KR20090080822A - Emotion based item search service provision method, database construction method and search server - Google Patents

Abstract

The present invention relates to a method for providing an emotion-based item search service, a database construction method, and a search server.

The present invention provides a search server for providing an emotion-based item search service, comprising: a database storing items, vocabulary association information, and item index information; A database construction management unit for generating lexical linkage information by linking an emotional vocabulary and an insensitive vocabulary, and generating item index information by linking an emotional vocabulary or an insensitive vocabulary with items; And receiving a search keyword, extracting an emotional or non-emotional vocabulary corresponding to the search keyword using the vocabulary association information, and using one or more items linked to the extracted emotional or non-emotional vocabulary using item index information. It provides a search server for providing an emotion-based item search service characterized in that it comprises a search management to provide a search.

According to the present invention, the search server expands the search keyword to an emotional or non-emotional vocabulary associated with the search keyword to search for an item, thereby providing a user with abundant items appropriate for a desired feeling or emotion.

Description

Method for providing emotion based item search service, database construction method and search server {Method and Server for Searching Items and Constructing Database based on Sensitivity}

The present invention relates to a method for providing an emotion-based item search service, a database construction method, and a search server. More specifically, by searching for related items by expanding the search keyword with the emotional or non-emotional vocabulary associated with the input search keyword, the user is not simply a color or a form, but a rich item appropriate for the feeling or emotion desired by the user. The present invention relates to an emotion-based item search service providing method, a database construction method, and a search server.

Nowadays, there are many websites that search and provide items which are contents such as images, videos, and music that the user wants. In order to provide such items, the search server of the website receives a search keyword input from the user and corresponds to the corresponding item. Search and provide.

The conventional search server for the item search and provision service searches and provides only items that are directly matched in one dimension to the search keyword received from the user terminal. For example, if the search keyword is an insensitive vocabulary such as 'fallen leaves', It only provides images that contain 'fallen leaves' that match the 'fallen leaves'.

However, although a user using the item search and providing service inputs an insensitive vocabulary of 'fallen leaves' as a search keyword, an item linked to an emotional vocabulary such as 'solitude', 'loneliness' and 'loneliness' associated with the 'falling leaves' In the case where it is desired to be provided, the conventional item searching and providing service has a problem in that it cannot provide an item for an emotional vocabulary associated with an insensitive vocabulary. In addition, even if a search keyword is entered as a search keyword of 'loneliness' and a user wants to receive an appropriate image of an emotional feeling as an item, the conventional item search and provision service provides an insensitive vocabulary related item associated with the emotional vocabulary. There is a problem that can not.

The present invention for solving the above-mentioned problems, by expanding the search keyword to an emotional or non-emotional vocabulary associated with the input search keyword to search for related items, it is not a simple color or shape to the user, but a feeling or emotion desired by the user Its main purpose is to search for and provide appropriate rich items for.

According to an aspect of the present invention for achieving the above object, a search server for providing an emotion-based item search service, comprising: a database storing items, vocabulary association information and item index information; A database construction management unit for generating lexical linkage information by linking an emotional vocabulary and an insensitive vocabulary, and generating item index information by linking an emotional vocabulary or an insensitive vocabulary with items; And receiving a search keyword, extracting an emotional or non-emotional vocabulary corresponding to the search keyword using the vocabulary association information, and using one or more items linked to the extracted emotional or non-emotional vocabulary using item index information. It provides a search server for providing an emotion-based item search service characterized in that it comprises a search management to provide a search.

According to another aspect of the present invention for achieving the above object, in a method of providing an emotion-based item search service by the search server, (a) lexical linking information and the emotional relationship is linked to the emotional vocabulary and the emotional Generating item index information associated with a vocabulary or the insensitive vocabulary; (b) receiving a search keyword; (c) extracting the emotional or non-emotional vocabulary corresponding to the search keyword using the vocabulary association information; (d) searching for one or more items associated with the emotional vocabulary or the insensitive vocabulary using the item index information; And (e) providing the one or more items.

According to another aspect of the present invention for achieving the above object, in a method of building a database based on emotion, (a) search for a vocabulary using document data, the search by the emotional vocabulary, insensitive vocabulary and the Extracting a lexical relationship between the emotional vocabulary and the insensitive vocabulary; (b) generating and storing the lexical association information by linking the emotional vocabulary having the lexical relationship with the insensitive vocabulary; (c) identifying attribute information of an item and extracting the emotional word or the non-emotional word matching the attribute information using the vocabulary association information; And (d) generating and storing the item index information in association with the extracted emotional vocabulary or the insensitive vocabulary.

According to another aspect of the present invention for achieving the above object, (a) receiving a first emotional vocabulary; (b) selecting an insensitive vocabulary associated with the received first emotional vocabulary; And (c) searching for an item using the insensitive vocabulary as a keyword.

As described above, according to the present invention, the search server expands the search keyword to an emotional or non-emotional vocabulary associated with the input search keyword and searches for related items. It is effective to search for and provide abundant items appropriate for emotions.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. First of all, in adding reference numerals to the components of each drawing, it should be noted that the same reference numerals are used as much as possible even if displayed on different drawings. In addition, in describing the present invention, when it is determined that the detailed description of the related well-known configuration or function is apparent to those skilled in the art or may obscure the gist of the present invention, the detailed description thereof will be omitted.

1 is a schematic diagram of an entire system for providing an emotion-based item search service according to the present invention.

As shown in FIG. 1, the entire system for providing an emotion-based item search service according to the present invention is an item search service through a wired / wireless communication network 110 from a terminal 120 and a terminal 120 requesting an item search service. It includes a search server 100 for receiving a request to search for the corresponding item to the terminal 120.

In the embodiment of the present invention, the item means wired / wireless content. Content refers to images, audio, video, flash files, etc. created by users or companies, and can be combined with text in some cases. In addition, this item may be provided to the user for free or for free.

The terminal 120 inputs a search keyword when an item search service is requested to the search server 100 through the wired / wireless communication network 110. The search keyword may be an emotional vocabulary or an insensitive vocabulary. Here, the terminal 120 is connected to a wired or wireless communication network 110 such as the Internet, a mobile communication network, and can be networked with the search server 100, a mobile communication terminal, a desktop computer, a laptop, a hand-held computer, etc. It can include both.

In the above-mentioned 'insensitive vocabulary', for example, fallen leaves, sea, mountains, schools, etc. may be examples of vocabulary not containing human emotion. In addition, the 'emotional vocabulary' may include loneliness, loneliness, joy, warmth, coolness, and the like as a vocabulary containing human emotion.

When the search server 100 receives a request for an item search service through the search keyword from the terminal 120, the search server 100 detects the received search keyword and searches for and provides an item corresponding to or associated with the search keyword to the terminal 120.

The search server 100 searches for an item corresponding to a search keyword in two steps. The first step is to extract a vocabulary (emotional vocabulary, an insensitive vocabulary) associated with the search keyword. The step is to search for an item associated with the extracted vocabulary (emotional vocabulary or insensitive vocabulary).

 For the first step, the search server 100 extracts semantic relations between a plurality of vocabularies and vocabularies from web documents on the Internet and vocabulary related programs such as WordNet, and generates linkage information between these vocabularies. Should be If the linkage information between the vocabulary having a predetermined vocabulary relationship is generated, the speed of extracting the vocabulary associated with the search keyword can be increased. In the present invention, the linking information between the vocabulary having a predetermined vocabulary relationship is called 'lexical linking information' or 'lexical network index information'.

For example, referring to the vocabulary linking information with reference to FIG. 5, when there is an insensitive vocabulary of 'fallen leaves', a emotional vocabulary of' loneliness' that can be associated with it can be extracted, and ' Emotional vocabulary such as loneliness and loneliness can be further extracted. Thus, 'falling leaves', 'loneliness', 'loneliness', 'loneliness', etc. are linked to each other by the association of the association word and similar words to generate the lexical link information. If the input search keyword is 'loneliness', 'falling leaves', 'loneliness', and 'solitude', which are linked emotional words or insensitive words, may be extracted from the generated vocabulary linking information.

Extracting the vocabulary (emotional vocabulary, insensitive vocabulary) associated with the above-described search keyword has the effect of generating a plurality of search keywords in addition to the input search keyword, and thus does not search for an item by the input search keyword. By using both the entered search keyword and the related vocabulary, it is possible to search for items based on richer emotions.

In addition, for a second step of searching for an item using the extracted vocabulary (emotional vocabulary, insensitive vocabulary) associated with the above-described search keyword, the search server 100 grasps the attribute information of the stored item, The vocabulary (emotional vocabulary or non-emotional vocabulary) related to this attribute information should be found and associated with the item, and the associated information should be prepared in advance. If link information between the vocabulary and the item is generated in this way, the speed of searching for the corresponding item may be increased. In the present invention, the association information between the vocabulary and the item is referred to as 'item index information'.

Referring to FIG. 6, when the item index information is described as an example, if the attribute information of 'Item A' is related to one of 'fallen leaves', 'loneliness', 'solitude', and 'loneliness', 'Item A' May be associated with vocabulary such as' falling leaves', 'loneliness',' solitude ', and' loneliness' (emotional vocabulary, insensitive vocabulary), and item A 'as shown in FIG. Item index information associated with 'solitude' and 'loneliness' may be generated. If the item index information is prepared in advance, the words 'lost', 'lost', and 'solitude' extracted from the vocabulary linking information about the search keyword 'loneliness' are 'lost', 'lost' in the item index information. You can search for all items, including item A, which is associated with loneliness and loneliness.

The items mentioned above are content that includes one or more in images, audio and video. For example, an item may be content such as a skin image, a video, and music that may be associated with a specific emotional vocabulary.

2 is a block diagram of a search server 100 that provides an emotion-based item search service according to the present invention.

As shown in FIG. 2, the search server 100 that provides an emotion-based item search service includes a database 210, a database construction manager 220, and a search manager 230.

The database 210 stores item, lexical association information that is information associated with vocabulary having a predetermined lexical relationship, and item index information, which is information associated with a vocabulary and items. Through the vocabulary linking information, it is possible to know which vocabulary is associated with which vocabulary relationship, and through the item index information, it is possible to know which vocabulary similar to the attribute of the item is linked.

The database construction management unit 220 generates vocabulary linking information and item index information stored in the database 210. The vocabulary linking information is generated by linking the emotional vocabulary and the insensitive vocabulary, and the emotional vocabulary or the insensitive vocabulary is linked with the item. To generate the item index information.

The above-described database construction manager 220 extracts a vocabulary including emotional vocabulary and insensitive vocabulary from various document data such as web documents and vocabulary related programs on the Internet in order to generate vocabulary linking information, and the vocabulary relation between the extracted vocabularies Automatically extract the lexical linkage information. In this case, the vocabulary may be extracted from a web document on the Internet using data extraction techniques such as data mining and data warehouse. Also, WordNet, which is a list of semantic vocabulary of English developed by Princeton University's Cognitive Science Institute, can be used to extract lexical relations between emotional vocabulary, insensitive vocabulary and vocabulary through vocabulary related programs.

In addition, the database construction management unit 220 may perform a translation function between languages of the emotional words or the insensitive words. If the extracted words are English, they are translated into Korean.

The vocabulary including the emotional vocabulary and the insensitive vocabulary extracted to the database construction management unit 220 may be parts of speech such as nouns, verbs, adjectives, and the like. The vocabulary relations among the vocabulary may be synonymous, synonymous, antonym, upper or lower, It may be one or more in the relations of the fish, homonym, global, partial, relationship, modal, companion, association, relational, and the like.

When explaining the vocabulary relationship between the vocabulary extracted by the database construction management unit 220 as follows.

The synonym relationship mentioned above is a relationship between two or more vocabularies with similar meanings, and a synonym relationship is a relationship between two or more vocabularies that have different forms but the same meaning, and the antonym relationship belongs to the same field of meaning and the opposite meaning among them. The relationship between the vocabulary representing. For example, 'pretty' and 'beauty' are synonymous or synonymous, and 'fast' and 'slow' are the opposite.

If the first vocabulary is one kind of the second vocabulary, the second vocabulary is referred to as the upper term of the first vocabulary, and the first vocabulary is referred to as the lower vocabulary of the second vocabulary. At this time, the relationship between the two vocabulary is called a parent word relationship or a child word relationship. For example, 'fruit' and 'apple' are in shark or subword relationships.

If the first vocabulary and the second vocabulary have the same superior word, the second vocabulary is said to be the equal word of the first vocabulary, and the two vocabularies are said to be equivalent. For example, 'baseball' and 'soccer' are equivalence relations with the same synonyms of exercise.

If the first vocabulary is part of the second vocabulary, then the second vocabulary is referred to as the entire word of the first vocabulary, and the first vocabulary is referred to as the partial vocabulary of the second vocabulary. These two vocabularies are said to be in full or partial terms. For example, 'bicycle' and 'pedal' are either whole or partial.

When an action called a second vocabulary acts as a first vocabulary in a specific situation, the second vocabulary is said to be an aspect of the first vocabulary, and the two vocabularies are said to be in a symbolic relationship. For example, 'stutter' and 'speak' are modal relations.

If the second vocabulary is necessarily performed when the first vocabulary is acted, the second vocabulary is said to accompany the first vocabulary, and the two vocabularies are said to be accompanying words. For example, sleep and snoring are accompanying words.

When a vocabulary is associated with one vocabulary, the two vocabulary are associated with each other. When 'patriotism' is associated with 'Taegeukgi', 'Taegeukgi' and 'patriotism' are associated with each other. When two words can be related to a certain meaning or action, the two words can be referred to as relations. For example, 'study' and 'library' are related.

In addition, the database construction manager 220 extracts the emotional or non-emotional vocabulary corresponding to the attribute information of the item using the generated vocabulary association information, and generates the item index information in association with the corresponding item, wherein the attribute information corresponds to the corresponding vocabulary. At least one of text information, image information, image information, and the like representing the item.

The search manager 230 receives a search keyword from the terminal 120 and extracts an emotional or insensitive vocabulary corresponding to the input search keyword using the lexical association information stored in the database 210, and then extracts the search keyword from the terminal 120. The stored item index information is used to search the database 210 for one or more items associated with the emotional or non-emotional vocabulary extracted previously and provide the terminal 120 to the terminal 120.

In addition, the search manager 230 may calculate and provide a accumulated search ranking for the searched item.

The above-mentioned "items" are contents including one or more in images, audio, and video. For example, an item may be content such as a skin image, a video, and music that may be associated with a specific emotional vocabulary.

According to the search server 100 that provides the emotion-based item search service according to the present invention described above, a user searches by searching for a related item by expanding a search keyword with an emotional or non-emotional vocabulary associated with an input search keyword. It is effective to search and provide not only simple colors or shapes, but also abundant items appropriate for a feeling or emotion desired by a user.

3 is a flowchart illustrating a method of providing an emotion-based item search service according to the present invention.

As shown in FIG. 3, the search server 100 provides an emotion-based item search service in which the lexical association information and the emotional or non-emotional vocabulary are linked to each other. Generating item index information (S300), receiving a search keyword from the terminal (S302), and extracting an emotional or insensitive vocabulary corresponding to the received search keyword using the vocabulary association information generated in step S300. Searching for one or more items associated with the emotional or non-emotional vocabulary using the item index information generated in step S304, S300 (S306), and providing one or more items retrieved in step S306 to the terminal ( S308).

In the above-described step of generating the vocabulary linkage information at step S300, the step of searching for a vocabulary using document data such as a web document and a vocabulary related program on the Internet, the emotional vocabulary, the insensitive vocabulary, the emotional vocabulary and the insensitive vocabulary by searching Extracting a lexical relationship of the vocabulary, and generating lexical linkage information by linking an emotional vocabulary having an lexical relationship such as a synonym relationship and an antonym relationship with an insensitive vocabulary.

In addition, the procedure of generating the item index information in the above-described step S300, the step of identifying the attribute information of the item, extracting the emotional vocabulary or insensitive vocabulary matching the attribute information of the item and the extracted emotional vocabulary or insensitivity Generating item index information by associating the vocabulary with the item.

According to a search method for providing an emotion-based item search service according to the present invention, an item is searched by expanding a search keyword with an emotional or non-emotional vocabulary associated with an input search keyword, thereby providing a simple color or shape to a user. Rather, it may provide a search service for an appropriate item related to a feeling or emotion desired by the user.

4 is a flowchart illustrating a database construction method for providing an emotion-based item search service according to the present invention.

Referring to FIG. 4, in a method of constructing an emotion-based database, searching for a vocabulary using document data, and extracting a vocabulary relationship between an emotional vocabulary, an insensitive vocabulary, and an emotional vocabulary and an insensitive vocabulary according to the search (S400) ), Generating and storing vocabulary association information by linking the emotional and non-emotional vocabulary having a predetermined vocabulary relationship (S402), grasp the attribute information of the stored item, and use the stored vocabulary association information to store the attribute of the item. Extracting an emotional or insensitive vocabulary matching the information (S404) and generating and storing item index information in association with the stored vocabulary or the emotional vocabulary extracted in step S404 (S406). .

The aforementioned document data may be a web document on the Internet or a document in a lexical related program.

The search server 100 constructs a database for storing the vocabulary association information and the item index information through the above-described method, so that the search keyword can be extended to the emotional or insensitive vocabulary associated with the input search keyword. Then, the item can be searched by the expanded search keyword. As a result, the search server 100 may provide the user with a search service for an appropriate item related to a feeling or emotion desired by the user as well as a simple color or form.

In addition, the emotion-based item retrieval method according to the present invention may further include a process of checking how the insensitive vocabulary matches the emotional vocabulary when there are a large number of insensitive vocabularies. Accordingly, the entire process of the emotion-based item searching method according to the present invention may be configured as follows.

The entire process of the emotion-based item retrieval method according to the present invention comprises the steps of receiving a first emotional vocabulary, selecting an insensitive vocabulary associated with the received first emotional vocabulary, and an item using the non-emotional vocabulary as a keyword. Searching.

The method may further include selecting a second emotional vocabulary associated with the first emotional vocabulary before selecting the insensitive vocabulary. In this case, the step of selecting the insensitive vocabulary may include the first emotional vocabulary or the first emotional vocabulary. 2 Choose an insensitive vocabulary associated with the emotional vocabulary.

In addition, the step of selecting the insensitive vocabulary may include selecting a first insensitive vocabulary having a high linkage index using the vocabulary linkage information and a first emotional vocabulary and a second emotional vocabulary associated with the first insensitive vocabulary. Comparing the vocabulary and selecting the first insensitive vocabulary when the first emotional vocabulary and the second emotional vocabulary are positive.

In addition, the step of selecting the insensitive vocabulary may include selecting the first insensitive vocabulary having the highest linkage index and the second insensitive vocabulary having a lower linkage index than the first insensitive vocabulary using the vocabulary linkage information; Comparing the first emotional word and the second emotional word associated with the first insensitive word, when the first emotional word and the second emotional word have a negative relationship, the second insensitive word is not selected without selecting the first insensitive word. Selecting the vocabulary.

If the keyword of the searched item is a first insensitive vocabulary and a second insensitive vocabulary, and the first insensitive vocabulary is associated with the first emotional vocabulary in the step of selecting the insensitive vocabulary. The method may further include determining whether to include an item in a search result based on a linkage index for the first emotional word and the second insensitive word using the vocabulary linkage information.

In addition, after the above-mentioned searching for the item, the searched item is displayed on the user display, and an insensitive vocabulary for searching for the corresponding item is displayed.

Detailed description thereof is as follows.

Since a plurality of emotional vocabulary and insensitive vocabulary can exist, it can be distinguished into first and second. When the search server 100 receives the first emotional vocabulary, the search server 100 selects an insensitive vocabulary associated with the received first emotional vocabulary. The process of selecting this may be the lexical association information described above. In addition, inversely related to FIG. 5, a plurality of insensitive vocabularies may be linked to one emotional vocabulary. For example, an insensitive vocabulary of 'solitude' may be associated with (fall leaves, 0.8), (tear, 0.7) (glow, 0.6), (sul, 0.3), (cake, -0.5), etc. . The number (association index) here shows how relevant the emotional vocabulary is. For cakes, -0.3 indicates an insensitive vocabulary as opposed to the emotional vocabulary of loneliness.

When a plurality of insensitive vocabulary related to the first emotional vocabulary is found using the vocabulary linkage information, a highly relevant insensitive vocabulary may be found using the linkage index described above. In the example of loneliness, 'falling leaves' are chosen. Then, the leaves can be searched again to see how closely the leaves are related to the emotional vocabulary. For example, the first emotional vocabulary is compared with other emotional vocabularies mapped to the insensitive vocabulary, using data mapped with the multiple emotional vocabulary for the insensitive vocabulary of FIG. 5. For example, if (Lonely, 0.8) and (Lonely, 0.9) are mapped to the fallen leaves, the first emotional vocabulary of 'loneliness' is compared to the second emotional vocabulary of 'lonely' (positive words, synonyms). , Leaves, etc.), 'fallen leaves' is selected as an insensitive vocabulary.

Meanwhile, the case where (eye, 0.7) is mapped to the emotional vocabulary of 'joy' is as follows.

If the unemotional vocabulary mapped to the emotional vocabulary of 'joy' is 'eye', examine whether the 'eye' is properly mapped to the emotional vocabulary of 'joy'. In other words, as a result of examining the emotional vocabulary mapped to the word 'eye', (joy, 0.5) and (sorrow, 0.6), joy and sadness are negative relationships (opposite words), so 'eye' is the insensitive vocabulary of the search target. You can choose another insensitive vocabulary instead of selecting.

In addition, when there are two or more keywords in the item, it is possible to select whether to include them in the search results by the corresponding keywords.

For example, an insensitive vocabulary of 'fallen leaves' is selected from the keyword 'solitude', so that a large number of items having 'fallen leaves' as a keyword may be selected. In this case, it is possible to determine an association of the keyword of each item with the emotional vocabulary.

For example, if the keywords of the first item are 'fall leaves' and 'tear', the word 'tear' also has a high index of association with 'solitude', and thus the first item may be included as a search result. However, if the keyword of the second item is 'falling leaves' or 'cake', since the 'cake' has a linkage index opposite to 'solitude', the second item may be excluded.

An item may be searched based on the selected insensitive vocabulary, and the retrieved item may be displayed as an item associated with the emotional vocabulary input to the user. In this case, it may be displayed to the user which insensitive vocabulary is selected.

The above description is merely illustrative of the technical idea of the present invention, and those skilled in the art to which the present invention pertains may make various modifications and changes without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention but to describe the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. The protection scope of the present invention should be interpreted by the following claims, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of the present invention.

1 is a schematic diagram of an entire system for providing an emotion-based item search service according to the present invention;

2 is a block diagram of a search server providing an emotion-based item search service according to the present invention;

3 is a flowchart illustrating a method of providing an emotion-based item search service according to the present invention;

4 is a flowchart illustrating a method for constructing a database for providing an emotion-based item search service according to the present invention;

5 is an exemplary diagram of lexical association information for providing an emotion-based item search service according to the present invention;

6 is an exemplary diagram of item index information for providing an emotion-based item search service according to the present invention.

<Description of Symbols for Main Parts of Drawings>

100: search server

110: wired or wireless network

120: terminal

210: database

220: database deployment management unit

230: search management

Claims (16)

In a search server that provides an emotion-based item search service, A database storing items, lexical association information, and item index information; A database construction management unit generating the vocabulary linking information by linking an emotional vocabulary and an insensitive vocabulary, and generating the item index information by linking the emotional vocabulary or the insensitive vocabulary with the item; And Receiving a search keyword, using the vocabulary association information to extract the emotional vocabulary or insensitive vocabulary corresponding to the search keyword, and using the item index information one or more linked to the extracted emotional vocabulary or insensitive vocabulary Search management unit for searching and providing items Search server providing an emotion-based item search service comprising a. The method of claim 1, And the item is content including at least one of an image, audio, and video. The method of claim 1, The database construction management unit, And extracting the emotional vocabulary, the insensitive vocabulary, and the vocabulary relation between the emotional vocabulary and the insensitive vocabulary from the document data to generate the vocabulary linking information. The method of claim 3, wherein The lexical relation may include at least one of a synonym relation, a synonym relation, an antonym relation, an upper term relation, a lower term relation, an equivalence relation, a general term relation, a partial term relation, a aspect term relation, an accompanying term relation, an association term relation, and a relation term relation. Search server providing an emotion-based item search service comprising a. The method of claim 1, The database construction management unit, Search server for providing an emotion-based item search service, characterized in that for performing the language translation function of the emotional vocabulary or the insensitive vocabulary. The method of claim 1, The database construction management unit, The emotional word or the insensitive word corresponding to the attribute information of the item is extracted using the vocabulary association information to be generated as the item index information in association with the item, wherein the attribute information is text information representing the item. And a search server for providing an emotion-based item search service, characterized in that at least one of image information and image information. In the method for the search server to provide an emotion-based item search service, generating a vocabulary linking information in which an emotional vocabulary and an insensitive vocabulary are linked, and item index information in which the emotional vocabulary or the insensitive vocabulary and an item are linked; (b) receiving a search keyword; (c) extracting the emotional or non-emotional vocabulary corresponding to the search keyword using the vocabulary association information; (d) searching for one or more items associated with the emotional vocabulary or the insensitive vocabulary using the item index information; And (e) providing the one or more items Emotion-based item search service providing method comprising a. The method of claim 7, wherein Step (a) is, (a1) searching for a vocabulary using document data; (a2) extracting the emotional vocabulary, the insensitive vocabulary, and the lexical relationship between the emotional vocabulary and the insensitive vocabulary by the searching; And (a3) generating the vocabulary linking information by linking the emotional vocabulary having the vocabulary relationship with the insensitive vocabulary Emotion-based item search service providing method characterized in that for generating the lexical linkage information, including. The method of claim 7, wherein Step (a) is, (a4) identifying attribute information of the item; (a5) extracting the emotional or non-emotional vocabulary matching the attribute information of the item using the vocabulary association information; And (a6) generating the item index information by linking the extracted emotional or non-emotional vocabulary with the item Emotion-based item search service providing method comprising a. In the emotion-based database construction method, (a) searching for a vocabulary using document data and extracting an emotional vocabulary, an insensitive vocabulary, and a lexical relationship between the emotional vocabulary and the insensitive vocabulary; (b) generating and storing the lexical association information by linking the emotional vocabulary having the lexical relationship with the insensitive vocabulary; (c) identifying attribute information of an item and extracting the emotional word or the non-emotional word matching the attribute information using the vocabulary association information; And (d) generating and storing the item index information in association with the extracted emotional or non-emotional vocabulary; Emotion-based database construction method comprising a. (a) receiving a first emotional vocabulary; (b) selecting an insensitive vocabulary associated with the received first emotional vocabulary; And and (c) searching for an item using the insensitive vocabulary as a keyword. The method of claim 11, Before step (b) above, Selecting a second emotional vocabulary associated with the first emotional vocabulary, In the step (b), the emotion-based item retrieval service providing method comprising selecting an insensitive vocabulary associated with the first emotional vocabulary or the second emotional vocabulary. The method of claim 11, Step (b) is, Selecting a first insensitive vocabulary having a high linkage index using the lexical linkage information; and Comparing the first emotional vocabulary and the second emotional vocabulary associated with the first insensitive vocabulary and selecting the first insensitive vocabulary when the first emotional vocabulary and the second emotional vocabulary are in a positive relationship step Emotion-based item search service providing method comprising a. The method of claim 11, Step (b) is, Selecting a first insensitive vocabulary having the highest link index and a second insensitive vocabulary having a linked index lower than the first insensitive vocabulary using the vocabulary link information; And Comparing the first emotional word and the second emotional word associated with the first insensitive word, when the first emotional word and the second emotional word have a negative relationship, the first insensitive word is not selected. Selecting the second insensitive vocabulary without Emotion-based item search service providing method comprising a. The method of claim 11, If the keyword of the searched item is a first insensitive vocabulary and a second insensitive vocabulary, and in the step of selecting the insensitive vocabulary, the first insensitive vocabulary is associated with a first emotional vocabulary, And determining whether to include the item in a search result based on a linkage index for the first emotional word and the second non-emotional word using lexical linkage information. How to Provide Search Services. The method of claim 11, After step (c), And displaying the retrieved item on a user display and displaying the insensitive vocabulary for allowing the item to be searched.

Images