KR100828560B1 - Method for recommending words associated with an object of search and system for executing the method - Google Patents

Abstract

The present invention relates to a method for recommending a word associated with a search object and a system for performing the method. According to an aspect of the present invention, there is provided a method of recommending a word associated with a search object, the method comprising: measuring recommendation information about a word included in a search object, calculating a word weight for the word based on the recommendation information, and the word weight Recommending the word as a query for the search subject based on the < RTI ID = 0.0 >

Book Search, Words, Word Suggestions, Weights

Description

METHOD FOR RECOMMENDING WORDS ASSOCIATED WITH AN OBJECT OF SEARCH AND SYSTEM FOR EXECUTING THE METHOD}

1 is a view showing an example of a method for searching the text content of a book in the prior art.

2 is a flowchart illustrating a method of recommending a word associated with a search object according to an embodiment of the present invention.

3 is a flowchart illustrating a method of measuring recommendation information according to an embodiment of the present invention.

4 is a flowchart illustrating a method of calculating word weights according to an embodiment of the present invention.

5 is an example of a web page for recommending words included in the text content of a book.

6 is an example of a search results page provided for the selection of suggested words.

7 is a block diagram illustrating an internal configuration of a word recommendation system according to another embodiment of the present invention.

8 is a block diagram illustrating an internal configuration of a word weighting unit according to another exemplary embodiment of the present invention.

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

700: word recommendation system

701: recommended information measuring unit

702: word weighting unit

703: word recommendation

801: first word weight generator

802: second word weight generation unit

803: word weight calculation unit

The present invention relates to a method for recommending a word associated with a search object and a system for performing the method. More particularly, the word occurrence frequency of the word for a word included in the search object, a document of a document including the word. The present invention relates to a method of calculating a word weight using a length and the like, and to recommend the word for each document to be searched based on the word weight and a system for performing the method.

A search engine is a software or website that makes it easy to find material on the Internet. At this time, the search engine serves to provide data on the query of the user through the website.

In a book search using such a search engine, the related art provides a function of searching for a book by searching a book through a query for a book of interest and re-searching the text content of the found book. 1 is a view showing an example of a method for searching the text content of a book in the prior art.

Reference numeral 100 denotes a web page for searching again within the body content of the searched book. Such a web page may provide the user with a portion 102 of body content that includes the user's query entered through the search field 101. In addition, a link to the first page of the table of contents 103 may be set in the table of contents 103 of the book so that the user may check the contents of the text, or the page moving button 104 may be used to select the previous or next page. Button 105 may be provided.

However, in a book search using such a search engine, users may have difficulty in selecting a query for a search when searching a book because they do not know what words are in the book. That is, when trying to search the text content of the book to obtain information about the book or to find desired information in the book, there is a problem in that it is difficult to find out what query to use.

The present invention proposes a new technique related to a method for recommending a word associated with a search object and a system for performing the method, in order to solve the above problems of the prior art.

An object of the present invention is to facilitate a search for a search target by recommending a word included in the search target to a user when performing a search within the search target.

Another object of the present invention is to calculate a word weight for a word included in the document for each document included in the search target and recommend the word to the user in the order of the high word weight.

Another object of the present invention is to calculate the word weight based on the frequency of appearance of the word, the document length of the document and the number of documents in which the word appears in the entire document, etc. To have a higher word weight for.

In order to achieve the above object and to solve the above-mentioned problems of the prior art, a method for recommending a word associated with a search object according to an embodiment of the present invention, measuring the recommendation information for the words included in the search object Calculating a word weight for the word based on the recommendation information; and recommending the word as a query for the search target based on the word weight.

According to one aspect of the invention, the recommendation information is (1) the document length of the document containing the word, (2) the total number of documents for the entire document corresponding to the search object, (3) the frequency of word appearance of the word (4) the word appearance document number, which is the number of documents in which the word appears, or (5) the average length of documents for the entire document.

According to another aspect of the invention, the step of measuring the recommendation information for the words included in the search object, the step of measuring the word appearance frequency, the document length and the word appearance document number through the document, the document And calculating the total document length by using the document length measured for each of the plurality of documents, and calculating the average document length based on the total document length and the total number of documents.

According to another aspect of the invention, the step of measuring the word appearance frequency, the document length and the word appearance document number through the document, measuring the document length for any one of the entire document A first step, a second step of measuring the frequency of word appearance for each word included in the document, a third step of updating the number of word appearance documents for each word, and the first to third steps in the entire document A fourth step may be repeated.

According to another embodiment of the present invention, a word recommendation system for recommending a word associated with a search object may include: a recommendation information measuring unit configured to measure recommendation information about a word included in a search object; A word weighting unit for calculating a word weight and a word recommending unit for recommending the word as a query for the search target based on the word weight.

Hereinafter, various embodiments of the present disclosure will be described in detail with reference to the accompanying drawings.

2 is a flowchart illustrating a method of recommending a word associated with a search object according to an embodiment of the present invention.

In operation S210, a word recommendation system for recommending a word associated with a search object as a query for the search object measures recommendation information about the word included in the search object. In this case, the recommendation information is (1) the document length of the document including the word, (2) the total number of documents for the entire document corresponding to the search object, (3) the frequency of word appearance of the word, (4) the It may include the word appearance document number which is the number of documents in which the word appeared or (5) the document average length for the entire document.

In addition, the search object may include a text content of a book composed of a plurality of documents, and the text content may be classified into the document based on the table of contents of the book. In addition, the word may include an extraction result extracted from the search target using a noun extractor.

The method of measuring the recommendation information will be described in more detail with reference to FIG. 3.

In operation S220, the word recommendation system calculates a word weight for the word based on the recommendation information. That is, the word recommendation system may calculate the word weight using the document length, the total number of documents, the frequency of word appearance, the number of word appearance documents, or the average length of documents. The step S220 of calculating the word weight based on the recommendation information will be described in more detail with reference to FIG. 4.

In step S230, the word recommendation system recommends the word as a query for the search target based on the word weight. At this time, step (S230) is a step of listing the words included in the document for each document included in the search target based on the word weight (S231), the word weight among the words listed in the order of high Extracting the word by a predetermined number (S232) and when the document is exposed to a search page, providing the extracted word as a query for the document (S233). It may include.

That is, the word recommendation system calculates the word weight for each word by using the recommendation information for each of a plurality of documents in which the body content is classified through the table of contents for searching in the result of the text content of the book. The word included in the document selected by the user may be recommended according to the word weight.

As such, in the word recommendation method, when a search is performed within the search target through the same process as in steps S210 to S230, the search for the search target is facilitated by recommending a word included in the search target to the user. It can be done. An example of recommending a word using the word recommendation method and an example of providing a search result through the recommended word will be described in detail with reference to FIGS. 5 and 6.

3 is a flowchart illustrating a method of measuring recommendation information according to an embodiment of the present invention. As illustrated in FIG. 3, steps S310 to S330 may be performed by being included in step S210 described with reference to FIG. 2.

In step S310, the word recommendation system measures the word appearance frequency, the document length, and the word appearance document number through the document. In this case, as shown in FIG. 3, step S310 may include steps S311 to S316.

In step S311, the word recommendation system measures the document length for any one of the entire documents. In this case, the document length may be measured through the byte size of the character string of the document. Such document length may be repeatedly accumulated through the step S314 of repeatedly performing the steps S311 to S313 and used to measure the entire document length.

In step S312, the word recommendation system measures the frequency of word appearance for each word included in the document. That is, the word appearance frequency may be measured by measuring the number of times included in the document for each word.

In step S313, the word recommendation system updates the word appearance document number for each word. In this case, the word appearance document number may include the number of documents in which the word appears for each word included in the entire document, and updating the number of word appearance documents for each word (S313) includes the document for each word. Or counting the number of times or adding a document in which the word appears to a document list associated with the word.

In the steps S311 to S313, the word appearance frequency of the extracted words may be measured while extracting the words included in the document one by one using the noun extractor described in FIG. 2. The number of appearance documents can be updated. The document length may be measured by checking the byte size of the character string of the document stored at the time of generating the document.

In step S314, the word recommendation system measures the document length and the word occurrence frequency for the entire document to repeat steps S311 to S313 for the entire document, and the word appearance document Update the count. At this time, the word recommendation system performs steps S311 to S313 and then performs the entire document.

In step S315, the word recommendation system measures the number of times the steps S311 to S313 are repeatedly performed as the total number of documents. That is, each time steps S311 to S313 are repeated in step S314, the total number of documents may be measured by measuring the number of repetitions.

In step S316, the word recommendation system checks the number of the last word appearance document updated. The word appearance document number may be checked through the number of documents counted for each word as described in step S313 or through the number of documents included in the document list associated with the word.

In operation S320, the word recommendation system calculates an entire document length using the document length measured for each document. That is, the total document length may be calculated by repeatedly obtaining the document length accumulated as the total document length through the repeating of step S311 or by summing the document lengths measured and stored for each document, respectively.

In step S330, the word recommendation system calculates the document average length based on the total document length and the total number of documents. More specifically, the ratio of the total document length to the total number of documents can be calculated and measured as the document average length.

As described above, through steps S310 to S330, (1) the document length of the document including the word, (2) the total number of documents for the entire document corresponding to the search target, and (3) the word of the word. The recommendation information including the frequency of appearance, (4) the number of word appearance documents which is the number of documents in which the word appeared, or (5) the average length of documents for the entire document, can be measured, and using the recommendation information The word weight described with reference to FIG. 2 may be calculated.

4 is a flowchart illustrating a method of calculating word weights according to an embodiment of the present invention. As illustrated in FIG. 4, steps S410 to S430 may be performed by being included in step S220 described with reference to FIG. 2.

In step S410, the word recommendation system generates a first word weight for the document based on the word occurrence frequency, the document length of the document, and the total number of documents. In this case, the generating of the first word weight (S410) includes calculating the ratio of the document length of the document length to the document length for each document and adding the word appearance frequency (S411) and the word appearance frequency. The method may include generating a ratio of the sum of word appearance frequencies as the first word weight (S412). The first word weight may be expressed as Equation 1 below.

In step S420, the word recommendation system generates a second word weight for the document based on the number of word appearance documents and the total number of documents. In this case, the word appearance document number may include the number of documents in which the word appears for each word included in the entire document, and generating the second word weight (S420) includes the number of the word appearance document number. It may be a step of generating a ratio of the total number of documents as the second word weight. The second word weight may be expressed as Equation 2 below.

In operation S430, the word recommendation system calculates the word weight for each document based on the first word weight and the second word weight. In this case, the word recommendation system may calculate the word weight through a multiplication operation of the first word weight and the second word weight. That is, the word weight is given by Equation 3 below.

As such, the word weight may be calculated for the words included in the document for each document included in the search target, and the words may be recommended to the user in the order of the high word weights.

Further, the word is higher for the word with the higher frequency according to the document length by calculating the word weight based on the frequency of appearance of the word, the document length of the document, the number of documents in which the word appeared in the entire document, and the like. It can be weighted. This word weight can be summarized as shown in Equation 4 below.

By calculating the word weight for each document for the entire document as described above, the word for each document can be recommended to the user through the word weight for the word included in the document.

5 is an example of a web page for recommending words included in the text content of a book.

Reference numeral 500 denotes a web page for recommending words included in the text content by using the text content of the book as a search target. As described with reference to FIGS. 2 through 4, the web page may include a search field 501 for searching in the body content, and is currently present to the user to conveniently provide such a search in the body content to the user. The word 503 included in the document 502 appearing on the web page may be extracted and provided using word weights.

That is, the user can easily perform the search of the document 502 by using the words 503 so recommended. At this time, the search of the document 502 may be performed by entering a word 503 through the search field 501 or selecting a desired word 504 among the recommended words 503. For example, the user may click on a word (“sequence”) 504 to obtain a search result for the word 504.

6 is an example of a search results page provided for the selection of suggested words.

Reference numeral 600 denotes a search result page that provides a user with a search result 601 for the word (“sequence”) 504 described in FIG.

That is, the contents of the book are classified into a plurality of documents through the table of contents of the book, and the words included in the document 504 selected in the document are recommended by recommending words included in the document selected by the user through word weights. It may serve as a search result 601. In addition, the user selects a desired search result 602 from among a plurality of search results 601, so that the user can obtain a desired search result from the text content more conveniently.

As described above, when a search is performed within the search target through the book recommendation method described with reference to FIGS. 2 to 6, the user can easily search for the search target by recommending a word included in the search target to the user. The word weight may be calculated for the words included in the document for each document included in the search target, and the words may be recommended to the user in the order of the high word weights.

Further, the word is higher for the word with the higher frequency according to the document length by calculating the word weight based on the frequency of appearance of the word, the document length of the document, the number of documents in which the word appeared in the entire document, and the like. It can be weighted.

7 is a block diagram illustrating an internal configuration of a word recommendation system according to another embodiment of the present invention. In this case, the word recommendation system 700 for recommending a word for a search target to the user may include a recommendation information measuring unit 701, a word weighting unit 702, and a word recommending unit 703.

The recommendation information measuring unit 701 measures recommendation information about a word included in a search target. In this case, the recommendation information is (1) the document length of the document including the word, (2) the total number of documents for the entire document corresponding to the search object, (3) the frequency of word appearance of the word, (4) the It may include the word appearance document number which is the number of documents in which the word appeared or (5) the document average length for the entire document.

In addition, the search object may include a text content of a book composed of a plurality of documents, and the text content may be classified into the document based on the table of contents of the book. In addition, the word may include an extraction result extracted from the search target using a noun extractor. That is, the recommendation information measuring unit 701 may measure the recommendation information for each of a plurality of documents in which the main content is classified according to the table of contents.

To this end, the recommendation information measuring unit 701 measures the word appearance frequency, the document length, and the word appearance document number through the document and calculates the total document length using the document length measured for each document. The recommendation information may be measured by calculating the document average length based on the total document length and the total number of documents. In this case, the word appearance frequency may be measured for each word included in the document while sequentially processing the plurality of documents one by one.

In addition, the word appearance document number may be measured by counting the number of documents for a specific word for each document or by adding a document in which the word appears to a document list associated with the word. That is, after repeating the above process for the entire document, the word appearance document number may be confirmed by checking the number of the counted documents or the number of the documents included in the document list. In addition, the recommendation information measuring unit 701 may measure the number of iterations and confirm the total number of documents.

The word weighting unit 702 calculates a word weight for the word based on the recommendation information. That is, the word weight for each document may be calculated using the recommendation information measured by the recommendation information measuring unit 701 and used as a reference for recommending a word as a query for the document. The word weighting unit 702 for calculating such word weights will be described in more detail with reference to FIG. 8.

The word recommender 703 recommends the word as a query for the search target based on the word weight. To this end, the word recommendation unit 703 lists the words included in the document for each document included in the search object based on the word weights, and writes the words in the order of the higher word weights among the listed words. When the word is extracted by a predetermined number and the document is exposed to a search page, the extracted word may be provided as a query for the document.

8 is a block diagram illustrating an internal configuration of a word weighting unit in another embodiment of the present invention. As illustrated in FIG. 8, the word weighting unit 702 may include a first word weight generation unit 801, a second word weight generation unit 802, and a word weight calculation unit 803.

The first word weight generator 801 generates a first word weight for the document based on the word occurrence frequency, the document length of the document, and the total number of documents. In order to generate the first word weight, the first word weight generation unit 801 calculates a ratio of the document average length of the document length to each word, adds the word appearance frequency, and adds the word appearance frequency. The ratio of the sum of word occurrence frequencies may be calculated and generated as the first word weight.

The second word weight generator 802 generates a second word weight for the document based on the number of word appearance documents and the total number of documents. In this case, the word appearance document number may include the number of documents in which the word appears for each word included in the entire document, the second word weight generation unit 802 is the total number of the word appearance document number The ratio of the number of documents may be calculated and generated as the second word weight.

The word weight calculator 803 may generate the first word weight generator 801 through the first word weight generator 801 and the second word weight generator 802 through the second word weight generator 802. The word weight may be calculated for each document. That is, the word weight calculator 803 may calculate the word weight through a multiplication operation of the first word weight and the second word weight.

As such, when a search is performed within the search target through the word recommendation system 700 according to the present invention, the user can easily search for the search target by recommending a word included in the search target to the user. The word weight may be calculated for the words included in the document for each document included in the search target, and the words may be recommended to the user in the order of the high word weights. Further, the word is higher for the word with the higher frequency according to the document length by calculating the word weight based on the frequency of appearance of the word, the document length of the document, the number of documents in which the word appeared in the entire document, and the like. It can be weighted.

Embodiments according to the present invention can be implemented in the form of program instructions that can be executed by various computer means can be recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the media may be those specially designed and constructed for the purposes of the present invention, or they may be of the kind well-known and available to those having skill in the computer software arts. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, 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 not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

In the present invention as described above has been described by the specific embodiments, such as specific components and limited embodiments and drawings, but this is provided to help a more general understanding of the present invention, the present invention is not limited to the above embodiments. For those skilled in the art, various modifications and variations are possible from these descriptions.

Therefore, the spirit of the present invention should not be limited to the described embodiments, and all the things that are equivalent to or equivalent to the claims as well as the following claims will belong to the scope of the present invention. .

According to the present invention, when a search is performed within a search target, the user can easily search for the search target by recommending a word included in the search target to the user.

According to the present invention, the word weight is calculated for a word included in the document for each document included in the search target, and the word is recommended to the user in the order of the high word weight.

According to the present invention, the word weight is calculated according to the document length by calculating the word weight based on the frequency of appearance of the word, the document length of the document, the number of documents in which the word appears in the entire document, and the like. Can have a higher word weight.

Claims (20)

In the method for suggesting a word associated with a search object, Measuring recommendation information for a word included in a search target; Calculating a word weight for the word based on the recommendation information; And Recommending the word as a query for the search target based on the word weight Including, The step of recommending the word as a query for the search target based on the word weight, Listing words included in the document for each document included in the search object based on the word weights; Extracting the word from the listed words by a predetermined number in order of increasing the word weight; And If the document is exposed to a search page, providing the extracted words as a query for the document Word recommendation method comprising a. The method of claim 1, The recommendation information includes (1) the document length of the document including the word, (2) the total number of documents for the entire document corresponding to the search object, (3) the frequency of word appearance of the word, and (4) the word. A word appearance document number, which is the number of document appearances, or (5) a document average length for the entire document. The method of claim 2, The step of measuring the recommendation information for the words included in the search target, Measuring the word appearance frequency, the document length, and the word appearance document number through the document; Calculating a total document length using the document length measured for each document; And Calculating the document average length based on the total document length and the total number of documents Word recommendation method comprising a. In the method for suggesting a word associated with a search object, Measuring recommendation information for a word included in a search target; Calculating a word weight for the word based on the recommendation information; And Recommending the word as a query for the search target based on the word weight Including, The recommendation information includes (1) the document length of the document including the word, (2) the total number of documents for the entire document corresponding to the search object, (3) the frequency of word appearance of the word, and (4) the word. Word appearance document number, which is the number of documents that appeared, or (5) a document average length for the entire document, The step of measuring the recommendation information for the words included in the search target, Measuring the word appearance frequency, the document length, and the word appearance document number through the document; Calculating a total document length using the document length measured for each document; And Calculating the document average length based on the total document length and the total number of documents Including, The step of measuring the word appearance frequency, the document length and the word appearance document number through the document, A first step of measuring the document length for any one of the entire documents; A second step of measuring the word occurrence frequency for each word included in the document; Updating the number of word appearance documents for each word; And A fourth step of repeating the first to third steps with respect to the entire document Word recommendation method comprising a. The method of claim 4, wherein The word appearance document number includes the number of documents in which the word appears for each word included in the entire document, The third step, Counting the number of documents for each word, or adding a document in which the word appears to a document list associated with the word. The method of claim 4, wherein The step of measuring the word appearance frequency, the document length and the word appearance document number through the document, A fifth step of measuring the number of times the first step to the third step is repeated as the total number of documents Word recommendation method characterized in that it further comprises. The method of claim 4, wherein The step of measuring the word appearance frequency, the document length and the word appearance document number through the document, A fifth step of confirming a number of the last word appearance document updated Word recommendation method characterized in that it further comprises. In the method for suggesting a word associated with a search object, Measuring recommendation information for a word included in a search target; Calculating a word weight for the word based on the recommendation information; And Recommending the word as a query for the search target based on the word weight Including, The recommendation information includes (1) the document length of the document including the word, (2) the total number of documents for the entire document corresponding to the search object, (3) the frequency of word appearance of the word, and (4) the word. Word appearance document number, which is the number of documents that appeared, or (5) a document average length for the entire document, The step of calculating a word weight for the word based on the recommendation information, Generating a first word weight for the document based on the word occurrence frequency, the document length of the document and the total number of documents; Generating a second word weight for the document based on the number of word appearance documents and the total number of documents; And Calculating the word weight for each document based on the first word weight and the second word weight Word recommendation method comprising a. The method of claim 8, The step of generating a first word weight based on the word appearance frequency, the document length of the document and the total number of documents, Calculating a ratio of the document length to the average document length of each document and adding the word appearance frequency to each word; And Generating a ratio of the word occurrence frequency to the summed word appearance frequency as the first word weight Word recommendation method comprising a. The method of claim 8, The word appearance document number includes the number of documents in which the word appears for each word included in the entire document, The step of generating a second word weight based on the word appearance document number and the total document number, Generating a ratio of the number of word appearance documents to the total number of documents as the second word weight. The method of claim 8, The step of calculating the word weight for each document based on the first word weight and the second word weight, And calculating the word weight through a multiplication operation of the first word weight and the second word weight. delete In the method for suggesting a word associated with a search object, Measuring recommendation information for a word included in a search target; Calculating a word weight for the word based on the recommendation information; And Recommending the word as a query for the search target based on the word weight Including, The search target includes a text content of a book composed of a plurality of documents, And the text content is classified into the document based on the table of contents of the book. In the method for suggesting a word associated with a search object, Measuring recommendation information for a word included in a search target; Calculating a word weight for the word based on the recommendation information; And Recommending the word as a query for the search target based on the word weight Including, The word is a word recommendation method, characterized in that it comprises an extraction result extracted from the search target using a noun extractor. A computer-readable recording medium in which a program for executing the method of any one of claims 1 to 11, 13 or 14 is recorded. In the word recommendation system for recommending a word associated with a search object, A recommendation information measuring unit configured to measure recommendation information about a word included in a search target; A word weighting unit that calculates a word weight for the word based on the recommendation information; And A word recommender for recommending the word as a query for the search target based on the word weight Word recommendation system comprising a. The method of claim 16, The recommendation information includes (1) the document length of the document including the word, (2) the total number of documents for the entire document corresponding to the search object, (3) the frequency of word appearance of the word, and (4) the word. A word appearance document number, which is the number of document appearances, or (5) a document average length for the entire document. The method of claim 17, The recommendation information measuring unit, A measuring unit measuring the word appearance frequency, the document length, and the word appearance document number through the document; An entire document length calculator configured to calculate an entire document length using the document length measured for each document; And A document average length calculator for calculating the average document length based on the total document length and the total number of documents Word recommendation system comprising a. The method of claim 17, The word weighting unit, A first word weight generator configured to generate a first word weight based on the word occurrence frequency, the document length of the document, and the total number of documents for the document; A second word weight generator configured to generate a second word weight based on the word appearance document number and the total document number for the document; And A word weight calculator configured to calculate the word weight for each document based on the first word weight and the second word weight Word recommendation system comprising a. The method of claim 16, The word recommendation unit, A listing unit for listing the words included in the documents for each document included in the search object based on the word weights; A word extracting unit which extracts the word from the listed words by a predetermined number in order of increasing the word weight; And When the document is exposed to a search page, a query providing unit that provides the extracted word as a query for the document. Word recommendation system comprising a.

Images