JP2003099461A - Twin engine system of information retrieval home page and information retrieving method using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently and flexibly retrieve a large quantity of document information stored in a server machine. SOLUTION: In a document retrieval system 1 to extract a document conforming to retrieval conditions among a plurality of documents as retrieving objects, it is provided with a first retrieval processing part 13 to retrieve the documents by a first retrieving means and a second retrieval processing part 15 to retrieve the documents by a second retrieving means having intended use different from the first retrieval processing means 13, a result which is retrieved and found by the first retrieval processing part 13 is stored and the documents are retrieved by succeeding the result by the second retrieval processing part 15.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a search using a computer, and more particularly to a twin engine method for an information search home page provided with two search processing units having different purposes in one information search system.

[0002]

2. Description of the Related Art In recent years, the spread of computer systems has been remarkable, and the amount of information handled has been increasing with the spread. In addition, since the number of opportunities to handle electronic documents has increased recently, whether or not a desired document can be quickly searched greatly affects the work efficiency.

As conventional document searching means, category searching means for searching by narrowing down categories, heading, publication date, ISBN (International Stan)
dard Book Number), a condition search means for performing a search by extracting a document that matches the specified search condition from the document data stored in the authors, etc. There is a full-text search means for searching for a document including "," and a document is searched by using one of these search programs.

[0004]

For example, when searching for a "book" in which "how to grow" of "written in 2001" and "guisuisu" is described from a large amount of documents by a condition search means, a user Can search by specifying conditions such as "2001", "book", "Uguisu". However, this method can extract the "book" of "Uguisu" written in "2001", but it is necessary to check for each book whether or not the content of "bringing up" is described in the extracted document. I can only do it. If a large number of documents are extracted as a search result, it takes a lot of time to search for a desired book.

When searching for a document similar to the above with the full-text search means, the user selects "2001",
You can search by specifying keywords such as "book", "Uguisu", and "raising". However, since the whole sentence is searched from all the document data, it takes a long time to search, and when the number of input keywords is small, a large amount of documents are extracted as a search result, and there is a problem that the user is confused.

The present invention has been made in order to solve the above-mentioned problems of the prior art, and its purpose is to efficiently and flexibly retrieve document information stored in large quantities in a server machine. To be able to do it.

[0007]

In order to achieve the above object, a twin engine method of an information search home page according to the present invention is a document search system for extracting a document matching a search condition from a plurality of documents to be searched. In the first search means, a first search processing unit for searching a document by the first search means,
The first search processing unit is provided with a second search processing unit for performing a search by a second search means having a different purpose, the result obtained by the search by the first search processing unit is stored, and the result is stored. It is characterized in that the second search processing unit subsequently carries out a search. The first search processing unit is an XML search processing unit.

Further, the first search means comprises a category search means for performing a search by narrowing down the categories and a condition search means for performing a search by extracting a document that matches an instructed search condition. The second searching means is characterized by being a full-text searching means for searching a document including a word or phrase for which search is instructed by referring to all the sentences of each document.

Further, in the information retrieval method having the category retrieval line, the condition retrieval line and the full-text retrieval line, first one of the retrieval lines is selected to obtain the retrieval result, and when the desired information result cannot be obtained. , The search is carried out by taking over the immediately preceding information result to another search line.

[0010]

DETAILED DESCRIPTION OF THE INVENTION 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 system adopting the twin engine method of the information search home page of the present invention. As shown in FIG. 1, the system according to the present embodiment stores information, an information storage / retrieval device 1 that retrieves information instructed by a user, a network control device 2 that relays a network, and a network. The communication line 3 connects the control devices 2 to each other and the user side terminal device 4.

The information storage / retrieval apparatus 1 has a control processing unit 11 and an XML (eXtensible Markup Lang) for document information.
uage) a structured and stored search book information storage base 12, and an XML information search processing unit 13 that searches documents stored in the search book information storage base 12,
A search full-text information storage base 14 in which full-text of document information is stored, a full-text search processing unit 15 that searches for documents stored in the search full-text information storage base 14, and browses as a result of the search. It is composed of a book information storage base 16, 17, and 18 in which document information necessary for that is stored.

Further, as shown in FIG. 2, the control processing unit 11
Is a search method selection unit 21 for selecting an arbitrary search method from search methods such as category search, condition search, and full-text search.
A search condition input unit 22 that inputs search conditions according to the selected search method, a data transfer unit 23 that transfers data when transferring search results, and a category that is selected when performing a category search A category selection unit 24 and a search keyword input unit 2 that receives a keyword input from the user-side terminal device 4 when performing a conditional search or a full-text search.
5, a search result display unit 26 for transmitting the searched results to the user side terminal device 4, and a document required for browsing the search results from the book information storage bases 16, 17, 18 Data transfer unit 2 for transferring to side terminal device 4
7 and.

Further, as shown in FIG. 3, the data transfer unit 2
3 is a search engine takeover unit 3 that takes over the search engine
1, a search engine information receiving unit 32 that receives the search information when taking over, a storage base takeover unit 33 that takes over the storage base, and a stored information receiving unit 34 that receives the stored information at the time of takeover. An output information takeover unit 35 that takes over the results, an input information receiving unit 36 that receives a search keyword or the like input from the user side terminal device 4 when performing a search, and an output information receiving unit that receives the results when taking over the results And a part 37.

Next, referring to the flow charts shown in FIGS. 4, 5 and 6, the operation of the present embodiment configured as described above will be described. In this embodiment, there are three lines, a category search line, a condition search line, and a full-text search line. When starting a search, first, a search method is selected (step 1). When performing a category search, go to step 2, and when performing a condition search, step 1 in FIG.
1 to proceed to step 22 of FIG. 6 when performing a full text search.

First, the category search line will be described. When performing a category search, first, a related category is selected from the categories of hierarchy 1 (step 2) and the search is performed (step 3). User side terminal device 4
Displays the search result in the first hierarchy (step 4). If the number of displayed results is small (NO in step 5) and there is no desired document in the search results (n <= 2: none in step 6), the process returns to step 2 and the category of hierarchy 2 is entered. Select the relevant category from
Repeat steps 3 to 6.

When the desired document is found in step 6 (YES in step 6), the details of the document are displayed (step 7), and the full text search line shown in FIG. 6 is taken over. On the contrary, if there are too many results displayed when the category search is performed (YES in step 5), or if the desired document is not obtained even if the category of hierarchy 3 is selected (n = 3 in step 6: no ), The user can select another search method (step 8). The results obtained by the category search are stored (steps 9 and 10), and are carried over to FIG. 5 in the case of the conditional search and to FIG. 6 in the case of the full-text search.

Next, the condition search line shown in FIG. 5 will be described. When the conditional search is specified in step 1 or when the conditional search is specified in step 8, the conditional search is performed. When performing a condition search, the user first inputs one or more search conditions (step 11). Here, if there is inherited data (YES in step 12), the inherited data is inserted (step 13), and if there is no inherited data (NO in step 12), only the search condition is used and the search is performed. Is executed (step 14). The search result is displayed on the user side terminal device 4 (step 15).

When the number of displayed results is small (NO in step 16) and there is a desired document in it (YES in step 17), details of the document are displayed (step 18), and It is inherited by the full-text search line shown in 6. On the contrary, if there are too many results displayed when the conditional search is performed (YES in step 16), or if the desired document cannot be obtained (NO in step 17),
The user can select another search method (step 19). When a different search condition is added and the condition search is performed again, the obtained search result is stored (step 2
0), and steps 11 to 17 are repeated. When performing a full-text search based on the obtained search results, the obtained search results are stored (step 2
1), it is succeeded to FIG.

Next, the full-text search line shown in FIG. 6 will be described. If full-text search is specified in step 1, or if full-text search is specified in steps 8 and 19, full-text search is performed. When performing a full-text search, first
The user inputs one or more search keywords (step 22). Here, if there is inherited data (step 2)
Insert the inherited data into (YES in step 3) (step 2
4) If there is no inherited data (NO in step 23), only the search keyword is used as it is to execute the search (step 25). The retrieved result is the user terminal device 4
Is displayed (step 26).

If the number of displayed results is large (YES in step 27), or if the number of displayed results is small (NO in step 27) and there is no book to be found (NO in step 28). The obtained search result is stored (step 33), and the processing is continued to step 22 to step 28. If the desired book is found (YES in step 28), the details of the document are displayed (step 29), and the user inputs the necessary search keyword (step 30), and the search in the desired book is performed. By being executed (step 31), the search keyword position is displayed on the screen (step 32).

Next, an example of actually using the system of the present embodiment to search for a "book" of "bringing up" of "guisuisu" written in "2001" will be briefly described. First, a category search is performed by selecting a natural science from the hierarchy 1 category "natural science, history, geography, travel ...". This is not enough to find the desired document, so Tier 2 "Biology, Mathematics, Physics ..."
Biology is further selected, and botany is further selected from Tier 3 "Botany and zoology" to repeatedly search (steps 2 to 6). It is assumed that 30 documents are found here. In order to narrow down the search results of these documents, conditional search is selected as another search method (step 8). 30 searched
The data for each document is stored here (step 9).

In the condition search, keywords such as "2001" and "book" are input (step 11). Since 30 documents have been searched by the category search, this result is taken over, and the document matching the input keyword is searched from these 30 documents (step 14). 4 here
Although a number of books have been found, if it is desired to further narrow down the search results, full-text search can be used as another search method (step 19). The contents of the four books, which are the search results at this time, are stored (step 20).

In the full-text search, a full-text search is performed for the "bringing up" of the "guisuisu" with respect to the contents of the four succeeded books.
(Steps 22 to 25). If two books are found as a result of searching here, these two books are "200
It is a "book" of "raising" of "Uguisu" written in 1st year. By inputting a necessary keyword for these two books and performing a full-text search, a desired page of a desired book can be browsed.

Although three search methods of category search, condition search, and full-text search are adopted as the search method of the present invention, it goes without saying that other search methods may be used and the present invention is not limited to the above embodiment.

[0025]

As described above, according to the twin engine method of the information search home page of the present invention, the first search means is used in the document search system for extracting the documents matching the search condition from the plurality of documents to be searched. The first search processing unit includes a first search processing unit that searches for a document and a second search processing unit that searches using a second search unit having a different purpose from the first search processing unit. Since the result obtained by the search is stored and the result is taken over and the search is performed by the second search processing unit, it is possible to efficiently and flexibly search a large amount of stored information.

Since the first search processing unit is the XML search processing unit and the document is structured, the user can search for a desired document by a simple operation.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a system adopting a twin engine method of an information search home page of the present invention.

FIG. 2 is a functional block diagram showing in detail an internal configuration of a control processing unit shown in FIG.

FIG. 3 is a functional block diagram showing in detail an internal configuration of a data transfer unit shown in FIG.

FIG. 4 is a flowchart showing the operation of a category search line.

FIG. 5 is a flowchart showing the operation of a condition search line.

FIG. 6 is a flowchart showing the operation of a full-text search line.

[Explanation of symbols]

1 Information storage and retrieval device 2 Network controller 3 communication lines 4 User side terminal device 11 Control processing unit 12 Search book information storage base 13 XML information search processing unit 14 Full-text information storage base for retrieval 15 Full-text search processing section 16, 17, 18 Library information storage base

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Yuichiro Iino             3-4-1 Yurakucho, Chiyoda-ku, Tokyo             Machine Industry Co., Ltd. (72) Kenichi Suganuma             3-4-1 Yurakucho, Chiyoda-ku, Tokyo             Machine Industry Co., Ltd. (72) Inventor Kenji Watanabe             5-58 Shimonagaya, Konan-ku, Yokohama-shi, Kanagawa Prefecture             Kashima Media Bind Co., Ltd. F term (reference) 5B075 KK02 KK07 ND02 NK02 QS01                       QS07 UU40

Claims (4)

[Claims] 1. A document search system for extracting a document that matches a search condition from a plurality of documents to be searched, a first search processing section for searching a document by a first search means, and a first search. A second search processing unit that performs a search by a second search unit having a different purpose from the processing unit is provided, the result obtained by the search by the first search processing unit is stored, and the result is succeeded to the second search processing unit. Twin engine method of information retrieval home page characterized by performing retrieval in the retrieval processing section of. 2. The twin engine method for the information search home page according to claim 1, wherein the first search processing section is an XML search processing section. 3. The twin engine method of the information search home page according to claim 1 or 2, wherein the first search means matches a category search means for performing a search by narrowing down categories and an instructed search condition. The second search means comprises a full-text search means for searching for a document including a search-indicated phrase by referring to all the sentences of each document. Twin engine method for information retrieval homepage featuring 4. An information search method having first and second search processing sections, wherein the first search processing section has a category search line and a condition search line, and the second search processing section has a full text. It has a search line, first selects one of the above search lines to obtain the search result, and when the desired information result is not obtained, the previous information result is taken over by another search line to perform the search. How to search for information.