JP2937519B2 - Document search device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

 [Object of the invention]

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document search apparatus for performing a full text search for extracting a document including an input keyword.

[0002]

2. Description of the Related Art In recent years, documents have been digitized and a large amount of document data has been distributed. When extracting what the user needs from this large amount of document data, a search method of performing a search by inputting a keyword composed of a character string has become mainstream. The search by this keyword is mainly
The method is roughly classified.

(1) A method in which a keyword is previously assigned to all documents, and a document to which a keyword input by a user is assigned is extracted.

(2) A method using a full-text search to determine whether a document contains a keyword input by a user.

In the above method (1), by limiting the number of keywords assigned to a document, a character string matching process with a keyword input by a user is minimized, and a high-speed search is realized. I have. However, a keyword must be assigned to a document in advance, and if a keyword that has not been assigned is specified by a user, it is impossible to extract the document. Moreover,
The task of assigning a keyword to a document is a burden on the document creator, and the method of determining the keyword may be left to the document creator, making it difficult to ensure keyword uniformity.

In the method (2), a document containing a character string input by a user is extracted, so that there is little omission in extraction.

In the above methods (1) and (2), when a document that satisfies the condition is found, the document containing the character string input by the user is listed and displayed, and the search result is displayed to the user. give. The user is required to select necessary documents from the documents output in a large amount by scrolling the display screen to determine in order whether each document is suitable for the purpose. At this time, the user is not shown how the keyword appears in the document.

[0008]

As described above, in a conventional full-text search apparatus, documents including a keyword input by a user are listed and displayed in the order of file storage in a search result obtained from the search apparatus. However, there is a problem that the user must read all the documents listed and displayed.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and has as its object to provide a document search apparatus which outputs a search result by estimating the importance of the appearance position of a character string including a keyword in a document. I do.

[Structure of the Invention]

[0011]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention provides a document search apparatus for extracting a document including an input keyword in a text, an input means for inputting a keyword, and a title for the document. A document dividing unit that divides the document into a preamble sentence, a text, etc., a character string matching unit that determines whether or not the keyword is included in the document, and a character string matching unit that determines whether the keyword is included in the document. Priority calculating means for calculating the priority of the search document based on the appearance position of the document and the document division information by the document dividing means; and search result output means for outputting the documents in the order of the priority obtained by the priority calculating means. It is characterized by having.

[0012]

Since the present invention is constructed as described above, the priority of the retrieved document is determined based on the position (for example, title, preamble, or text) where the keyword input from the input means appears in the document. By calculating the degrees and outputting the documents as search results in the order of the calculated priorities,
An efficient document search is performed.

[0013]

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

FIG. 1 is a block diagram showing the configuration of a document search apparatus according to one embodiment of the present invention.

In FIG. 1, reference numeral 1 denotes an external storage device which is composed of, for example, a floppy disk device or a hard disk device, and stores already created document data.
The document data read from the external storage device 1 is stored in a document data memory 2 composed of, for example, a dynamic RAM. One piece of document data includes a text data portion including only text information in a document and a non-text data portion including image data, format information, and the like.

Reference numeral 3 denotes an input unit for inputting a search keyword, a command, and the like, and is constituted by, for example, a keyboard, a mouse, and a device for controlling them.
Under the control of the control unit 4, for example, a VRAM,
The bit information stored in the VRAM is displayed on a display unit 5 comprising a CRT display for displaying as dot rows. The display unit 5 further displays search results, document data stored in the document data memory 2, and the like.

The control unit 4 comprises a RAM which stores a system program and is used as a buffer memory and a CPU which executes a control operation. The control unit 4 is connected to each of the above-mentioned devices or each of the devices described later by a bus, and controls each device. And controls and processes such as data transfer between devices. The control unit 4 includes buffers and counters necessary for control and processing.For example, the number of document data stored in the external storage device 1 is stored in a document number storage buffer (not shown). .

6 represents a text data portion of the document data,
As shown in FIG. 2, the document dividing unit divides the document into a title part 7, a preamble sentence part 8, and a body part 9, and from the document data stored in the document data memory 2 to the line feed code and the punctuation point in one sentence unit. Call as Document division
6 is the title part from the first sentence of one text document to the preceeding sentence of which the end of the sentence is a punctuation mark7, "Introduction" or "Introduction"
The preamble of a sentence that includes a phrase indicating that the content of the text document starts to be specifically described, such as the preamble sentence 8 in the abstract part, and the final sentence of the text document from the sentence following the preamble sentence Is determined as the body part 9 and the document is divided. This document division information is stored in the control unit 4
Is output to

A matching unit 10 calls one sentence from the document data memory 2 in which one text document data is stored, and determines whether or not the sentence includes the keyword input from the input unit 3. On the other hand, reference numeral 11 denotes a partial value table. The partial value table 11 stores calculated values for ranking documents required by the user.
The calculation value is set according to which part of the document, that is, which part of the document division part, the keyword input from is present. For example, as shown in FIG. 3, "10" is set for the title part 7, "5" is set for the preamble part 8, and "2" is set for the body part. If the keyword is included in the text document, the matching unit 10 separates the value corresponding to the position where the keyword appears based on the document division information output from the document division unit 6 to the control unit 4 into parts. It is called from the value table 11 and added to the calculated value storage buffer 12. FIG. 4 shows an example of the contents of the calculated value storage buffer 12 when the addition to the calculated value storage buffer 12 for a plurality of text document data is performed from the first sentence to the last sentence of the text document data. The calculated value stored in the calculated value storage buffer 12 indicates the output priority of the search result when outputting the search result.

Reference numeral 13 denotes a search result output unit.
Refers to the contents of the calculated value storage buffer 12 storing the calculated values for each document, and displays the documents on the display unit 5 in the order of the calculated values, that is, in order of priority, as shown in FIG. 5, for example.

Next, a specific processing operation of the document retrieval apparatus having the above configuration will be described with reference to a flowchart showing the processing flow of FIG.

First, buffers and counters in the control unit 4
Then, the calculation value storage buffer 12 is initialized, and subsequently, a plurality of keywords composed of character strings for search are input from the input unit 3 by the user. (Step S1,
Step S2).

When the keyword input is completed, the data of one text document is read into the document data memory 2 from the external storage device 1 in which a plurality of document data are stored. 1
When the text document is read, the ICHI flag, which is the document division information, is reset, and a text counter N (not shown) in the control unit 4 is set to "1". (Step S
3, step S4).

Subsequently, from the document data memory 2, the first sentence delimited by a line feed code or a period, for example, the first sentence 14 shown in FIG. 2 (hereinafter referred to as a specific text document, see the text document shown in FIG. 2) Control) 4
Are read into the document dividing unit 6 and the matching unit 10 via (Step S5). When the first sentence 14 is read,
Since the ICHI flag has been reset in step S4, the process proceeds to step S8 through steps S6 and S7 in which it is checked whether the ICHI flag is the text part 9 and the preamble sentence part 8.

In step S8, since the first sentence 14 has been read, the document division unit 6 determines that the sentence 14 is the title portion 7.
Based on the document division information, the matching unit 10
The value “10” corresponding to the title portion 7 is called from the partial value table 11, and the ICHI flag is set to “10”. Subsequently, in step S9, it is checked in step S9 whether the read sentence has a terminating period,
Since the first sentence 14 has no punctuation, the process proceeds to step S10.
Move to

In step S10, the matching unit 10 checks whether or not the read sentence includes the keyword input from the input unit 3 by a matching method. here,
Assuming that the keyword is a character string "document search device", the first sentence 14 includes a character string "document search device", so that the process proceeds to step S11.

In step S11, the matching unit 10
Is added to the calculated value storage buffer 12. That is, the value of the ICHI flag “10” is added to file [1], that is, the area of “document 1”. After addition, step S12
Is checked to see if there is one more sentence to read. In the example of FIG. 2, since there is one sentence to be read, the process returns to step S5.

In step S5, the second sentence 15 is read and the same processing as described above is performed. However, since this sentence 15 does not include the keyword "search", step S11 is skipped (ie, Since no keyword is included in the second sentence 15, addition to the “document 1” area is not performed), and the process proceeds from step 10 to step 12, and further proceeds from step S 12 to step S 5 again. Return.

Returning to step S5, the third sentence 16 is read and the same processing as described above is performed. However, since the third sentence 16 has an end-of-sentence point, the process is executed in step S9.
Then, control goes to a step S13.

In step S13, since the third sentence 16 that has been read is a sentence following the title portion 7 and there is a period at the end of the sentence, the document dividing section 6 determines that it is the preamble sentence portion 8. Based on the document division information, the value “5” corresponding to the preamble sentence portion 8 is called from the partial value table 11 via the matching unit 10, and the ICHI flag is set to “5”. Value "5" corresponding to the preamble sentence part 8 to the ICHI flag
Is completed, the process moves to step S10.

In step S10, since the third sentence 16 that is read includes the character string "document search device", the process proceeds to step S11. Step S11
In, addition to the calculated value storage buffer 12 is performed, but since the ICHI flag is set to "5",
“5” is added to the area of “document 1”. By this addition, the content of the calculated value storage buffer 12 becomes “15”. The calculated value storage buffer 12 stores calculated values for all documents for each document, and the search result output priority is determined based on the calculated values. This step S1
When the processing of 1 is completed, the processing similarly returns from step S12 to step S5 in order to read one sentence of the following preamble sentence part 8.

When one sentence of the preamble sentence portion 8 after the third sentence 16 is read, the processing after step S7 is partially different from the above processing. That is, in step S7, ICHI
Since the flag is set to the value “5” corresponding to the preamble sentence part 8, the processing shifts to step 14 instead of step S8.

At step S14, it is checked whether or not the read sentence contains a character string "Introduction" indicating the body part 9. If the read sentence does not include the character string “Introduction”, the process proceeds to step S10, and the same process as above is repeated.

If the read sentence is the first sentence 17 of the body part 9, the process proceeds to step S15. In step S15, since the word “Introduction” is included, the document dividing unit 6 determines that the document is the text part 9.
Based on the document division information, the value “2” corresponding to the text part 9 is called from the partial value table 11 via the matching unit 10, and the ICHI flag is set to “2”. When the setting of the value “2” corresponding to the text part 9 to the ICHI flag is completed, the process proceeds to step S10.

The processing in steps S11 and S12 following step S10 is the same as the above-described processing. However, when the body part 9 includes a keyword, "2" set in the ICHI flag is replaced with "2". Calculation value storage buffer 12
To the “document 1” area. In the example sentence shown in FIG. 2, since the keyword is not included in the body part 9, the addition of "2" is not performed.

When one sentence of the body part 9 is read, the ICHI flag is set to a value "2" corresponding to the body part 9 in step S15.
Therefore, the processing skips the processing steps in the middle from step S6 and moves to step S10.

When the above operation is repeated up to the last sentence of the text portion 9, there is no more sentence to be read, and the search for "document 1" ends. At this time, the calculated value storage buffer 12
, A calculated value as shown in FIG. 4 is stored in the area of “document 1”.

When there is no more sentence to be read, the process returns from step S16 to step S3 to read the next document, for example, "document 2".

Returning to step S3, the text document of "document 2" is read from the external storage device 1 into the document data memory 2 in the same manner as described above, and further, in step S4, the ICHI flag as document division information is reset. At the same time, the text counter N in the control section 4 is set to "2". After step S5, the same processing as the above processing is repeated.

As described above, when the search processing for all text documents stored in the external storage device 1 is completed, that is, when the value of the text number storage buffer in the control unit 4 matches the value of the text counter N, the processing is started. Is Step S1
The process moves from step 6 to step S17.

In step S17, the search result output unit 13 is activated, and the contents of the calculated value storage buffer 12 are referred to. Assuming that, for example, the contents shown in FIG. 4 are stored in the calculated value storage buffer 12, the search result output unit 13 displays the text documents in the order shown in FIG. Displayed at 5 and prioritized output.

As described above, an efficient document search is performed by preferentially outputting a document having a high calculated value in the calculated value storage buffer 12 as being close to the document desired by the user.

In the above embodiment, the matching unit 10
It is determined whether a keyword is included in the read sentence and added to the calculated value storage buffer 12 based on the document division information and the position where the keyword appears, but the matching unit 10 includes the keyword. The same effect can be achieved even if the configuration is divided into a character string matching unit that determines whether or not the keyword is included and a priority calculation unit that adds the calculated value to the calculated value storage buffer 12 based on the document division information and the position where the keyword appears. Is done.

Further, in the above embodiment, the document is designated by the title part 7,
The document is divided into the foreword part 8 and the body part 9, but the invention is not limited to this.For example, the document is divided into a postscript part and the like, or divided into the title part 7 and the body part 9. Addition / deletion of the number of divisions may be arbitrarily set.

Further, in the above embodiment, the document is designated by the title part 7,
Although the document is divided in the order of the preamble sentence portion 8 and the body portion 9 in this order, the present invention is not limited to this, and it is needless to say that the document division order can be changed arbitrarily according to the technical field to which the document belongs.

Further, the present invention is not limited to the above-described embodiment, and it is needless to say that various modifications can be made without departing from the gist of the present invention.

[0047]

As described in detail above, according to the document search apparatus of the present invention, the importance of the keyword in the document is estimated in accordance with the position of the keyword in the document, and the output of the search result is performed. By setting priorities, efficient document search can be performed, and as a result, the user's labor required for searching for a target document in a document database can be significantly reduced. It is enormous.

[Brief description of the drawings]

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

FIG. 2 is a diagram illustrating an example of division of a title, a preamble, and a body of one document.

FIG. 3 is a diagram showing an example of the contents of a partial value table for storing values for prioritizing search results;

FIG. 4 is a diagram illustrating an example of the contents of a calculated value storage buffer that stores a calculated value for each document.

FIG. 5 is a diagram showing an output order of search results.

FIG. 6 is a flowchart showing an outline of a processing flow.

[Explanation of symbols]

3 ... input unit (input means) 6 ... document division unit (document division means) 10 ... matching unit (character string matching means, priority calculation means) 13 ... search result output unit (search result output means)

Continuation of the front page (56) References JP-A-58-44536 (JP, A) JP-A-2-224069 (JP, A) JP-A-3-294963 (JP, A) JP-A-2-44465 (JP, A) , A) JP-A-58-50071 (JP, A) Kageura Gorge et al., “Full-text search system considering the logical structure of documents”, Bulletin of the National Center for Science Information Systems, Vol. 49-58 (September 30, 1990) (58) Field surveyed (Int. Cl. ⁶ , DB name) G06F 17/30

Claims (1)

(57) [Claims] An input unit for inputting a keyword; a document dividing unit for dividing a document into a title, an introductory sentence, a text, and the like; A character string matching unit that determines whether the keyword is included in the document, and a search based on the appearance position of the keyword determined by the character string matching unit in the document and the document division information by the document division unit A document search device comprising: a priority calculation unit that calculates a priority of a document; and a search result output unit that outputs documents in the order of the priority obtained by the priority calculation unit.