JP2002297661A - Web retrieving device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a Web retrieving device which makes it possible to grasp the trends, correlations, etc., of retrieved information and to perform data mining by performing the retrieval by using a plurality of attributes and presenting the retrieval result multi-dimensionally. SOLUTION: This device is equipped with: a retrieving means 3a which retrieves the information of an object Web page to be retrieved by using a plurality of attributes as retrieval conditions; and a presenting means 3b which presents the retrieval result multi-dimensionally according to the values of the attributes.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a Web search apparatus for quickly and accurately searching Web pages provided on the Internet or the like.

[0002]

2. Description of the Related Art A large amount of information is uploaded on the Internet, and necessary information can be obtained by performing a keyword search on a search site or by using a category provided by the search site. In general, conditions are narrowed down.

[0003] In addition, many meta search sites have recently been launched due to dissatisfaction with full-text search engines due to keywords. The meta search is to search a plurality of search engines in parallel and display the integrated result, and to use a robot type or directory type search engine as a back end.

[0004] In such sites, search engines,
Many search services, such as books and CDs, are classified according to purpose and search is performed by assigning keywords to an appropriate engine.
The obtained result is embedded in an HTML document and presented to the user.

[0005] When presenting the search results, many sites provide a function of ranking search data by normalized scores and sorting based on the number of hits, the number of references, and the like.

[0006] Regarding the page contents, there are those that return the information of the search engine as it is, and those that obtain the actual page and return some filtered information.

[0007]

When performing a search using the above-described search engine, a user (searcher) grasps the characteristics of a plurality of search sites and uses them properly, or uses a complicated and detailed search option. Often it is necessary to set and repeat the search.

[0008] In order to obtain desired information from Web pages searched in large numbers, duplicate Web pages are required.
Pages and unnecessary information must be excluded.

On the other hand, in a meta search site, a search is performed in a category prepared by the site without specifying a complicated search formula. Such a category is uniformly classified, and is not necessarily classified. The desired information is not always found. In addition, search site categories are manually created, and there is a limit to quickly and appropriately classifying a large amount of information. In addition, there was the possibility that the creator would enter into it.

When tracing the information of a related Web page from a Web page, there is only fixed link information described in the Web page, and the Web according to the purpose and use is provided.
Page b could not be referenced.

In view of such circumstances, the present invention performs data mining by performing a search using a plurality of attributes and presenting the search results in a multi-dimensional manner so that the tendency and correlation of the searched information can be grasped. The purpose of the present invention is to provide a Web search device that can perform the search.

Also, a URL (Uniform Resource Locato)
An object of the present invention is to provide a Web search device capable of performing a search quickly and accurately by performing a search using the attribute of r).

[0013]

To achieve the above object, the present invention employs the following means.

The Web search apparatus of the present invention searches for a Web page using a plurality of attributes and presents the search results in a multidimensional manner, making it easy to grasp the search results.

Further, by using the attributes included in the URL, it is possible to quickly perform a search using a plurality of attributes.

[0016] A Web search apparatus of the present invention provides a search means for searching for information of a Web page to be searched using a plurality of attributes as search conditions, and presents the search result in a multi-dimensional manner in accordance with the value of the attribute. And a presentation unit that performs

As a result, a Web search device capable of performing a search quickly and accurately is achieved.

In the Web search device, it is preferable to use, as the attribute, at least one of a directory name, a file name, an organization name, an organization address, an organization classification, and an organization location, which are attributes of a URL (Uniform Resource Locator). is there.

For example, the directory name and file name of the URL often indicate the contents of the stored information, and the value of this attribute is news, prod.
ucs, books, reports, etc., and can be regarded as having information such as news, products, books, and reports.

Therefore, by performing a search using the URL attribute, an appropriate search result can be obtained quickly.

In the Web search device, the attributes include a keyword, tag information, and a link destination UR.
It is preferable to use at least one of L, link source URL, link name, and update date.

Further, in the Web search apparatus, the search means may perform a keyword search on information of a Web page to be searched, and perform a search using the attribute on a result of the keyword search. good.

[0024] Thus, multi-dimensional display can be performed on the information narrowed down by an arbitrary keyword, and for example, it is possible to evaluate an approach situation of each company regarding the target technical term.

[0024] In the Web search device, it is preferable that the presenting means presents a URL of information selected by a searcher among the search results shown in the multi-dimension.

Thus, when the searcher selects necessary information from the search results shown in multi-dimensions, the URL of this information is presented, and the necessary Web page based on the URL can be browsed. .

In the Web search device, the presenting means presents a search result using the attribute by using a part of the attribute, and enables the attribute used for the presentation to be arbitrarily switched.

Thus, it is possible to evaluate a search result shown in multiple dimensions from different viewpoints.

In the present invention, the above-mentioned components can be combined as much as possible.

[0029]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a conceptual diagram of a search using the Web search device of the present invention, and FIG. 2 is a diagram showing a display example of search results by the Web search device.

The Web server 1 shown in FIG. 1 provides a Web page via the Internet regardless of domestic and overseas. The keyword extracting device 2 includes a keyword extracting unit 2a that accesses the Web server 1 and extracts a keyword from information of a Web page, and a storage unit 2b that stores the keyword in association with the URL of the extracted Web page. Have.

Then, the Web search device 3 has at least U
A search unit for searching for a Web page using a plurality of attributes including an RL attribute, and a presentation unit for presenting the search results in a multidimensional manner are provided.

The URL of the Web page is, for example, “h
http: // www. ○○○○○. co. jp / new
s / zisin / higai. html / ", access means: // host name.domain name / path name (directory name) / file name /".

The domain names include a first level domain mainly representing a country name, a second level domain mainly representing a business type or an organization, a third level domain mainly representing an organization name, and the like. Therefore, it is possible to determine from the domain name which organization provides this Web page, and in the above example, it can be seen that the Japanese company XXX is providing this Web page.

Further, directory names and file names below the domain name often indicate the contents of stored information. In the above example, news = news, zi
Since sin = earthquake and higai = damage, it is understood that the file is related to the news of the damage caused by the earthquake.

Therefore, the Web search apparatus of the present embodiment evaluates the search results using the following attributes indicating the properties of the Web page. (1) Attribute and attribute value of URL information ([Attribute: Attribute value]) [Domain name: URL of top page] [Third and higher domains: Organization names such as company name, organization name] [Second domain: co , Go, ne, ac, etc.] [First domain: location of jp, uk, etc.] [Contents: products, books, reports, links, etc.] (2) Attributes and attribute values of in-page information [Keywords: [Keyword in title or text] [Update date: update date] [Link destination URL: URL described as link destination] [Link source URL: URL described as link destination on another page] [Link name: anchor tag] [Character string enclosed in []] [Tag information: Contents described using tags of HTML and XML files and tags themselves (specific examples will be described later)] (3) Attributes and attribute values of search results [Number of hits: Number of pages searched and hit based on other attributes] [Number of links (portal level): Number of links included in pages of search results] [Number of linked (popularity): Number of links The number of links that specify the page as the link destination] [Sort by update date: update date and time of the search result page] Since the URL is described with alphanumeric characters and specific symbols, the attribute value is English or Roman characters, Abbreviations are often used, and when using kanji or kana as attribute values,
[News = news] [products, seihin = products] [book
s, hon = book] [touden, tepco = TEPCO], etc., a dictionary is created in advance that associates notation with alphanumeric characters and symbols with words in kanji and kana, and discriminates against this dictionary. . For HTML sites that have a uniform HTML tag format, such as news sites, various attributes of Web pages can be acquired by using tags.

[0037]

[Outside 1] Various attributes can be obtained from the above HTML by using the tags described below.

[Outside 2] Therefore, data with various attributes (XML files) can be obtained.

○ Acquisition attribute (XML file) <Data><Genre><Title><Origin><Publishingdate><Text><Heading><Information information (URL, TEL, etc.) published in the article> Also, in general Web pages, "heading tags"
It is possible to guess the contents of the page to some extent without reading the text by using a character string using <h1> or the like or a tag specifying a display method such as font or color. If it is set to automatically jump to the next page a few seconds later, it is treated as a meaningless page. It is possible to assign an attribute as to whether the page is heavy or not depending on whether the image is embedded. <Example 1> Next, in the present example apparatus, the number of hits for technical terms (keywords) to be searched is classified by category and by domain of a competitor, and is evaluated by displaying it in a multidimensional manner. explain. First, the keyword extracting device 2 accesses the Web server 1 via the Internet, and the keyword extracting means 2a extracts the keyword of each Web page and the head of each Web page text. UR
L is stored in the storage unit 2b in association with L. And W
A search condition is input to the search means 3a of the eb search device 3.

FIG. 2 shows a display example of the search condition input screen. For example, "camera" is set as a search target word (keyword), and x is one of the dimension axes when multidimensional evaluation is performed. On the axis is the manufacturer name (organization name) "Company X, Company Y,
"Company Z" is set, the content of information "product, open source, standardization, technical document, R &D" is set on the z axis (dimension axis), and the number of hits is set on the y axis (dimension axis).

At this time, the category (contents of information) is manually selected, but may be automatically generated based on information in a Web page.

For example, if there are buttons such as "product information", "FAQ", and "news" on the top page of a company or the like, the information of the link destination URL of this button may be information belonging to the category of the button name. Since the user can understand, the button name is selected as a category, and information on the link destination URL is extracted as information on this category.

The title of a Web page or a word (keyword) included in the head of the page may be automatically extracted as a category. Under these conditions, the storage unit 2b
When the information stored in is searched, first, all the information including "camera" in the keyword is extracted. Then, this information is classified for each combination of attributes set for each dimension axis.
In this example, the combination of the attribute value of the x-axis and the attribute value of the z-axis “X
Company, product, company Y, product, company Z, product, company X, open source, company Y, open source, and
A search is performed, and the number of hits is set as a y-axis value, and the search result is displayed in a multidimensional manner. FIG. 3 is an example of this display.
As a result, it is possible to grasp the status of the approach to each company regarding the search target word (camera). This x, z
When the searcher selects (clicks) a bar that seems to contain necessary information from among the bars indicated for each combination of the attribute values of the axis, the presentation unit 3b displays the URL of the corresponding Web page and the body of the URL. Are displayed in a list on the monitor 3c. The searcher then reads the top of each page,
When a necessary page is selected, the presentation unit 3b sets the UR
Based on L, a connection is made to a Web server 1 via the Internet to acquire information on a target Web page and display it on the monitor 3c. At this time, the display of the Web page may be performed by the presentation unit 3b itself, but the specified URL is transmitted to another Web browsing software (browser), and the Web page is displayed.
b It may be displayed by browsing software. As described above, in the present embodiment, it is possible to search for a target page while classifying the information into appropriate categories and grasping the overall image of related information.

In particular, a large amount of information can be presented in a simple manner, such as by visually displaying the correlation and structure of information, and an appropriate search result can be obtained with a small number of searches (almost once).

Further, facts and structures such as a certain pattern, a trend (trend), a model, and a correlation between data can be grasped from the retrieved information such as the status of an approach to each company. That is, data mining can be performed. In this example, the search result based on the search target word is displayed in a multidimensional manner. However, the present invention is not limited to this. All the information stored in the storage unit 2b is displayed in a multidimensional manner, The result extracted by only the combination may be used, or the result extracted by the combination of the attributes may be used in combination with another search method such as a keyword search. Also,
In this example, an attribute value is set for each dimension axis. However, an attribute may be specified, and an attribute value may be extracted from target information and used. For example, instead of a specific company name (attribute value) on the x-axis,
"Organization name" is set, and the organization names (domains at the third level or higher) included in the URL of the target information are extracted, and some of the top companies in the order of the number of hits are [Y company, X company, Q company, Others]. <Embodiment 2> Next, a description will be given of an example in which a site map of a domain, such as a company that is engaged in the technology, for each search target word is shown for each category. This example is the first embodiment described above.
The method for displaying the keyword after the keyword search is different from the method for selecting the URL after the keyword search, and the other configurations are the same. The description of the same configuration as that of the first embodiment is partially omitted. FIG. 4 is a display example of a search condition input screen. For example, “printer” is set as a search target word (keyword), and the upper and lower sides of the space, that is, y are set.
Set the "content" attribute on the axis (dimension axis),
That is, the organization names “xxx”, “yyy”, and “z” are displayed on the x-axis (dimension axis).
zz ”is set, and before and after the space, that is, on the z axis (dimension axis),
"Sort by update date" was set so that new items are displayed in the foreground and old ones are displayed in the back. When the information stored in the storage unit 2b is searched under these conditions, first, all information including "printer" as a keyword is extracted. Next, a directory name and a file name as attribute values of “contents” are extracted from the URL of the extracted information. In this example, product, papers, ar
title, rd, std, news, report,
price etc. were extracted. And the attribute value of the x axis and z
Combination of axis attribute values "xxx, product", "y"
yy, product "," zzz, product "
"Xxx, papers", "yyy, papers"
And search is performed, and the update date of the extracted page is set as the y-axis value, and the search result is displayed in a multidimensional manner. FIG. 5 is an example of this display. Actually, different colors are assigned to each site. In addition, U
The RL is not displayed as it is, but in the form of site name (organization name) / category (content).
The figure shows an example where attention is paid to the papers category. Above and below the product, article, r
Although the d and std categories are also displayed, in order to browse other categories as well, the category of interest is switched, and the target category is searched by working through the three-dimensional space (switching the viewpoint).

When the attribute of the search result is set as set on the z-axis, the attribute may be switched and displayed. For example, the attribute of the y-axis may be switched from the sorting by the update date to the number of links or the number of links to be displayed, and the portal degree or the popularity degree may be evaluated. If a desired category is found, by selecting this "site name / category", the presentation means 3b displays the URL
Then, it connects to the Web server 1 via the Internet based on, acquires the information of the target Web page, and displays it on the monitor 3c. As described above, in the present embodiment, the search results are displayed from different viewpoints, and the configuration of the category (site map) relating to the target site can be grasped in detail, thereby improving the searchability.

In particular, a large amount of information can be displayed from various viewpoints, and an appropriate search result can be obtained with a small number of searches (almost once).

Further, by displaying the retrieved information in a multi-dimensional manner, facts and structures such as a certain pattern, a trend (trend), a model and a correlation between data can be grasped from the information. That is, data mining can be performed. <Other Embodiments> In the above embodiments, Web pages on the Internet have been described. However, the present invention is not limited to this, and may be applied to any network such as an intranet that can specify information by URL. Further, it is desirable that the URL is previously extracted by the keyword extracting device as described above, but the present invention is not limited to this, and the Web search device may directly search the Web server. Furthermore, in the above-described embodiment, the search result is shown in three dimensions, but the present invention is not limited to this, and two-dimensional (for example, one using two of the above-described x, y, and z axes), Dimensions (for example, dimensions obtained by expressing the number of hits, the update date, and the like in terms of color shading, size, and the like, in addition to the three dimensions indicated by the x, y, and z axes described above) may be used. It should be noted that the Web search device of the present invention is not limited to the illustrated example described above, and it goes without saying that various changes can be made without departing from the spirit of the present invention.

[0048]

As described above, according to the present invention,
By performing a search using a plurality of attributes and presenting the search results in a multi-dimensional manner, it is possible to provide a Web search device capable of grasping trends and correlations of the searched information and performing data mining.

Further, by performing a search using the attribute of the URL, it is possible to perform a search quickly and accurately.
b A search device can be provided.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a Web search using a search device of the present invention.

FIG. 2 is a diagram illustrating an example of a screen for inputting search conditions according to the first embodiment.

FIG. 3 is a diagram illustrating an example in which search results of the first embodiment are displayed in a multi-dimensional manner.

FIG. 4 is a diagram illustrating an example of a screen for inputting search conditions according to the second embodiment.

FIG. 5 is a diagram showing an example in which search results of the second embodiment are displayed in a multidimensional manner.

[Explanation of symbols]

 Reference Signs List 1 Web server 2 Keyword extraction device 2a Keyword extraction device 2b Storage device 3 Search device 3a Search device 3b Presentation device 3c Monitor

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Keizo Uchiyama 1-3-1 Uchisaiwaicho, Chiyoda-ku, Tokyo Tokyo Electric Power Company (72) Inventor Itzuko Tsuzuka 1-1-3 Uchisaiwaicho, Chiyoda-ku, Tokyo East Inside Keio Electric Power Company (72) Inventor Akio Yasuda 6-14-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo Shinjuku Green Tower Building 19F Inside the Peace Information Center Co., Ltd. (72) Misa Onuma 6-14 Nishi-Shinjuku, Shinjuku-ku, Tokyo -1 Shinjuku Green Tower Building 19F Peace Information Center Co., Ltd. (72) Inventor Naoko Makita 6-14-1, Nishi Shinjuku, Shinjuku-ku, Tokyo Shinjuku Green Tower Building 19F Peace Information Center Co., Ltd. F-term (reference) 5B075 NK02 NK31 NK46 PP22 PQ02

Claims (6)

[Claims] A search unit that searches for information of a Web page to be searched using a plurality of attributes as search conditions; and a presentation unit that presents the search result in a multi-dimensional manner in accordance with the value of the attribute. Web search device characterized by the following. 2. The method according to claim 1, wherein the attribute is a URL (Uniform Reso
2. The Web search device according to claim 1, wherein at least one of a directory name, a file name, an organization name, an organization address, an organization classification, and an organization location, which are attributes of the urce locator, is used. 3. The Web search device according to claim 1, wherein at least one of a keyword, tag information, a link destination URL, a link source URL, a link name, and an update date is used as the attribute. . 4. The method according to claim 1, wherein the search unit searches for a Web to be searched.
4. The Web search device according to claim 1, wherein a keyword search is performed on page information, and a search using the attribute is performed on a result of the keyword search. 5. The UR of information selected by a searcher among the search results shown in the multidimensional,
5. The Web search device according to claim 1, wherein L is presented. 6. 6. The method according to claim 1, wherein the presenting means presents a search result using the attribute by using a part of the attribute, and enables the attribute used for the presentation to be arbitrarily switched. Web search device according to any one of the above.