KR100682552B1 - System, apparatus and method for providing a weight to search engines according to situation of user and computer readable medium processing the method - Google Patents

Abstract

The present invention relates to a system for providing search results through a plurality of search engines according to a user's search request. In particular, the present invention applies a weight to each of the search engines according to the current situation of the user, and applies the user to the search results. And an apparatus for applying a context variable to perform an update operation of the weight according to the reaction of the.

Search, search engine, weight, constant variable, update

Description

SYSTEM, APPARATUS AND METHOD FOR PROVIDING A WEIGHT TO SEARCH ENGINES ACCORDING TO SITUATION OF USER AND COMPUTER READABLE MEDIUM PROCESSING THE METHOD}

1 is a diagram showing a keyword search screen in a general search site.

2 is a diagram illustrating a system for assigning a weight to each search engine according to a user's situation according to the present invention.

3 is a block diagram showing a detailed structure of an apparatus for applying context variables according to an embodiment of the present invention.

4 is a signal flow diagram illustrating a procedure for assigning a weight to each search engine according to a user's situation according to the present invention.

5 is a flowchart illustrating a detailed procedure of weighting a search result according to an embodiment of the present invention.

6 is a flowchart illustrating a procedure for detecting a situation similar to the current situation according to an embodiment of the present invention.

7 is a flowchart illustrating a detailed procedure of weight update according to an embodiment of the present invention.

<Explanation of symbols for the main parts of the drawings>

100: keyword search box 110: search result information

120: search related information 200: user terminal

210: network 220: web server

230: device for applying the situation variable 240: search engine

300: user transceiver 310: search transceiver

320: weight applying unit 330: situation variable recognition unit

340: Weight extraction unit 350: Situation variable weight database

360: weight update unit

The present invention relates to a method for providing search results from a plurality of search engines, and more particularly, to a system, apparatus, method, and a computer readable computer for implementing the search engine weights according to a user's situation. A recording medium.

In the modern information society, computers and the Internet have emerged as essential tools for new communication as a medium for acquiring information. In general, for efficient information retrieval through the Internet, it is possible to input a search word through portal sites (or search sites) equipped with a search engine.

1 is a diagram illustrating a keyword search screen in a general search site. Referring to FIG. 1, when a keyword (for example, deep impact) to be searched is input to a keyword search box 100 provided in a web page of a search site, the search engine is provided. Search through a large number of information linked on the Internet through, and provide information related to the search keyword as a search result (110). At this time, the plurality of search results are sorted and output according to a specific method set in the search site. For example, priority is provided based on the degree of relevance to the corresponding keyword to the searched documents or the registration date of the searched documents. In addition, information 120 about the searched document may be provided together.

Since the exposure position of the search result or the accuracy of the search result reflects the reliability of the search site, the design of the search engine and the search result output algorithm becomes a very important factor in the search site.

On the other hand, the search engines currently used in the Internet may be used in a plurality of different systems depending on the technology used. For example, a search engine may have various methods depending on a search technique or a provider, and each of the plurality of search engines may display different search results. In addition, recently, a method of integrating and providing search results from a plurality of different search sites in one search site is also provided. Accordingly, as shown in FIG. 1, a search is performed through various other search sites according to a corresponding search request, and then the results are summarized and provided.

In such a case, the plurality of results are collected by using a predetermined appropriate method. For example, they can be collected using a method of simply merging lists, or by identifying how common each item is in different systems and prioritizing them according to their frequency.

However, the above conventional methods of collecting information are static methods that do not consider the user's situation regardless of the kind. Although the type of information desired by the user may be searched through the same keyword, it may vary depending on the environment or situation of each user. Therefore, conventional methods that do not consider the user's situation have a problem in that accuracy and reliability of information are lowered by applying the same information collection method to all users.

An object of the present invention is to provide information in consideration of a user's situation in order to increase the accuracy of information provided when providing information to a user through a plurality of search engines.

In addition, an object of the present invention is to store the response information according to the user's situation, and to provide information considered according to the user's situation by using the pre-stored information in the future request for a search of another user.

The system of the present invention for achieving the above object; In a system for providing a search result through a plurality of search engines in response to a user's search request, determining a plurality of situations in which the user's disposition to the search results can be varied, the current situation of the user among the situations And applying a weight to each of the search engines, and applying a situation variable to update the weights according to a user's selection of the search results for each search engine.

The apparatus of the present invention for achieving the above object; An apparatus for providing a search result through a plurality of search engines according to a user's search request, the apparatus comprising: determining a plurality of situations in which a user's disposition to the search result may be changed, and the current situation of the user among the situations Means for applying a weight for each search engine according to; And means for performing the updating of the weight according to a user's selection of the search result for each search engine.

The method of the present invention for achieving the above object; A method of providing a search result through a plurality of search engines according to a user's search request, the method comprising: receiving a search keyword from a user and requesting a keyword search according to the received search keyword from the plurality of search engines; Receiving a search result from each of the search engines and recognizing a current situation variable of the user; Retrieving information on the same or similar situation variables previously stored in a database based on the recognized current situation variables; Extracting a weight for each search engine in the same or similar context as the current situation as the search result, and reflecting the weight according to the extracted situation in the search results for each search engine; And rearranging the search results of the respective search engines according to the reflected results and transmitting the rearranged results to the terminal of the corresponding user.

Meanwhile, information for assigning a weight to each search engine according to the user's situation may be stored in a recording medium readable by a server computer. Such recording media includes all kinds of recording media on which programs and data are stored so that they can be read by a computer system. Examples include Read Only Memory (ROM), Random Access Memory (RAM), Compact Disk (CD), Digital Video Disk (DVD) -ROM, Magnetic Tape, Floppy Disk, Optical Data Storage, etc. It also includes implementations in the form of (eg, transmission over the Internet). In addition, these recording media can be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

The present invention utilizes the implicit feedback of a user to find a search engine that is more suitable for a given user's situation in a method of providing a search result for a request to the user's Internet search. In other words, we propose a method to effectively determine which search engine selects a result from a given search result as a more reliable result.

In this case, when a user selects one of the search results when providing a search result from a plurality of systems, the selected item may be regarded as a more suitable item for the user's needs. The system you provide (ie, the search engine) is more suitable for your situation. Therefore, the system stores information on which system each item comes from, and then increases the weight of the system that provided the item whenever the user selects a predetermined item. Then, when a user of the same or similar situation later requests a search result, the user can provide a more accurate final result by collecting the results from each system by using the stored weights.

Meanwhile, the present invention sets one or more situation variables in order to quantify the situation to be reflected in the search result. The situation variable means that the same user changes depending on the situation. According to the present invention, examples of what can be set as a user's context variable may be the following items.

1. The field (category) in which the user searches-e.g. electronics, pedagogy, inline skating, etc.

2. Usage time-What time of day do users use?

3. Day of the week

4. Weather

5. Location or region

6. Your current interests

7. The time a user sees a page

That is, the situation variables presented above are factors influencing the user's response to the search results. For example, since the propensity for a search result is different between a user who searches in the daytime and a user who searches in the middle of the day, it is better to reflect such usage time information as a situation variable according to the present invention. Reasonable information can be provided.

On the other hand, in the present invention, when collecting information from each system, the system records which system (or search engine) each item came from. Then, if the user selects one of the items after the aggregated results are presented to the user, the weight of the system that produced the item (the search engine that provided the search result) is increased and the weight of the system that is not. Decreases. The changed weight is stored in a database that stores 'user's contextual weight' along with the context variable of the current user. Accordingly, the situation variable information continuously changes as the user selects the presented result.

In addition, when a user requests a search, the system first requests a result generation from each system, and checks the situation variables of the current user to check whether there is data for the same situation in the previously stored database. If the same situation exists, the weight for each system in that situation is read and reflected in the result collection. On the other hand, if there is no identical situation, the similar situation is found as much as possible and reflected in the same way. However, if it is determined that there is no similar situation, each system is weighted as having the same weight and the results are collected.

Hereinafter, a detailed description of a preferred embodiment of the present invention will be described with reference to the accompanying drawings. In the following description, detailed descriptions of well-known functions or configurations will be omitted when it is determined that the detailed descriptions of the known functions or configurations may unnecessarily obscure the subject matter of the present invention.

2 is a diagram illustrating a system for assigning a weight to each search engine according to a user's situation according to the present invention.

Referring to FIG. 2, the system according to the present invention may include a user terminal 200, a network 210, a web server 220, an apparatus for applying context variables 230, and a search engine 240. As a subject who performs a search for the user terminal 200 and receives the information, the corresponding information is provided according to the situation of the user terminal 200. The network 210 may be a wired or wireless network, and any network including a LAN, a WAN, and the like may be used, but it is preferable to implement the network.

A user accesses the web server 220 through the network 210 as the user terminal 200. The web server 220 may be understood as a portal site (or search site) that provides information according to the search of the user. In addition, as another implementation method, it may be a general website linked to the portal site or the like and providing information at the request of a user.

Meanwhile, according to the present invention, the information provided by the user's request is made through a plurality of search engines 240. At this time, the search results provided through the plurality of search engines 240 are rearranged by applying the weight for each search engine in the contextual variable applying device 230 provided according to the present invention as described above. That is, the information on the situational response that appears when each user performs a search is accumulated and stored in the situation variable applying device 230, and the weight of each search engine 240 in the same or similar situation is requested when the user requests the information. Confirm and apply.

Therefore, the apparatus 230 for applying the situation variable according to the present invention serves to apply the weight of each system (search engine) according to the pre-stored situation to the search result, and also through the user's response to the search result. The previously stored weights in the situation are updated.

Hereinafter, the situation variable applying apparatus 230 will be described in more detail with reference to FIG. 3.

3 is a block diagram showing the detailed structure of the apparatus for applying the situation variable 230 according to an embodiment of the present invention. Referring to FIG. 3, the situation variable applying apparatus 230 includes a user transceiver 300, a search transceiver 310, a weight applier 320, a context variable recognizer 330, and a weight extractor 340. ), The situation variable weight database 350 and the weight updater 360. The configuration may be modified into other configurations having the same function according to the implementation method thereof.

As described above, each component of the situation variable applying apparatus 230 may be weighted according to a user's response to the search result and a weighting operation for each search engine that provides a search result when the user requests a search according to the present invention. Perform the update operation. On the other hand, the bold line in Figure 3 represents the flow of information used in the process of updating the weight according to the user's response.

First, a process of providing a search result by applying a weight to each search engine in consideration of the user's situation according to the present invention will be described.

Information (eg, keyword information) requested by the user is transmitted to the user transceiver 300 through the web server 220. The user transceiving unit 300 transmits the search request information of the user received through the web server 220 to the search transceiving unit 310 and the search result to which the weight is applied from the weight applying unit 320. Receives and serves to provide back to the user through the web server 220.

The search transceiver 310 receives search request information including a search keyword from the user transceiver 300 and performs a search according to the search keyword. That is, a search operation is performed from the plurality of search engines 240, and a search result according to the search operation is received from each search engine 240. The search results received from the plurality of received search engines 240 are transmitted to the weight applying unit 320 to apply contextual weights according to the present invention.

On the other hand, the situation variable recognition unit 330 recognizes and generates various predetermined situation variables when a user requests a search. For example, when the situation variable used for weighting is time, search field, or use site, the user transceiver 300 notifies the situation variable recognition unit 330 that a search request has been received from the user. Send the search keyword. Accordingly, the situation variable recognition unit 330 is notified of the search request from the user transmission / reception unit 300, the time at which the request is notified, the search keyword requesting the search, and the access site to which the user accesses the search. Will generate information.

The current situation information of the user who requested the current search generated by the situation variable recognizer 330 is provided to the weight extractor 340 to be used for weight extraction, and is provided to the weight updater 360 to be used for weight update.

The weight extracting unit 340 reads past situation information stored in the situation variable weight database 350 and compares the present situation information with the current situation information provided from the situation variable recognition unit 330. You find the situation. As described above, the previously stored weight information for each search engine is extracted in the same or similar past situation, and the extracted weight information for each search engine is transmitted to the weight applying unit 320.

Finally, the weight applying unit 320 receives a search result for each search engine from the search transmitting and receiving unit 310, and the angle suitable for the current situation provided from the weight extracting unit 340 to the received search results The search results are finally rearranged by providing weight information for each search engine for each search engine.

As such, the search result to which the weight is applied according to the current situation is transmitted to the user transceiver 300 and provided to the corresponding user terminal 200 through the web server 220.

At this time, the user checks the provided search result and selects or views any information to indicate a response to the provided information. The response seen by the user is transmitted to the user transceiver 300 through the web server 220, and the user transceiver 300 provides the weight updater 360 to use this for weight update.

Meanwhile, the weight updater 360 obtains pre-stored weight information on the situation from the situation variable weight database 350 and updates the new weight value through the response information of the currently received user. In this case, the weight updater 360 receives the search result information for the corresponding search from the search transceiver 310, obtains the current situation information from the situation variable recognition unit 330, and uses it in the weight update process. Done. Finally, the new weight calculated by the weight updater 360 is stored in the context variable weight database 350 and used as information for the next user's search. A specific example of the weight update will be described later.

The detailed structure of the system and apparatus according to the present invention has been described above with reference to FIGS. 2 and 3. Hereinafter, the procedure according to the present invention will be described in detail with reference to FIGS. 4 to 7.

4 is a signal flow diagram illustrating a procedure for assigning a weight to each search engine according to a user's situation according to the present invention. Referring to FIG. 4, a user who wants to search first accesses a searchable website or a website linked to a search site through his user terminal 200, and performs a search by inputting a search keyword. . At this time, the search keyword information input by the user is transmitted to the situation variable application device 230 through the web server 220 (S401, S402).

The situation variable application device 230 performs a search operation on the corresponding search keyword to each of the plurality of search engines 240 according to the search request. That is, when a plurality of search engines request a search (S403, S405), each search engine 240 provides a search result according to the search request to the situation variable applying device 230 (S404, S406). Done.

As described above, the situation variable application device 230 extracts weights for each search engine in a situation similar to or identical to the current situation by comparing the situation variable value of the current situation with previously stored situation variable values. To me. The weight of each search engine in the extracted current situation is reflected in the received search results of each search engine, and finally the search results are rearranged (S407).

The rearranged final search result is provided to the corresponding user terminal 200 through the web server 220 (S408 and S409). At this time, the user checks the rearranged search result information according to the situation through the user terminal 200, and responds to the corresponding information by clicking or viewing specific information. The user's response information is transmitted to the situation variable application device 230 through the web server 220 (S410 and S411). The situation variable application device 230 updates and stores the weight value according to the current situation by using the transmitted response information of the user (S412).

Hereinafter, a procedure of assigning a weight to each search engine to a search result in the context variable applying apparatus 230 according to the present invention will be described in more detail.

5 is a flowchart illustrating a detailed procedure of assigning a weight to a search result according to an exemplary embodiment of the present invention. Referring to FIG. 5, when the situation variable applying apparatus 230 receives a search keyword from the user (S501), as described above, a request for a keyword search according to the search keywords received by the plurality of search engines is performed (S502). Done.

Then, the situation variable applying device 230 receives a search result from each search engine (S503), and recognizes the current situation variable of the user (S504). At this time, the information on the same or similar situation variables previously stored in the database is searched based on the recognized current situation variables (S505). As a result of the search, the weights of the respective search engines in the same or similar context as the current situation are extracted, and the weights according to the extracted situations are reflected in the search results for each search engine (S506). Finally, the search result for each search engine is rearranged according to the reflected result and transmitted to the terminal of the corresponding user (S507).

A method of searching for a past situation similar or identical to a previously stored current situation from the recognized current situation in FIG. 5 will be described in more detail with reference to FIG. 6.

6 is a flowchart illustrating a procedure for detecting a situation similar to the current situation according to an embodiment of the present invention. Referring to FIG. 6, when it is assumed that there are M past situations previously stored in the situation variable weight database 350 and there are N situation variables set for each situation (S601), first, the first stored past situation (that is, m). = 1)) and comparison (S602). That is, the current situation is compared with the past m situation for each situation variable (S603).

The procedure for comparing the current situation with the past m situation for each situation variable is performed from the situation variable of n = 1 to the Nth situation variable in sequence (S604). That is, the similarity is calculated for the n situation variable in the m situation (S605). The method of calculating the similarity will be described later in the description of specific embodiments.

Subsequently, while increasing n (S606), the similarity calculation for all the situation variables is performed (S607). When the similarity calculation for all the situation variables in the m situation is completed, the similarity calculated for each situation variable in the m situation is added (S608).

At this time, by repeating the procedure to the M situation by increasing the value of m (S609, 610), the similarity for each situation variable in all situations is added up. Finally, by sorting according to the finally summed values for each situation (S611), the past situation that is the same as or similar to the current situation is searched for.

On the other hand, in the present invention as described above, in addition to the application of the pre-stored weights, it is possible to maintain more reliable information by continuously updating the weight value in the situation according to the response of each user.

7 is a flowchart illustrating a detailed procedure of weight update according to an embodiment of the present invention. Referring to FIG. 7, when the search result is presented to the user requesting the search as described above (S701), the corresponding user checks the search result and clicks or views specific information. Will react. The user's response information is transmitted to the situation variable applying device 230, and the situation variable applying device 230 detects this (S702).

The situation variable application device 230 adjusts the weight for each search engine for the situation according to the user's response (S703). At this time, the adjustment of the weight may be performed by adjusting the weight according to the selected item and the weight according to the skipped item. That is, when the user selects or clicks a specific item from a search result, since the selected item becomes reliable information in the current situation, the weight of the search engine including the selected item is increased. In addition, since it is preferable that the item that is located above the clicked item but cannot be selected and skipped is set to be lower than the preset weight, the weight of the search engine including the skipped item is reduced.

Finally, the adjusted weight value is normalized to a value between a specific range (for example, 0 to 1) (S704).

The system, apparatus and methods according to the present invention have been described in detail above. Hereinafter, specific examples will be described in detail to help understand the present invention.

<Example>

<How the system recognizes your situation variable>

The method of recognizing the user's situation variables is determined according to the type of each situation variable.

1. If the situation variable is 'time' or 'day of the week': The web server automatically assigns a time stamp every time a user's request comes in, so information about time and day can be collected automatically.

2. If the situation variable is 'the time a user views one page': The time difference from the time a user requests a page until the request for the next page is regarded as the time the user reads the page. Of course, when the session ends without a request for the next page, the page becomes the last page, so it is difficult to know exactly when the page is read.

3. If the situation variable is 'type of site': If the system according to the present invention is provided in the form of a service for multiple sites (eg, Yahoo, Interpark, etc.) It is also important information. At this time, there are many ways to know what field the current user is using, but there are generally explicit and implicit methods. As an example of an explicit method, when a user visits a site and makes a page request, it is possible to know what the site is (eg, a portal site, a general shopping mall, an online cafe, etc.) by looking at the characteristics of the site. Therefore, if you categorize each type of site in advance, you can find out the sites of similar fields as the site visited by the current user. An example of an implicit method is that, when comparing the similarity between the contents of pages of one site and the contents of pages of another site, an area of the site can be implicitly classified according to the number of similar pages.

4. When the situation variable is 'field (keyword)': It can be inferred from the keyword that the user uses for searching to find out what field the user is currently searching for. There are a number of ways to leak a field from a keyword, for example: The keyword entered by the user is searched in the keyword hierarchy built in advance. Such a hierarchical structure has a hierarchical structure of several levels, and as the upper hierarchy goes up, the keywords are classified into a small number of categories through major classification. For example, there is only one keyword at the top, the second tier has 7-9 major categories, and the third tier has 4-7 middle categories for each of these major categories, so there are 40-50 keywords in total. The first tier has 150-400 keywords, and the lower tier has more keywords. If there are no keywords in the keyword hierarchy, synonyms can be found. In addition, if a keyword is found, the keyword of a certain level (second or third tier) is taken as the search field of the current user, moving up from the position of the keyword.

5. If the situation variable is 'location': If the user is using a mobile device and the GPS device is attached to the device, the user's device can receive the exact location of the user.

6. If the situation variable is 'region': If the location information cannot be collected, the region where the user uses the computer may be collected from the IP address of the user's computer or the address provided when the user registers.

7. When the situation variable is 'user's current interest': The user's current interest is also a very important situation variable, but there are many ways to collect it. You can share it. An example of an explicit method is to ask the user's current interests and collect them each time they enter the site. As an example of the implicit method, when a user visits the site and uses it for a while, the keyword inputted by the user or extracted from the page visited by the user is extracted. If you compare these keywords with keywords extracted from past users stored in the system and calculate similarity, you can see that they had similar interests with the current user.

<How to retrieve information about same or similar situation variables in DB>

Whenever a user makes a request to the web server, the above-described situation variables of the user are collected. There may be several ways to search for the same or similar situation variables as the current user's situation variables.

1. The user's situation variables are compared with the situation variables stored in the DB one by one and the scores are assigned according to the similarity. For example, if the current user's time is 1:00 pm, the user who is close to 1:00 pm in the past users is given a high score, such as '1', and the farther away, the lower score is closer to '0'. If this is expressed as Equation 1, it is as follows.

Therefore, a value of 1 is assigned to a situation of the user used at the same time as the current user, and a value of 0 is assigned to a situation of a user having a time difference of 12 hours from the current user.

If the situation variable does not have a continuous value such as time, but has an intermittent value such as 'field' or 'site type', it can be given a value of '1' if a match is found and '0' if there is a mismatch. For example, if the site used by the current user is an internet shopping mall, if the site type is an internet shopping mall among the data in the context variable DB, a value of 1 is assigned and other sites are assigned a value of 0.

2. The total score is calculated by summing the points given for each situation variable. Table 1 below shows an example of score addition. Any number of situation variables can be used, but the example assumes that only three are used, and that there are three search engines.

Situation Variable 1 (Time) Situation Variable 2 (Site Type) Situation Variable 3 (Sector) Search engine1 weights Search engine 2 weights Search engine 3 weighting Past situation 1 5:30 am shopping mall camera 0.5 0.9 0.6 Past situation 2 11:30 am newspaper politics 0.9 0.4 0.7 ... ... ... ... ... ... ... Past situation m 5:30 pm Portal Culture and art 0.8 0.7 0.9

If the current user inputs a keyword in the field of culture and art on the newspaper site at 5:30 am, the user's similarity score with each past situation in the situation variable DB is calculated as follows.

Past score of 1 = 1 (time) + 0 (site type) + 0 (field) = 1

Past score 2 = 0.5 (hours) + 1 (site type) + 0 (field) = 1.5

Past score m = 0 (hours) + 0 (site type) + 1 (sector) = 1

3. You can take a predetermined number from above, sorted by highest total score.

<Display method of search result screen>

There are several ways to aggregate the results generated by the various search systems, but two examples are as follows.

As a first example, after a list is generated in each individual search system, the order may be determined according to how many systems each item is searched for. In this case, if there is a weight score proposed by the system of the present invention, the weight may be weighted up using this weight.

As a second example, each search engine typically calculates a goodness of fit score when listing. Standardize these goodness-of-fit scores so that they can be compared with each other and average the scores that each item receives from the individual systems. In this case, if there is a weight score suggested by the system of the present invention, the weighting average of the goodness of fit may be weighted using the weight.

The method of displaying the collected information in the system of the present invention may be implemented in two ways. The first is to present a list of each individual system separately on one screen, and present the final result collected from the system separately. The second is a method of presenting only the final aggregated results according to the present invention without listing each individual system separately. Of course, in the second case, the user can present a list of each individual system separately if desired.

On the other hand, the length of the list generated by each system is theoretically unlimited, but it is common to display 10 or 20 screens at a time in consideration of the user's information processing ability. The final list collected according to the invention can also be displayed in the same way. If you want, you can show more items on the next screen.

<Weight adjustment method>

The context variable weight database described above stores not only the variable values of individual situations but also the weights of individual systems in each situation. For example, if there are three separate systems, each situation will have three weights corresponding to each system. Initially, the weight is given equally to all systems, but as described above, the weight is modified according to the user's response.

After the user is presented with the final aggregated information, the user clicks on the item that he / she thinks appropriate. In this case, if there is an item skipped from the list, the item is inappropriate, so the individual system recommending the item decreases the weight and the clicked item is a suitable item, so it is preferable to increase the weight of the individual system generating the item. If there is data about the behavior after the user clicks on one of the lists, the weighting of the clicked items may be differentiated depending on how long the user has viewed each item. For example, even when you increase the weight of the individual system that created the clicked item, long-running items can have a larger weight increase, and for a short time, the item can have a relatively small weight increase.

A specific embodiment of the weight adjustment method is as follows.

1. When a user selects an item from the list: The weight for each search engine is increased by the weight change value below. If there is no item in a list of search engines, the weight for that search engine is reduced by a small number (eg 0.01).

2. If the user skips an item in the list: Reduce the weight for each search engine by the weight change value below. If there is no item in a list of search engines, the weight for that search engine is increased by a small number (eg 0.01).

For example, if the item was placed second in a search engine's results, the weight change would be 1/2 = 0.5.

Weight change number = 1 / (rank of corresponding item in search results of each search engine)

For example, suppose that the search result is generated and presented to the user as shown in Table 2 below using the situation variable DB above. Suppose a user is searching for a camera at a shopping mall at 5:30 (past situation 1 in the situation variable DB).

Original search results Final search results Search engine 1 Search engine 2 Search engine 3 1st Webpage a Webpage b Webpage c Webpage a 2nd Webpage b Webpage a Webpage d Webpage b ... ... ... ... ... 10th place Web page x Webpage y Webpage z Web page x

At this time, if the user skips the webpage a and selects the webpage b in the final search result presented to the user, the weight of each search engine is adjusted as follows.

1.Adjustment for selecting web page b:

Weight of search engine 1 = 0.5 (current weight) + 1/2 = 1

Weight of search engine 2 = 0.9 (current weight) + 1/1 = 1.9

Weight of search engine 3 = 0.6 (current weight)-0.01 = 0.59 (no b in the list)

2. Adjustments for skipping webpage a:

Weight of search engine 1 = 1.0 (current weight)-1/1 = 0

Weight of search engine 2 = 1.9 (current weight)-1/2 = 1.4

Weight of search engine 3 = 0.59 (current weight) + 0.01 = 0.6 (there is no a in the list)

On the other hand, as described above, when sorting the data from each search engine with the weights calculated above, it is desirable to add one step that is finally normalized. That is, according to the above-described weight adjustment method, the weight of each search engine in the situation variable weight DB may be 0 or less or 1 or more, so that the weight is normalized before multiplying the weight by the fitness score calculated by each search engine, and the weight is between 0 and 1. It is desirable to be in.

For example, if the weight of each search engine is 0.6, -0.3, 1.5, it is normalized using the following equation.

Weight 1 = (0.6 + 0.3) /1.8 = 0.5

Weight 1 = (-0.3 + 0.3) /1.8 = 0

Weight 1 = (1.5 + 0.3) /1.8 = 1

Using Equation 3, the minimum is always 0 and the maximum is 1.

On the other hand, in the embodiment of the present invention has been described with respect to specific embodiments, of course, various modifications are possible without departing from the scope of the invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined not only by the appended claims, but also by those equivalent to the claims.

According to the present invention, more accurate and highly reliable information is provided by providing a search result by applying a weight to a search engine that provides highly reliable information in the same or similar situation according to the environment or situation of each user upon request of a search. There is an advantage that can be provided.

Claims (15)

In a system for providing search results through a plurality of search engines in response to a user's search request, Determine a plurality of situations in which a user's disposition to the search result can be varied, apply a weight to each search engine according to the current situation of the user among the situations, and apply a search result for each search engine And a situation variable applying device for updating the weight according to a user's selection. The apparatus of claim 1, wherein the apparatus for applying situation variables is: A user transceiving unit configured to receive search request information of the user through a web server, receive a search result to which the weight is applied, and provide the search result to the user through the web server; A search transmission / reception unit configured to receive search request information including a search keyword from the user transceiver and perform a search from the plurality of search engines according to the search keyword; A situation variable recognition unit for recognizing and generating various predetermined situation variables when the user requests a search; A situation variable weight database storing weight information of a plurality of search engines for each situation in the past; A weight extraction unit which obtains past situation information stored in the situation variable weight database and extracts past situation identical or similar to the current situation by comparing the current situation information provided from the situation variable recognition unit; And Receiving a search result by receiving the search results for each search engine from the search transmission and reception unit, and providing weight information for each search engine for each search engine suitable for the current situation provided by the weight extraction unit to the received search results. And a weighting unit for each search engine according to a user's situation. The apparatus of claim 2, wherein the situation variable applying device is A search engine according to a user's situation, further comprising: a weight updater which retrieves pre-stored weight information of the situation from the situation variable weight database and updates the new weight value with the currently received user's response information; Star weighting system. The system of claim 1, wherein the information about the situation is set by at least one situation variable. The method of claim 4, wherein the situation variable includes any one selected from a search time, a day of the week, a time of viewing a page, a type of a searched site, a field of a search keyword, a user's location, a region, and a user's current interests. System for assigning a weight for each search engine according to the user's situation, characterized in that one or more. An apparatus for providing a search result through a plurality of search engines in response to a user's search request, Means for determining a plurality of situations in which a user's disposition to the search result can be varied and applying weights to the respective search engines according to the current situation of the user among the situations; And And a means for updating the weight according to a user's selection of the search result for each search engine. The apparatus of claim 6, wherein the means for applying the weights is: A user transceiving unit configured to receive search request information of the user through a web server, receive a search result to which the weight is applied, and provide the search result to the user through the web server; A search transmission / reception unit configured to receive search request information including a search keyword from the user transceiver and perform a search from the plurality of search engines according to the search keyword; A situation variable recognition unit for recognizing and generating various predetermined situation variables when the user requests a search; A situation variable weight database storing weight information of a plurality of search engines for each situation in the past; And A weight extracting unit for extracting past situation information stored in the situation variable weight database and extracting a past situation identical or similar to a current situation by comparing the current situation information provided from the situation variable recognition unit; Means for performing the update operation of the weight, Receiving a search result by receiving the search results for each search engine from the search transmission and reception unit, and providing weight information for each search engine for each search engine suitable for the current situation provided by the weight extraction unit to the received search results. Apparatus for applying a context variable to the weight of each search engine according to the user's situation, characterized in that it comprises a rearranged weight applying unit. The method of claim 7, wherein the means for performing the update of the weight, A search engine according to a user's situation, further comprising: a weight updater which retrieves pre-stored weight information of the situation from the situation variable weight database and updates the new weight value with the currently received user's response information; Apparatus for applying situation variables to give a weight to each star. 7. The apparatus of claim 6, wherein the information on the situation is set by at least one situation variable. The method of claim 9, wherein the situation variable is any one selected from a search time, a search day, a time a user views a page, a type of a searched site, a field of a search keyword, a user's location, a region, and a user's current interests. Apparatus for applying a context variable according to the user's situation, characterized in that the weighting for each search engine. In the method for providing search results through a plurality of search engines in response to a user's search request, Receiving a search keyword from a user and requesting a keyword search according to the received search keyword from the plurality of search engines; Receiving a search result from each of the search engines and recognizing a current situation variable of the user; Retrieving information on the same or similar situation variables previously stored in a database based on the recognized current situation variables; Extracting a weight for each search engine in the same or similar context as the current situation as the search result, and reflecting the weight according to the extracted situation in the search results for each search engine; And And rearranging the search results of the respective search engines according to the reflected results and transmitting the rearranged results to the terminal of the corresponding user. 12. The method of claim 11, wherein after the search results for each search engine are rearranged according to the reflected results and transmitted to the terminal of the corresponding user, Checking, by the user, the search result and selecting specific information; Detecting response information of a user who has selected the specific information; And And adjusting the weights of the respective search engines for the corresponding situations according to the user's response. 12. The method of claim 11, wherein the information about the situation is set by at least one situation variable. The method of claim 13, wherein the situation variable is any one selected from a search time, a search day, a time a user views a page, a type of a searched site, a field of a search keyword, a user's location, a region, and a user's current interests. The weighting method for each search engine according to the user's situation, characterized in that above. A computer-readable recording medium containing a program capable of performing the method according to any one of claims 11 to 14.

Images