JP4912142B2 - Search system - Google Patents

Description

  The present invention relates to a search system that displays a search result corresponding to a search keyword input by a user.

  Keyword search is widely used when searching for document or image data or data published on a network. In a keyword search, a search keyword input by a user is used as a search condition, and data including a search keyword from inside a document, tag data previously given to an image, or inside data that is open to the public on a network is included. Searched for. When data is found, items such as characters and images are displayed as search results, and the user can obtain data and URL by selecting them.

  However, the search keyword input by the user is not necessarily appropriate as a search condition for displaying the search result desired by the user. For this reason, there is a method that uses a dictionary database in which words are associated with each other, such as a thesaurus, and searches for related terms and phrases related to a search keyword as a search condition. By using a dictionary database for the search, even if the search keyword and the data do not completely match, it is possible to hit the search (searching for the searched data) with the content data including the related terms of the search keyword. .

  However, when using a dictionary database, a search is performed using a search condition that is many times the number of search keywords entered by the user, so the number of items displayed as a search result increases, and among the displayed items Since many items unnecessary for the user are included, it becomes difficult to find the items that the user needs. In addition, when an upper limit is set for the number of items to be displayed, necessary items may not be displayed.

For this reason, in addition to the data including the search keyword input by the user, data that is highly related to the search keyword and does not include the search keyword is hit (Patent Document 1), or the search keyword is related to a related phrase or relationship. It has been proposed to display a relationship between words (Patent Document 2) to prevent omission of search.
JP 11-328194 A JP 2004-287827 A

  However, the search method as described above, for example, when searching for an image, when the user cannot clearly recognize the search keyword and needs to search based on his own image or ambiguous memory, Since it is necessary to repeatedly enter a search keyword until a search result desired by the user is obtained, a search based on images lacks convenience.

  The present invention solves the above-described problems, and an object of the present invention is to provide a search system that can improve convenience when a user performs a search that relies on an image.

In order to achieve the above object, a search system according to the present invention is a search system in which a number of items corresponding to a search keyword input by a user and its related terms are displayed as search results. A dictionary database to which the degree of abstraction is assigned, a content database in which the data of the item is stored, and the items associated with the search keyword and the related phrase are read from the content database, and a predetermined number of items a search processing means for displaying the items as a search result, the referring to the dictionary database to identify the abstraction of the search keyword entered by the user, thus the concept of the search keyword narrower, the increasing the number of displayed correlated to said item search keyword together , In which a search processing means for the relevant word display number of association has been the item is decreased by increase of the items that are associated with the search keyword to display the search results.

The item is an image, and the search processing means displays a search image associated with the search keyword and a related image associated with the related phrase as a search result within a predetermined display area. In this case, it is preferable that a display area for displaying the search image and the related image is changed in the predetermined display area in accordance with a change in the display number of the search image and the related image .

Search system of the present invention, the search concept of Keywords thus narrows the search with increasing number of displayed associated item to the keyword correspondence to find the number of displayed items that are associated with the related phrases keywords because it reduces only the items increment of which is, when the image of the user is ambiguous by increasing the number of displayed items that are associated with the related phrases and more specifically an image of the user, the image of the user specific In this case, it is possible to increase the number of items displayed in association with the search keyword to clarify the user's image, and it is possible to improve convenience when the user performs a search depending on the image. .

In addition, since the display area for displaying the search image and the related image is changed within the predetermined display area in accordance with the change in the display number of the search image and the related image , the related image is noticeable when the user image is ambiguous. When an image is specific, each image can be displayed so that the search image is conspicuous, and a search that is easy for the user to understand can be performed.

  As shown in FIG. 1, the search system 10 includes a CPU 11. The CPU 11 executes the program 13 in the ROM 12 and controls the operation of the entire search system 10. The RAM 14 is used as a work memory when the program 13 is executed. In the search system 10, in addition to the CPU 11, ROM 12, and RAM 14, a content database 15 that stores content data used to display search results, a dictionary database 16 that associates search keywords with related phrases, and search history data that will be described later are stored. A history database 17, a keyboard and a mouse, an input device 18 for inputting a search keyword, and a display 19 for displaying a search result are connected to each other.

  The CPU 11 includes a search processing unit (search processing means) 11a. When the user inputs a search keyword using the input device 18, the search processing unit 11a accesses the dictionary database 16 and searches for related terms related to the search keyword.

  As shown in FIG. 2, in the dictionary database 16, the search keyword and the related phrase are associated with each other. The related terms are associated with the search keyword in descending order of the degree of association (degree of association) with the search keyword. The degree of association is initially set, and is corrected as appropriate in detail as will be described later.

  The association of related terms will be described in detail. In the dictionary database 16, “Athletic meet”, “Autumn leaves”, “Sweet potato”, etc. correspond to the search keyword “autumn” in the descending order of relevance. It is attached. Similarly, for the search keyword “Sakura”, “Somei Yoshino”, “Hanami”, “Sakura”, etc. are associated as related terms in the descending order of relevance. , “Hot spring”, “dialect”, “ramen”, “Kagoshima”,. In the table in the figure, the related words / phrases for each search keyword are shown so that the related words / phrases with a high degree of relevance are in the upper row.

  The search keyword is assigned an abstraction level according to the breadth of the concept. The degree of abstraction is any value from “1” to “36”, and the value increases as the concept of the search keyword becomes wider. In the dictionary database 16, an abstraction level “33” is assigned to the search keyword “autumn”, an abstraction level “21” is assigned to the search keyword “sakura”, and an abstraction level “10” is assigned to the search keyword “Kyushu”. An abstraction level is also assigned to each related phrase.

  The content database 15 stores a plurality of types of image data. Each image data is attached with a plurality of tag data, and a search keyword and a related phrase are assigned to each tag data. Accordingly, each image data is associated with the search keyword and the related phrase. Then, the search processing unit 11a reads image data of an image (item) to be displayed as a search result from the content database 15 and displays it from the display device 19. The image displayed as the search result is an image representing a photograph.

  Each time the user inputs the related phrase displayed as a search result as a search keyword again, search history data indicating the search keyword at this time and the success or failure of the search is written in the history database 17. Then, each time the search history data is written, the search processing unit 11a corrects the degree of association of the related words / phrases in the dictionary database 16 and rearranges the related words / phrases. Specifically, the search processing unit 11a increases the degree of relevance of related words / phrases that are frequently searched again by the user.

  Next, the flow of processing when searching for an image will be described with reference to FIGS.

  As shown in FIG. 4, when the user inputs a search keyword with the input device 18, the search processing unit 11 a refers to the dictionary database 16 to identify the abstraction level of the search keyword. Thereafter, the search processing unit 11a calculates the display number (number of items) of the search image associated with the search keyword and the display number (number of items) of the related image associated with the related phrase according to the degree of abstraction. .

  More specifically, as shown in FIGS. 5 and 6, for example, 36 windows 19 b are displayed in one screen in the image display area 19 a of the display device 19. The search processing unit 11a distinguishes the 36 windows 19b into a window 19b in which the search image 30 is displayed and a window 19b in which the related image 31 is displayed. The number of windows 19b displaying the search images 30 is such that the number of windows 19b displaying the search images 30 increases as the abstraction of the search keywords decreases, that is, as the concept of the search keywords decreases. A keyword abstraction level is assigned.

  In this embodiment, for example, the number of windows 19b that display the search image 30 is “1” → the abstraction level of the search keyword is “36”, and the number of windows 19b that display the search keyword is “2” → the abstraction level of the search keyword is “ 35 ”,“ 3 ”windows 19b for displaying the search keyword → the abstraction level of the search keyword is“ 34 ”, and the number of windows 19b for displaying the search image 30 and the abstraction level are inversely proportional. An abstraction level is assigned to the number of windows 19b displaying the image 30. A table in which these are associated is stored in the ROM 12, and the search processing unit 11 a refers to the ROM 12 and determines the number of windows 19 b for displaying the search image 30. When the number of windows 19b displaying the search image 30 is determined, this number is subtracted from the total number of windows 19b to determine the number of windows 19b displaying the related images 31. With these processes, the display area of the search image 30 and the display area of the related image 31 change in the display area 19a as the display number of the search image 30 and the related image 31 changes.

  When the number of windows 19b for displaying the related image 31 is determined, the number of related terms used to display the related image 31 is determined. The number of related phrases is set in advance by the number of windows 19b displaying the related images 31, and is set so that the number decreases as the number of windows 19b displaying the related images 31 decreases. A table in which the number of windows 19b displaying the related image 31 and the number of related words is associated is stored in the ROM 12, and the search processing unit 11a refers to the ROM 12 to determine the number of related words. When the number of related phrases is determined, the determined number of related phrases is extracted from the dictionary database 16 in descending order of the degree of association. When the extraction of the related phrase is completed, the image data corresponding to the search keyword and the related phrase is read from the content database 15, and the search image 30 and the related image 31 are displayed in each window 19b.

  With these processes, in the example of FIG. 5, the number of windows 19b for displaying the search image 30 is four, and the number of windows 19b for displaying the related images 31 is 32. In the window 19b that displays the related images 31, four related images 31 are displayed for each of the eight types of related words. In the display area 19a, a window 19b for displaying the search image 30 is arranged at the center, and a window 19b for displaying the related image 31 is arranged so as to surround the periphery. In addition, characters indicating the search keyword (“Autumn” in the figure) are displayed together with the search image 30, and characters indicating the related words (“1” to “8” in the figure) are displayed together with the related image 31. Thereby, it is displayed to a user whether the image relevant to which word is displayed.

  In the example of FIG. 6, there are 16 windows 19 b for displaying the search image 30 and 20 windows 19 b for displaying the related images 31. In the window 19b displaying the related images 31, four related images 31 are displayed for each of the five types of related words. In the display area 19a, a window 19b for displaying the search image 30 is arranged at the center, and a window 19b for displaying the related image 31 is arranged so as to surround the periphery. Further, characters indicating the search keyword (“sakura” in the figure) are displayed together with the search image 30, and characters indicating the related words (“1” to “5” in the figure) are displayed together with the related image 31. Thereby, it is displayed to a user whether the image relevant to which word is displayed.

  Thereafter, when the displayed image is not an image desired by the user and the user re-enters the search keyword, it is determined whether or not the re-entered search keyword is a related phrase being displayed. If the re-entered search keyword is not a related phrase, the same processing as described above is executed again. If the re-entered search keyword is a related phrase being displayed, the search keyword and its related phrase re-extracted according to the abstract level of the search keyword by the same process as described above are re-extracted. Image is displayed. In this case, the search history data is stored in the history database 17, and the degree of association in the dictionary database 16 is corrected after the search is completed.

  When image data exceeding the number of windows 19b displaying the search image 30 or the number of windows 19b displaying the related image 31 is stored in the content database 15, the image to be displayed is selected at random, or The selection method may be set as appropriate, such as selecting an image with a relatively large number of selections from the top, or preferentially displaying an image searched for overlapping with other related phrases.

  In the above embodiment, the display area of the search image and the display area of the related image are changed in one screen according to the number of each image by changing the number of windows 19b, but the window 19b is not used. In this case, the size of the entire display area 19a is set to “100%”, for example, and the result obtained by dividing this by the abstraction level of the search keyword is used as the display area of the search image, and the remaining display area is set as the display area of the related image. Depending on the number of images, it may be changed within one screen. In this case, specifically, when “autumn” described above is input as a search keyword, the abstraction level of “autumn” is “33”, so the size of the search image display area 30 is (100/33). ) = 3.3%. Further, the size of the remaining display area is the size of the related image display area 31, and the size of the related image display area 31 is 100−3.3 = 96.7%.

  In the above embodiment, the items used for displaying the search results are displayed as images. However, the items may be displayed as characters or marks, or may be represented by music.

  In the above embodiment, the image data of the content database 17 is set in advance, but the user may arbitrarily associate the image data with the tag data. In this case, the administrator of the search system only confirms these associations.

  In the above-described embodiment, search images corresponding to the display number determined by the abstraction level of the search keyword are displayed. If the search image associated with the abstraction level exceeds the display number, the search image or the search keyword is displayed. Double-clicking may increase the number of search images displayed.

  In the above embodiment, only the search in which the number of items changes according to the degree of abstraction is performed. It may be possible to select.

It is a block diagram which shows the electric constitution of a search system. It is explanatory drawing which shows the structure of a dictionary database. It is explanatory drawing which shows the structure of a content database. It is a flowchart which shows the flow of the process at the time of searching image data. It is explanatory drawing which shows the specific structure of a display area when a search keyword with high abstraction is input. It is explanatory drawing which shows the specific structure of a display area when a search keyword with low abstraction is input.

Explanation of symbols

10 Search system 11 CPU
11a Search processing unit (search processing means)
16 Dictionary database 19a Display area 30 Search image 31 Related image

Claims (2)

In a search system in which a number of items corresponding to a search keyword entered by a user and its related terms are displayed as search results,
A dictionary database in which the search keyword is assigned an abstraction level according to the size of the concept;
A content database storing the data of the item;
A search processing means for reading out the items associated with the search keyword and the related phrase from the content database, and displaying a predetermined number of items as a search result, wherein the user refers to the dictionary database. identifying the abstraction of the search keyword entered, therefore the concept of the search keyword narrower, with increasing number of display of the items that are associated with the search keyword, which is associated with the related phrases A search system comprising: a search processing means for displaying the search result by reducing the display number of the item by an increment of the item associated with the search keyword . The item is an image;
The search processing unit displays the search image and the related image when the search image associated with the search keyword and the related image associated with the related phrase are displayed as a search result in a predetermined display area. The search system according to claim 1 , wherein a display area for displaying the search image and the related image is changed in the predetermined display area in accordance with a change in the display number .

Images