JPH0785006A - Method and device for retrieving information - Google Patents

Abstract

PURPOSE:To easily retrieve a target file and further to improve a user interface by inputting a retrieval conditional expression with a character string in a natural language form, interpreting the meaning of this character string and retrieving a storage medium while referring to a file managing table and a directory based on the result of this interpretation. CONSTITUTION:A thesaurus dictionary 22 stores a keyword group for file retrieval and the relation of those keywords. A reserved word dictionary 23 stores the meaning and meaning contents of specified character strings previously decided for file retrieval. A logical arithmetic table 24 stores the logical arithmetic contents of specified character strings to be used for Japanese. When the retrieval conditional expression is inputted with the character string in the natural language form, an OS 20 and an application program 21 interprete the meaning of this character string while referring to these thesaurus dictionary 22, reserved word dictionary 23 and logical arithmetic table 24. Based on the result of this interpretation, the OS 20 and this program 21 retrieve the target file while referring to a file managing table 25 and a file storage area 26.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information retrieving method and apparatus, and more particularly to an information retrieving method and apparatus for finding a target file from a large number of files stored in a storage medium.

[0002]

2. Description of the Related Art In recent years, external storage devices, particularly hard disk devices, magneto-optical disk devices, etc., have become widespread due to their large capacity and low price. When such a large-capacity storage device becomes widespread, it becomes possible to store a large amount of document files and the like therein. Further, since the operation of taking in and out a storage medium such as a floppy disk, which has been necessary up to now, is not necessary and high-speed reading and writing are possible, the operating environment is improved. However, if an extremely large amount of documents are stored, it is difficult to find the target document file.

[0003]

Generally, an electronic device running an OS capable of managing files in a hierarchical structure (tree structure),
For example, each level (commonly called a directory)
It is usual to manage files by assigning a name suitable for classifying the files to and storing the document files created under each hierarchy. However, it is the operator who decides under which hierarchy the created document should be stored, and it is difficult to discover later that the operation is not appropriate. Further, even if the document is stored in the correct location, assuming that the number of document files stored under that layer expands,
It is inevitable that finding the desired document will be difficult.

[0004]

SUMMARY OF THE INVENTION The present invention has been made in view of the above prior art, and aims to provide an information retrieval method and an apparatus therefor which can easily find a target file and which has an excellent user interface. is there. In order to solve this problem, the information search method of the present invention includes the following steps. That is, in a file search method for searching for a desired file from a storage medium that stores a plurality of information files and a character string for identifying the file, the search condition expression is a natural language character string. Interpret the input process and the input natural language character string,
And a search step of searching the storage medium.

The information retrieval device of the present invention has the following configuration. That is, in a file search device that searches for a desired file from a storage medium that stores a plurality of information files and a character string for specifying the file, the search condition expression is a natural language character string. Interpret the input means and the input natural language character string,
And a search means for searching the storage medium.

[0006]

In the process or apparatus of the present invention, a sentence in a natural language form is input and a search is performed according to the meaning of the input sentence.

[0007]

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 is a block diagram of the information processing apparatus in the embodiment. In the figure, 1 is a CPU that controls the entire apparatus, 2 is a ROM that stores a boot program and character fonts, 3 is an R that is used for storing information to be edited and used as a work area of the CPU 1.
AM. Reference numeral 4 is a keyboard for giving various instructions and inputting a character string, and 5 is a pointing device such as a mouse for designating an arbitrary position on the display screen. Reference numeral 6 is a VRAM that develops an image corresponding to the information displayed on the display screen one-to-one, and 7 is a display device that displays the information developed in the VRAM 6. An external storage device (for example, a hard disk device) 8 stores an operating system (OS), various application programs, and various information described later.

The outline of the operation of the embodiment having the above-mentioned structure is that the user inputs the usual conversation type search information for the file stored in the external storage device 8 and searches for the corresponding file. Specifically, the search is performed by inputting a search condition in a text format (natural language format) that is normally used, such as "Document for computer last year". The stored contents of the external storage device 8 are shown in FIG.
The external storage device 8 stores the following information including an OS 20 and various application programs (document editing program, figure practice program, etc.) 21.
Reference numeral 22 is a thesaurus dictionary that stores a keyword group used for file search and their relation, 23 is a reserved word dictionary that stores the meaning of a specific character string predetermined for file search and its meaning content, Reference numeral 24 is a logical operation table that stores the logical operation contents of a specific character string used in Japanese, and 25 is a file management table that manages various files. Reference numeral 26 is a file storage area for storing information created and edited by executing various application programs as a file, and 27 is a dictionary for kana-kanji conversion.

In the above various information, the thesaurus 2
A conceptual diagram of No. 2 is shown in FIG. The thesaurus dictionary 22 has a tree structure as shown in the figure, and indicates that, for example, the words "language" and "academic" are included below (lower) the word "culture". Then, a word that literally represents the concept of the subordinate is connected to the subordinate of each word.
In such a dictionary structure, when a user inputs a specific character string to perform a search, the character string and all character strings located under the character string are searched for a file. . Reserved word dictionary 2 in the embodiment
The structure of 3 is shown in FIG. The reserved word dictionary 23 manages character strings related to time as illustrated. In the figure, “xx” indicates that an arbitrary character string is located.

For example, the character string "kara" is used to represent a period between the time when the form is represented by the first parameter and the time when the form is represented by the second parameter. Also, "last year" means subtracting "1" from the current year (clocked by a clock (not shown) of the present device). Further, it is assumed that "document" and "figure" are located as parameters following the character string "last year". In the embodiment, the character strings stored in the reserved word dictionary 23 have a priority order. Here, the priority order is managed by the storage order, but a number indicating the order may be held. However, "no document" of the character string "xx document (graphic)" is defined as having the lowest priority. As described above, the reserved word will 23 is used for interpreting the time-limited character string input by the user and recognizing its content. The reserved word dictionary 23 allows the user to arbitrarily register a new character string.

The contents of the logical operation table 24 are shown in FIG. It is necessary to create a more limited search character string by combining a plurality of independent search words (a character string whose search content is known only by the character string). In order to construct this in natural language, it is necessary to connect individual independent search terms by some means. This logical operation table manages the character string for the connection. In the figure,
For example, the character string “de,” is a connection character string for connecting two independent condition character strings, and when a character string satisfying the first condition character string and the second condition character string is searched for ( Used for AND). Further, in the case of the condition character string "Computer, last year's document" described above, two logical products of the keyword "computer" preceding the character string "and" and the following "last year's document" are calculated. It means to take. Therefore, it means exactly the same as "Last year, computer documents". However, when viewed as Japanese, the former is more natural. The user can freely add a new character string also to the logical operation table 24. The priority order is based on the order of storage, but a priority number may be added, as in the reserved word dictionary 23.

Next, the file management table 25 of the embodiment
The contents of the above are shown in FIG. As shown in the figure, each file is composed of its name, creator, and keyword information. When newly saving the information created during the execution of a certain application program, the user has to arbitrarily decide the file name and save it. At that stage, enter the creator and keywords. Ask. Then, the file name, creator, and keyword input here are added to the file management table. As a matter of course, when deleting a file stored in the external storage device 8, the information of the corresponding file is deleted from the file management table 25. The OS manages the processing for new saving and file deletion processing.

The keyword can be arbitrarily determined by the user, but it needs to be a word managed in the thesaurus dictionary 22 described above. Therefore, it is assumed that most of the words are registered in the thesaurus dictionary 22. Note that there is no guarantee that the thesaurus dictionary 22 will function as a complete dictionary, so that the user can freely register words in this dictionary. In the case of registering a word, the operation of displaying a list of character strings of the highest order (the highest order concept), selecting any one of them, and displaying the character string of the concept thereunder is repeated. Then, when the character string at the desired stage is displayed, the character string to be registered may be input while the character string is selected. Further, when the file is newly saved, it is also checked whether the input keyword exists in the thesaurus dictionary 22. Then, when it does not exist, or when the user gives an instruction to that effect, the registration process is performed. Since the registration process itself is as described above, its explanation is omitted.

In general, the keywords are words that express the characteristics of the document well, and are often words that are frequently used in the document. Therefore, regardless of the figure, in the case of a document, the most frequently used words in the document are treated as keywords, or some of the most frequently used words are displayed and the user specifies some of them. You may make it register what you did as a keyword. Searching for frequently used words in a document is known per se, and a description thereof will be omitted here.

Before and after the explanation, in the operating system operating on the apparatus of the embodiment, when the file is stored in the external storage device 8, the update date is held in the directory area for managing the file. Therefore, the search according to the search information specified by the date and time in the reserved word dictionary 23 described above is performed based on the date information held in the directory. However, the file management table may have an item of date information at all.
Therefore, the present invention is not limited by the type of OS.

Next, an example of an actual search process based on the above management information will be described with reference to FIGS. When an instruction to perform a search process is given using the keyboard 4 or the pointing device 5, a window for search as shown by reference numeral 70 in FIG. 7 is displayed. 71 in the figure
Is a cursor linked to the pointing device 5, 72 is a field for inputting a search character string, and 73
Is a text cursor indicating the position of the character string input from the keyboard 4.

Now, in the state of FIG. 7, the user inputs a condition character string for searching for a target file into the field 72 using the keyboard. The user can freely set the keyword (a character string for representing the file stored in the file management table), and therefore he / she does not know what is input. However, since the character string used in the conditional expression regarding time is limited, a logical button for the date is provided as shown in the figure. This makes it easy to input a conditional expression related to time without getting used to it. Although the creator may be regarded as a keyword, if the name of each individual is registered in the thesaurus dictionary 22, the efficiency of search and management will deteriorate. Therefore, as described above, the file management table 25 And manage it as an independent item. In order to distinguish it from general keywords, a button for inputting "creator" is prepared as shown in the figure. Although the input form of the "creator" has not been described, it is assumed here that the input is made in the format of "creator is xxxxxx". Therefore, for example, in the state of FIG.
When the button 75 is instructed by the cursor 71 and a physical button provided on the pointing device 5 is pressed (hereinafter, referred to as clicking this operation), a message window indicating who the creator is is displayed, Enter only its name. As a result, the character string is automatically entered in the field 72 in the format of “creator is xxxxxx”. Of course, the connection between the character string "creator is" and the character string (personal name) following it is high.

Now, assume that the button 74 is clicked in the state of FIG. As a result, a window for time (period) limitation as shown in FIG. 8 is displayed in a superimposed manner. The user clicks on one of the items. here,
It is assumed that the third item “#### to ##” is selected.
In this case, since two parameters are required, an example of a character string that can be the first parameter is displayed, for example, the message “Please input in the format XX year (XX month)” is displayed and the first parameter is displayed. Enter the parameters of.
Then, similarly, the second parameter is input.

Thus, the search window is in the state shown in FIG. Again, the buttons 74 and 75 are provided for a user who is not familiar with them, and similar character strings may be directly input from the keyboard 4 as long as the operation mode can be realized. Now, in the state of FIG.
When “execute” displayed in the search window is clicked, a search based on the character string input in the field 72 is executed.

As a result, if a large number of files exceeding a preset fixed value are found, a warning message is displayed (a warning message is also displayed if no files are found). By seeing such a message, the user knows that the conditional expression must be changed or restricted. Therefore, a new condition is added subsequently to the previously inputted condition character string (when the condition character string is changed, the former condition character string is erased). Here, suppose, for example, the user inputs ", with respect to FEP" from the keyboard 4. FIG. 10 shows this state.

When the button 76 is clicked again to restart the search, if three files having a fixed value or less are searched, a list of those file names is displayed as shown in FIG. It The user clicks any desired file name with the cursor 71 to display information (information in the file management table and directory) about the file in a window as shown in FIG. When it is determined that the target file has been found by repeating this operation, the predetermined operation is performed to read the file into the RAM 3 and the application that created the file is activated. If the search expression contains keywords in addition to the date / time information and creator information, and it is found that there are many search results (when the number of files equal to or greater than a fixed value is searched), A warning message as shown in 13 is displayed. And the entered keyword (in this case "electronic technology")
A list of subordinate concept character strings is displayed with reference to the thesaurus dictionary 22 to prompt the user to limit the words (keywords). Here, for example, by selecting “hardware” and performing a predetermined operation, it is possible to display a list of words that are subordinate concepts of hardware. In any case, if you find a word that describes the file you're looking for, indicate so. As a result, the keyword part in the search character string used in the immediately preceding search process is replaced with the specified word. Of course, the character string may be directly input from the keyboard 4.

The above processing is achieved by processing according to the flowcharts shown in FIGS. Hereinafter, description will be made step by step. First, in step S1, an input screen (window) for initial conditions for search is displayed, and in the next steps S3 and S11, it is determined whether or not the buttons 74 and 75 have been clicked.

If the button 74 (button for inputting date and time conditions) is clicked here, the process proceeds to step S.
Proceed to step 5 to display the date and time item list and have one selected. After this, the process proceeds to step S7, a window for inputting parameters for the selected item is displayed, and the user is prompted to input it. Then, in step S9, a natural language character string is created based on the selected item and the input parameter, and is output to the search condition input field 72 in the window.

When the button 75 (button for inputting the creator's name) is clicked, step S1
Proceed to step 3 to display a window prompting you to enter the creator's name, and have the user enter the name in the window. Then, the process proceeds to step S15 to generate a character string of the format "creator is xxxxxx" and outputs it to the search condition input field 72. When the buttons 74 and 75 are not clicked and the character input by the keyboard is detected, the process proceeds to step S17 to directly input the search character string in the search condition input field 72.

In any case, the process proceeds to step S19, and the above process is repeated until it is determined that the button 76 (button for instructing search execution) is clicked. When the button 76 is clicked, the process proceeds to step S21 to analyze the character string input in the search condition input field 72 and construct a logical expression so that the CPU can easily understand it. If it is determined by this analysis that a grammatical error has occurred, the process proceeds from step S23 to step S25 to display that there is an error in the character string of the search condition, and returns to step S3. If the analysis is completed normally, the process proceeds to step S27 to execute the search process based on the analysis result.

In steps S29 and S31, it is determined as a result of the search whether or not the target file has been found, and whether or not the number of files has been narrowed down to a preset number or less. In the following, if one file could not be found (Fig. 15) and too many files were found (Fig. 1)
6), and the case where the number of files within the allowable range can be found (FIG. 17) will be described in order.

If no one is found, the process proceeds to step S41. Here, first, a message indicating that no items have been found is displayed, and in step S43, it is determined whether or not information regarding time is included in the search character string at that time. Then, if the information on the time is included, the process proceeds to step S45 to display a message prompting the change of the time and the expansion of the range. Further, in step S47, it is determined whether or not the search character string includes keyword information. If no keyword information is included, the process returns to step S3. If the keyword information is included, the process proceeds to step S49, a list of character strings of superordinate concepts for the keyword is searched for by referring to the thesaurus dictionary 22, and one of them is selected (selected. It is also possible to select a higher concept of the character string).

Normally, there is one keyword of a superordinate concept to one keyword, but in the embodiment, there may be a plurality of keywords. For example, computer languages such as "C" and "LISP" can be used literally as a subordinate concept of "computer" and also as a subordinate concept of language. However, when the keyword input as the search character string exists in a plurality of places in the thesaurus dictionary 22, the logical sum of the results searched by each is output as the final search result.

When a keyword of a superordinate concept is selected or input from the keyboard 4, the process proceeds to step S51, the keyword used in the immediately preceding search process is replaced with the keyword selected in step S49, and the process returns to step S3. On the other hand, if there are too many search results, the process proceeds to step S61, and a message indicating that there are too many is displayed. Then, in step S63, it is determined whether or not the search character string at that time includes information regarding time. If the information about time is included, the process proceeds to step S65 to display a message prompting change of time / reduction of range.

In step S67, it is determined whether or not the search character string contains keyword information. If no keyword information is included, the process returns to step S3. If the keyword information is included, the process proceeds to step S69, a list of character strings of subordinate concepts for the keyword is searched for by searching the thesaurus dictionary 22, and one of them is selected (selected. It is also possible to select a subordinate concept of the character string). Thus, when a new keyword is selected or input, the process proceeds to step S71, the keyword used in the immediately preceding search process is replaced with the keyword selected in step S69, and the process returns to step S3.

As a result of the search, if the number is narrowed down to a certain degree or less, the process proceeds to step S81. Here, a list of candidates that match the search conditions is displayed. Then, the process proceeds to the next step S83, and one of them is selected. When this selection is made, the details of the selected file are read from the file management table 25 and the directory, and the user is notified. The user looks at this content and determines whether the document is a target document. If it is determined that the document is not the target document, steps S87 to S8 are performed.
The process of returning to 1 and selecting any one of the remaining documents and confirming the contents is repeated. If it is determined that the target document can be found in this way, an instruction to that effect is given. As a result, the process proceeds from step S87 to step S89, and the corresponding application program is started according to the attribute (document or graphic) of the file selected at that time, and the specified file is read. Perform editing processing.

In the above process, it is described that the application program is started when the target file is found, but the present invention is not limited by the usage method. For example, the file name list may be printed.

Further, as a search condition character string, a character string having only one parameter, for example, a character string "xx years ago" is used as a message when a large number of files are found, "xx years ago when? ]
Etc., display a message such as the first half of the year, broadening, what month, or from month to month etc., select one of them, select one of them, and if necessary You may input a parameter and reconstruct a search condition character string. As described above, according to this embodiment, it is possible to input the search condition in natural language. In addition, if no corresponding file is found or there are too many files, it gives a guideline on how to reconstruct the search condition, so create a search condition character string efficiently by an appropriate method. It becomes possible to do. Although the information to be searched is a document or a graphic in the embodiment, the present invention is not limited to this.

[0034]

As described above, according to the present invention, it is possible to provide an information retrieval method and an apparatus therefor which can easily find a target file and which has an excellent user interface.

[Brief description of drawings]

FIG. 1 is a block diagram of an information processing apparatus according to an embodiment.

FIG. 2 is a diagram showing information stored in an external storage device according to an embodiment.

FIG. 3 is a conceptual configuration diagram of a thesaurus dictionary in the embodiment.

FIG. 4 is a diagram showing a data structure of a reserved word dictionary in the embodiment.

FIG. 5 is a diagram showing a structure of a file management table in the embodiment.

FIG. 6 is a diagram showing a structure of a logical operation table in the embodiment.

FIG. 7 is a diagram showing an example of a state of a display screen of search processing in the embodiment.

FIG. 8 is a diagram showing an example of a state of a display screen of search processing in the embodiment.

FIG. 9 is a diagram showing an example of a state of a display screen of search processing in the embodiment.

FIG. 10 is a diagram showing an example of a state of a display screen of search processing in the embodiment.

FIG. 11 is a diagram showing an example of a state of a display screen of search processing in the embodiment.

FIG. 12 is a diagram showing a display example of details of one file as a search result in the embodiment.

FIG. 13 is a diagram showing an example of a state of a display screen of search processing in the embodiment.

FIG. 14 is a flowchart illustrating a search process according to the embodiment.

FIG. 15 is a flowchart illustrating a search process according to the embodiment.

FIG. 16 is a flowchart illustrating a search process according to the embodiment.

FIG. 17 is a flowchart illustrating a search process according to the embodiment.

[Explanation of symbols]

 1 CPU 2 ROM 3 RAM 4 keyboard 5 pointing device 6 VRAM 7 display device 8 external storage device

Claims (4)

[Claims] 1. A file search method for searching for a desired file from a storage medium storing a plurality of information files and a character string for identifying each file, wherein the search condition expression is in a natural language form. An information search method comprising: a step of inputting a character string; and a search step of interpreting the input natural language character string to search the storage medium. 2. If, as a result of the search in the search step, there are no corresponding files or a large number of files, the time-limited words contained in the natural language character string used for the search or the contents of the target file are displayed. The information retrieval method according to claim 1, further comprising a display step of displaying a message that suggests an enlargement direction or a reduction direction of a word to be expressed. 3. A search condition expression is placed in a file search device for searching a desired file from a storage medium storing a plurality of information files and a character string for specifying the file, and the search condition expression is a natural language. An information retrieving apparatus comprising: a unit for inputting a character string in a form; and a retrieving unit for interpreting the input natural language character string to retrieve the storage medium. 4. If, as a result of the search by the search means, there are no corresponding files or a large number of files, the time-limited words contained in the natural language character string used for the search or the contents of the target file are displayed. The information retrieval apparatus according to claim 1, further comprising a display step of displaying a message that suggests an expansion direction or a reduction direction of a word to be expressed.