KR101349969B1 - System and method for proving recommendation query - Google Patents

Abstract

A recommendation query providing system and method are disclosed. In the recommendation query providing system, a user interface unit receiving a query input by a user, a query determination unit determining an attribute of the received query, an index-based recommendation query search and a non-index-based recommendation query based on the determination result of the query A search method selection unit for selecting a recommended query search method corresponding to the received query word during the search; and a recommended query search unit searching for a recommended query according to the selected recommended query search method, wherein the user interface unit searches for the searched recommended query word. May be provided to the user.

Description

SYSTEM AND METHOD FOR PROVING RECOMMENDATION QUERY}

Embodiments of the present invention relate to a system and method for providing a recommended query, which automatically completes a query input from a user when providing a search service and provides the user with a completed query as a recommended query.

Recently, due to the spread of communication networks such as high-speed Internet and the rapid increase of users of high-speed communication network, development of new services and diversification of service items through the communication network are in progress. The most common service among the services using such a communication network is a search service.

When a query is input from a user, a search service corresponds to a search result corresponding to the input query (for example, a web site including the input query, an article including the input query, or an image having a file name including the input query). ) Is a service that provides the user.

Recently, various services such as a service for providing a recommended query for an input query and a service for providing a query associated with the input query have been provided in the search service in order to maximize user convenience. When the user inputs some phonemes or syllables constituting the query, the recommended query providing service provides the user with a completed query including the input phonemes or syllables.

As one of the technologies related to such a query providing service, Korean Patent Laid-Open Publication No. 2010-0069118 (published June 24, 2010), "Query query database construction method and query query method using the query database" query log Representative query and heteromorphic query are selected by normalizing the data, and click documents and the query log data corresponding to click log data are classified by interest topics, and the representative query and heteromorphic query words are indexed and stored together with the information on the topic of interest. A query index database construction method is described.

However, since the query recommendation method basically provides a recommendation query using an index database, when a query composed of a consonant that is a phoneme unit that is the basis of syllables is input from a user, this form is due to the variety of query words that can be combined into a consonant. There is a problem that the recommended query cannot be provided because all the input queries are not indexed for each type in anticipation, and the capacity of the database is inevitably increased.

Accordingly, even when a first query is input, a method of searching and providing a query stored in a database in real time may be considered to provide a recommended query. However, in this method, there is a problem that the recommendation query is slower than the method of indexing the recommendation query due to the heavy load on the server by generating the recommendation query by searching the database for each input query in real time.

Accordingly, there is a demand for a method for providing a recommended query that can provide a recommended query even though a first query is input.

There is provided a recommendation query providing system and method capable of providing a recommendation query with a faster response speed even when the user inputs a query composed of only the first constellation or a query composed of the consonants and letters.

The recommended query providing system includes a user interface unit for receiving a query input by a user, a query determination unit for analyzing the property of the received query to determine whether the received query is an initial query, and an index based on the determination result of the query. A search method selection unit for selecting a recommendation query search method corresponding to the received query word based on a search for a recommended recommendation query and a search for a non-index based recommendation query, and a search query search unit for searching a recommendation query according to the selected recommendation query search method; The user interface unit may provide the searched recommendation query word to the user.

delete

According to one side, the initial query may be a query including one or more Hangul query consisting of the initial except the last letter.

According to another aspect, the search method selection unit selects the non-index-based recommended query search method as the recommended query search method corresponding to the received query when the received query is determined to be an initial query, and the received query is If it is determined that the query is not the first query, the index-based recommendation query search method may be selected as a search method corresponding to the received query word.

According to another aspect, the non-index based recommendation query search method may be a method of automatically generating the candidate query by automatically completing the received query based on a graph in which the query words are arranged according to the initial input order.

According to another aspect, the index-based recommendation query search method may be a method of searching for candidate query words corresponding to the received query words among candidate query words classified and indexed according to phoneme units, syllable units, or suffixes. .

According to another aspect, the method may further include a recommendation query selection unit that selects a recommendation query to be recommended to the user from among the candidate query based on the priority of the candidate query.

According to another aspect, the priority may be determined according to a search frequency of the candidate query.

In the method for providing a recommendation query, the recommendation query providing system provides a recommendation query, the method comprising: receiving a query input by a user, analyzing an attribute of the received query, and determining whether the received query is an initial query; Selecting a recommendation query search method corresponding to the received query word among an index-based recommendation query search and a non-index-based recommendation query search based on the determination result of the query, and search for a recommendation query word according to the selected recommendation query search method And providing the searched recommendation query to the user.

Even when the user enters a query consisting of only the first consonant or a query containing a mixture of both initial and letter, the user can select the search method that corresponds to the attribute of the input query during index-based recommendation query search and non-index-based recommendation query search. By providing a, even when a user inputs a query in various ways, the recommendation query can be provided and the recommendation query can be provided at a faster response speed.

1 is an exemplary diagram for describing a method of searching for an index-based recommendation query according to an embodiment of the present invention.
2 is an exemplary view for explaining an index-based recommendation query search method according to an embodiment of the present invention.
3 is a block diagram illustrating a recommendation query providing system according to an exemplary embodiment of the present invention.
4 is an exemplary view illustrating a screen on which a recommendation query is provided through a user interface unit according to an embodiment of the present invention.
5 is a flowchart illustrating a recommendation query providing method according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The recommendation query providing system according to the present invention provides a recommendation query by selecting a search method corresponding to a query received from a user from a non-index-based recommendation query search method and an index-based recommendation query search method.

The non-index-based recommendation query search method is a method of generating a candidate query by automatically completing a query received from a user based on a graph in which the query words are arranged according to the initial input order. A candidate query corresponding to a query received from a user is searched for among candidate candidates classified and indexed by syllable units or suffixes.

1 is an exemplary diagram for describing a method of searching for an index-based recommendation query according to an embodiment of the present invention. FIG. 1 illustrates a structure in which query words "next", "crucible", "lunch box", "Dobongsan" and "Dobonggu", which begin with the initial "C", are arranged on the graph. In the figure, "S" indicates a starting point at which the query word is input and "E" indicates a point at which input of the query word is terminated. Hereinafter, a method for searching for a recommended query using a graph based on the first Korean Hangul is described as an example. However, the present invention is not limited thereto, and the recommended query using a graph based on an English consonant or a vowel is described. You can also explore.

First, referring to the initial entry order of each query with reference to the drawings, the initial consonants of the query "next" is "c" and "b", the initials of the query "drug" "c", "ㄱ" and " B ", and the initiality of the query word" lunch "is" c "," g ", and" d ". In addition, the initiality of the query word "Dobongsan" is "c", "ㅂ", and "ㅅ", and the initiality of the query word "Dobonggu" is "c", "ㅂ", and "a".

In this case, as shown in FIG. 1, since all five query words start with the letter "c", it can be seen that the five query words are recorded on one root up to the letter "c" on the graph.

However, the query "next" enters the initial "o" after the initial "c", but the query "done" enters the initial "a" after the initial "c", It can be seen that the query term "dogini" branches after the initial "C" and the query terms "Dobong-gu" and "Dobongsan" branch after the initial "ㅂ".

Accordingly, the non-index-based recommendation query search method is a real-time process of generating a plurality of candidate query words by automatically completing a query word received from a user based on a graph in which query words are arranged according to the initial input order as described above. Search for candidate queries for the recommended query, so even if the user enters the query in a variety of ways, such as a first-word only query, a first- and first-word mixed query, or an English consonant or vowel only You can provide a query.

For a more detailed description of the non-index based candidate query search method, please refer to Korean Patent Publication No. 10-1016519 (Registration date February 14, 2011) of the applicant "Recommended query method and system" and Korea registration Refer to Patent Publication No. 10-1023911 (March 14, 2011) "Referred query method and system" can be referred to.

2 is an exemplary view for explaining an index-based recommendation query search method according to an embodiment of the present invention.

As shown in FIG. 2, in the case of searching for a recommended query based on an index, for example, the query index database may include a phoneme index and a search frequency for each phoneme unit, a query word, and a query word. For example, the question mark "K" is recorded as the query word, "Kookmin Bank", "game", etc., and the question mark "B" is recorded as the query word "Naver", "Nexon", etc. For the unit "c", "next", "dokdo", etc. may be recorded as the query words. And, for the subunit "party", "next", "donkey", etc. may be recorded as the query words. In addition, each query may be recorded in the query index database in the order of its search frequency, or alternatively, it may be recorded in the query index database in alphabetical order.

In the drawing, for example, the candidate query is indexed in the query index database in units of phonemes. However, the present invention is not limited thereto, and the candidate query may be classified and indexed according to syllable units or suffixes.

Accordingly, the index-based recommendation query search method can quickly search for and provide a recommendation query by searching for a candidate query corresponding to the query received from the user based on the queries recorded in the query index database as described above.

For a more detailed description of such an index-based candidate query search method, see Korean Patent Application Publication No. 10-0707725 (published April 9, 2007) of the present applicant. Computer-readable recording medium having recorded thereon a program for execution.

3 is a block diagram illustrating a recommendation query providing system according to an exemplary embodiment of the present invention.

Referring to the drawings, the recommendation query providing system 300 may include a user interface 310, a query determination unit 320, a search method selection unit 330, a recommendation query search unit 340, and an index-based candidate query generation unit ( 350, the index-based candidate query generator 360 and the recommended query selector 370 may be included.

The user interface 310 receives a query input by a user through a user terminal such as a PC, a tablet PC, a smart phone, etc., and transmits the query to the query determination unit 320 while searching for the recommendation query providing system 300. The recommendation query is provided to the user through the user terminal.

The query determining unit 320 determines the attribute of the query received from the user interface 310. For example, the query determining unit 320 may provide a user with a recommendation query even when the user inputs a query in various ways, such as when a user inputs a query composed of only the first consonants or a query including a mixture of first and letters. The received query received from the interface 310 may be analyzed to determine whether the query is an initial query. Here, the initial query word may be a query word including one or more Hangul queries composed of the first letter except the last letter.

For example, when the query received from the user is "b", "yes", etc., since the query may provide a recommendation query using an index-based query search method, the query determination unit 320 may select the query. May not be judged by a query. On the other hand, if the query received from the user is "n", "n", etc., the query may include one or more Hangul queries consisting of first words except the last letter, and thus may be determined as a first one.

The search method selection unit 330 selects a recommendation query search method corresponding to the query received during the index-based recommendation query search and the non-index-based recommendation query search based on the attributes of the query determined by the query determiner 320. For example, when the received query is determined to be an initial query, the search method selection unit 330 selects the non-index based recommendation query search method described with reference to FIG. 1 as the recommended query search method corresponding to the received query. If it is determined that the received query is not the first query, the index-based recommendation query search method described with reference to FIG. 2 may be selected as the search method corresponding to the received query.

The recommended query search unit 340 searches for the recommended query according to the recommended query search method selected by the search method selection unit 330.

The non-index based candidate query generation unit 350 may provide a plurality of candidate query terms for providing a recommended query for the received query using a non-index based recommendation search based on the query received through the user interface 310. Create

The index-based candidate query generator 360 may query a plurality of candidate queries for providing a recommendation query for the received query by using the index-based suggestion search method based on the query received through the user interface 310. Search in the index database.

For example, when the recommended query search method selected by the search method selection unit 330 is a non-index based suggestion search method, the recommended query search unit 340 receives an input from the user to the non-index based candidate query word generation unit 350. The query may be transmitted to generate a plurality of candidate query. On the other hand, if the recommended query search method selected by the search method selection unit 330 is an index-based search term search method, the candidate indexed in the query index database by transmitting the query input from the user to the index-based candidate query generation unit 360. In the query, a plurality of candidate queries corresponding to the query input from the user may be searched.

The recommendation query selecting unit 370 selects a recommendation query to be recommended to the user among the plurality of candidate query words searched by the recommendation query searching unit 340 based on the priority of each query. In this case, the priority of each query may be determined according to a search frequency of each candidate query.

4 is an exemplary view illustrating a screen on which a recommendation query is provided through a user interface unit according to an embodiment of the present invention.

Referring to FIG. 4, when a user inputs an initial query such as “ㅇ ㅇ” into the search box, the system for providing a query query according to the present invention may include a plurality of “Oriental Cement”, “German Time”, “Great Cinema”, and the like. You can provide a recommended query.

Accordingly, the system for providing a recommendation query system according to the present invention is an index-based recommendation query search method and a non-index-based recommendation even when a user inputs a query composed of only a first consonant or a query mixed with a consonant and a letter due to the configuration described above. It is possible to provide a suggested query by selecting a search method that corresponds to the attribute of the input query during query search. Therefore, even if a user inputs a query in various ways, it is possible to provide a suggested query but faster than entering a general query. You can provide a suggested query with a response rate.

5 is a flowchart illustrating a recommendation query providing method according to an embodiment of the present invention.

The recommendation query providing system according to the present invention first receives a query input by the user from the user terminal in order to provide the recommendation query for the query input by the user (S510). In order to select the optimal recommended query search method according to the attribute of the received query, the attribute of the received query is determined (S520).

Thereafter, among the index-based recommendation query search and the non-index-based recommendation query search based on the determined attribute of the query, a recommended query search method corresponding to the received query is selected. As an example, if the received query is the initial query, the recommendation query providing system may search for a plurality of candidate queries to be provided as the recommendation query in a non-index based recommendation query search method (S540). However, when the received query is not the initial query (S530), a plurality of candidate queries to be provided as the recommended query may be searched through the index-based recommendation query search method (S550).

Thereafter, the recommendation query providing system may select recommendation query words to be finally provided to the user based on the priority of the searched candidate query words (S560), and transmit the selected recommendation query words to the user terminal so as to be provided to the user. S570).

The method for providing a recommendation query according to the present invention may be implemented in the form of program instructions that can be executed by various computer means and recorded in a computer readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

As described above, the present invention has been described by way of limited embodiments and drawings, but the present invention is not limited to the above embodiments, and those skilled in the art to which the present invention pertains various modifications and variations from such descriptions. This is possible.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined by the equivalents of the claims, as well as the claims.

300: recommendation query system
310: user interface unit
320: query determination unit
330: navigation method selection unit
340: recommended query search unit
350: non-index based recommendation query generator
360: index-based recommendation query generator
370: recommendation query selection unit

Claims (17)

A user interface to receive a query input by a user;
A query determiner that determines whether the received query is an initial query by analyzing attributes of the received query;
A search method selection unit for selecting a recommendation query search method corresponding to the received query from an index-based recommendation query search method and a non-index based recommendation query search method different from the index-based recommendation query search method; And
Recommended query search unit for searching for a recommended query in accordance with the selected recommended query search method
Lt; / RTI >
The search method selection unit,
If the received query is determined to be an initial query, the non-index based recommendation query search method is selected as the recommended query search method corresponding to the received query, and the index is determined to be not the first query. Selects a recommendation query corresponding to the received query as a search method
The user interface unit,
And providing the searched recommendation query to the user. delete The method of claim 1,
The initial query word,
Suggested query providing system, characterized in that the query containing at least one Hangul query consisting of the initial concluding the last letter. delete The method of claim 1,
The non-index based recommendation query search method,
And a method of generating a candidate query by automatically completing the received query based on a graph in which query terms are arranged in an initial order of input. The method of claim 1,
The index-based recommendation query search method,
And a candidate query word corresponding to the received query word among candidate query words classified and indexed according to a phoneme unit, a syllable unit, or a suffix. The method according to claim 5 or 6,
And a recommendation query selection unit for selecting a recommendation query to be recommended to the user among the candidate queries based on the priority of the candidate query. The method of claim 7, wherein
The priority order may be,
System for providing a recommended query, characterized in that determined according to the search frequency of the candidate query. In the method for providing a recommendation query system providing a recommendation query,
Receiving a query input by a user;
Analyzing the attributes of the received query to determine whether the received query is an initial query;
Selecting a recommendation query search method corresponding to the received query from an index-based recommendation query search method and a non-index-based recommendation query search method different from the index-based recommendation query search method based on the determination result for the query;
Searching for a recommended query in accordance with the selected recommended query search method; And
Providing the searched recommendation query word to the user
Lt; / RTI >
Selecting the recommended query search method
If the received query is determined to be an initial query, the non-index-based recommendation query search method is selected as the recommended query search method corresponding to the received query, and if it is determined that the received query is not the first query, the index And selecting a recommendation query corresponding to the received query as a search method. delete 10. The method of claim 9,
The initial query word,
Method for providing a recommendation query, characterized in that the query containing at least one Hangul query consisting of the initial consonants except the last letter. delete 10. The method of claim 9,
The non-index based recommendation query search method,
And a method of generating a candidate query by automatically completing the received query based on a graph in which query terms are arranged in an initial order of input. 10. The method of claim 9,
The index-based recommendation query search method,
And searching for a query corresponding to the received query among candidate queries that are classified and indexed according to a phoneme unit, a syllable unit, or a suffix. 10. The method of claim 9,
After the searching step,
And selecting a recommendation query to be recommended to the user from among the candidate query words based on the priorities of the plurality of candidate query words searched according to the selected recommendation query search method. 16. The method of claim 15,
The priority order may be,
The method of claim 1, wherein the candidate query is determined according to a search frequency of the candidate query. A computer-readable recording medium having recorded thereon a program for performing the method for providing a recommendation query according to any one of claims 9, 11, or 13 to 16.

Images