JP2014134920A - Design document retrieval system, construction method of design document retrieval system and program therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically extract retrieval information from design documents so as to retrieve required design documents without missing, and to retrieve the design documents by using the retrieval information.SOLUTION: A design document retrieval system comprises: a character string extraction section for extracting a character string from a design document file; a specification item name collection section for collecting specification item names from a collection of design document files classified for each design document type, while using the extracted character string; a specification data collection section for collecting specification data from a collection of design document files classified for each design document case, while using the extracted character string; and a specification item name-specification data relation extraction section in which the design document files, the specification item names collected by the specification item name collection section and the specification data collected by the specification data collection section are made correspondent and stored in a specification database.

Description

  The present invention relates to a technology for managing and retrieving design-related documents created when designing a product or a plant.

  Design-related documents (hereinafter referred to as “design documents”) created prior to product manufacturing and plant construction often describe one target from multiple viewpoints. As an example, in the construction of a plant such as a power plant, various design documents with different viewpoints are made for one structure. For example, in a piping design drawing, a structural diagram showing the physical structure, a piping system diagram showing how the pipes are connected, a sensor arrangement drawing showing how the sensors are arranged on the piping, sensor data There is a data flow diagram showing how to transmit the information.

  Design documents are used for manufacturing and construction purposes. In addition, the design document may be referred to during maintenance and inspection. Furthermore, a new design document may be created based on an existing design document. In the work using these design documents, it is necessary to search all the plurality of design documents and browse them in relation to each other.

  Design documents often include drawings as well as text written in natural language. In addition, since it is created by various systems including a CAD system, a printed design document is often stored in a format captured by a scanner or the like instead of an electronic file itself containing created data. This is because it is possible to view any drawing as long as it is in a format captured by a scanner.

  For this reason, text search techniques such as full-text search cannot retrieve sufficient search results for design documents. Therefore, as disclosed in Patent Document 1, it is common to search information such as keywords in advance for each design document, and search using such information.

JP 2010-49562 A

"Algorithm Introduction" ISBN: 978-476903357 Chapter 15

  By adding search information such as keywords to each design document, it becomes possible to retrieve the necessary design document.

  However, the input of search information is left to the person who registers the design document. For this reason, when there is omission of registration of search information, or when search information that was not initially assumed is required, it is difficult to search all related design documents. At this time, it is difficult to apply a general document search program to a design document with a lot of drawing data.

  As a countermeasure, a method of automatically extracting search information such as keywords can be considered. However, in terms of term extraction from documents that have been studied for general documents, there are many automatic extractions based on statistical properties. The terms that should be keywords in the design document are not necessarily statistically characteristic terms.

  Therefore, from the viewpoint of keyword extraction for associating design documents, a method for extracting terms from the viewpoint of whether documents can be associated with each other is necessary. Therefore, there is a demand for a method of automatically extracting keywords passed between design documents and automatically extracting keywords by appropriately classifying them.

  An object of the present invention is to automatically extract search information from a design document so that a necessary design document can be searched without omission, and to perform a search using the search information.

In order to solve the above problems, for example, the configuration described in the claims is adopted.
The present application includes a plurality of means for solving the above problems. To give an example, a design document database for storing a design document file, and a specification for registering the relationship between the design document file, specification item name, and specification data. A database, a search query reception unit that receives an input of a search query, a design document search unit that searches the specification database based on a matter name or a specification item name input as a search query, and the design based on a search result The design document search system includes a design document file output unit that outputs a design document file from a document database.

  In the design document search system of the present invention, the extracted character string is further extracted from a character string extraction unit that extracts a character string from the design document file and the set of design document files classified for each type of design document. A specification item collection unit that collects the specification item name using the extracted character string from a set of the design document files classified for each item of the design document. A specification item name-specification data relationship extraction unit that associates the design document file with the specification item name collected by the specification item name collection unit and the specification data collected by the specification data collection unit and stores them in the specification database And may be provided.

  Further, the design document search system construction method of the present invention includes a step of extracting a character string from a design document file, a step of classifying the design document file according to a design document type and a design document item, and the design The specification item names are collected from the set of design document files classified for each type of document using the extracted character strings, and the extracted from the set of design document files classified for each item of the design document. Collecting specification data using character strings, associating the collected specification item names with the collected specification data, and specifying the relationship between the design document file, the specification item names, and the specification data And registering as a database.

  The program of the present invention is a program for causing a computer to construct a design document search system, which includes a step of extracting a character string from a design document file, and the design document file includes a type of design document and a design document. Collecting specification item names by evaluating the similarity between character strings using the extracted character strings from the step of classifying according to the project and a set of design document files classified for each type of design document Collecting the specification data by evaluating the similarity between the character strings using the extracted character string from the set of design document files classified for each case of the design document, and the collected A step of associating a specification item name with the collected specification data, and a relationship between the design document file, the specification item name and the specification data as a specification database Is a program for executing the steps of recording.

  According to the present invention, by automatically extracting the specification item name and the specification data from the design document, it becomes possible to add the search information that has not been given in advance as the search information. The burden of searching can be reduced.

The functional block diagram which shows an example of the search system of the design document of this invention. The figure which shows an example of the apparatus structure of the search system of the design document of this invention. The figure which shows the data structure of a design document file. The figure which shows the flow of a specification item name collection process. The figure which shows the total table of a specification item name collection process. The figure which shows the output table of a specification item name collection process. The figure which shows the flow of a specification data collection process. The figure which shows the aggregation table of a specification data collection process. The figure which shows the output table of a specification data collection process. An example of string mapping. The figure which shows an example of the table | surface of similarity evaluation of a character string. Result of matching example string. The figure which shows the flow of a specification item name-specification relation data extraction process. The figure which shows the aggregation table of specification item name-specification relation data. The figure which shows the table of a design document database. The figure which shows the table of a specification database. The figure which shows the relationship between a company asset management system and the search system of a design document. Formula 1 for obtaining the character string similarity.

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

  FIG. 1 shows a functional block diagram of the design document search system of the present invention. This system performs two processes: a process for extracting specification item names and specification data from a design document, and a process for retrieving a design document. The process of extracting the specification item name and the specification data from the design document includes a design document file input unit 101, a design document file reading unit 102, a character string extraction unit 103, a design document file classification unit 104, a character string similarity evaluation unit 105, The specification item name collection unit 106, the specification data collection unit 107, the specification item name-specification data relationship extraction unit 108, the design document database 109, and the specification database 110 are executed. The process for searching for a design document is executed by the search query receiving unit 111, the design document search unit 112, the design document file output unit 113, the design document database 109, and the specification database 110.

  In addition, as shown in FIG. 2, the present invention has a central processing unit 201 that has a central processing unit and processes information by a storage program method, a main storage device 202 composed of a random access memory, a processing target document and processing results. The system operates on an apparatus such as a computer, which includes an external storage device 203 that stores a dictionary, an input device 204 that is used to input a document, and an output device 205 that outputs information processing results such as a created dictionary. The external storage device 203 stores a database 2031 and a dictionary 2032. The input device 204 includes a CD-ROM reader 2041, a DVD reader 2042, a keyboard 2043, and the like. The output device 205 includes a CD-ROM writing device 2051, a DVD writing device 2052, a display 2053, and the like. In addition, the central processing unit 201 may be connected to another information processing apparatus 207 via the network 206.

  The design document search system can be realized by executing a predetermined program loaded on the main storage device by the central processing unit and executed by the central processing unit in the computer. This predetermined program is input from a storage medium storing the program via an input device or from a network and directly loaded onto the main storage device or temporarily stored in an external storage device. From the main storage device.

  The invention of the program according to the present invention is a program that is incorporated in a computer and operates as a design document search system. The design document search system shown in the block diagram of FIG. 1 is configured by incorporating the program of the present invention into a computer.

  The design document file input unit 101 receives an electronic file input from the outside of the system in the form of a storage medium such as a DVD or a CD-ROM by the input device 204 and stores it in the external storage device 203. On the external storage device 203, a design document database 109 and a specification database 110 are constructed. The design document file is an electronic file that is created by a CAD system, a word processor, or the like, or a design document printed on paper is captured by a scanning process. It is assumed that the character string contained therein is character-coded. The character code is not particularly limited. In the case of a scanned file, it is assumed that the character code is recognized from the character image by the optical character reading program.

  The design document file is given a unique symbol for identification. Hereinafter, a unique symbol for identification is called a file identifier. In the design document file, a file identifier, a design document type, and a design document item are described and can be read.

  The design document file reading unit 102 reads the design document stored in the design document database 109 built on the external storage device 203 into the main storage device 202.

  The character string extraction unit 103 extracts a character string from the design document file. It is assumed that the extracted character string is stored in the main storage device 202 in association with the file identifier, the design document type, the design document item, and the position in the file. The position in the file is represented by a page number and two-dimensional coordinates in the page. In addition, the character strings are appropriately divided according to the line feed code and the size of the spatial distance between the character strings.

  The design document file classification unit 104 classifies the design document file according to the design document type and the design document item. Using a set of design document files classified for each type, the specification item name collection unit 106 collects specification item names. Using the set of design document files classified for each case, the specification data collection unit 107 executes specification data aggregation processing. When data is sent from the design document file classification unit 104 to the specification item name collection unit 106 and the specification data collection unit 107, they are included in the file identifier, design document type, design document item, and design document in the format shown in FIG. Assume that string information is sent. The character string information includes the character string itself, the page on which the character string appears, and the X and Y coordinates of the page. FIG. 3 shows data of one design document file.

  Here, an example of the design document file is “Piping diagram” as the type of design document, “Plumbing work for power plant A” as the type of design document, and “Piping name” as the specification item. As specification data, “AS-1 (symbol representing piping)”...

  The specification item name collection unit 106 evaluates the similarity between character strings extracted from the same type of design document, for example, a piping diagram, and executes a process of collecting specification item names using the result. At this time, the character string similarity evaluation unit 105 is used to evaluate the similarity between character strings extracted from the design document.

  The flow of the specification item name collection process is shown in FIG. First, a predetermined number of design document files of the same type are read (S401). The predetermined number is not particularly limited. What is necessary is just to set suitably according to the processing performance of a processing apparatus. If the two sets of all design document files are compared, the process is completed (S402). If there are two sets not processed, two design document files are taken out (S403), and the two files are compared. Character strings are extracted from each, and the similarity of all character strings is compared using the character string similarity (S404).

  The comparison result is stored in the tabulation table shown in FIG. 5 (S405). Character strings evaluated as similar by the character string similarity evaluation unit 105 are aggregated into one record (one horizontal line), and the frequency is stored for each character string. The sum of the frequencies of all strings is stored in the total frequency part. There may be multiple file identifiers. In FIG. 5, for example, the character string 1 is “AAA” and the character string k is “AAa”. Character strings “BBB”, “CCC”, and the like are also tabulated in a similar record.

  When all files have been processed, the total frequency of the tabulation table is checked, and only records whose values are n% or more of the number of files are selected (S406). n can be set arbitrarily. A character string having the maximum appearance frequency is selected for the selected record, and the character string, for example, “AAA” is set as a standard character string (S407). A specification item name list is output in the format of FIG. 6 (S408).

  The specification data collection unit 107 evaluates the similarity between character strings extracted from the design document of the same item, and executes a process of collecting specification data using the result. At this time, the character string similarity evaluation unit 105 is used to evaluate the similarity between character strings extracted from the design document.

  The flow of the specification data collection process is shown in FIG. First, a predetermined number of design document files of the same item are read (S701). The predetermined number is not particularly limited. What is necessary is just to set suitably according to the processing performance of a processing apparatus. Further, the specification item name list created by the specification item name collection process is read (S702). If the two sets of all design document files are compared, the process is completed (S703). If there are two unprocessed sets, two design document files are taken out (S704), and the two files are compared. Character strings are extracted from each of them, and the similarity of all the character strings is compared using the character string similarity (S705).

  The comparison result is stored in the aggregation table shown in FIG. 8 (S707). Character strings evaluated as similar by the character string similarity evaluation unit 105 are aggregated into one record (one horizontal line), and the frequency is stored for each character string. There may be multiple file identifiers. When all files have been processed, the total frequency of the aggregation table is checked, and only records whose values are n% or more of the number of files are selected (S708). n can be set arbitrarily. The selected record is checked, and if there is the same character string as the specification item name in the specification item name list, the record is excluded (S709). Further, a character string having the maximum appearance frequency is selected for the selected record, and the character string is set as a standard character string (S710). Finally, the specification data list is output in the format of FIG. 9 (S711).

  The character string similarity calculation in the character string similarity evaluation unit 105 is performed using DP matching. Since this calculation method is disclosed in many books such as Non-Patent Document 1, it will not be described in detail here.

  Here, the calculation method will be briefly described using the correspondence of the character strings shown in FIG. 10 as an example. Match two strings for simplicity. In the case of collating three or more character strings, all the combinations need only be calculated, and will not be described here.

  The cost is a numerical value indicating how different two character strings are. The calculation is performed using the number of operations necessary to transform one character string into the other character string. As operations, insertion, deletion, and replacement of characters are considered. A cost is assigned to each operation, and the costs are totaled for the necessary operations. Here, it is assumed that -2 points are given when a character is inserted, deleted, or replaced, and 2 points are given when they match.

  In DP matching, as shown in FIG. 11, the character strings to be compared are associated with columns and rows, the scores are managed in a two-dimensional table, and the scores are calculated in turn in the table. In the example of FIG. 11, “document creation method using document parts” is associated with a row, and “document creation method using document reuse” is associated with a column. Suppose that the position of a cell in the table is expressed using rows and columns. The square in the nth row and the mth column is represented by (n, m). Note that both rows and columns start from 1. The score S (n, m) of the cell (n, m) is calculated by Equation 1 in FIG. At this time, the cell used to calculate the score is stored.

  For example, the value of S (12,13) is such that the twelfth character of the row is “saku” and the thirteenth character of the column is “saku”. Therefore, since the second term is 10, the third term is 10, and the third term is 10, the largest one term is selected and becomes 14. At this time, it is stored that the score of the square (12, 13) is calculated from the value of the square (11, 12). However, when the score becomes 0, all of the stored contents are deleted.

  The score of each cell in the score table represents the similarity of the character string corresponding to the cell traced until the score of the cell is calculated. When this score is calculated, the correspondence between the character strings can be obtained by following the reverse order of storing the used squares. By following the correspondences in order from the cell with the highest score, it is possible to obtain the correspondences in descending order of similarity. At this time, it is possible to prevent the portion including the same character string from being extracted many times by preventing the cell once traced from being traced twice. In FIG. 11, the value of (15, 16) is 20 and is the maximum. Follow the squares you followed to get this value. The result of associating the character strings is shown in FIG.

  In this embodiment, whether or not the character strings are similar is determined using Equation 1, but it is also possible to determine using a dictionary. By using a dictionary, it is possible to determine that synonyms having completely different notations such as “collation” and “matching” are the same.

  The specification item name-specification data relationship extraction unit 108 associates the specification item name collected in the specification item name collection process with the specification data collected in the specification data collection process in each design document.

  The flow of processing is shown in FIG. The design document file, specification item name list, and specification data list are read (S1301, S1302, and S1303). The character string of the design document file is checked, and a character string including the character string included in the specification item name list as a partial character string is selected (S1304). It is checked whether or not the length of the selected character string is longer than the character string of the specification item name list included as a partial character string (S1306). n is an arbitrary number, but may be about 5. If it is long, the character string including the specification item name is subjected to dependency analysis, and the destination of the character string matching the specification item name is extracted. If the character string of the destination is in the specification data list (S1307), it is registered in the tabulation table of FIG. 14 (S1308). If not included, the spatial proximity relationship when printing is checked (S1309). The spatial proximity relationship at the time of printing can be realized by calculating the Euclidean distance using the XY coordinates shown in FIG. 3 and checking whether the value is equal to or less than the threshold value. If a character string having a close spatial relationship is in the specification data list (S1310), it is registered in the tabulation table of FIG. 14 (S1308). If all the specification item names have been processed (S1305), the specification item name-specification relationship data is output (S1311), and the process ends.

  The design document database 109 stores all design document files. This database can be constructed on the external storage device 203 using a program such as a relational database, an XML database, or a file server. In this embodiment, it is assumed that the table is constructed on the relational database. The table is constructed as shown in FIG.

  In the specification database 110, the relationship between the specification item name, the specification data, and the design document file that are collected from the design document and related to each design document file is registered. This database can be constructed on the external storage device 203 using a program such as a relational database or an XML database. In this embodiment, it is assumed that the table is constructed on the relational database. It builds with the table shown in FIG.

  The search query receiving unit 111 receives an input of a query for searching for a design document. A query is input using an input device 204 such as a keyboard. The search assumed in the present invention specifies a project name, for example, “Plumbing construction of power plant A” and a specification item name, for example, “piping name”, and a design document file including the specification item name in the designated item It is considered that these design document files are arranged for each file having a common specification item name and specification data and presented to the user. Therefore, it is suitable for the user to input a query by selecting an item to be searched from the item list and specification item name list displayed in the menu format. However, there is no problem even if it is a method of searching by inputting an arbitrary character string using the keyboard.

  The design document search unit 112 searches the file identifier, specification item name, and specification data stored in the specification database 110 for the combination of the case name, specification item name, and specification data input as the search query. Such search processing is easy to implement using a general relational database or XML database, and will not be described in detail here.

  The design document file output unit 113 reads the design document file from the design document database 109 using the stored file identifier, and outputs it to the output device 205 such as a display. At this time, it is possible to easily display the related design documents by outputting the design documents having the same specification data in close proximity.

  The retrieved design document can be used in cooperation with an enterprise asset management (EAM) system. Device data such as a plant is registered in the corporate asset management system. By associating this data with the equipment described in the design document, it is possible to support efficient equipment maintenance.

  FIG. 17 shows an example of a related configuration of the enterprise asset management (EAM) system 301 and the design document search system 302 of the present invention. The corporate asset management system 301 includes data 305 such as operation data of the device A, maintenance history, failure history, and the like. The design document search system 302 includes a design document file 306 such as a layout drawing of the device A in the setting document database 303. When planning maintenance work of device A based on the operation data 305 of device A, a “document A layout diagram” 306 that is a design document including a description of device A is retrieved using the design document retrieval system 302. Then, the layout diagram 306 of the device A is displayed. As a result, it is possible to provide support for efficient device maintenance.

  As described above, it has been explained that search information is automatically extracted from a design document by the design document search system shown in FIG. 1 and can be efficiently searched using this information.

101 Design Document File Input Unit 102 Design Document File Reading Unit 103 Character String Extraction Unit 104 Design Document File Classification Unit 105 Character String Similarity Evaluation Unit 106 Specification Item Name Collection Unit 107 Specification Data Collection Unit 108 Specification Item Name-Specification Data Relationship Extraction Unit 109 Design document database 110 Specification database 111 Search query reception unit 112 Design document search unit 113 Design document file output unit 201 Central processing unit 202 Main storage unit 203 External storage unit 204 Input unit 205 Output unit 206 Network 207 Information processing unit 301 Companies Asset management system 302 Design document search system 303 Design document database 305 Enterprise asset management system data 306 Design document search system design document file

Claims (15)

A design document database for storing design document files;
A specification database for registering the relationship between the design document file, specification item names, and specification data;
A search query receiving unit that receives input of a search query;
A design document search unit that searches the specification database based on a case name or a specification item name input as a search query;
A design document search system comprising: a design document file output unit for outputting a design document file from the design document database based on a search result. In the design document search system according to claim 1,
The design document file output unit is a design document search system for outputting a design document file having common specification data in close proximity. In the design document search system according to claim 1,
The design document search system includes a file identifier for identification, a type of design document, and a design document item in the design document file. In the design document search system according to claim 1,
A design document search system in which the specification database includes a file identifier for identification, specification items, and specification data. The design document search system according to claim 1, further comprising:
A character string extraction unit for extracting a character string from the design document file;
A specification item name collection unit that collects the specification item name using the extracted character string from the set of design document files classified for each type of design document;
A specification data collection unit that collects the specification data using the extracted character string from the set of design document files classified for each case of the design document;
A specification item name-specification data relationship extraction unit that associates the design document file with the specification item name collected by the specification item name collection unit and the specification data collected by the specification data collection unit and stores the specification data in the specification database. A design document search system. The design document search system according to claim 5, further comprising:
A character string similarity evaluation unit for evaluating the similarity between character strings,
The specification item name collection unit collects the specification item name using the evaluation of the similarity of the character string by the character string similarity evaluation unit,
The specification data collection unit is a design document search system that collects the specification data by using the character string similarity evaluation performed by the character string similarity evaluation unit. The design document search system according to claim 6,
The character string similarity evaluation unit is a design document search system that evaluates the similarity of character strings using a DP matching method or a dictionary. In the design document search system according to claim 5,
The character string data extracted from the design document file includes a file identifier for identification, a design document type, a design document item, and character string information.
The character string information includes a character string itself, a page on which the character string appears, a design document search system including the X coordinate and Y coordinate of the page. In the design document search system according to claim 5,
The specification item name-specification data relationship extraction unit is a design document search system that associates the specification item name with the specification data based on a destination of a character string that matches the specification item name. In the design document search system according to claim 5,
The specification item name-specification data relationship extraction unit is a design document search system that associates the specification item name with the specification data based on a spatial proximity relationship when printed. Extracting a character string from the design document file;
Classifying the design document file according to the type of design document and the project of the design document;
Collecting a specification item name from the set of design document files classified for each type of design document using the extracted character string;
Collecting specification data using the extracted character string from a set of design document files classified for each case of the design document;
Associating the collected specification item names with the collected specification data;
A method for constructing a design document search system, comprising: registering a relationship between the design document file, the specification item name, and the specification data as a specification database. In the construction method of the design document search system according to claim 11,
The step of collecting the specification item names evaluates the similarity between character strings, collects the specification item names,
The step of collecting the specification data is a method for constructing a design document search system for evaluating the similarity between character strings and collecting the specification data. In the construction method of the design document search system according to claim 11,
The step of associating the specification item name with the specification data is a design document search for associating the specification item name with the specification data based on a destination of a character string that matches the specification item name. How to build a system. In the construction method of the design document search system according to claim 11,
The step of associating the specification item name with the specification data includes a design document search system that associates the specification item name with the specification data based on a spatial proximity relationship when printed. Construction method. A program for causing a computer to construct a design document search system,
Extracting a character string from the design document file;
Classifying the design document file according to the type of design document and the project of the design document;
From the set of design document files classified for each type of design document, using the extracted character string, evaluating the similarity between the character strings and collecting specification item names;
From the set of design document files classified for each case of the design document, using the extracted character string, evaluating the similarity between the character strings and collecting specification data;
Associating the collected specification item names with the collected specification data;
A program for executing a step of registering a relationship between the design document file, the specification item name, and the specification data as a specification database.

Images