KR20120009817A - System and method for providing search service - Google Patents

Abstract

PURPOSE: A search service system and a method thereof are provided to separate a search word of multiwords into morphemes and apply a token score to the morphemes, thereby offering a best matching search result. CONSTITUTION: A search word processing unit(120) separates a search word of a user terminal into morphemes to match a token score to the morphemes. The search word processing unit selects a morpheme combination consisting of the morphemes based on the token score. A database(150) stores the token scores by a morpheme. A search unit(110) searches based on the morpheme combination. A search result unit(140) provides a webpage including a search result to the user terminal. The search word processing unit selects one with the highest token score among the morpheme combinations.

Description

Search service provision system and method {SYSTEM AND METHOD FOR PROVIDING SEARCH SERVICE}

The present invention relates to a search service providing system and method.

Recently, when searching for information desired by a search, there is an increasing tendency to input multiple terms consisting of a plurality of terms as search terms. As such, when multiple terms are input, the search engine does not know which of the multiple terms should be considered as the most important search term or whether multiple terms should be considered as a single term, thereby failing to provide accurate search results.

There is a technique described in the Republic of Korea Patent Publication No. 2007-0090014 as a prior art for solving this problem. This prior art tracks a search for a period of time after a multi-term search that includes a specific term and then deletes any one of the multiple terms in the next search, and the number of times a full multi-term search that includes the specific term and The probability of deleting a word is calculated as the ratio of the number of times a word is deleted from the search. In the prior art, when multiple terms are input, if the difference between the deletion probabilities of two specific terms is greater than a threshold value, the term having a high deletion probability is discarded and the search is performed based on a term having a low deletion probability. That is, the prior art selects a term using a deletion probability calculated based on a term deleted from a subsequent search among successive search terms.

However, in a real search environment, in order to find a desired search result, a user may not search only by deleting a specific term but search by a certain term and then add a new term in the next search. In addition, there is a case where a specific term is deleted after several searches instead of deleting the term after the multiple search. The prior art has a problem that it does not accurately reflect the intention of the user in this case.

An object of the present invention is to provide a search service providing system and method that can provide accurate search results for a multi-term search query.

According to one aspect of the present invention, a search service providing method of a search service providing system is provided. According to the search service providing method, the search service providing system receives a search word from a user terminal, divides the search word into a plurality of morphemes, and matches token scores of the corresponding morphemes to each morpheme. And the search service providing system selects a morpheme combination consisting of at least one morpheme among the plurality of morphemes based on the token score, performs a search with the morpheme combination, and includes a search result performed by the morpheme combination. A web page is provided to the user terminal.

After generating a plurality of first morpheme combinations from the plurality of morphemes, and setting a combination of token scores of the morphemes included in each first morpheme combination to the token scores of the first morpheme combinations, the plurality of first morphemes The first morpheme combination having the largest token score among the combinations may be selected as the morpheme combination.

The number of morphemes included in each morpheme combination may be one less than the number of the morphemes.

The search service providing system determines whether to perform further search from the search results performed by the morpheme combinations, and when the further search is necessary, generates a plurality of second morpheme combinations from the morphemes, and assigns to each second morpheme combination. The combination of the token scores of the included morphemes may be set as the token scores of the corresponding second morpheme combinations, and a search may be performed by selecting a second morpheme combination having the largest token score among the plurality of second morpheme combinations.

The number of morphemes included in the second morpheme combination may be smaller than the number of morphemes included in the first morpheme combination.

The search service providing system may determine whether an additional search is performed from a search result performed by the search word after performing the search with the search word, and may divide the search term into a plurality of morphemes when the search further requires.

The token score of each morpheme may be a value indicating the degree to which the morpheme was maintained in a search for a certain period of time.

The search service providing system receives a search term from a plurality of sessions, and when each session ends, generates a search term pair consisting of two search terms that are in a subset relationship from the search term of the session, and selects a search term pair in each session term. After extracting a first search term pair including a morpheme of, the token score of the predetermined morpheme may be calculated from the first search term pair of the plurality of sessions.

In this case, a first search word pair consisting of two search terms each including the predetermined morphemes in the first search word pair of the plurality of sessions is selected as a second search word pair, and the number of the first search word pairs and the second search word are selected. The token score may be calculated based on the number of pairs.

According to another aspect of the present invention, a search service providing system including a search word processor, a database, a search unit, and a search result provider is provided. The search word processor divides the search word from the user terminal into a plurality of morphemes, matches token scores of the corresponding morphemes with each morpheme, and selects a morpheme combination consisting of at least one morpheme among the plurality of morphemes based on the token scores. do. The database stores the token scores by morpheme. The search unit performs a search by the morpheme combination, and the search result providing unit provides a web page including a search result of the search unit to the user terminal.

According to a method of providing a search service according to another aspect of the present invention, the search service providing system receives a search word from a plurality of sessions, and when each session ends, the search service providing system receives two search terms having a subset relationship from the search word of the corresponding session. Generate a search term pair, select a first search term pair including a predetermined morpheme from the search term pairs of the plurality of sessions, and then score the token of the predetermined morpheme based on the first search term pair of the plurality of sessions Calculate The search service providing system divides a search word received from a user terminal into a plurality of morphemes, matches the token scores with the plurality of morphemes, respectively, and then stores at least one of the plurality of morphemes based on the token scores. A morpheme combination consisting of morphemes is selected, and a search is performed on the morpheme combinations to provide a web page including a search result to the user terminal.

According to an embodiment of the present invention, the multi-term search term may be divided into morphemes, and then the morpheme token score may be applied to provide search results that most closely match the intention of the multi-term search term input by the user.

According to another embodiment of the present invention, when a search result corresponding to a multi-term search word is sufficient, a process of resetting the search word by separating the search word into morphemes may not be performed, thereby reducing the load on the system.

According to another embodiment of the present invention, by separating the multi-term search terms in stages, the search results that best match the search intention through the separation process that follows when the first search does not find a search result that matches the user's search intention Can be provided.

1 is a schematic block diagram of a search service providing system and its surroundings according to an embodiment of the present invention.
2 is a schematic flowchart of a token scoring method according to an embodiment of the present invention.
3 and 4 are diagrams showing examples of search word pairs in sessions, respectively.
5 to 7 are schematic flowcharts of a search service providing method according to an exemplary embodiment of the present invention, respectively.
8 is a schematic diagram of a search service providing system according to another embodiment of the present invention.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

Throughout the specification, when a part is said to "include" a certain component, it means that it can further include other components, without excluding other components unless specifically stated otherwise.

1 is a schematic block diagram of a search service providing system and its surroundings according to an embodiment of the present invention.

Referring to FIG. 1, the search service providing system 100 is connected to the user terminal 300 through the network 200, and the search unit 110, the search term processing unit 120, the score calculation unit 130, and the search are performed. The result providing unit 140 and the database 150 is included.

When the user accesses a web page provided by the search service providing system 100 through the user terminal 300 and inputs a search word, the search unit 110 searches for the search word. In this case, the search unit 110 may search by using an internal search engine or an external search engine. If the search result 110 does not have or lacks a search result that matches the search word, the search word processor 120 separates the search word into a plurality of morphemes through morphological analysis of the search word. Here, morphological analysis means dividing a search word into the smallest units having meanings.

The search term processor 120 matches the token score to each morpheme by using the morpheme token score stored in the database 150, and sets at least one morpheme combination having a high token score as a new search word. In addition, the search unit 110 searches for the search word reset by the search word processing unit 120. The search result providing unit 140 provides the user terminal 300 with a web page in which the search results performed on the initial search word and the reset search word are arranged.

The score calculation unit 130 analyzes a user's search word input in a plurality of sessions for a certain period, calculates a token score for each morpheme, and stores the token score in the database 150. Here, the session refers to a period in which a user of the user terminal maintains a search intention while inputting a search word. The score estimator 130 may calculate the token score as a value calculated for the various sessions to the extent that each morpheme was not deleted from the multiple searches of each session. That is, the score calculation unit may calculate the degree to which each morpheme is maintained as a token score of the corresponding morpheme in the search of several sessions. To this end, the score calculation unit 130 generates a search term pair consisting of two search terms that are in a subset relationship from a plurality of search terms inputted during the session when each session ends, and includes a predetermined morpheme in the search term pairs of several sessions. Token scores of the corresponding morphemes may be calculated as a ratio of the number of search term pairs and the number of search term pairs in which both search terms include a predetermined morpheme.

As such, the search service providing system 100 may newly set a search word corresponding to a user's search intention based on a token score indicating a degree of being maintained without being deleted in various sessions, thereby providing an accurate search result to the user. .

Next, a search service providing method in the search service providing system will be described in detail with reference to FIGS. 2 to 7.

First, a token score calculation method according to an embodiment of the present invention will be described in detail with reference to FIGS. 2 to 4.

FIG. 2 is a schematic flowchart of a token scoring method according to an embodiment of the present invention, and FIGS. 3 and 4 are diagrams showing examples of search word pairs in sessions, respectively.

Referring to FIG. 2, the score calculation unit 130 of FIG. 1 receives a search word from a plurality of user terminals (S210), and if a search word is not input from each user terminal for a predetermined time, the session of the corresponding user terminal ends. It is determined that (S220). If the session is not terminated, the check calculation unit 130 continues to receive a search word from the user terminal.

If it is determined that the session is terminated, the score calculation unit 130 generates a search term pair from the search terms input during the session (S230), and one search term among the generated search term pairs corresponds to a subset of other search terms. A search word pair is selected (S240). For example, if "Ganghwado", "Ganghwado pension", "Ganghwado gourmet", "Ganghwado travel", and "Ganghwado family pension" were entered as search terms during a session by a user (hereinafter referred to as "session 1"), As shown in FIG. 3, a pair of search terms in a subset relationship is "Ganghwado-Ganghwado Pension", "Ganghwado-Ganghwado Restaurant", "Ganghwado-Ganghwado Travel", "Ganghwado-Ganghwado Family Pension", "Ganghwado Pension-Ganghwado Family Pension" "This exists. In addition, if "Reinforcement Pension", "Reinforcement Family Pension", "Family Pension", Indoor Pension "is entered as a search term during another session (hereinafter referred to as" Session 2 "), as shown in FIG. The search term pairs in the subset relationship include "reinforcement pension-reinforcement family pension" and "family pension-reinforcement family pension".

The score calculation unit 130 calculates the token score for each morpheme based on the search term pairs of the various sessions collected for a certain period of time, and stores the calculated token score in the database (150 of FIG. 1) (S250). For example, as shown in Table 1, the score calculation unit 130 may calculate the token score by the ratio of the number of search word pairs remaining in the corresponding morpheme to the total number of search word pairs in which the corresponding morphemes are included for each morpheme.

For convenience of explanation, it is assumed that sessions collected for a period of time are sessions 1 and 2 shown in FIGS. 3 and 4, respectively. Then, the search term pairs containing the morpheme "Ganghwado" are "Ganghwado-Ganghwado Pension", "Ganghwado-Ganghwado Restaurant", "Ganghwado-Ganghwado Travel", "Ganghwado-Ganghwado Family Pension", "Ganghwado Pension-Ganghwado Family Pension", There are a total of seven, such as "Reinforcement Pension-Reinforced Family Pension" and "Family Pension-Reinforced Family Pension". The two pairs of these search term pairs have "Reinforcement Strength" and "Reinforcement Pension" -Ganghwado restaurant, "Ganghwado-Ganghwado travel", "Ganghwado-Ganghwado family pension", "Ganghwado pension-Ganghwado family pension", "Ganghwado pension-Ganghwado family pension" exist. In this case, the score calculator 130 may calculate a token score of "strength" as 6/7, that is, 0.857. Similarly, in Sessions 1 and 2, there are five search term pairs that contain the stem "Pensions" and three search term pairs that have "Pensions" left in both terms, so the score calculation 130 calculates the token score of "Pensions". 3/5, or 0.6. In the sessions 1 and 2, since there are four search term pairs containing the stem family "family" and one search term pair with "family" left in both terms, the score calculation unit 130 calculates the token score of "family". Calculate as 1/4, 0.25. Also, in sessions 1 and 2, there is one search term pair that includes the phrase "travel", and there are 0 search term pairs that have "travel" left in both terms, so the score calculation 130 calculates the token score of "travel". Calculate 0.

morpheme Number of query pairs Number of query pairs remaining Token score Ganghwado 7 6 0.857 pension 5 3 0.6 family 4 One 0.25 Travel One 0 0

As such, the score calculator 130 generates search term pairs for each session in various sessions collected for a certain period of time, and calculates a token score for each stem of the search term based on the search term pairs of the various sessions. Since the token score is generated based on a session in which the user inputs a search word with a search intention, the entire search intention of the user in the session may be reflected. That is, whether a user deletes a specific term from a search term or adds a term to a search term during a session, and whether the deletion or addition of a specific term occurs in the next search or after a few searches, the user's search during the session. Intent can be reflected in the token score.

Next, a method of providing a search service using the token score generated as described above will be described in detail with reference to FIGS. 5 to 7.

5 to 7 are schematic flowcharts of a search service providing method according to an exemplary embodiment of the present invention, respectively.

Referring to FIG. 5, a user accesses a web page provided by a search service providing system (100 of FIG. 1) through a user terminal (300 of FIG. 1) and inputs a search word (S510). After the search term processing unit 120 of FIG. 1 of the search service providing system 100 divides the search term into a plurality of morphemes through morpheme analysis of the search term received from the user terminal 300 (S520), tokens corresponding to each morpheme are provided. The score is found in the database (150 in FIG. 1) and matched to the corresponding morpheme (S530). The query processing unit 120 generates a morpheme combination consisting of at least one morpheme from the plurality of morphemes (S540), and compares the token scores of the morpheme combinations and selects the morpheme combination having the highest token score as the new search word (S550). In this case, the morpheme combination may be a combination between adjacent morphemes among a plurality of morphemes. For example, when a search term is separated from N morphemes, two morpheme combinations consisting of adjacent (N-1) morphemes may be generated from the N morphemes. The token score of the morpheme combination may be a combination of token scores of the morphemes included in the morpheme combination, for example, the sum of the token scores.

For example, when a user inputs a search term "Ganghwado Pension Trip" through the user terminal 300, the query processing unit 120 is a morpheme combination of "Ganghwado Pension" and "Pension Trip" from "Ganghwado Pension Trip". Can be generated. In this case, referring to the example of Table 1, the token score of the "reinforcement pension" is "0.857 + 0.6", that is, 1.457, and the token score of "pension trip" is "0.6 + 0", that is, 0.6, the query processing unit 120 Selects "reinforcement pension" as a new search term.

The search unit 110 performs a search on the new search term set by the search term processing unit 120 using a search engine (S560), and the search result providing unit (140 of FIG. 1) searches the search result of the search unit 110. Providing the arranged web page to the user terminal 300 (S570).

In this way, the search service providing system 100 may provide a search result that best matches the intention of the multi-term search word input by the user.

Next, referring to FIG. 6, unlike the embodiment described with reference to FIG. 5, the search unit 110 first searches for a search word input from the user terminal 300 using a search engine (S512). It may be determined whether S satisfies a predetermined condition (S514). The predetermined condition is a condition indicating whether a search result related to the search word is sufficient. For example, the predetermined condition may be a condition in which a search result whose relevance to the search word is equal to or greater than a threshold is a predetermined number or more.

If the search result of the search unit 110 does not satisfy a predetermined condition, the search service providing system 100 performs the process described with reference to steps S520 to S570 of FIG. 5. In this case, the search result providing unit 140 sorts the search result performed by the search unit 110 as the initial search term and the search result performed by the search term reset by the search term processing unit 120 on a web page, and sorts the corresponding web page. It may be provided to the terminal 300 (S571).

Meanwhile, when the search result of the search unit 110 satisfies a predetermined condition, the search result providing unit 140 sorts the search result performed by the search unit 110 as the first search term on the web page and displays the web page. It may be provided to the user terminal 300 (S580). In this case, the search word processor 120 may not perform the steps S520 to S571.

By doing so, the search service providing system 100 may not perform a process of resetting the search terms by separating the search terms into morphemes when the search results corresponding to the multi-term search terms are sufficient, thereby reducing the load on the system. have.

Next, referring to FIG. 7, unlike the embodiment described with reference to FIGS. 5 and 6, before providing a web page to the user terminal 300 (S571 and S570), the search unit 110 may generate a total search result. In operation S562, it may be determined whether the search result performed by the initial search word and the search result performed by the search word reset by the search word processor 120 satisfy a predetermined condition (S562).

If the total search result does not satisfy the predetermined condition, the search word processor 120 generates the morpheme combination again by reducing the number of morphemes forming the morpheme combination (S564, S541). Thereafter, the search service providing system 100 performs a search based on the newly generated morpheme combination to determine whether the total search result satisfies a predetermined condition (S550, S560, and S562). The search service providing system 100 may repeat the processes of steps S541, S550, S560, S562, and S564 until a sufficient search result is obtained.

For example, when the search word is divided into N morphemes, the search term processing unit 120 performs a search by selecting a morpheme combination consisting of (N-1) morphemes from the N morphemes, and then the total search result is a predetermined condition. If it is not satisfied, the search can be performed by selecting a morpheme combination consisting of (N-2) morphemes from N morphemes. If the total search result after the search also does not satisfy a predetermined condition, the search term processing unit 120 selects a morpheme combination consisting of (N-3) morphemes and performs a search. You can continue searching while reducing the number of.

On the other hand, while repeating the processes of steps S541, S550, S560, S562, and S564, if the total search result satisfies a predetermined condition, the search result providing unit 140 displays a web page in which the total search results are sorted. 300) (S571). On the contrary, even if the total search result does not satisfy the predetermined condition, when the search through the morphological separation is performed a predetermined number of times, the search service providing system 100 stops the additional search and displays the web page where the final search result is sorted. It may be provided to the user terminal 300.

As such, the search service providing system 100 divides the multi-term search terms step by step, so that even if the first search does not find a search result that matches the user's search intention, the search service that best matches the search intention is performed. Can be provided.

At least some functions of the search service providing system and method according to the embodiment of the present invention described above may be implemented in hardware or software coupled to the hardware. Hereinafter, an embodiment in which a search service providing system and method are coupled to a computer system will be described in detail with reference to FIG. 8.

8 is a schematic diagram of a search service providing system 800 according to another exemplary embodiment of the present invention. The search unit 110, the search term processing unit 120, and the score calculation unit 130 described with reference to FIGS. 1 to 5 are described. , A system that can be used to perform at least some of the functions of the search result providing unit 140 and the database 150.

Referring to FIG. 8, the search service providing system 800 may include a processor 810, a memory 820, at least a storage device 830, an input / output (I / O) interface 840, and a network interface 850. ).

The processor 810 may be implemented as a central processing unit (CPU), other chipsets, microprocessors, or the like, and the memory 820 may include dynamic random access memory (DRAM) and rambus DRAM (rambus DRAM). DRAM, RDRAM), synchronous DRAM (synchronous DRAM, SDRAM), static RAM (RAM, etc.) may be implemented as a medium such as RAM. The storage device 830 may include a hard disk, a compact disk read only memory (CD-ROM), a CD rewritable (CD-RW), a digital video disk ROM (DVD-ROM), a DVD-RAM, and a DVD-RW disk. The optical disk may be implemented as a permanent or volatile storage device such as an optical disk such as a blu-ray disk, a flash memory, or various types of RAM. The I / O interface 840 also allows the processor 810 and / or memory 820 to access the storage device 830, and the network interface 850 may be the processor 810 and / or memory 820. Allows access to the network 200.

In this case, the processor 810 may store program instructions for implementing at least some functions of the functions of the searcher 110, the search term processor 120, the score calculator 130, and the search result provider 140. ) And the functions of the database 150 may be located in the storage device 830 to control the operations described with reference to FIGS. 1 to 7. The program command may be stored in the storage device 830 or may be stored in another system connected to a network.

The processor 810, the memory 820, the storage device 830, the I / O interface 840, and the network interface 850 shown in FIG. 8 may be implemented in one computer or distributed in a plurality of computers. It may be implemented.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concepts of the present invention defined in the following claims are also provided. It belongs to the scope of rights.

Claims (21)

A search service providing method of a search service providing system,
Receiving a search word from a user terminal,
Separating the search term into a plurality of morphemes,
Matching each morpheme's token score with that morpheme,
Selecting a morpheme combination consisting of at least one morpheme among the plurality of morphemes based on the token score,
Performing a search with the morpheme combinations, and
Providing a web page including a search result performed by the morpheme combination to the user terminal.
Search service providing method comprising a. In claim 1,
The selecting step,
Generating a plurality of first morpheme combinations from the plurality of morphemes,
Setting a combination of token scores of the morphemes included in each first morpheme combination to the token scores of the first morpheme combinations,
Selecting a first morpheme combination having the largest token score among the plurality of first morpheme combinations as the morpheme combination
Search service providing method comprising a. In claim 2,
The combination of token scores is a sum of token scores. In claim 2,
And the number of morphemes included in each morpheme combination is one less than the number of the morphemes. In claim 2,
Determining whether to perform further search from the search result performed by the morpheme combination;
Generating a plurality of second morpheme combinations from the plurality of morphemes when the further search is necessary,
Setting the combination of the token scores of the morphemes included in each second morpheme combination as the token scores of the corresponding second morpheme combinations,
Selecting a second morpheme combination having the largest token score among the plurality of second morpheme combinations, and
Performing a search with the selected second morpheme combination
Search service providing method further comprising. In claim 5,
And a number of morphemes included in the second morpheme combination is less than a number of morphemes included in the first morpheme combination. In claim 6,
And a difference between the number of morphemes included in the second morpheme combination and the number of morphemes included in the first morpheme combination is one. In claim 5,
And the web page further comprises a search result performed by the selected second morpheme combination. In claim 1,
Performing a search with the search term, and
Determining whether an additional search is performed from the search results performed by the search term;
More,
The separating may include separating the search term into a plurality of morphemes when the additional search is required.
How to Provide Search Services. The method according to any one of claims 1 to 9,
The token score of each morpheme is a value indicating the degree to which the morpheme has been maintained in a search for a period of time. The method according to any one of claims 1 to 9,
Receiving a search word from each of the plurality of sessions,
Generating a search term pair consisting of two search terms that are in a subset relationship from each other at the end of each session,
Extracting a first search term pair including a predetermined morpheme from the search term pair of each session, and
Calculating the token score of the predetermined morpheme from the first search term pair of the plurality of sessions
Search service providing method further comprising. In claim 11,
The calculating step,
Selecting a first search term pair consisting of two search terms each including the predetermined morpheme in the first search term pair of the plurality of sessions as a second search term pair, and
Calculating the token score based on the number of the first search word pairs and the number of the second search word pairs;
Search service providing method comprising a. In claim 12,
And the token score corresponds to a ratio of the number of the second search word pairs to the number of the first search word pairs. A search word processor that separates the search word from the user terminal into a plurality of morphemes, matches token scores of the corresponding morphemes with each morpheme, and selects a morpheme combination consisting of at least one morpheme among the plurality of morphemes based on the token scores;
A database storing the token scores according to the morphemes;
A search unit which performs a search using the morpheme combinations, and
Search result providing unit for providing a web page including the search results of the search unit to the user terminal
Search service providing system comprising a. The method of claim 14,
And the search word processing unit selects a morpheme combination having the largest token score among a plurality of morpheme combinations generated from the morphemes. The method of claim 14 or 15,
The token score of each morpheme is a value indicating the degree to which the morpheme was maintained in the search for a period of time. The method of claim 14 or 15,
Receiving a search term from a plurality of sessions, and when each session ends, generates a search term pair consisting of two search terms that are in a subset relationship from the search term of the corresponding session, and includes a first morpheme in the search term pair of each session. And a score calculation unit for calculating the token score of the predetermined morpheme based on the search term pairs, and storing the token score in the database. The method of claim 17,
The score calculation unit calculates the token score based on the number of the first search term pairs of the plurality of sessions and the number of second search term pairs in which two search terms each of the first search term pairs includes the predetermined morpheme. Search service provision system. A search service providing method of a search service providing system,
Receiving a search word from each of the plurality of sessions,
Generating a search term pair consisting of two search terms that are in a subset relationship from each other at the end of each session,
Selecting a first search term pair including a morpheme from the search term pairs of the plurality of sessions,
Calculating the token score of the predetermined morpheme based on the first search term pair of the plurality of sessions,
Receiving a search word from a user terminal,
Dividing the search word from the user terminal into a plurality of morphemes;
Matching the token scores to the plurality of morphemes, respectively;
Selecting a morpheme combination consisting of at least one morpheme among the plurality of morphemes based on the token score,
Performing a search with the morpheme combinations, and
Providing a web page including a search result performed by the morpheme combination to the user terminal.
Search service providing method comprising a. The method of claim 19,
The calculating step,
Selecting a first search term pair consisting of two search terms each of which includes two search terms in the first search term pair of the plurality of sessions as a second search term pair, and
Calculating the token score based on the number of the first search word pairs and the number of the second search word pairs;
Search service providing method comprising a. The method of claim 19,
The selecting step,
Generating a plurality of morpheme combinations from the plurality of morphemes, and
Selecting a morpheme combination having the largest token score among the plurality of morpheme combinations
Search service providing method comprising a.

Images