KR20050078655A - Dynamic keyword extraction and processing system - Google Patents

Abstract

When the user navigates and searches the Internet, the present invention prepares the search word and related website information which is expected to be searched or navigated by the user in advance based on the information of the website currently being searched or visited. It is intended to facilitate further searching and retrieval of related information. Specifically, it extracts the keyword you are currently browsing based on the URL you entered and the URL of the website you're visiting, and refines it to produce related words that are more likely to be searched or navigated by you in the future. And preparing a list of related sites in advance and preparing and displaying them for immediate delivery when the user needs navigation to another site or additional search results. Therefore, the extraction and processing of the keyword and the display of the result are processing systems that simultaneously process and prepare dynamically in parallel with the user's navigation. The present invention relates to a method of extracting a keyword, a method of processing a keyword, and a pre-associated word using a keyword. A method for preparing a list of web sites and executing a search in advance to display such keywords and related words, related sites, search results for the related keywords and related words when necessary, and a system for performing the search.

Description

Dynamic Keyword Extraction and Processing System

In the existing web browser (S230) or the toolbar, when the user inputs a string such as a keyword or a URL in the address bar, only the connection or search for the corresponding site is supported. However, the URL input by the user or the web browser S230 is not used for a new key search or connection to a website by automatically processing the URL string displayed in the address window when the website is accessed. That is, in the conventional method, the URL or keyword inputted by the user was used only for accessing the website, and the URL displayed as a result of accessing the site was only used to indicate the address of the site, and was not reprocessed. The present invention extracts a keyword from the URL or keyword input by the user and the URL output in the address window when the web browser (S230) is connected to the website, and processes and utilizes the keyword to the related site expected in the future. It provides a way to make navigation and search easier.

In the present invention, a string such as a URL or a keyword that a user inputs in a web browser address window S210 to access a website, and a URL of the corresponding site displayed in the address window while the web browser S230 is connected to the website. We focused on the fact that the user has important information related to the information he is looking for. Typically, a domain name or URL will consist of words that are most relevant to what the site offers. Therefore, by processing URLs that are entered or displayed in the address bar, keywords can be extracted in reverse, and by reprocessing the extracted keywords in various forms, users can prepare in advance the search words that users will use or the related sites that they expect to navigate. have. The present invention proposes a method of extracting a keyword from a URL, a method of reprocessing the extracted keyword, and a method of enabling a user to use it for future navigation or additional search.

The present invention has been conceived in that the string constituting the URL is selected from among the names that most reflect the characteristics of the website or from the most relevant words. Looking at the structure of URL, four domains are connected and are as follows.

The top level domain of a URL contains information about the nature of the domain (military, company, education, country, etc.).

-Second level domains contain information of personality (group, company, network, school, etc.).

3rd level domain contains organization name and site keyword

4th level domain contains information about the name of the server

time

URL of the National Assembly: www.assembly.go.kr

3rd level domain: assembly

Second level domain: go

Top-Level Domains:

In this case, the third-level domain 'assembly' is extracted as a keyword, and the second-level domain 'go' represents the category of the website of the government agency, and the top-level domain 'kr' represents the country of Korea. The extracted keyword is assembly and there are 'parliament' as translation words. There are similar words 'parliament' and 'congress' and related words are 'parliamentary record', 'parliamentary schedule', 'parliamentary member', 'general election', 'legislature'. Search results of related sites through search engines, such as 'National Library of Congress', 'History of National Assembly', 'Introduction of the National Assembly's Chairman', and 'National Assembly Budget Monitor' Can be.

Also, for www.yahoo.com, the top-level domain 'com' represents the nature of the domain, and because it's not a country domain, there's no second-level domain, and the third-level domain is' yahoo ', so the keyword is a third-level domain' yahoo 'can be extracted.

In addition, when using the keyword service, a user may input the keyword itself into the address window from the beginning. In this case, the keyword is a keyword input by the user and the keyword of the user is displayed in the keyword display window S220 as it is.

In some cases, it is difficult to think that the keywords are properly extracted from the URL. In other words, when constructing URLs or domain names, the purpose is to keep them as short as possible, or to preempt good words so that the URLs are made of suboptimal words, circumventive words, or extremely unrelated strings. There may be. Two complementary methods can be used for this.

First, for abbreviations or abbreviated words, it is possible to store keywords that can be expected from abbreviations and abbreviated strings separately from the dictionary DB to compensate for the difficulties in searching the dictionary DB when the strings come. Can be used.

Second, a method of extracting a keyword from a name and a comment on a corresponding URL output from a search engine by transferring a URL input or displayed in a web browser address window S210 to a specific search engine as it is. According to this method, keyword extraction accuracy can be improved even for sites that use URLs or domain names with strings irrelevant to the nature of the web site.

In order to implement the above two complementary measures, the definitions of these strings should be prepared in the database so that the system can expand them from the abbreviations or abbreviations to the keywords. You need a program module that sends and retrieves information from that site.

An object of the present invention is to extract a keyword from a function that was not provided in the existing web browser (S230) or the toolbar, that is, a string input by the user or a web page currently being visited, and at the same time using the extracted keyword By providing a word or URL that is expected to be searched in the future, to make the user easier to navigate or search to the associated site that is expected in the future. According to the present invention, there is provided a method of extracting a keyword from a URL input by a user, a method of extracting a keyword from a keyword or an input string input by the user, and a web of the user as a result of the URL or keyword input by the user. How to extract keywords from the URL displayed in the browser or Internet toolbar, i.e. from the displayed URL, Extract keywords from the Title, a field in the Meta Tags of the web page the user is currently visiting There is a way.

Extracting keywords from the above URLs is excluded from the claims of the present invention. However, in the present invention, the keyword input by the user is already a keyword by itself, and extracts the keyword from a title, which is a field of the meta tag of the web page the user is currently visiting, and directly inputs the user's input string, Alternatively, the title string extracted from the title of the web page currently being visited is obtained as a keyword. The user's input string is used as a keyword in itself and does not require a keyword extraction process. For the input string entered by the user for search or navigation, especially in Korean, the keywords can be further refined through morphological analysis, except for surveys and suffixes. Suffixes or probes also often have a new meaning as a whole string. Therefore, the present invention uses the input string as it is or 'as is' as an keyword without undergoing a separate morphological analysis or a separate processing process for the input string. In other words, the input string that the user enters "Patent Office" or "Blue House" or "National Assembly" as a web browser address bar or Internet toolbar is a keyword in itself and is a related word that the user is likely to further search in the future. It is used to provide them or related websites.

In the HTML document, each field of Meta Tags, that is, title, description, keyword, etc., is used by the search engine to classify the document category and search group separately from the body of the document. . Web page authors who consider the classification of search engines will enter appropriate words because Description and Keyword are invisible but have the characteristics of index terms. However, the title is the upper border of the web browser, and in the case of Microsoft Internet Explorer, the title of the document is displayed in a string of a certain length in the blue upper border of the web browser. Because it is the first item to appear in, it will be entered as a string that best describes the content of the web page. However, "Welcome to the homepage" or "Welcome to the homepage", etc., are parts that have little meaning as keywords in terms of extraction and processing. desirable. In addition to characters, symbols and special characters before or after the title or in the middle of a string, such as rhombus, heart symbol, spade symbol, clover symbol, exclamation point (!), Equal sign (=), etc. The same special character or symbol also has no meaning as a keyword. With the exception of these conventional words and special characters or symbols, all remaining strings or words are candidates for keywords.

You don't have to have only one keyword for a particular web page. When the homepage of the National Assembly and the Blue House used as an example of the present invention is removed, the keywords "Korean National Assembly" and "Blue House" are extracted, respectively. In the case of the homepage of the JPO (http://www.kipo.go.kr), if the code is removed from the title of "The 21st century's first-class intellectual property powerhouse realization! The keyword candidate is extracted from the keyword candidate, and the entire title string, that is, the sentence "the 21st century first-class intellectual property realization patent office" is a string composed of compound words, but as such, the entire string of the title can be used as a keyword.

Words that are most likely to be used by the user for future searches or Internet navigation from the user's direct input or keywords obtained by the above method are retrieved from "dictionary DB" and "related word DB" and displayed on the user's web browser or toolbar. Presented. From the dictionary DB, synonyms, antonyms, translation words, and translation strings for the entered words or strings are searched and displayed in the user's web browser or toolbar. In the related word DB, a list of words related to a keyword obtained from a keyword or a title by a user's direct input is searched and displayed. The related word DB is a database in which one or more related keywords are stored in pairs for a particular keyword, as the purpose of the present invention. Dictionary DB is also a database of synonyms, antonyms, and translation words for keywords.

The system proposed by the present invention for the keyword extraction and processing and expansion to the related word and related site list is composed of the following modules.

(i) a program module for receiving a character string input or displayed in a web browser address bar or a toolbar, the module displaying a keyword from a user's character string input or a title of a corresponding web site in a web browser address window S210; S10),

(ii) a program module for separating input strings and extracting keywords, comprising: a module for extracting keywords from a user input string or a title of a visiting web page or a third level domain of a URL of a visiting web page; , (S20)

(iii) a program module for processing the extracted keywords, wherein the module finds similar words, synonyms, translation words, and related words for the keyword (S30).

(iv) a database for processing keywords and a module for searching related words with keywords, the module for searching and providing dictionaries, synonyms, synonyms, and related words in a dictionary database or a related word database of a specific server or an external server; (S40)

(v) a user interface program module for defining a method of processing and displaying a keyword, wherein the module provides a selection of whether to display only the keyword or how to process the keyword and display the keyword (S50).

(vi) a program module for executing a search or connection for a string, and a module for executing a web browser S230 in a manner desired by a user

It consists of six modules in total. (S60)

In this case, the location of each execution module may vary depending on how the program is implemented. The entire module and the database listed above can be located on the user's terminal, or the whole can be placed on a specific external server or a part of the user can be located. It can also be deployed in the terminal and distributed to some of the external servers.

The operation of each module listed above is the same as the described order, but the `(v) user interface program module defining the manner in which the keyword is processed and displayed '(S50) is a step in which the user uses the system of the present invention. If there is no user's method definition, it is executed by the system predefined method. The definition of the user's method allows the user to select items to be displayed to the user about the synonyms, antonyms, related words, and translation words for the keyword in response to the user's input. A user may select not to display a word, and a user who needs a translation word may select a language of a translation word for an input word.

In addition, the `(iv) database for processing keywords and program modules for searching related words by keyword '(S40) describes the database and two program modules for searching them together. All of the databases on a particular server are possible. A program module that searches for related words for a keyword is ready to search a search engine or a specific database as soon as the keyword is extracted from the URL, to search for words or related sites related to the extracted keyword and display them in the user's toolbar or address bar. do.

The extracted keyword and related words of the keyword can be displayed in a keyword display window (S220), a toolbar or a pop-up window. Ideally, a keyword display window (S220) is created next to an address window to extract extracted keywords and related words. It is preferable to list them and search can take time, so it is desirable to add relevant words or websites to the display as they are searched.

Program module that accepts the strings entered or displayed in the `(i) address bar above in the same way as the user input, even if the URL is displayed in the address bar of the web browser by a visit to the website, not by the user. It is accepted as a character string by the program described in the item '(S10)', where the extraction and processing of keywords, the searching of related words or related web sites, and the display of processed results are repeated according to the user's navigation.

The program module for executing a search or linking to the '(vi) string' above (S60) reduces the procedure and time for visiting the search engine while the user is navigating, and the user is able to adapt the system and method proposed by the present invention. As a result, the processed result displayed in the keyword display window S220, the toolbar, or the pop-up window can be selected to immediately search or immediately open the web site using the corresponding keyword.

In the present invention, the user inputs a keyword or URL every time in the web browser address window (S210) in order to connect to the website, and also re-enters the URL from the beginning when the connected site is not a desired site or needs to be moved to another related site. It also reduces the inconvenience of having to search related sites from search engines and processes and presents keywords and related keywords from the sites that users have visited once. It can pre-search for highly probable websites and prepare related words in advance on the user's device. This saves time and effort for URL input, repetitive search and navigation.

1 shows each module and role of the system of the present invention

2 shows an example of an execution screen

3 is a flowchart showing a keyword extraction and processing flow.

4 is a block diagram of a system of the present invention.

S10, S20, S30, S40, S50, S60: Each module constituting the system.

Claims (10)

When you navigate and search the Internet, you can prepare additional search and navigation on the current page by preparing in advance the search words and related website information that you are expected to search or navigate based on the information of the website you are currently browsing or visiting. In dynamic keyword extraction and processing to facilitate retrieval of relevant information, Extracting a keyword from the input string when the user inputs the keyword in an address bar or a toolbar of the web browser; Searching for a synonym, antonym, a related word, and a translation word in the extracted keyword in a dictionary DB or a related word DB; Displaying the retrieved synonyms, antonyms, related words, and translated words in a user's web browser or toolbar; Dynamic keyword extraction and processing system comprising a. As a result of the website connection or navigation, the keyword is extracted from the URL string of the currently visited website displayed in the address bar of the web browser (S230) or the toolbar and displayed in the keyword display window or the toolbar as it is, or in a database and a search engine. Dynamic keyword extraction and processing system characterized in that by searching through the displayed results displayed in the processed form. The method of claim 1, wherein the dynamic keyword extraction and processing system is A program module for accepting a string input or displayed in an address window; A program module for separating input strings and extracting keywords; A program module for processing the extracted keyword; Dictionary DB and related word DB for processing a keyword; A user interface program module for defining a manner in which a user processes and displays a keyword; A program module for performing further refinement or concatenation of strings displayed in processed form according to a user definition or basic definition; Dynamic keyword extraction and processing system comprising a. The method of claim 3, wherein the arrangement of each program module and database, Located in all of the user's terminal; Located on an external server connected to the Internet; A method of separately separating and positioning the terminal of the user on an external server connected to the Internet; Dynamic keyword extraction and processing system comprising a. The method according to claim 3, If the string displayed in the address bar of the web browser or the toolbar does not have a '.' (Dot), it is determined that the string is a keyword; If there is a '.' (Dot), it is determined as a URL; The character string determined as the URL is classified into a top level domain, a 2nd level domain, a 3rd level domain, and a 4th level domain; Dynamic keyword extraction and processing system characterized in that the extraction of the keywords in the third-level domain, the category information of the local information or website in the second-level domain and the top-level domain. The method of claim 2, wherein the keyword display window or toolbar includes a keyword display window or a toolbar newly added as a plug-in to the web browser in which the user is working, and is a pop-up window displayed separately from the web browser in which the user is working. Dynamic keyword extraction and processing system comprising a. The method of claim 2, wherein the form of processing is displayed by including a translated word of a keyword, antonym of a keyword, a similar word of a keyword, and displaying a name or keyword of a site searched through a search engine as a keyword to a user. The category information extracted from each level domain of the string is also included in the target to be processed and displayed, and the displayed string is displayed as a pop-up menu or a toolbar menu to perform a function of searching and opening. Dynamic keyword extraction, including the ability to display new searches or searches using strings, and the ability for users to add, delete, or correct words of their choice to the processed string. Processing system. The search engine of claim 7 includes utilizing a DB in a user terminal itself, a DB for a specific search engine, or a DB of a specific server, and when searching the user in a processed form, the search result of the search engine is intact. Or processing and displaying, including a method of immediately performing a search at the stage of keyword extraction and processing in parallel with the navigation of the user's website, even if the user does not search for the keyword. A dynamic keyword extraction and processing system comprising a method of preparing a list of pre-search results for a keyword immediately when a user clicks on the keyword display window or toolbar. In the method of extracting a keyword, when it is difficult to extract the keyword only by the third-level domain string constituting the URL string, (i) a method of extracting a keyword for the string of the URL using the abbreviation DB and the abbreviation DB; ii) extracting words that make up the meta title string as keywords if the meta title is present in the HTML source of the website page the user is visiting. (iii) using the most frequently used words as keywords in the website page the user is visiting. Extraction method (iv) Dynamic keyword extraction and processing system comprising the method of extracting a keyword from the name and comment for the URL by sending the entire URL string entered or displayed in the web browser address bar to the search engine . The extraction and processing of keywords and the search and collection of search results with the words associated with the keywords are processed in parallel with the user's web browser or Internet browsing tools. It includes a method of storing a temporary storage directory, a file, a DB, and the like before execution, and displaying the prepared data by searching in advance when the user needs a search result, a related word, or a related site list. Dynamic keyword extraction and processing system.