KR20020062464A - A method of searching the Web using meta search engines and the system thereof - Google Patents

Abstract

PURPOSE: A system and a method for searching the Internet information using a meta search engine are provided to output only the web pages much related to the personal peculiarities of a user, to output the same search word in the different arrangement order according to the personal peculiarities of the user, and to reduce outstandingly the range of the search result by filtering and outputting only the web pages related to the user. CONSTITUTION: The meta search engine mounted to a user computer receives the search word from the user(S301). The meta search engine extracts equivalent, thesaurus, antonym, and non-thesaurus of the search words by referring to a user search DB(S302) and creates the first, second, and third search formula(S303). The meta search engine requests the searching concerned to the first, second, and third search formula to several remote search engines through multi-threading(S304). The several remote search engines transfer the search result to the user computer(S305). The user computer arranges the search result in the user-related order(S306) and outputs the search result to the user(S307). The meta search engine adds the user-related equivalent, thesaurus, antonym, and non-thesaurus to the user search DB(S308).

Description

A method of searching the Web using meta search engines and the system

The present invention relates to an information search method and a meta search engine system on the Internet using a meta search engine. More specifically, the present invention relates to a user's search terms inputted in consideration of personal characteristics of the user such as occupation, age, and interests. The present invention relates to an information search method and a meta search engine system on the Internet using a meta search engine that first outputs a relevant web page.

Internet users can obtain the required information through many websites residing on the network. To do this, the user needs to know the address of a website that can provide the information they require, but in reality it is impossible to know a large number of addresses. Therefore, many search engines are currently provided to enable a simple operation such as entering only an index word to find a desired website (or web document). Currently, search engines can be classified into directory search engines, robot search engines, and meta search engines according to their types. Directory search engines collect and process documents manually to build a database and supply collected data in response to user queries. Robot search engines are a way that software search robots collect and process web documents. Currently, almost all search engines belong to this format. The meta search engine does not have its own search engine module, and provides searched data by using existing search engines.

In general, when a user enters a search word and then executes a search command, addresses of a plurality of websites (or web documents) are provided to the user. In some cases, a short description may be provided along with the address, and the user may view the contents and attempt to access the corresponding site. In addition, the user attempts to connect to the searched site until the user obtains satisfactory information. The user may waste a lot of time obtaining the information required in this process.

On the other hand, even in the case of inputting the same search word, the content of the information to be originally searched may differ for each person according to the user's professional characteristics, desired expertise or taste. For example, using the index "housing" as an example, if the user is an architect, a prospective buyer, a businesswomen, or a real estate agent, the contents of the information required by each person will be different. However, when the same index word is input, almost the same search data is provided regardless of the user's characteristics and personality. This is also an inefficient factor for search parties.

The present invention has been made to solve the above problems, and provides a method of searching for information on the Internet using a meta search engine that preferentially outputs only web pages highly relevant to the user in consideration of personal characteristics such as the user's job, age, and interests. It aims to do it.

Another object of the present invention is to provide an information retrieval method on the Internet using a meta search engine that outputs the same search word in a different arrangement order according to a user's personal characteristics.

Another object of the present invention is to provide an information retrieval method on the Internet using a meta search engine that significantly reduces the range of search results compared to a conventional search engine by filtering and outputting only web pages that are closely related to a user.

1 is a system configuration diagram of the first embodiment of the present invention.

2 is a diagram showing an example of the configuration of a user search word DB according to the present invention;

3 is a flowchart showing operation of the first embodiment of the present invention.

4 is a schematic diagram illustrating a process of sorting search results according to the present invention.

Fig. 5 is a flowchart showing the operation of the second embodiment of the present invention.

6 is a system configuration diagram of the third embodiment of the present invention.

* Description of the symbols for the main parts of the drawings *

110, 120, 130: search engine server

170: web server

140, 150, 160: user's computer

In order to achieve the above object, an information search method on the Internet using the meta search engine system according to the first aspect of the present invention comprises the steps of receiving one or more search terms from a user, the idiosyncratic word that is very closely related to the search terms, From the search term, the search condition is further referred to by referring to a user search word DB that stores data about a related word that is somewhat related to the search word, an opposite word that is not related to the search word, and a non-related word that is not related to the search word. Generating a plurality of additional search expressions, transmitting the plurality of search expressions to a plurality of robot search engine servers, and receiving search results for the plurality of search expressions from the plurality of robot search engine servers. Arranging the list in a predetermined rank, and outputting search results arranged in the predetermined rank to a user And monitoring which web page the user has stayed in for a long time among the search results, and modifying the user search word DB using the information, wherein the user search word DB is stored in the computer of the user. Stored.

An information retrieval method on the Internet using the meta search engine system according to the second aspect of the present invention includes the steps of: a user connecting to a web server through a user computer, logging in, transmitting a search word from the user computer to a web server; , Data stored about a codfish that the web server is very closely related to the search word, a related word that is somewhat related to the search word, an opposite word that is not related to the search word at all, and an unrelated word that is not generally related to the search word. Generating a plurality of search expressions further including a search condition from the search word by referring to a user search word DB; transmitting, by a web server, the plurality of search expressions to a plurality of robot search engine servers; Receives search results for the plurality of search expressions from the plurality of robot search engine servers, and Arranging by rank; transmitting, by the web server, the search results arranged in the predetermined rank to the user computer; and, by the web server, monitoring which web page the user has stayed in for a long time among the search results; Modifying the user search word DB using the information, wherein the user search word DB is stored in the web server.

The information search method on the Internet using the meta search engine system according to the third aspect of the present invention includes the steps of a user connecting to a web server through a wireless internet terminal, transmitting a search word to the web server from the wireless terminal, User search terms in which data is stored on the server for which the server is very closely related to the search term, related words to some extent related to the search word, opposite words not related to the search word at all, and unrelated words generally unrelated to the search word. Generating a plurality of search expressions further including a search condition from the search word by referring to a DB; transmitting, by a web server, the plurality of search expressions to a plurality of robot search engine servers; Receive a search result for the plurality of search expressions from a plurality of robot search engine server, the predetermined rank Opening the web server, transmitting the search results arranged in the predetermined order to the wireless Internet terminal, and the web server monitoring which web page the user stayed in the search results for a long time; And modifying the user search word DB using information, wherein the user search word DB is stored in the web server.

First, prior to describing the present invention in detail, the terms used in the present specification, such as counterparts, related words, unrelated words, and opposite words will be described.

The parallel word refers to a word that is determined to be closely related to the word that the user wants to search.

A related word refers to a word that is determined to be closely related to a word to be searched by a user.

The opposite word refers to a word that is determined to be completely unrelated to a word to be searched by a user.

An unrelated word is a word that is determined to be largely unrelated to a word to be searched by a user.

In general, the match word and the related word will be the words that appear frequently on the web pages that the user wants to find among the search results for the search term, and the opposite and unrelated words are the web pages that are not related to the user at all in the search results. Will be a word that often appears in. There may be a plurality of parallel, related, opposite, and unrelated words for a search word.

Which words will be matched, relative, irrelevant, or inverse of a search term depends on the user's past search history, for example, which site the user has been in for a long time, and which site has been bookmarked. The decision is made with reference to information such as whether or not it was done. Therefore, a matched word, a related word, a non-related word, or a counter word for a search word is a material reflecting personal characteristics such as a user's job, an interest, an age, and a taste. The method of determining the counterpart, related word, etc. will be described later in detail.

The parallel, related, opposite, and unrelated words for each search term the user has searched are stored in the user search word database on the user's computer. Since the user search word DB is gradually formed and developed according to the search trace of each user, the user search word DB includes different contents according to the occupation, age, field of interest, and gender of each user.

Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

The first embodiment implements the meta search engine system according to the present invention centered on a client computer, and a schematic configuration diagram is shown in FIG.

As shown in Fig. 1, the meta search engine system according to the present invention includes a plurality of search engine servers, each user computer, and a web server. The plurality of search engine servers, each user computer, and the web server 170 are connected through the Internet.

1, only three search engine servers 110, 120, and 130 and three user computers 140, 150, and 160 are shown for convenience.

The plurality of search engine servers 110, 120, and 130 refer to, for example, servers of robot search engines such as Lycos, Yahoo, and AltaVista. These robotic search engines typically have a spider program that navigates to and reads the content of a large number of web pages, a huge indexing program called a catalog from the read pages, and indexes them after receiving a user's search request. It is composed of a program that transmits the search results to the user in comparison with.

The user first accesses the web server 170 and registers as a member, and downloads and installs the meta search engine program according to the present invention on the user computers 140, 150, and 160. The user computer may be a personal computer, a workstation, a notebook computer, or the like. As the program is installed, user search words DBs 141, 151, and 161 are generated in the user computers 140, 150, and 160 for storing information related to the search word searched by the user. The structure of the user search word DBs 141, 151, and 161 is shown in FIG. As shown in FIG. 2, in the user search word DB, a parallel 202, a related 203, an opposite 204, and an unrelated 205 are stored for each search word 201.

The web server 170 keeps the user search terms DB (141, 151, 161) of all the users registered as members, and if the contents of each user search term DB (141, 151, 161) change, the user computer 140, 150, 160 notifies the web server 170 immediately, so that the web server 170 stores the latest contents of each user search word DB (140, 150, 160).

3 is a flowchart illustrating a process of searching for information by a user by the meta search engine system shown in FIG.

As shown in FIG. 3, the meta search engine installed in the user computer first receives a search word from the user (S301).

Subsequently, the equivalent, related, counter, and unrelated words corresponding to the search word are extracted with reference to the user search word DB stored in the user's computer (S302), and first, second and third search expressions are generated therefrom (S303). ).

If the search word entered by the user does not exist in the user search word DB of the user's computer, the meta search engine of the present invention connects to a web server and requests that the matched word, related word, opposite word, and unrelated word corresponding to the search word be transmitted. . Since the web server 170 stores a user search word DB of each user who has registered as a member, it may be checked whether other users than the user have searched the search word. In this way, the web server finds another user who has previously searched for the search word, and transmits the matching, related, inverse, or unrelated word related to the search word stored in the user search word DB of the user to the user computer.

At this time, in searching whether other users have searched the search word, only a specific group of users may be searched, without having to search all users who are members. In other words, depending on which field of words is searched the most, each user can belong to any one group, and the range of users to be searched can be limited to only users belonging to the same group. Which group each user belongs to can be determined by the web server by building a dictionary of the most frequently occurring words from each topic according to a number of topics, such as sports, computers, securities, travel, entertainment, and job search. have.

Primary, secondary, and tertiary search expressions are the processing of user-entered search terms, and the matching, related, opposite, and unrelated terms of the user search word database reflecting personal characteristics such as the user's occupation, age, and interests. In addition to the search condition (search word) input by the user, information about the user is added.

The forms of the first, second and third order search expressions are as follows. (n is a natural number)

<First search expression>

{Searcher} AND {matcher ₁ AND matcher ₂ AND ... matcher _n } AND {matcher ₁ OR matcher ₂ OR ... matcher _n } AND NOT {counterweight ₁ OR matcher ₂ OR ... matcher _n } AND NOT {unrelated ₁ AND unrelated ₂ AND ... unrelated _n }

<Secondary search expression>

{Query} AND {comparable word 1 and word 2 equivalent and comparable control .. n} AND NOT {bandaeeo bandaeeo ₁ OR ₂ OR ... _n} bandaeeo

<Third search expression>

{search word}

As in the above equation, the first search expression includes a web page that includes both a search word and a parallel word, one or more related words, no counter words, and zero to n-1 non-related words. To search. In addition, the second search expression is to search for a web page that includes both a search word and a parallel word, the opposite word is not included at all, and the third search expression is for searching a web page including a search word. In the present embodiment, the first, second, and third order search equations are defined as above, but various other methods are possible.

Since the search conditions are more difficult among the 1st, 2nd and 3rd search expressions, if a web page that satisfies the 1st search expression is found, it is very likely that the web page contains the information the user wanted to find. high.

Since the meta search engine according to the present invention outputs to the user in the order of the meta search result by the first search expression, the meta search result by the second search expression, and the meta search result by the third search expression, the user is irrelevant. You don't have to visit a lot of sites, just find the information you need by visiting the sites that are the top priority.

In addition, in the conventional search engine, the search results entered by the user are large enough to be millions, but, as in the present invention, after processing the search query to further include a search condition suitable for each user characteristic, As a result, the search results are significantly reduced beyond comparison. Thus, the user can search for information very efficiently and quickly.

When the first, second, and third search expressions are generated, the meta search engine according to the present invention is multi-threaded to multiple search engines located at remote locations as shown in FIG. 2. Request a search for the third search formula (S304). Here, multi-threading refers to progressing simultaneously by time division, rather than sequentially performing any task. In FIG. 2, three search engines SE ₁ , SE ₂ , and SE ₃ are requested to be searched for convenience.

The various search engine servers transmit the search results for each of the first, second and third search expressions to the user's computer (S305). At this time, several search engine servers may transmit search results arranged in order of the most relevant information to the user's computer by their ranking algorithm.

The user computer arranges the search results transmitted from the various search engine servers in a certain order from the ones that appear to be highly relevant to the user (S306) and outputs them to the user (S307).

The user will visit the web pages listed earlier in the search results displayed on the screen of the user's computer, and find the desired information.

The meta search engine according to the present invention monitors which site the user has visited and how long the user stayed in the search results, and further stores the matching, related, opposite and irrelevant words in the user search query database using such information. In operation S308, the display returns to a mode in which a new search word may be input again (S301).

Hereinafter, a process of extracting a matched word, a related word, an opposite word, an unrelated word for a search term by using information about which site the user visited and how long the user stayed in the search result (S308 of FIG. 3). It demonstrates concretely.

For example, a web page numbered 1, 2, 3, 4, 5, in this order, displays the screen of the user's computer along with the site title, URL (Universal Resource Locator), and a brief description. Let's say that the output is

The meta-search engine according to the present invention did the user click on the URL of each web page, how long did the user stay if clicked, added the web page to favorites, output the web page to the printer, and the contents of the web page? Whether it is stored in the storage device or block copy of the contents of the web page.

Here, the bookmark refers to a menu in a Microsoft Internet Explorer browser, but is not limited thereto, and includes a menu of another browser corresponding thereto.

If a user clicks on a web page for a long time, adds it to a bookmark, saves it, etc., this means that the web page is a properly searched site according to the user's intention.

Therefore, based on these criteria, each web page determines whether the user wants to pass or fail.

For example, suppose that only 1 and 5 of 1, 2, 3, 4, and 5 sites are passed, and 3 words of each webpage appear at a frequency above a certain value for convenience (for each word A ~ I). . Here, the word means a noun that determines the nature of the content of a web page, such as 'sports', 'travel', 'webmaster', 'language', 'computer', 'go', etc. Nouns that do not affect the nature of the content of the web page, such as 'below' or 'first', or nouns that have suffixes of verbs such as 'goodness' are excluded.

Site 1 (passed): A, B, C

Site 2 (Failed): A, G, H

Site 3 (Failed): A, C, G

Site 4 (Failed): B, G, I

Site 5 (passed): A, D, E

In the above example, if you look for a match, relative, antonym, or unrelated term for a search term:

First, the words, A, B, C, A, D, and E, which appear at a predetermined frequency or more, may be determined to be closely related to the search word. Among these, A, which is a word that overlaps a certain frequency or more, is determined to be particularly closely related to the search word, so A may be a counterpart. The related words are words A, B, C, A, D, and E except for A, which is the counterpart, B, C, D, and E.

On the other hand, in the failed web page, words A, G, H, C, B, I that have a certain frequency or more, except for the counterpart and the related words, G, H, I can be determined to be largely unrelated to the search word. have. Among them, G, which is a word appearing overlapping more than a certain frequency, may be determined to have no relation with the search word, and thus G may be the opposite word. In addition, the unrelated words are words except for G, which are opposite words among the above words, G, H, and I, and H and I.

The extracted equivalent, related, opposite, and unrelated words are added to the corresponding item in the user search word DB. In this way, the user search word DB is gradually developed to reflect the personal characteristics of the user as the user searches.

Hereinafter, a method (S306 of FIG. 3) in which the meta search engine according to the present invention receives a search result from various search engine servers, that is, a list of web pages and arranges them in a predetermined order will be described in detail.

The search results received from the various search engine servers by the meta search engine according to the present invention are as shown in FIG. That is, from the SE ₁ search engine server 110, a search result for the first search expression (R ₁₁ ), a search result for the second search expression (R ₂₁ ), and a search result for the third search expression (R ₃₁ ). Received from the SE ₂ search engine server 120, the search results for the first search expression (R ₁₂ ), the search results for the second search expression (R ₂₂ ), the search results for the third search expression (R ₃₂ ) Received from the SE ₃ search engine server 130, the search results for the first search expression (R ₁₃ ), the search results for the second search expression (R ₂₃ ), the search results for the third search expression (R ₃₃ ) Is sent.

The meta search engine according to the present invention sorts the search results (R ₁₁ , R ₁₂ , R ₁₃ ) to generate a search result (R ₁ ), and sorts the search results (R ₂₁ , R ₂₂ , R ₂₃ ) to search results. generating a (R _2), and the search result _{(R 31, R 32, R} 33) that generated an alignment search result (R ₃₎ next, a search result _{(R 1, R 2, R} 3) by rearranging the The final search result R is displayed on the screen of the user's computer.

Hereinafter, a method of generating a search result R ₁ by _first sorting the search results R ₁₁ , R ₁₂ , and R ₁₃ for the first search expression will be described.

For example, suppose that the URL list for the first search expression is received from the three search engine servers 110, 120, and 130 shown in FIG. 1 as follows.

From search engine server 110: URL ₁ , URL ₂ , URL ₃ , URL ₄

From search engine server 120: URL ₅ , URL ₁ , URL ₆ , URL ₃

From search engine server 130: URL ₁ , URL ₇ , URL ₅ , URL ₈

The above URL ₁ to URL ₈ are arranged according to the following criteria, and the following first, second and third criteria are applied in order.

First, sort by the frequency of URLs.

In the example above, URL ₁ appears three times, URL ₃ , URL ₅ twice, URL ₂ , URL ₄ , URL ₆ , URL ₇ , and URL ₈ appear once. Therefore, the alignment result applying the primary criterion is as follows.

URL _1- > URL ₃ , URL _5- > URL ₂ , URL ₄ , URL ₆ , URL ₇ , URL ₈

Second, if the URLs appear the same, they are sorted according to search engine priorities.

If the search engines are prioritized in the order of SE ₁ , SE ₂ , and SE ₃ , the sorting result using the secondary criteria is as follows.

URL _1- > URL ₃ , URL _5- > URL ₂ , URL _4- > URL _6- > URL ₇ , URL ₈

Third, in the same search engine, the higher priority is given priority.

For example, according to the SE ₃ search engine, since URL ₇ takes precedence over URL ₈ , URL ₇ is arranged before URL ₈ . The sorting result applying the cubic criteria is as follows.

URL _1- > URL _3- > URL _5- > URL _2- > URL _4- > URL _6- > URL _7- > URL ₈

Sorting the search results (R ₂₁ , R ₂₂ , R ₂₃ ) of FIG. 3 to generate a search result (R ₂ ), sorting the search results (R ₃₁ , R ₃₂ , R ₃₃ ) to search results (R ₃ ) It is also generated by the above criteria.

Next, the process of sorting the search results (R ₁ , R ₂ , R ₃ ) to generate the final search result (R) will be described. The search results (R ₁ , R ₂ , R ₃ ) are overlapping parts of each other. since this may be, the final search results of the R ₁ and the first output, next, the remainder obtained by subtracting the portion is the same as R ₁ and the output from R _2, which finally overlaps with R _1, R ₂ from R ₃ The rest of the output is subtracted.

Hereinafter, a second embodiment of the present invention will be described.

In the second embodiment, the meta search engine system according to the present invention is implemented around a server computer. The difference from the first embodiment is that when a user requests a search for a search word after accessing a web server through a user computer, the web server performs all the processes in steps S302 to S306 of FIG. Is that. In addition, unlike the first embodiment, the user search word DB of each user who has subscribed as a member is not stored in each user's own computer, but only in the server computer.

The process in the second embodiment will be described in detail with reference to FIG.

The user connects to the web server (S501) and, if a member, logs in by inputting a member ID and password (S503). If it is not a member registration state, by first pressing the member registration button, and enters personal information such as name, member ID, address, telephone number, and completes the membership registration (S502).

After the user logs in, the user inputs a search word and clicks a search button (S504). This search term is transmitted to a web server (S505), the web server generates first, second and third search expressions by referring to the user search word DB of the user, and then sends the generated search expressions to various robot search engine servers. Request a search. When the web server receives the search results from the robot search engine server, the search results are determined based on the search results, and then the web page including the sorted search results is transmitted to the user computer (S506).

The web server then monitors for events such as which URL in the search results the user clicked, how long did they stay if clicked, added the web page to favorites, output the web page to the printer, and so on. Upon termination of the web browser associated with the search result, the content of the event is received. Using this content, the web server extracts the parallel, related, opposite and unrelated words by the above-described process, and adds them to the corresponding item in the user search word DB of the user (S507).

Hereinafter, a third embodiment of the present invention will be described.

The third embodiment implements the meta search engine system according to the present invention in a wireless Internet terminal environment in which a WAP browser such as a mobile phone and a smart phone is embedded. Also in the third embodiment, as in the second embodiment, the user search word DB that stores the information about the search word is not stored in the terminal, but is stored and managed only in the web server.

6 is a schematic system configuration diagram of the third embodiment. As shown in FIG. 6, an element called a WAP gateway 603 exists between the wireless terminal 601 and the web server 602 to enable communication between a user device and a web server supporting WAP.

The operation of the third embodiment will be briefly described below.

A user connects to a web server using a wireless terminal. At this time, since the WAP gateway has user information for each telephone number of the wireless terminal, the user's login process is unnecessary. The web server transmits a screen for inputting a search word to the user's terminal. When the user inputs a search word on the screen of the wireless terminal and presses the send button, the search word is transmitted to the web server. The web server generates first, second and third search expressions by referring to the user search word DB of the user stored therein, and performs a meta search using various robot search engines.

Since the third embodiment is also focused on the web server, the process of sorting the search results after the meta search, and monitoring the user's behavior to extract the equivalent, relative, opposite, and unrelated words and add them to the user search query DB is necessary. The procedure is substantially the same as in the second embodiment.

Although the present invention has been described using the examples, the scope of the present invention is not limited to the specific embodiments, and should be interpreted by the appended claims.

According to the meta search engine of the present invention, only web pages highly relevant to the user are preferentially output in consideration of personal characteristics such as the user's job, age, and interests, and the arrangement order of the same search word according to the user's personal characteristics. Since it outputs differently, there is an effect that allows the user to search information very efficiently and quickly. In addition, since the search is processed after the search to further include the search conditions for the individual user characteristics, the search results are significantly reduced compared to the conventional search engine.

Claims (10)

In the information retrieval method on the Internet using a meta search engine, Receiving one or more search terms from a user; A user search query DB that stores data about a co-ordinated word that is very closely related to the search word, a related word that is somewhat related to the search word, an opposite word that is not related to the search word at all, and an unrelated word that is not generally related to the search word. Generating a plurality of search expressions further adding a search condition from the search word; Transmitting the plurality of search expressions to a plurality of robot search engine servers; Receiving search results for the plurality of search expressions from the plurality of robot search engine servers and arranging them in a predetermined order; Outputting search results arranged in the predetermined order to a user; Monitoring which web page the user has stayed in for a long time among the search results, and modifying the user search word DB using this information; The user search word DB is an information search method on the Internet using a meta search engine stored in the user's computer. In the information retrieval method on the Internet using a meta search engine, A user accessing a web server through a user computer and logging in; Transmitting a search word from the user computer to a web server; A user whose data is stored on a web server that is closely related to the search term, a related term that is somewhat related to the search term, an opposite term that is not related to the search term at all, and an unrelated term that is not generally related to the search term. Generating a plurality of search expressions further adding a search condition from the search word with reference to the search word DB; Transmitting, by a web server, the plurality of search expressions to a plurality of robot search engine servers; Receiving, by a web server, search results for the plurality of search expressions from the plurality of robot search engine servers and arranging the search results in a predetermined order; Transmitting, by the web server, the search results arranged at the predetermined rank to the user computer; Monitoring, by the web server, which web page the user has stayed in for a long time among the search results, and modifying the user search word DB using this information; The user search word DB is an information search method on the Internet using a meta search engine stored in the web server. In the information retrieval method on the Internet using a meta search engine, Accessing a web server through a wireless internet terminal; Transmitting a search word to the web server in the wireless terminal; A user whose data is stored on a web server that is closely related to the search term, a related term that is somewhat related to the search term, an opposite term that is not related to the search term at all, and an unrelated term that is not generally related to the search term. Generating a plurality of search expressions further adding a search condition from the search word with reference to the search word DB; Transmitting, by a web server, the plurality of search expressions to a plurality of robot search engine servers; Receiving, by a web server, search results for the plurality of search expressions from the plurality of robot search engine servers and arranging the search results in a predetermined order; Transmitting, by the web server, the search results arranged in the predetermined order to the wireless Internet terminal; Monitoring, by the web server, which web page the user has stayed in for a long time among the search results, and modifying the user search word DB using this information; The user search word DB is an information search method on the Internet using a meta search engine stored in the web server. The method of claim 1, wherein the modifying the user search word DB comprises: Among the search results, the user judges that the web page has passed a web page in which the user has stayed for a predetermined time, added a bookmark, outputted to a printer, saved, or block copied, etc., and other web pages failed. Determining that; Extract the words appearing more than a certain frequency from the passed web pages, and the words appearing overlapping more than a certain frequency among them is determined as the counterpart of the search word, and the user search word DB by associating the matched word with the search word. Adding to; Determining a word other than a word determined as the equivalent among words that appear more than a certain frequency in the passed web pages as a related word of the search word, and adding the related word to the user search word DB in association with the search word; The counterpart is extracted from the web pages that are not the counterpart or the related word, and a word that appears at a predetermined frequency or more is determined. Adding to the user search word DB in association with the search word; The non-correlated word is determined as a non-related word of the search word, except for the counterpart, the related word, and a word that appears more than a certain frequency in the failed web pages, and the non-related word is associated with the search word. Searching for information on the Internet using a meta search engine comprising the step of adding to the search term DB. The method of claim 4, wherein the counterpart, the related word, the opposite word, and the unrelated word consist only of words that affect the nature of the contents of the web page. The method of any one of claims 1 to 3, wherein generating a plurality of search expressions from a search word Loading a matched word, related word, related word, unrelated word related to the search word from the user search word DB; Generating, as a first search expression, a condition for searching for a web page including both the search word and the parallel word, one or more related words, no counter words, and no non-related words; Generating, as a second search expression, a condition for searching a web page including both the search word and the counterpart and the counterpart at all; And generating a condition for searching a web page including the search term as a third search expression. The method of any one of claims 1 to 3, wherein the transmitting of the plurality of search expressions to the plurality of robot search engine servers is performed on the Internet using a meta search engine that is multi-threaded. How to retrieve information. The method of claim 1, wherein the receiving of the search results for the plurality of search expressions from the plurality of robot search engine servers and arranging the search results in a predetermined order comprises: Sorting each of the search results for each search expression; Reordering the search results for each of the sorted search expressions, wherein the reordering step includes: first sorting the search results of the search expressions having a difficult search condition, and deleting the URLs having a lower priority when the URLs are duplicated. Information retrieval method on the Internet using meta search engine. The method of claim 8, wherein sorting by the search results for the search expression is as follows. (a) prioritizing the most frequent occurrences of URLs; (b) prioritizing the search engines having higher priority if the frequency of URLs is the same among the sorting results of step (a); And and (c) sorting according to the ranking in the search engine if the search engines have the same priority among the sorting results of step (b). In the meta search engine system for searching for information on the Internet, A user search word DB for storing information related to a search word input by a user through a predetermined web browser, and transmits a search word to servers of a plurality of robot search engines according to a user's search request; A plurality of user computers that sort and output the search results received from the server of the user; A web server for storing the latest contents of the user search word DB included in the plurality of user computers, and transmitting information related to the search word by referring to the contents of the user search word DB of other users according to a request of the user computer; And Meta search engine system including a server of a plurality of search engines for transmitting a search result in response to a search request of the user computer.