JPH11353313A - Information retrieving method, information automatic classifying method, and information analyzing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To retrieve and analyze information by making good use of a data base system and to display the information out on a screen in a variety of styles that a user desires by classifying the retrieved information according to classification fields and storing the classified information in the data base of the user. SOLUTION: A classifying program 41 executed by a central processor 40 classifies information read sequentially out of a disk loaded in a CD-ROM drive 50. Namely, the classifying program 41 confirms whether or not there is a specific classification field in each pieces of information read out of the disk and selects at least one classification field. Once classification fields are selected, words and a document having regularity are detected according to them and classification is carried out on the basis of them. Then items of a solution subject and a solving means are extracted as a classification field and patent information data having contents reconstituted are stored in a hard disk drive 30.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information retrieval method, an information automatic classification method, and an information analysis method. More specifically, a user database system is connected to an information source to retrieve desired information. The searched information is automatically classified and stored in a database system, and based on various parameters such as search conditions, analysis conditions, and sampling conditions input and set by a user using the database system. The present invention also relates to an information search method, an information automatic classification method, and an information analysis method capable of searching and analyzing information.

[0002]

2. Description of the Related Art Generally, when constructing an information database, a user manually searches for and analyzes desired information from various information sources and exchanges the information with a database held by the user. Thereafter, the processed information is stored in the database, and is output by a calculation in the database according to a search condition set by a user.

For example, when constructing a patent information database, a user accesses an external information database on a computer communication or an information source such as a CD-ROM in which patent information is recorded to search and output desired information. After that, analysis was performed, and the data was manually processed and stored one by one so as to conform to the user database.
When searching, classifying, and outputting desired information using the database constructed as described above, information is displayed on a screen in a desired form and output using keywords or appropriate search conditions. I was

[0004]

However, in such a conventional method of searching and analyzing information using a database, a user can directly search, classify, further analyze, and analyze a large amount of information. Since the information is manually input into the database, there is an inconvenience that the information is extremely troublesome.

[0005] In addition, since the output conditions for the user to search and output desired information using the database are limited, manual operations such as editing must be performed in order to output the desired information. Required and extremely cumbersome.
In addition, when performing a search operation again, the previous search result cannot be used, so that the search must be performed from the beginning.

[0006] The present invention has been made in view of such conventional problems, and has as its object to provide an information classification method capable of automatically classifying information recorded in an information source and storing the information in a user database. And It is another object of the present invention to search and analyze information using a database system for storing the classified information, and to display and output the information in various forms desired by a user based on the searched and analyzed information. Is to provide.

[0007]

In order to achieve the above object, in an information automatic classification method according to the present invention, search target information recorded in an information source is selected, and a classification field for the information is set. Searching for the search target information and determining whether or not the classification field is present in the information; automatically classifying the searched information based on the classification field; and And storing the information in the user's database.

[0008] In the information search method according to the present invention, a search condition for searching for predetermined information data from an information source is input and set, and the search condition based on the input and set search condition is set. Outputting numerical data related to the search result of the information data to the screen, and forming the content of the information data for each search result corresponding to the numerical data based on the character content already specified from the information content belonging to one or more fields. And outputting the contents as the contents of the plurality of fields.

Further, in the information analysis method according to the present invention, a step of inputting and setting a search condition for searching for information data of a recording medium, and a search result of the information data based on the input search condition Outputting the numerical data on the screen to the screen, outputting the statistical numerical value on the analysis result of the search information data based on the input and set analysis conditions in a graphic form on the screen, and outputting the statistical data on the screen. Analyzing and outputting the contents of the information data corresponding to the graphic from the information contents belonging to one or more fields to a plurality of field contents formed based on the already specified character criterion. .

[0010]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a flowchart illustrating an information automatic classification method according to the present invention. As shown in the figure, first, the user selects a patent technical field to be searched (S1). Next, the user determines the type of the information source used for the search (S2). That is, it is determined whether the information source is an external patent information database accessible via the Internet or a personal computer communication, a CD-ROM storing patent information, or an optical disk (S3). Here, if the information source is not a recording medium such as a CD-ROM, an optical disk, and a magnetic tape, but is an external database that can be accessed only through computer communication, or cannot be directly exchanged with a user database. If it is the information source, the desired information recorded in the information source is copied or downloaded and stored in a separate recording medium connected to the user's database (S4).

Next, it is determined whether or not the classification field to be classified by the user exists in the information source (S5). If not, the user examines the information and checks his / her own information. A classification field is entered in the corresponding position of the information so that it can be exchanged with the database (S6). Here, the classification field means a field in which the user attempts to classify the data in the information source. For example, when the user attempts to classify the patent information by field, the object of the invention and the embodiment , Effects, claims,
Abstract fields, drawings, International Patent Classification (IPC) numbers, application numbers, publication numbers, patent numbers, and the like are classification fields. When the preparation for classifying information by field is completed in this way, the information is automatically classified into predetermined classification fields according to a predetermined information classification program programmed to be exchanged with a user database (S7).

Next, it is determined whether or not the classified information data is reliable (S8), and is stored in a database (S9). At this time, the information classification program can be arbitrarily programmed by a user. For example, if the information in the information source includes both image information and text information, the image information and the text information can be programmed so that they can be classified and stored separately. If the information source includes only text information, the corresponding image information can be copied or downloaded from an external information source and stored so that it can be exchanged with the database.

As described above, in the network configuration of the integrated database system according to the embodiment of the present invention utilizing the classified and stored information, as shown in FIG. (For example, using a web browser), a network interface 10 for exchanging data with PC1, PC2,..., And an RO storing system initialization and OS driving programs
M20, a memory bank 21 for loading a program, a central processing unit (CPU) 40 for executing a program loaded in the memory bank 21 to perform classification, storage, and retrieval of information data. It includes a hard disk drive (HDD) 30 storing information data and storing a program for classification / search, and a CD-ROM driver 50 into which a CD-ROM storing patent information is inserted and mounted. It is configured.

Hereinafter, a process of searching and analyzing information using the integrated database system according to the present invention will be described. First, when an information source, for example, a CD-ROM is inserted into the CD-ROM driver 50 to drive the system, the search engine 42 program is loaded from the hard disk device 30 to the memory bank 21.

Next, the classification program 41 executed by the central processing unit 40 performs a classification operation on information sequentially read from the disk inserted into the CD-ROM driver 50. That is, the classification program 41 checks whether or not there is a specific classification field for each piece of information read from the disc, selects at least one or more classification fields. Based on this, words and sentences having regularity are detected and classified based on the detected words and sentences.

For example, in the case of Japanese patent information data, the field of the object of the invention is a TO text and a B
Since it is structured in a predetermined form like a Y sentence, it is used to classify the object field of the invention as a TO sentence as a solution problem and a BY sentence as a solution means. In the case of U.S. patent information data, in the entire text of the abstract,
Classify the first sentence as a solution problem, and classify other sentences as a solution.

In the case of Korean patent information data, for a material having an abstract portion, the abstract portion is classified as a problem to be solved.
Classify the claims as a solution, but if the material has no summary, classify the first sentence of the claim as a solution, and classify the other sentences as a solution and analyze It can be used as basic data for

As described above, the items of the solutions and the solutions are extracted and classified as the classification fields, and the patent information data whose contents are reconstructed as shown in FIG. Since the classification program 41 performs a process of extracting and classifying a corresponding sentence based on the classification field described above with respect to all information data read from the CD-ROM disk, all the patent document data recorded on the disk is shown in FIG. As shown in (1), the data is reprocessed and stored separately in a corresponding storage area of the hard disk device 40, thereby constructing a database for the integrated information management system.

Hereinafter, with respect to the information search method according to the present invention,
A description will be given based on FIGS. As described above, a patent document using a database of an integrated information management system constructed to have a classification item based on a classification field extracted by a reference word or a sentence structure selected from the contents of one or more items When searching for (S4
1) First, the user executes a web browser (WEBROWSER) installed in the search terminal PC to execute the search server 10 in which the integrated information management database is constructed.
0 (S42), an initial screen as shown in FIG. 5 is displayed. For this purpose, the search server 100 uses the search engine 42 and the DB, which are programs for searching and analyzing the materials of the search terminal PC. The management program 43 is loaded from the hard disk device 30 to the memory bank 21 and executed.

Here, when a connection is made from the search terminal PC, the search engine 42, which is an execution program in the search server 100, executes a corresponding search through the network interface 10 for the JABA APPLET constituting the integrated information management system. Download to terminal. Next, when the connection to the search server 100 is terminated and a search becomes possible, the user inputs a user name and a password on an initial screen as shown in FIG. When the search server 100 determines that the search is performed, the search server 100 provides an initial screen for explaining the outline of the system as shown in FIG. 6, and when the user requests a patent search, the patent as shown in FIG. A search menu screen is provided.

Next, the user inputs and sets the search condition of the material data to be searched (S43), and searches for the patent information data to be searched, the technology classification code (IPC code or separately set). Code) and search terms (keywords) (for example, search target: US patent, technology classification code: G1)
1B *, search term: video, etc.).

Next, the search engine 42 program loaded from the hard disk device 30 into the memory bank 21 and driven by the central processing unit 40 makes an inquiry about the condition value set by the user, and has already read the inquiry. The DB management program 43 searches the hard disk device 30 for information data corresponding to the input and set query to be transmitted to the DB management program 43 loaded into the memory bank 21.

Next, the DB management program 43
The hard disk device 30 retrieves the patent information data corresponding to the received inquiry from the hard disk device 30 (S44). That is, the DB management program 43 searches the huge amount of patent information data stored in the hard disk device 30 for the input data. Then, it is confirmed whether or not the information data satisfying all the set search conditions (when the search condition is “AND” search in FIG. 7) exists in the hard disk device 30.

Next, as a result of the search, patent data which satisfies all of the search conditions set and entered by the user, that is, the term "video" is included.
When the US patent information data of 11B * is detected from the hard disk device 30 (S45), the DB management program 43 determines the number of search results, that is, the total number of searched information data through the network interface 10. Therefore, the number of patent information data that matches the condition that the user wants to search is output and displayed as numerical data on the screen of the search terminal PC (S46).

Next, after visually confirming the numerical data based on the patent search result output on the screen, the user stores the search result in a file form (S4).
7), the contents of the “search contents” window in which the conditions set by the user on the patent search menu screen as shown in FIG. 7 are converted and displayed in a search formula are, for example, (search target = US patent document) And (search word = “video”) and (search field = G
11B *) is moved to the “search type” window of the file storage window, and when the user selects “store list” after inputting the file name and the description of the information content related thereto, The search engine 42 program which has received the transmission of the selected contents sets the hard disk drive 30 as the received file name using the search formula from which the search result was obtained and the index data capable of indexing the information data obtained therefrom.
(S48), and the storage area is a directory specified by the user who has logged in through the initial screen of FIG. Therefore, when the same user logs in to search for a patent in the future, desired information can be obtained by simply inputting a file name for information data relating to the same search condition.

For example, when the search result information stored in the file format as described above is selected, the search formula in the selected file and the index data of the search information are stored in the memory bank 2.
1, when the user adds a new search condition to the loaded and displayed search condition, for example, when "digital" is added as a search word,
The search engine 42 program includes a new search formula (search object = US patent material) in which “digital” is added to the previous search formula and (search word = “video & digital”).
And (search field = G11B *) to request the DB management program 43 to search, so that information data of a search result can be obtained more quickly.

The search result information data obtained as described above is newly stored in the hard disk device 30 by designating a new file name as described above, so that it can be used in the subsequent search. Hereinafter, an information analysis method according to the present invention, which is performed using the materials obtained by the above-described search conditions, will be described with reference to FIGS.

First, when the user selects a patent analysis for the patent information data searched as described above in order to check whether there is a technical trend or a desired material (S51),
As shown in FIG. 9, a menu screen for patent analysis in which the corresponding technical field codes to which the information retrieved from the retrieval server 100 belongs is displayed in a tree structure.

Next, the user inputs and sets the analysis condition of the information data to be analyzed, that is, selectively selects the value from the input elements such as the technology classification code and the analysis period of the patent information data to be analyzed. For example, when “G11B” is input as the technology classification code and “96-97” is input as the analysis period, the analysis conditions set by inputting from the retrieved patent information data are set. In other words, the analysis process is performed only on the information data that satisfies the following condition, that is, only the patent information data including the G11B technology code as the technology classification code from 1996 to 1997 out of the G11A and G11B searched in FIG. 9 (S53). ).

Next, the search engine 42 program executed by the central processing unit 40 includes only information data which satisfies the analysis conditions set by inputting and analyzing the information data in the searched patent information data. , And re-creates a tree structure of an international patent classification code (IPC code) to which the selected information data belongs.
To the corresponding search terminal PC via the PC to provide the user with the international patent classification code for the information data based on the analysis condition updated in the form of a tree map (S5).
6) It is possible to access desired information step by step.

Next, the user selects G11B and 001 from among the international patent classification (IPC) codes output on the screen in the form of the tree map, and further selects 004 from among the subcodes 002, 003 and 004. When selected, the search engine 42 program has the selected 004 as a subcode from the searched patent information data, that is, selects only the patent information data belonging to the technical field of G001B001 / 004.
As shown in (1), the data on the selected patent information data is output in such a manner that the item of the solution and the solution and the patent number are sequentially output to the screen on the search terminal PC as classification fields included in the selected patent information data. Since the information is provided, the user can accurately confirm that the desired information data is obtained by looking at the contents of the solution problems and the solution means items presented by the information from the patent information data belonging to a specific technical field.

The analysis result can be constructed in a tree map form, for example, a pie map form as shown in FIG. 11, and in FIGS. 12 and 1.
The trend map (trend ma) shown in FIG.
The graphic data can be provided in a p) form and a bar map form as shown in FIGS. 14 and 15.

Also, when the analysis result of the search engine 42 program is configured in a graphic form, each area of the graphic form, for example, FIG.
A, b, c, d of 1 wherein the ratio is
It is possible to express the ratio of the technical fields belonging to the lower stage of the upper technical code by setting the upper technical code of the technical field code currently selected as the statistical value for the information data of the analysis result to 100%, or , Regardless of the currently selected level of the technology tree,
It is also possible to express the ratio of the technical fields corresponding to the same step on the tree of the currently selected technical field by setting the total number of retrieved total result data to 100%, for example,
From among the retrieved materials in which the international patent classification code has a tree map form as shown in FIG.
After selecting 005, which is the subcode of 11B, and requesting the analysis result in a pie map format as shown in FIG. 14, the search engine 42 program converts the analysis result to the technical field of G11B, which is the upper code of 005. The search material having the code is assumed to be 100% and transmitted to the form as shown in FIG. 16 and output on the screen, or a lower code of 005 out of all the technical classification codes of the searched material. G11A-X
It is also possible to make the data including XX up to 100% as a whole and transmit it to the form shown in FIG. 17 and output it to the screen.

Next, when the user selects a specific section or area from the graphic form, for example, the area a in FIG. 11 (S57), the central processing unit 40 executing the search engine 42 program causes the searched patent to execute. Selecting only the patent information data belonging to the selected section or area from the information data, extracting the outline information of the selected information data, for example, the filing date, the applicant, the name of the invention, etc .; The information is transmitted to the corresponding search terminal PC through the interface 10 (S58), and the user can easily obtain general information on information data included in a specific section or area selected from the analysis results provided in the graphic form. Can be confirmed.

After confirming specific material data desired by the user through the summary information output on the screen as described above or by the classification file (solution problem and solution means) item in FIG. Upon request (S59),
The central processing unit 40 extracts only the requested specific information data from the searched information data and transmits the extracted specific information data to the corresponding search terminal PC through the network interface 10, as shown in FIGS. Such detailed data is output on the screen (S60), and when the user requests enlargement of the screen, the central processing unit 40 displays the information shown in FIG.
Provides the enlarged drawing data as shown in FIG.

In the above embodiment, the above-described analysis process can be repeated for all the information data selected from a specific field or a specific year according to the searched information data or analysis conditions. If the number of pieces of information data is too large, after inputting and setting sampling conditions, for example, international patent classification code, year, etc. on the patent analysis menu screen as shown in FIG. S52) The search engine 42 program preferentially samples the searched patent information data based on the input and set sampling conditions (S54), and performs an analysis process only on the sampled information data. Do (S5
5) For example, the total search material is 100 in 1981, 19
1982 250 cases,. . . In the case of 1221 cases in 1996, when the user inputs “year category” as the sampling condition and 5% as the sampling ratio, and requests sampling of the information data, the input sampling ratio Is 5
5% per year for patent material data, that is, 2 cases for 1981 and 50 cases for 1982
Cases,. . . An arbitrary selection is made for every 24 cases in 1996, and the above-described analysis process is performed only on the patent information data thus selected.

In addition, it is also possible to specify the sampling for each technical field together with the sampling for the above-mentioned year. If the sampling is specified for each technical field in this way, 5% is arbitrarily selected from each year. Instead of selecting one by one, the sample materials of the search materials will be selected from the search materials for the relevant fiscal year so that they will be 5% for each nuclear technology field.

[0038]

As described above, in the information retrieval method, the information automatic classification method, and the information analysis method according to the present invention, the information data is provided by classifying the solution problems and the solution means items as a classification file. Since the data is stored in the form of a file in cooperation with the search formula of the material data, access to the desired material is further facilitated, and the desired data can be stored without repeatedly searching using the stored file. As a result, the information data can be quickly searched, and the sampling of the search result information data can be performed, so that the analysis information data is significantly reduced, and thus the analysis of the information data can be performed more quickly. There is an effect that is.

In addition, since the analysis result of the information data can be provided on the screen in various graphic forms such as a pie map, a tree map, a trend map, and a bar map, the analysis result of the information data is grasped. This is very easy.

[Brief description of the drawings]

FIG. 1 is a flowchart illustrating a process of storing information in a database system based on an information search and information automatic classification method according to the present invention.

FIG. 2 is a configuration diagram showing an entire network of the integrated database system according to the present invention.

FIG. 3 is a diagram illustrating an example of information data stored in a hard disk drive (HDD) of FIG. 2;

FIG. 4 is a flowchart illustrating an information search method using the system of FIG. 3;

FIG. 5 is a drawing (part 1) showing an initial screen provided by connection of the integrated information management system by a web browser.

FIG. 6 is a drawing (part 2) showing an initial screen provided by connection of the integrated information management system by a web browser.

FIG. 7 is a diagram showing a patent search menu screen provided by an integrated information management system for user information search.

FIG. 8 is a flowchart illustrating an information analysis method according to the present invention.

FIG. 9 is a diagram showing a state where the analysis result of the information data is output on a screen in a tree map form (part 1);
It is.

FIG. 10 is a drawing (part 2) illustrating a state in which the analysis result of the information data is output on a screen in a tree map form.

FIG. 11 is a diagram showing a state where an analysis result of information data is output on a screen in a pie map format.

FIG. 12 is a drawing (part 1) illustrating a state in which the analysis result of the information data is output on a screen in the form of a trend map.

FIG. 13 is a drawing (part 2) illustrating a state in which the analysis result of the information data is output on a screen in the form of a trend map.

FIG. 14 is a drawing (part 1) illustrating a state in which the analysis result of the information data is output to a screen in a bar map format.

FIG. 15 is a drawing (part 2) illustrating a state in which the analysis result of the information data is output on a screen in a bar map format.

FIG. 16 is a drawing (No. 1) illustrating a state in which the analysis result of the information data is output to the screen in a pie map format.

FIG. 17 is a diagram (part 2) illustrating a state in which the analysis result of the information data is output to a screen in a pie map format.

FIG. 18 is a drawing (part 1) showing a state in which detailed materials of specific patent information data are output on a screen.

FIG. 19 is a drawing (No. 2) showing a state where detailed materials of specific patent information data are output on a screen.

FIG. 20 is a diagram showing an enlarged view of a drawing included in the detailed material of the patent information data.

[Explanation of symbols]

 PC ... search terminal 10 ... network interface 20 ... ROM 21 ... memory bank 30 ... hard disk 40 ... CPU 41 ... classification program 42 ... search engine program 43 ... DB management program 50 ... CD-ROM driver

Claims (13)

[Claims] 1. A method for connecting a user's information database system to an information source that records or provides predetermined information by itself and searches and classifies desired information from the information source, wherein Selecting search target information and setting a classification field for the search target information; searching the search target information to determine whether the search target information includes the classification field; Automatically classifying the classified information based on a classification field, and storing the classified information in a database of the user in sequence. 2. The method according to claim 1, wherein the setting of the classification field is set based on at least one of a symbol, a character, a word, and a sentence described in a material of the search target information. Information retrieval and automatic classification method. 3. The step of determining whether a classification field is present includes determining whether the classification field does not exist.
2. The method according to claim 1, further comprising the step of writing a classification field at a predetermined position of the searched information material. 4. An information search method using a database, wherein a search condition for searching for predetermined information data from a recording medium is input and set, and the information based on the input and set search condition is set. Outputting numerical data relating to the data search result to a screen; and forming the contents of each search result information data corresponding to the numerical data from the information contents belonging to one or more fields based on a character reference already specified. And a step of classifying and outputting as contents of a plurality of fields. 5. The information retrieval method according to claim 4, wherein the character criterion is a criterion for classifying information contents belonging to the one or more fields into fields of a solution problem and a solution means. 6. An information analysis method using a database, wherein a search condition for searching for information data on a recording medium is input and set, and a search result of the information data based on the input search condition is provided. Outputting numerical data to a screen; outputting statistical values related to the analysis result of the search information data based on the input and set analysis conditions to a screen in a graphic form; And analyzing and outputting the contents of the corresponding information data from the information contents belonging to one or more fields into a plurality of field contents formed on the basis of a previously specified character criterion. 7. The material according to claim 6, wherein, when a specific area of the graphic form to be output on the screen is selected, information of the material data corresponding to a corresponding ratio of the selected area is output to the screen. Analysis method. 8. The information according to claim 6, wherein the graphic form output on the screen is expressed so as to have an area corresponding to the ratio of the search material data to the corresponding technical field code. Analysis method. 9. The information analysis method according to claim 8, wherein the ratio is expressed by setting the number of search material data corresponding to the upper technical field code to 100%. 10. The information analysis method according to claim 8, wherein the ratio is expressed by setting the number of all the retrieved result data to 100%. 11. An information search method using a database, wherein a search condition for searching for material data on a recording medium is input and set, and the material data based on the input and set search condition. Outputting the numerical data related to the search result to a screen, storing the data of the search result corresponding to the numerical data and the search condition in a file format, and reading the stored data in the file format. And a step of: adding new search conditions to the read search conditions and obtaining updated search result material data from the read search result material data. 12. An information analysis method using a database, wherein a search condition for searching for material data on a recording medium is input and set, and a search is performed based on the input and set search condition. Analyzing the sampled material data based on the input and set analysis conditions, and outputting a statistical value related to the analysis result in a graphic form to a screen. , An information analysis method of sequentially performing. 13. The information analysis method according to claim 12, wherein said sampling step selects and samples an international patent classification code or yearly ratio of material data.