JPH06301725A - Retrieval device for character-string of hierarchized document - Google Patents

Abstract

PURPOSE:To constitute the device so that not only each document retrieval result bus also a retrieval result of each stage can be obtained by a single retrieval by displaying stepwise or simultaneously the degree of coincidence of every document and the degree of coincidence of every character-string set of each stage. CONSTITUTION:A client machine 4 executes an access to a server 1, and by a retrieving part 11 of the server 1, the retrieval is executed to all documents. In this case, by generating self-correlation information and only collating it with a map, the retrieval can be executed at a high speed. A result of retrieval executed by the retrieving part 11 is sent as it is to a sum-up processing part 12. By this sum-up processing part 12, from the result of retrieval from the retrieving part 11, at every document, the degree of coincidence to a character-string set of each stage is summed up and the processing for setting the highest degree of coincidence to the degree of coincidence of its stage, and also, setting the degree of coincidence of the uppermost stage of each document to the degree of coincidence of its document is executed. Its result is sent to the client machine 4, and the document having the degree of coincidence exceeding a set threshold are displayed in the lump on its display.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character string search device for a hierarchical document, and in particular, a character string set in each stage is located on the lower side of the sequence.
In a storage device in which one or two or more layered documents each composed of one or two or more character string sets are stored in one document or two or more documents, each document and each column in each document are stored. The present invention relates to a character string search device for a hierarchical document for performing a fuzzy search including an incomplete match of a target character string with respect to a character string set.

[0002]

2. Description of the Related Art Conventionally, when checking whether or not a particular character string is included in each document in a storage device in which a large number of documents are stored, it is usual that all the character strings are included. Although an exact match search was performed to check only whether or not there was a case where the search could not be performed due to a subtle difference in the description, especially when searching for a long foreign word written in katakana. Further, there is a problem that the search becomes significantly slow when the total amount of documents to be searched is large, such as the number of documents to be searched.

Therefore, in Japanese Patent Application Laid-Open No. 4-326164 filed by the same applicant as the present applicant, the autocorrelation information of each character (code) is stored for each document at the same time when the document is stored, and the document is searched at the time of retrieval. By obtaining the autocorrelation information of each character in the character string and detecting the presence or absence of it, not only the presence or absence of the search character string in each search target document but also the degree of matching can be easily and quickly A search system that can be searched is disclosed.

The above system speeds up the search for a specific character string in each document. For example, one document is very large, and "title", "preface", "text 1", "text 2", "postscript". In the case of a hierarchical document divided into items such as, if it is possible to know which item has a desired character string, subsequent processing may be facilitated.
In addition, when there is no item with a high degree of matching, it may be important in determining the end of the search as to which item has a character string that matches the specific character string.

[0005]

SUMMARY OF THE INVENTION The present invention has been made in view of the problems of the prior art as described above, and its main purpose is simply to determine whether or not there is a specific character string in a document. In addition to the above, you can easily and quickly search for which item in each document to be searched has a character string that matches a specific character string. To provide a device.

[0006]

According to the present invention, the above-mentioned object is composed of a character string set which is hierarchized in one stage or two or more stages, and the character string set of each stage is on the lower side thereof. For each document and each character string set in each document in a storage device in which one or two or more layered documents composed of one or more character string sets are stored, A character string search device for a hierarchical document for performing a fuzzy search that also includes an incomplete match of a specific character string, wherein the specific character string is searched for all the characters of each document, and the degree of coincidence is searched. For each document, the search unit to determine, the highest degree of coincidence is calculated as the highest degree of coincidence by aggregating the degree of coincidence with respect to the character string set of each stage from the bottom side,
The document further includes a totalization processing unit that sets the degree of coincidence at the top of each document as the degree of coincidence of the document, and gradually calculates the degree of coincidence for each document and the degree of coincidence for each character string set at each stage of the document. The present invention is achieved by providing a character string search device for a hierarchical document, which is displayed simultaneously or simultaneously.

[0007]

In this way, for example, fuzzy search is performed for all the characters to be searched in a plurality of hierarchical documents divided by items, and the results are totaled from the character string set at the bottom, By further obtaining the search result of, and the search result of each document, the search result of each stage can be obtained together with the search result of each document by one search.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will now be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the system configuration of a server / client type workstation to which the present invention is applied. This system is a mass storage device 2
And a plurality of client machines 4 connected to the server 1 via a known network 3.

A large number of documents are stored in the storage device 2. Here, as shown in FIG. 2, each document is composed of a set of character strings hierarchically arranged in one stage or two or more stages, and the character string set in each stage is one or more in the lower stage. It consists of a layered document composed of a set of character strings.
In this embodiment, it is assumed that a large number of documents “application documents” are stored. This "application document" consists of "application", "specification" and "abstract", and "application" is
It consists of "document name", "reference number", "inventor", "patent applicant", "agent", etc. For example, "patent applicant" means "identification number", "zip code", "address". Or "place of residence", "name or name", "representative", etc. Further, the "specification" means "document name", "title of invention", "claims", "detailed description of invention",
"Detailed description of the invention" includes, for example, "industrial application field", "prior art", "problem to be solved by the invention", and "solve the problem". "Means for carrying out", "action", "example", "effect of the invention". Further, in the document stored in the storage device 2, autocorrelation information of each character (code) is simultaneously created as a map at the time of storage and stored as a kind of index together with the document and managed by the server 1.

As shown in FIG. 3, the server 1 searches for a specific character string from the above-mentioned autocorrelation information, and a search unit 11 for determining the degree of matching, and a search result from the search unit 11, For each document, the matching scores for the character string sets in each column are added up from the bottom, and the highest matching score is set as the matching score for that document, and the matching score at the top of each document is the matching score for that document. And a totalization processing unit 12 that is provided.

The outline of the operating procedure of this embodiment will be described below. A specific character string from a client machine 4,
For example, when performing a search for the character string “feedforward” on all the documents stored in the storage device 2, the character string is input as a “search key” from the client machine 4 and the threshold of the degree of coincidence described later is, for example, 70%. Set as above. Then, this client machine 4 accesses the server 1,
The search unit 11 of the server 1 searches all documents. At this time, since the autocorrelation information of each document is created and stored in advance as a map as described above, only the autocorrelation information of the character string "feedforward" is also created and collated with the above map for high speed. You can search. The speed of this search hardly depends on the capacity of the entire document, but depends on the length of the character string to be searched.

The search result obtained by the search unit 11 is sent to the totalization processing unit 12 as it is. From the search result from the search unit 11 in the totalization processing unit 12, the degree of coincidence with respect to each character string set for each document is aggregated for each document, and the highest degree of coincidence is taken as the degree of coincidence for that stage. In addition, processing is performed in which the top-level matching degree of each document is set as the matching degree of the document.
Then, as shown in FIG. 4, the result is sent to the client machine 4, and the documents having the degree of coincidence equal to or higher than the above-mentioned set threshold value are collectively displayed on the display. Then, the operator clicks the "sort" key in FIG. 4 with a pointing device such as a mouse to rearrange and display the images in descending order of the degree of coincidence. Then, the operator selects one of the displayed documents, for example, "levitation support device". Then, as shown in FIG. 5A, the “application”,
The degree of coincidence for each of the “specification” and the “summary” is sent from the server 1 to the client machine 4 and displayed. Next, for example, when “specification” is selected, as shown in FIG. 5B, “document name”, “invention title”, “claims”, “detailed description of invention” and “drawing” are shown. The degree of coincidence for each of the "brief explanations" is sent from the server 1 to the client machine 4 and displayed. Further, for example, when “Detailed description of the invention” is selected, as shown in FIG. 5C, “industrial application field”, “conventional technology”, “problem to be solved by the invention”, “problem to be solved” The degree of coincidence for each of "means for solving", "action", "embodiment" and "effect of the invention" is sent from the server 1 to the client machine 4 and displayed. In this way, the operator can gradually narrow down the part that includes the character string "feed forward" that he / she wants to search. For example, in the "means for solving problems" and "action" parts, the character string "feed "Forward" is included, but the character string "Feedforward" is included in the "Example" portion, and the character string "Feed Howard" is included in the "Effect of invention". It is effective when you do not do it and want to correct this.

In the present embodiment, the degree of coincidence for each document and the degree of coincidence for each character string set at each stage of the document are displayed stepwise, but if they can be displayed, they may be displayed simultaneously. Needless to say.

On the other hand, by inputting a plurality of specific character strings (search keys) to be searched on the screen as shown in FIG. 4, AND, O
It is also possible to easily perform a composite search under the conditions of R and ANDNOT. For example, when searching for the character string "differential", the character string "feedforward", and the character string "control" under the AND condition, there is only the character string "differential" in the "conventional technique" part, and When there is only the character string "feedforward" in the part of "" and only the character string "control" in the part of "issue to be solved by the invention", the character string "differential"
And the string "feedforward" and the string "control"
And are searched separately, and the matching degrees are added together and divided by the number of search character strings (3 in this case) (33% in this case) to be the search result under the AND condition. Also,
In the set "Detailed description of the invention" in the upper row, the result of dividing the matching degrees of the search character strings in the character string set in the bottom row by the number of search character strings (in this case, 3) (in this case, Is 100%) as the search result under the AND condition. That is, A
In the case of a composite search under the conditions of ND, OR, and AND, only the degree of coincidence of the search character string in the character string set at the bottom is required, and only one map of autocorrelation information as an index is required. . Where AND, O
It is needless to say that the method of outputting the composite search result under the conditions of R and ANDNOT is not limited to the above, and there are various methods depending on the application.
When searching with D, OR, ANDNOT conditions, and when all conditions are entered at once, AND, OR, ANDNOT
The degree of coincidence may be the same or different when the search is performed under the condition of.

[0016]

As is apparent from the above description, according to the character string search device for a hierarchical document according to the present invention, a specific character string is searched for for all the characters of a hierarchical document, and the matching degree is searched. Judgment For each document, the degree of coincidence for the character string set of each row is aggregated from the bottom, the highest degree of coincidence is taken as the degree of coincidence, and the degree of coincidence at the top of each document is the degree of coincidence of that document. Then, by displaying the degree of coincidence of each document and the degree of coincidence of each character string set of each stage of the document stepwise or simultaneously, the retrieval result of each stage together with the retrieval result of each document can be obtained by one retrieval. Thus, it is possible to easily and quickly search which item in each document to be searched has a character string that matches a specific character string.

[Brief description of drawings]

FIG. 1 is a block diagram showing a system configuration of a server / client type workstation to which the present invention is applied.

FIG. 2 is an explanatory diagram showing a structure of a hierarchical document stored in a storage device.

FIG. 3 is a block diagram showing a part of a functional configuration of a server and a client machine in a server / client type workstation to which the present invention is applied.

FIG. 4 is an explanatory diagram showing a display state of a display screen of a client machine.

5 (a) to (c) are explanatory views showing only the main part of FIG.

[Explanation of symbols]

 1 Server 2 Storage Device 3 Network 4 Client Device 11 Search Unit 12 Aggregation Processing Unit

Claims (1)

[Claims] 1. A character string set hierarchically arranged in one stage or two or more stages, and the character string set in each stage is composed of one or two or more character string sets on the lower stage side. A fuzzy search including an incomplete match of a specific character string is performed for each document and each character string set in each document in a storage device in which one or more hierarchical documents are stored. A character string search device for a hierarchical document for searching the specific character string for all the characters of each document, and a search unit for determining the degree of coincidence, and for each document, the lowest stage From the side, the degree of coincidence with respect to the character string set of each stage is totaled, the highest degree of coincidence is set to the degree of coincidence of that stage, and the highest degree of coincidence of each document is set to the degree of coincidence of the document. And the degree of agreement for each document and each of the documents. A character string search device for a hierarchical document, wherein the degree of coincidence for each set of character strings in columns is displayed stepwise or simultaneously.