character strings having a fixed length (for example, two characters) from the beginning of the character string are made into one group in the simplest manner. Instead of a fixed length, a variable length according to the length n of the character string (for example, n / 2, n * 2/5 + 1, etc.) may be used. As a more advanced method, a method such as sequential pattern mining is used to determine whether or not grouping is possible with a continuous character string not only from the beginning but also from the end, or a character string consisting of multiple characters selected from within the character string. to decide. Alternatively, the degree of similarity may be measured only by the ratio of the constituent character types regardless of the character string order. In addition to the above, the similarity between attribute names can be calculated using various methods that have been conventionally proposed as methods for calculating the similarity between character strings.

  For example, a character string common to all attribute names in the group is assigned to each classified group as a unique group name. Thus, for example, the attribute names “address”, “current address”, and “address / location” are integrated into the group “address”. The group to which each extracted attribute name belongs is registered in the attribute name list by the attribute name classification unit 202a.

  A value extraction unit 203 is a functional unit that extracts a specific value of each attribute from each cell specified by the document analysis unit 201 and counts the number of appearances for each attribute. That is, the extracted value, the document name and attribute name from which the value is extracted, and the number of occurrences of the value (the initial value is 1) are associated and registered in the value list. When an entry having the same value extracted from the same attribute name is already in the list, a new extraction source document name is added to the entry and the number of appearances is increased by one. For example, if there are three records including “Tokushima city” in the “address” attribute through all documents in the document storage unit 200, the number of occurrences of “Tokushima city” in the “address” attribute is three. At this time, for example, even if there are five records including “Tokushima city” in the “current address” attribute, this number is not added to the number of occurrences of “Tokushima city” in the “address” attribute.

  Reference numeral 204 denotes a document search unit, which is a functional unit that searches for documents in the document storage unit 200 that match a search condition input by a searcher. The document search unit 204 has a function of searching only a document that includes a specified character string in a specified attribute.

  An input / output unit 205 is a functional unit that receives input of various instructions and the like by a searcher from the keyboard 111 or the mouse 112, and generates various screens described in a flowchart described later and displays them on the display 110.

  Next, FIG. 3 is a flowchart showing a procedure for creating an attribute name list and a value list in the document search apparatus according to the embodiment of the present invention. It is assumed that the illustrated process is executed in advance before a document search process to be described later.

  First, the document analysis unit 201 sequentially analyzes the documents in the document storage unit 200 and specifies a table area in each document (step S301). Next, based on this analysis result, the attribute name extraction unit 202 extracts attribute names from each table in the document and counts the number of appearances (step S302). Further, the attribute name classification unit 202a classifies the attribute names extracted above into a plurality of groups (step S303).

  On the other hand, the value extraction unit 203 extracts a value from each table in the document based on the analysis result by the document analysis unit 201 and counts the number of appearances (step S304). Through these processes, the attribute name list and the value list described above are created and held by the attribute name extraction unit 202 and the value extraction unit 203, respectively.

  Next, FIG. 4 is a flowchart showing the procedure of the document search process in the document search apparatus according to the embodiment of the present invention.

  First, the input / output unit 205 displays a search screen as shown in FIG. 5 (step S401). As shown in the figure, the search screen has an attribute name input area 500 and a value input area 501. When the searcher knows exactly the attribute name and value to be specified, a desired character string is displayed in each of the above fields. What is necessary is just to input and press the search button 502. FIG. In other words, when the input / output unit 205 detects that the search button 502 is pressed (step S402: Yes), the document search unit 204 that has received an instruction from the input / output unit 205 specifies the attribute specified at that time. A document including a value is searched (step S403), and a list of search results is displayed by the input / output unit 205 (step S404).

  However, as described above, if the attribute names are slightly different in each table in the document storage unit 200 such as “address”, “current address”, and “address / location”, it is difficult to specify them without omission. Therefore, the searcher presses the attribute button 503 before executing the search, displays a list of attribute names appearing in the document in the document storage unit 200, and selects a desired attribute name from the list.

  In other words, when the input / output unit 205 detects that the attribute button 503 on the search screen in FIG. 5 is pressed (step S402: No, step S405: Yes), the attribute name extraction unit 202 holds the attribute name list. Based on this, a list of attribute names as shown in FIG. 6 is created and displayed (step S406).

  As shown in the figure, in the attribute name list, the attribute names extracted from all the documents in the document storage unit 200 are displayed in a list for each group (in the figure, the underlined character string means the above group name). The number of occurrences is displayed in parentheses next to each attribute name. Then, the currently selected attribute name is displayed in the search candidate area 600. By pressing the selection button 601 or the add button 602 in this state, the attribute name can be input to the attribute name input area 500 of the search screen in FIG. . That is, the input / output unit 205 that has detected the above operation (step S407: Yes) deletes the attribute name list of FIG. 6 (step S408) and displays the search screen of FIG. The attribute name selected in FIG. 6 is displayed (step S401).

  If an attribute name has already been input in the attribute name input area 500 of FIG. 5, when another attribute name is selected in FIG. 6 and the selection button 601 is pressed, only the selected attribute name is input. Displayed in area 500. On the other hand, when the add button 602 is pressed, the attribute name selected in FIG. 6 is added to the input attribute name.

  In addition, after waiting for the value button 504 to be pressed on the search screen of FIG. 5 and detecting that the value button 504 has been pressed (step S402: No, step S405: No, step S409: Yes), the input / output unit 205 sets the value. Based on the value list held by the extraction unit 203, a value list as shown in FIG. 7 is created and displayed (step S410).

  As shown in the figure, in the value list, the values of the attributes ("product" and "product name" in the example shown) that are input in the attribute name input area 500 when the value button 504 is pressed are displayed in a list. In addition, the number of occurrences of each value is displayed as a numerical value and a graph in parentheses. Then, the value being selected is displayed in the search candidate area 700. By pressing the selection button 701 or the add button 702 in this state, the value can be input to the value input area 501 of the search screen in FIG. That is, the input / output unit 205 that has detected the above operation (step S411: Yes) deletes the value list of FIG. 7 (step S412), displays the search screen of FIG. 5, and displays the search screen of FIG. The value selected in is displayed (step S401).

  If some value has already been input in the value input area 501 in FIG. 5, when another value is selected in FIG. 7 and the selection button 701 is pressed, only the selected value is displayed in the value input area 501. The On the other hand, when the add button 702 is pressed, the value selected in FIG. 7 is added to the input value.

  After inputting the attribute name and value by the procedure after step S405, the search button 502 is pressed in the same manner as when the character string is directly input (step S402: Yes), and the value specified for the specified attribute name is changed. The included document can be searched (steps S403 and S404).

  According to the embodiment described above, since a desired item can be selected from a list of attribute names and values actually used in each table in the search target document, what table is included in the search target document. Even a searcher who is not familiar with the structure of each table can perform an accurate search by avoiding a search leak caused by a subtle notation.

  In addition, attribute names are aggregated and displayed for each thing that is semantically identical or similar (strictly, those that are highly likely to be semantically identical or similar), so similar attribute names can be easily specified in a batch it can.

  The attribute name list shown in FIG. 6 is simply a list of groups in the order of the Japanese syllabary. For example, the number of times each attribute name is actually used for the search is stored as a history, A group containing a high attribute name or an attribute name that is frequently used in combination with an attribute name that has already been entered is displayed at the top, or a separate ranking screen is provided, and a frequently used attribute name or its belonging The groups may be displayed up to the top N. In the case of a table in which attributes are arranged not only in the column direction but also in the row direction, the attribute names in the column direction and the attribute names in the row direction may be distinguished (separated) and displayed.

  The value list shown in FIG. 7 is a list of values for a specific attribute name. For example, a button such as “all value list” is prepared on the search screen of FIG. You may be able to do it. Furthermore, in the above-described embodiment, the attribute names that completely include the specified value are listed. For example, the specified value is divided into semantic blocks by NLP (natural language processing), and any block is You may make it list up the attribute name contained in a value.

  As described above, according to the document search device, the document search method, and the document search program according to the present invention, a document search device and a document search that can search a character string by specifying a position in a document. In the method and the document search program, the position and the character string can be specified accurately and comprehensively by an easy operation by the searcher.

  The document search method described in this embodiment can be realized by executing a program prepared in advance on a computer such as a personal computer or a workstation. This program is recorded on a computer-readable recording medium such as the hard disk 105, the flexible disk 107, the CD-ROM, the CD-RW 109, the MO, and the DVD, and is executed by being read from the recording medium by the computer. The program may be a transmission medium that can be distributed via a network such as the Internet.

  As described above, the document search apparatus, the document search method, and the document search program according to the present invention are useful for searching a document having a table structure, and in particular, subtle notation of attribute names and values in the table. Suitable when there are a large number of tables with

It is explanatory drawing which shows an example of the hardware constitutions of the document search apparatus concerning embodiment of this invention. It is explanatory drawing which shows functionally the structure of the document search device concerning embodiment of this invention. It is a flowchart which shows the procedure of the production | generation process of an attribute name list and a value list in the document search apparatus concerning embodiment of this invention. It is a flowchart which shows the procedure of the document search process in the document search apparatus concerning embodiment of this invention. 6 is an explanatory diagram illustrating an example of a search screen created and displayed by an input / output unit 205. FIG. It is explanatory drawing which shows an example of the attribute name list screen produced and displayed by the input-output part. It is explanatory drawing which shows an example of the value list screen produced and displayed by the input-output part.

Explanation of symbols

100 bus 101 CPU
102 ROM
103 RAM
104 HDD
105 HD
106 FDD
107 FD
108 CD-RW drive 109 CD-RW
110 Display 111 Keyboard 112 Mouse 113 Network I / F
Reference Signs List 114 Communication cable 200 Document storage unit 201 Document analysis unit 202 Attribute name extraction unit 202a Attribute name classification unit 203 Value extraction unit 204 Document search unit 205 Input / output unit

Claims (7)

In a document search device that searches for an electronic document including a specified character string at a specified position,
Attribute name extraction means for extracting the name of each attribute in the table included in the electronic document;
Attribute name classification means for classifying the names extracted by the attribute name extraction means into groups of semantically identical or similar ones;
Attribute name display means for displaying the name classified by the attribute name classification means for each group;
A document search apparatus comprising:   2. The document search apparatus according to claim 1, wherein the attribute name classifying unit classifies the names extracted by the attribute name extracting unit into groups having the same or similar notation. Further, a value extracting means for extracting the value of each attribute whose name is extracted by the attribute name extracting means,
Value display means for displaying the value extracted by the value extraction means;
The document search apparatus according to claim 1 or 2, further comprising: In a document search method for searching for an electronic document including a specified character string at a specified position,
An attribute name extraction step of extracting the name of each attribute in the table included in the electronic document;
An attribute name classification step for classifying the names extracted in the attribute name extraction step into groups of semantically identical or similar ones;
An attribute name display step for displaying the name classified in the attribute name classification step for each group;
A document retrieval method comprising:   5. The document search method according to claim 4, wherein in the attribute name classification step, the names extracted in the attribute name extraction step are classified into groups having the same or similar notation. Furthermore, a value extraction step for extracting the value of each attribute whose name has been extracted in the attribute name extraction step,
A value display step for displaying the value extracted in the value extraction step;
The document search method according to claim 4, wherein the document search method includes: A document search program for causing a computer to execute the method according to any one of claims 4 to 6.

Images