JPH0567147A - Document retriever - Google Patents

Abstract

PURPOSE:To retrieve document data as it is without converting into a format proper to a retriever in advance by providing a data structure detecting means and a data structure generating means, and generating an index based on the result of the data structure generating means. CONSTITUTION:The data structure detecting means 13 detects the structure of data inputted based on information stored in a data structure storage means 14 according to an input document in advance. The data structure generating means 16 generates data structure information such as the start and end cut of a field record, etc., based on the result of the data structure detecting means 13, and synthesizes the data structure information with inputted data. An index generating means 18 generates the index based on data generated at the data structure generating means 16. Thereby, the retrieval can be performed without performing data formatting work for a document data base as in a conventional system.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document retrieving apparatus for performing full-text retrieval from an enormous computerized document database.

[0002]

2. Description of the Related Art In recent years, with the spread of word processors and personal computers and the practical use of character recognition by computers, the number of electronic documents created by these has increased. Therefore, interest in a document database for accumulating a large amount of document information and retrieving the document information as needed is increasing. In a conventional document database, when searching for a document, a keyword search using a keyword attached to each document is generally used. However, there are problems such as keyword addition work not keeping up with the increase of accumulated documents, keywords becoming obsolete over time, and it is not possible to support searches by keywords beyond the expectation of the database administrator, resulting in frequent omission of searches. was there.

From such a background, recently, a document database called a full-text database has attracted attention. In the full-text database, the search conditions given by the user and all the information of the stored documents are collated,
Output documents that meet the search conditions. In the search condition, a character string such as a sentence may be used in addition to a word such as a conventional keyword.

A conventional document retrieval device will be described below. FIG. 9 shows a conventional document retrieval device. Figure 9
In 901, 901 is a data input unit. A data storage unit 902 temporarily stores the data input by the data input unit 901. A search unit 903 and a search condition storage unit 904 are searched by the search unit 903 while being searched. Reference numeral 905 is a data output unit.

The operation of the document retrieval apparatus configured as described above will be described below. First of all, the data formatted for the data input section is transferred to the data input section 9
The data is input at 01 and stored in the data storage unit 902. For the stored data, the search unit 903 causes the search condition storage unit 9 to
The search is performed according to the search condition 04, and only the records that match the condition are output by the data output unit 905.

[0006]

However, in the above-mentioned conventional structure, since the data input section is simple, it is necessary to perform preprocessing such as conversion into a data format for the input section beforehand, and a newly acquired document database. There was a problem that it was not possible to search immediately from.

The present invention solves the above-mentioned problems of the prior art, and an object of the present invention is to provide an apparatus capable of retrieving a document by simply pre-registering format information unique to the document database and without converting the format of the document database.

[0008]

In order to achieve the above-mentioned object, the document retrieval apparatus of the present invention is, firstly, based on the information stored in the data input means for inputting document data and the data structure storage means. Data structure detecting means for detecting the structure of the data input by the input means, and data structure generating means for generating data structure information based on the result of the data structure detecting means and combining the data structure information with the input data. Index generating means for generating an index based on the data generated by the data structure generating means,
Search condition data storage means for storing search condition data,
And a data output means for outputting the search result, the search condition data storing means searching for the result of the data structure generating means based on the index created by the index creating means. There is.

Secondly, in the above structure, the data structure generating means generates search object information as to whether or not it becomes a search object together with the data structure information based on the result of the data structure detecting means, and inputs the data structure information and the search object information. It is configured to combine the data with the data.

[0010]

According to the present invention, in the above structure, firstly, the data structure detecting means detects the structure of the data inputted based on the information previously stored in the data structure storing means corresponding to the input document, and the data structure is generated. Means generates data structure information such as start and end phrase breaks of field records based on the result of the data structure detecting means, synthesizes the data structure information and the input data, and the index generating means is the data structure generating means. By creating an index based on the generated data, it is possible to search without formatting the document database as in the conventional case.

Secondly, in the above structure, the data structure generating means generates search object information as to whether or not it becomes a search object together with the data structure information based on the result of the data structure detecting means. By synthesizing the input data, high-speed search can be performed without performing formatting work.

[0012]

【Example】

(Embodiment 1) Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of a document search device according to a first embodiment of the present invention. In FIG. 1, 11 is a data input unit, and 12 is a data input unit 1.
1 is a first data storage unit that stores the data input, 13 is a data structure detection unit that detects the structure of the data stored in the first data storage unit 12, and 14 is a data structure detection unit 13 that detects the data structure. A data structure storage unit for storing a data structure, a data structure input unit 15 for inputting a data structure to the data structure storage unit 14, and a data structure detection unit 13 for reading data from the first data storage unit 12 17 is a data structure generation unit that adds structure information to the data line according to the detection result of
Is a second data storage unit that stores the data created in
Reference numeral 18 denotes an index generation unit that generates an index address for each data structure unit from the data generated by the data structure generation unit 16, and 19 denotes the index generation unit 18
An index storage unit that stores the index information generated in step 2, a search unit 20 that searches the second data storage unit 17, a search condition storage unit 21 that stores the conditions of the search unit 20, and a reference numeral 22 A data output unit that controls the output of document data based on the result of the search unit 20.

Reference numeral 30 is an input data line, 31 is a data line from the data input unit 11 to the first data storage unit 12, 32 is a data line read from the first data storage unit, and 33 is a data line. A data line from which the data structure detection unit 13 reads the data structure from the data structure storage unit 14,
Reference numeral 34 is a data line for the data structure input unit 15 to write the data structure to the data structure storage unit 14, 35 is a detection signal line for sending the result detected by the data structure detection unit 13 to the data structure generation unit 16, and 36 is Reference numeral 37 is a data line for sending the data added by the data structure generation unit 16 to the second data storage unit 17 and index generation unit 18, and 37 is a data line for writing the index information generated by the index generation unit 18 into the index storage unit 19. , 38 are data lines for reading the index information from the index storage unit 19, and 39 is from the search unit 20 to the second data storage unit 17.
Is a data line from which the search unit 20 reads the search condition from the search condition storage unit 21.
Reference numeral 41 is a data line for the data output unit 22 to read data from the second data storage unit 17, 42 is a match signal line for notifying the data output unit 22 whether the search unit 20 matches the condition, and 43 is a data line. It is a data line that outputs data from the output unit 22.

The operation of the document retrieval apparatus configured as above when a data string having the structure shown in FIG. 2 is input will be described. In FIG. 2, C (111) to C (LM
N) C as a character, and C (111) to C (L1
1), C (121) to C (L21) such that characters C are gathered in a unit of a field are fields, and fields and fields are gathered in a unit of a field are records. The number of characters in each field and the number of fields in each record may be different. Further, it is assumed that the character or character string for distinguishing the field and the record is included in the field and the record, respectively, and these characters or character strings are stored in the data structure storage unit 14 by the data structure input unit 15. To be done.

When such a data string is input to the data input unit 11, the data input unit 11 causes the first data storage unit 12 to operate.
Write data to. Simultaneously with writing, the data structure detection unit 13 detects a character or a character string for distinguishing a field from a record based on the information stored in the data structure storage unit 14. This character or string should be accepted as a regular expression. When the data structure detection unit 13 detects the data structure, the data structure generation unit 16 sequentially reads the data from the first data storage unit, adds the data structure information to the input data, and writes it to the second data storage unit 17, and at the same time. The index generation unit 18 generates an index address from the tag unit generated by the data structure generation unit 16 and writes the index address in the index storage unit 19.

FIG. 3 is an example in which data structure information is added, and shows the contents of the second data storage section and the index storage section. The result is that the tag portion of the second data storage unit adds the data structure information. The contents of the tag part are FS as shown in FIG.
Indicates the start of the field, and D indicates normal data. Further, FE indicates the end of the field, RS indicates the purchase of the record, RE indicates the end of the record, and K indicates a delimiter for separating the field from the field record. Furthermore, the contents of the index storage section are FS
The address of the letter with is attached.

The search unit 20 can read the index information from the index storage unit 19 and read the field to be searched from the second data storage unit based on the index information.

FIG. 5 shows two types of data, newspaper information and book information, as an example of input data, and an example of a data structure registered in the data structure memory for each data.

As described above, according to this embodiment, the structure of document data is registered in advance in the data structure storage unit by providing the data structure detection unit, the data structure storage unit, and the data structure generation unit, and the input is made. The data structure detection unit detects the data structure of the generated data, and the data structure generation unit adds the structure information to the signal lines other than the data line, thereby searching without formatting the document data. Is possible.

(Second Embodiment) A second embodiment of the present invention will be described below with reference to the drawings. FIG. 6 shows a document retrieval apparatus according to the second embodiment of the present invention, and FIG. 7 shows an example of the output of the data structure insertion unit of the document retrieval apparatus according to the second embodiment of the present invention.

In FIG. 6, 61 is a data input section, 62 is a first data storage section for storing the data input by the data input section 61, and 63 is the first data storage section 6
2 is a data structure detecting unit that detects the structure of the data stored in 2, 64 is a data structure storing unit that stores the data structure detected by the data structure detecting unit 63, and 65 is a data structure that is stored in the data structure storing unit 64. Reference numeral 66 denotes a data structure input unit for inputting, 66 is a data structure insertion unit for reading the data of the first data storage unit 62 and inserting structure information between the data according to the detection result of the data structure detection unit 63, 67
Is a second data storage unit that stores the data created by the data structure insertion unit 66, and 68 is an index generation unit that creates an index address for each data structure unit from the data created by the data structure insertion unit 66, 69 is an index storage unit that stores the index information generated by the index generation unit 68, 70 is a search unit that searches the second data storage unit 67, and 71 is a search condition storage that stores the conditions of the search unit 70. 72 is a data output unit that controls the output of the document data from the result of the search unit 70, and 73 is a data structure deletion unit that deletes the data representing the structure inserted by the data structure insertion unit.

Reference numeral 80 is an input data line, 81 is a data line from the data input unit 61 to the first data storage unit 62, 82 is a data line read from the first data storage unit, and 83 is a data line. The structure detection unit 63 is a data line for reading the data structure from the data structure storage unit 64, and
Reference numeral 4 is a data line for the data structure input unit 65 to write the data structure to the data structure storage unit 64, 85 is a signal line for sending the result detected by the data structure detection unit 63 to the data structure insertion unit 66, and 86 is a data line. A data line for sending the data inserted by the structure inserting unit 66 to the second data storage unit 67 and the index generation unit 68, and 87 for writing the index information generated by the index generation unit 68 in the index storage unit 69, Reference numeral 88 is a data line for reading index information from the index storage unit 69, 89 is a data line for reading the second data storage unit 67 from the search unit 70, and 90 is a search condition from the search condition storage unit 71 to the search unit 70. Reference numeral 91 is a data line for reading, and 91 is a data line for the data output unit 72 to read data from the second data storage unit 67. Yes, 92 is a match signal line that informs the data output unit 72 whether the search unit 70 matches the condition, 93 is a data line that outputs data from the data output unit 72, and 94 is a data structure deletion unit 73. It is a signal line in which the data inserted in step 3 is deleted.

First Embodiment 16 is different from the configuration of FIG.
In the data structure generation unit of No. 6, the data line is expanded and the data structure is added, whereas in the data structure insertion unit of No. 66, the data line is not expanded, and the structure information is inserted by inserting the special character code that can be distinguished. 63 is a point to delete the character inserted by the data structure insertion unit 66.

The operation of the document search apparatus configured as described above when the data shown in FIG. 2 is input will be described.
First, when the data string of FIG. 2 is input to the data input unit 61, the data input unit 61 writes the data in the first data storage unit 62. Simultaneously with writing, the data structure detection unit 63 detects a character or a character string for distinguishing a field from a record. This character or string should be accepted as a regular expression. When detected by the data structure detection unit 63, the data structure insertion unit 66 sequentially reads the data from the first data storage unit, inserts the data structure information into the input data, and writes it in the second data storage unit 67. An index address is generated by the index generation unit 68 from the data inserted at the same time and written in the index storage unit 69.

FIG. 7 shows an example in which data structure information is added, and shows the contents of the second data storage section and the index storage section. The content of the second data storage unit is the result of inserting the data structure information. In the inserted contents, the FS indicates the start of the field as shown in FIG. Further, the contents of the index storage section are provided with the addresses of characters with FS.

The search unit 70 can read the index information from the index storage unit 69 and the field to be searched from the second data storage unit.

As described above, according to the present embodiment, by providing the data structure detection unit, the data structure storage unit, and the data structure insertion unit, the structure of the document data is registered in advance in the structure data storage unit and input. The data structure detecting unit detects the data structure of the data, and the data structure inserting unit inserts the data structure into the data, so that the search can be performed without formatting the document data.

(Embodiment 3) A third embodiment of the present invention will be described below.

The configuration is the same as that of FIG. 1, but the difference from the first embodiment is that data collation / non-collation is generated in addition to the tag indicating the structure generated in the data structure generation unit 16. .. FIG. 8 shows an example of data generated by the data structure generation unit 16. The data section and the tag section in FIG. 8 are shown in FIG.
However, by adding 1-bit information in which the check mode ∘ indicates the check and the cross indicates the non-check, the search unit 20 reads the information and determines whether or not it is the search target.

As described above, by providing the mode indicating the collation / non-collation of the data generated in the data structure generation unit 16, it is possible to search from the data to be excluded from the collation target.

[0031]

As described above, according to the first aspect of the present invention, the document data is searched in advance by providing the data structure detecting means, the data structure generating means, and the index generating means for generating an index based on the result of the data structure generating means. It is possible to search from the document data as it is without converting to a device-specific format.

Further, as a second means, the structure generating means generates search target information as to whether or not it becomes a search target together with the data structure information based on the result of the data structure detecting means, and inputs the data structure information and the search target information. By synthesizing the generated data with each other, unnecessary data is excluded from the reference target, which enables higher-speed processing.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a document search device according to a first embodiment of the present invention.

FIG. 2 is a diagram of document data handled by the document search device according to the first embodiment.

FIG. 3 is a diagram showing contents of a second data storage unit and an index storage unit in the first embodiment.

FIG. 4 is a diagram showing the contents of a tag section generated by a data structure generation section in the first embodiment.

FIG. 5A is an exemplary diagram of newspaper information data handled by the document search device in the first embodiment. FIG. 5B is an exemplary diagram of book information data handled by the document search device in the first embodiment.

FIG. 6 is a configuration diagram of a document search device according to a second embodiment of the present invention.

FIG. 7 is a diagram showing the contents of a second data storage unit and an index storage unit in the second embodiment.

FIG. 8 is a diagram showing the contents of a second data storage unit and an index storage unit in the third embodiment.

FIG. 9 is a configuration diagram of a conventional document search device.

[Explanation of symbols]

 11 data input unit 12 first data storage unit 13 data structure detection unit 14 data structure storage unit 15 data structure input unit 16 data structure generation unit 17 second data storage unit 18 index generation unit 19 index storage unit 20 search unit 21 search condition 21 Storage unit 22 Data output unit 30 Data line 31 Data line 32 Data line 33 Data line 34 Data line 35 Detection signal line 36 Data line 37 Data line 38 Data line 39 Data line 40 Data line 41 Data line 42 Data line 42 Data line 43 Data line 61 data input unit 62 first data storage unit 63 data structure detection unit 64 data structure storage unit 65 data structure input unit 66 data structure insertion unit 67 second data storage unit 68 index generation unit 69 index storage unit 70 search unit 71 search condition 71 Memory 72 Data output Part 73 Data structure deletion part 80 Data line 81 Data line 82 Data line 83 Data line 84 Data line 85 Detection signal line 86 Data line 87 Data line 88 Data line 89 Data line 90 Data line 91 Data line 92 Data line 92 Data line 93 Data line 94 data line 901 data input unit 902 data storage unit 903 search unit 904 search condition storage unit 905 data output unit 910 data line 911 data line 912 data line 913 data line 914 match signal line 915 data line 916 data line

Front page continuation (72) Inventor Noboru Tamura 1006 Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. (72) Inventor ▲ Tsuru ▼ Ken Hayashi, Kadoma City, Osaka Prefecture

Claims (4)

[Claims] 1. Data input means for inputting document data,
Data structure detecting means for detecting the structure of the data input by the input means based on the information stored in the data structure storing means, and generating data structure information based on the result of the data structure detecting means and inputting the data structure information. Data structure generating means for synthesizing the generated data, index generating means for creating an index based on the data generated by the data structure generating means, search condition data storing means for storing search condition data, and the index generating means. A document retrieval device comprising: a retrieval unit for conducting a retrieval according to the retrieval condition data storage unit on the output of the data structure generation unit based on the created index; and a data output unit for outputting a retrieval result. 2. The document search device according to claim 1, wherein the data structure generating means inserts structure information of the data into the input data. 3. The document retrieval apparatus according to claim 1, wherein the data structure generating means adds the structure information of the data to the input data. 4. The document according to claim 1, wherein the data structure generating means generates search object information together with the data structure information based on the result of the data structure detecting means and synthesizes the data structure information, the search object information and the input data. Search device.