JP5313295B2 - Document search service providing method and system - Google Patents

Abstract

A method and a system for serving document exploration service are provided to form an accurately adjusted tree structure by renewing the tree structure based on the document browse inclination of the user who is using the document search service for searching documents classified in tree structure. A document delivery unit(215) provides an access medium to a document. A reaction information acquisition unit(216) obtains a user response information about the access medium. A tree structure updating unit(217) renews a tree structure in consideration of the response information. The document delivery unit provides a medium for visualizing the tree structure. The reaction information acquisition unit collects the response information of the user about the visualization medium. The tree structure updating unit cuts off the connection where the frequency of the response of the user is less than a designated value.

Description

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

  There are many documents on the web for diverse interests. The user can acquire information by transmitting the query information (query) for the information desired by the user to the search engine. However, it is very troublesome to input a question and answer for a subject that the user himself is interested in each time.

  On the other hand, in order to access the information desired by the user without going through the steps such as inputting questions, connect to a vertical site (vertical site (special field site)) or a blog that has expertise in a specific field. You can get the latest information on the fields.

  The level of information existing on such vertical sites and blogs is improving day by day, and vertical sites and blogs are developing as media that can acquire useful information that is the fastest and deepest in the field. is there.

  However, it is also inconvenient for the user to visit each site to browse information distributed on various vertical sites and blogs. As a supplement to this, vertical sites and blogs provide RSS feeds (Real Feed Syndication Feeds), and programs such as RSS readers (RSS Reader) are used to subscribe to them.

  However, each RSS feed provides information independently of each other and handles documents of the same content or very similar content as separate information, so that the user can improve efficiency in the process of searching and browsing information. Additional effort is required.

  SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a document search service that suitably updates the tree structure by considering user reaction information in providing a document search service that can search for documents classified by the tree structure. And providing a system.

  Another object of the present invention is to update the tree structure to reflect the document browsing tendency of the user who uses the document search service, thereby forming a more intuitive but accurately adjusted tree structure. To provide a document search service providing method and system.

  According to an aspect of the present invention, there is provided a method for providing a document search service capable of searching for documents classified in a tree structure, the step of providing an access means for a document, and the step of obtaining user response information for the access means And a method of providing a document search service including a step of updating the tree structure in consideration of reaction information.

  Providing access to the document can provide a means for visualizing the tree structure. User response information is obtained by collecting user response information for the means by which the tree structure of the document is visualized.

  The step of updating the tree structure by the document search service providing method can break the connection in which the user's reaction frequency is not more than a predetermined value in the tree structure.

  Further, in the step of updating the tree structure, the tree structure can be updated by further considering the question word information input by the document viewer.

  The tree structure used to provide the document search service includes the steps of classifying documents by subject, extracting the subject of the document, forming a cluster based on the extracted subject, and subjecting the cluster to the subject. Mapping to a predetermined directory belonging to a tree structure forming method.

  In such a tree structure forming method, the document includes at least one field, and in the step of extracting the title of the document, the title can be extracted in consideration of the attributes of the fields constituting the document.

  On the other hand, at the stage of forming a cluster, the extracted theme can be classified in syllable units, and a portion of the theme that is shared with other documents can be selected as a central concept candidate of the cluster. The central concept candidate is selected using n-gram analysis for the extracted subject.

  According to another aspect of the present invention, there is provided a system for providing a document search service capable of searching for documents classified in a tree structure, a document providing unit for providing access means for a document, and a user response to the access means. There is provided a document search service providing system including a reaction information acquisition unit that acquires information and a tree structure update unit that updates the tree structure in consideration of the reaction information.

  In the document search service providing system, the document providing unit can provide a means for visualizing the tree structure, and the response information acquiring unit is a user's reaction information with respect to the visualizing means provided by the document providing unit. Can be collected.

  In the document search service providing system, the tree structure update unit can break a connection in which a user's reaction frequency is equal to or lower than a predetermined value in the tree structure.

  In addition, the tree structure update unit can update the tree structure of the document by allowing the viewer of the document to further consider the question word information.

  The document search service providing system can include a tree structure forming unit that forms a tree structure. The tree structure forming unit includes a document classifying unit that classifies documents according to a subject, a subject extracting unit that extracts a subject of the document, a cluster forming unit that forms a cluster based on the extracted subject, and a predetermined cluster belonging to the subject And a directory mapping unit for mapping to the directory.

  On the other hand, the document includes at least one field, and the subject extraction unit can extract the subject in consideration of the attributes of the fields constituting the document.

  In addition, the cluster forming unit can classify the extracted subject in syllable units, and select a portion of the subject that is shared with other documents as a central concept candidate of the cluster. The central concept candidate is selected using n-gram analysis for the extracted subject.

  On the other hand, the document search service providing method of the present invention is performed by a computer and recorded on a computer-readable recording medium that records a program to be executed by the computer.

  Other aspects, features and advantages will be apparent from the accompanying drawings, the claims and the detailed description of the invention.

  According to a preferred embodiment of the present invention, when providing a document search service capable of searching for documents classified in a tree structure, it is possible to update the tree structure suitably by considering user reaction information. A document search service providing method and system can be provided.

  According to a preferred embodiment of the present invention, the tree structure is updated to reflect the user's document browsing tendency using the document search service, so that the tree structure can be adjusted more intuitively but accurately. Can provide a document search service providing method and system.

4 is a flowchart of a document search service providing method according to an embodiment of the present invention. 1 is a configuration diagram of a document search service providing system according to an embodiment of the present invention. FIG. It is the figure which illustrated the document tree structure before the update which concerns on one Example of this invention. It is the figure which illustrated the document tree structure after the update which concerns on one Example of this invention. It is the figure which illustrated the document search service provision screen before the update which concerns on one Example of this invention. It is the figure which illustrated the document search service provision screen after the update which concerns on one Example of this invention. It is the figure which illustrated the document search service provision screen containing the advertisement display area concerning one example of the present invention.

  Hereinafter, embodiments of a document search service providing method and system according to the present invention will be described in detail with reference to the accompanying drawings. The present invention is not limited to the specific embodiments, but should be understood as including all modifications, equivalents and alternatives included in the technical idea and scope of the present invention. In the description of the present invention, when it is determined that a specific description of a related known technique obscures the gist of the present invention, a detailed description thereof is omitted. Further, in the description based on the attached drawings, the same constituent elements or corresponding constituent elements are given the same drawing numbers, and redundant description thereof is omitted.

  FIG. 1 is a flowchart showing a processing flow of a document search service providing method according to an embodiment of the present invention, and FIG. 2 is a configuration diagram of a document search service providing system according to an embodiment of the present invention.

  1 and 2, a document search service providing system according to an embodiment of the present invention includes a search service providing server 210, a document classification unit 211, a subject extraction unit 212, a cluster formation unit 213, a directory mapping unit 214, The document providing unit 215, the reaction information obtaining unit 216, the tree structure updating unit 217, the original document database 221, the search service database 222, the reaction information database 223, the user terminal 230, and the tree structure forming unit 240 are included.

  The tree structure forming step (S110) is a step of forming a tree structure that expresses a connection relationship between each document provided by the document search service providing system to which the document search service providing method is applied. The tree structure forming step (S110) is performed by the tree structure forming unit 240.

  The tree structure forming step (S110) includes a step of classifying documents by subject (S112), a step of extracting the subject of each classified document (S114), and forming a cluster of documents based on the extracted subject. A step S116 and a step of mapping a cluster of documents to a directory (S118) may be included. The detailed operation of each step will be described below.

  The step of classifying documents by subject (S112) is a step in which the document classification unit 211 classifies each document in the original document database 221.

  The document classification unit 211 acquires information on the document from the original document database 221 and acquires information on the structure of the classification from the search service database 222. Based on the acquired information, the classification to which the document is matched is determined, and information on the matching relationship between the document and the classification is stored in the search service database 222.

  At this stage, the document classification unit 211 can classify the document by subject by determining whether the document includes a specific keyword and specific content using information included in the document.

  As an example, a document that is suitable for matching with the classification 'wine' contains a classification name 'wine' itself and a synonym for the classification name, and is closely related to 'wine'. It is determined in consideration of whether the document includes “sommelier”, “decanting”, and the like.

  On the other hand, when determining whether or not matching between a document and a classification is possible, it is possible to quantify whether or not the classification-related keyword is included and use it as a reference. As an example, when each keyword related to a specific subject is included, a predetermined score is given, and when the total of these scores exceeds a certain criterion, it can be determined that the keyword is matched with the corresponding classification.

  Note that it is not determined that one document is necessarily matched with only one classification in the document classification stage. As an example, if there is a classification of “Wine” and a classification of “Japanese comics”, a document that deals with appreciation reviews (reviews) on “Japanese manga”, which is a Japanese comic with wine as the subject, Not only the classification but also the classification of 'Japanese comics' is matched at the same time.

  The original document database 221 stores information for each original document that is classified and reconfigured by the search service providing server 210. These original documents are finally provided to the user terminal 230. The original document is collected on the web by a web robot or the like.

  On the other hand, each original document stored in the original document database 221 of the present invention can include a document having a predetermined attribute. As an example, structured documents used in vertical sites and blogs are used as original documents.

  Each structured document can be stored by dividing the content in the document into at least one region or section. These regions or partitions are named fields. As an example, a document such as a blog post may include a title field, a body field, a creation time field, a keyword field for the post, and the like.

  With respect to such a document, the creator can generate a document by inputting content corresponding to each field name, so that the field name and the corresponding content are usefully used in the subject extraction stage described later. The

  Further, in such a vertical site and blog, there is a possibility that the relationship between each document is structured. Such a structured relationship of documents appears in a directory form at the corresponding site.

  As an example, a vertical site with a movie theme can include directories such as 'Movie Appreciation Review', 'Movie Ranking' and 'Latest Releases' as a directory to classify documents on the site, It can have information about the directory that classifies the post.

  The directory names on such vertical sites and blogs are used as keywords related to the subjects handled by the sites. These keywords are used in the above-described document classification by subject, and are used to improve the accuracy of the classification.

  In the present invention, the term “document” is understood as a term used to refer to an electronically recorded document. A document is described using a markup language such as HTML and has an extension such as “* .htm”, but is analyzed as being limited to a file having a specific description format and extension. Absent.

  The search service database 222 stores information on the matching relationship between the document and the classification determined by the document classification unit 211. In other words, the search service database 222 can store the matching availability for each classification for each document, and can numerically store the inclusion availability for each classification-related keyword.

  On the other hand, the information storage method and form in the original document database 221 and the search service database 222 described above may be variously changed within the scope of the present invention.

  The step of extracting the subject of each document (S114) is a step in which the subject extracting unit 212 extracts the subject of each document stored in the original document database 221. Document title means a word, phrase or sentence that contains the document content and subject.

  The subject extraction unit 212 extracts the subject of each document using the document information stored in the original document database 221 and stores the extracted subject in the search service database 222.

  At this stage, the subject extraction unit 212 can extract the subject of the document using information included in each document. The structure of the document, the appearance frequency of each word included in the document, the attributes when the document is browsed by the user terminal 230, and the like are used as the criteria for determining the title.

  That is, the document information used in the process of extracting the title is understood as a concept including not only the content text directly included in the document but also information regarding the form in which the document is viewed on the user terminal 230. The

  As an example, a website such as a blog may include each structured document. These documents can store information in a field having at least one name. That is, text included in a field having a name such as 'title' or '(title)' can be selected as the title.

  As another example, when a document is browsed through a web browser or the like of the user terminal 230, the document is emphasized by being displayed larger than other contents in the document or expressed with a differentiated attribute. Text is also considered as a candidate phrase.

  The step of forming a cluster of documents based on the extracted title (S116) is a step in which the cluster forming unit 213 performs clustering based on the title information of the document.

  The cluster forming unit 213 forms a cluster of each document based on the document title information acquired from the search service database 222. Information about the formed cluster is stored in the search service database 222.

  A document cluster refers to a group of documents that share a central concept. A cluster of documents is formed in consideration of whether or not a common part exists in the subject of each document. The name of each cluster is named using this central concept.

  Among the document titles, a character string that is common to other documents is a candidate for the central concept of the cluster, and if the number of documents having a common character string is greater than or equal to a predetermined value, one independent cluster is formed. .

  As an example, if the subject of one document is 'Sommelier Tracking: Wine Etiquette-Wine to Enjoy Together' and the subject of the other document is 'Table Manner 5-Wine Etiquette', the common part of the two subjects The “wine etiquette” is extracted as a central concept.

  An n-gram analysis method can be used in the process of extracting the overlapping part in the document title as a central concept. In this case, the titles are separated in syllable units and recombined into a character string having a predetermined number of syllables.

  Of the character strings recombined in this way, the overlapping portion can be a candidate for the central concept. In the above example, 'wine' having two syllables and 'wine etiquette' having five syllables are extracted as candidates for the central concept.

  In addition, as described above, when there are many identical overlapping parts in the titles of each document, a process of determining one overlapping part as a central concept is required, but at this time, the number of overlapping syllables, overlapping The relationship between the part and the classification name of the document, the number of documents having the corresponding overlapping part, and the like are used as determination criteria, and one central concept can be determined.

  In the example described above, the overlapping portions are 'wine', 'etiquette' and 'wine etiquette'. In this case, “wine” is the same as “wine”, which is the subject of each document, and is not appropriate as the central concept of one cluster.

  Even in the aspect of the number of documents having overlapping portions, the number of documents sharing 'wine' is an excessively large value for forming a single cluster. In this way, when determining the central concept candidates of the cluster, it is required to limit the number of documents sharing the central concept candidates to a predetermined range.

  The length of the central concept candidate can also be taken into consideration. In the case of an overly short central concept candidate, the part of speech may be a particle or a general term that is inappropriate for use in a particular classification.

  On the other hand, in the case of a long central concept candidate, it is expected that the degree of association between the documents in the cluster sharing the candidate is high and the probability that noise is included is small, so the longest central concept candidate Candidates can be considered preferentially.

  In the case of the above example, since the remaining two central concept candidates are included in the “wine etiquette”, first, the longest “wine etiquette” can be considered as a candidate, and the number of documents sharing the “wine etiquette”, etc. If it is judged that the other criteria are satisfied, it is selected as a central concept constituting a single cluster.

  Also, if the number of documents that share the “etiquette” and the number of documents that share the “wine etiquette” are very similar among documents included in the subject of “wine”, a more specific “wine etiquette” It is efficient to select as the central concept.

  A cluster can be constructed based on the fact that a predetermined number or more of these central concepts are related. Information about the central concept forming a cluster and documents belonging to the cluster sharing the central concept is stored in the search service database 222.

  The step (S118) of mapping the document cluster to the directory is a step in which the directory mapping unit 214 maps the document cluster to the directory based on the central concept of each cluster.

  A directory means a sub-topic that can contain one or more clusters as a classification of documents, ie, subordinate concepts of the subject. As an example, if the subject matter to which the document was categorized is 'Wine', its directory includes 'Wine Origin', 'Wine History' and 'Wine Etiquette', etc. It is possible to include clusters in which place names such as 'Bordeaux' and 'Burgundy', known as production areas, are formed as central concepts.

  The directory mapping unit 214 acquires information on the directory structure and information on the cluster from the search service database 222, and determines a cluster to be mapped to each directory. The determined directory mapping information is stored in the search service database 222.

  The directory to which a cluster is mapped is determined by whether the central concept of the cluster includes a keyword associated with the directory.

  As an example, when the directory is “wine etiquette”, “etiquette” can be a keyword for determining whether or not to include the directory in the directory name. In the process of mapping a cluster formed for each document that has already been determined to fall under the 'wine' classification, it is efficient to map the directory with keywords excluding the classification name 'wine' itself.

  On the other hand, for these keywords, the corresponding keywords can be expanded using the dictionary enumeration method. 'Etiquette' is extended to keywords that have different synonyms, synonyms and notation languages.

  In this case, keywords such as 'respect,' 'equitette,' 'manner,' and 'manner' are considered as additional keywords for directory mapping. Through this, the efficiency of directory mapping can be improved.

  Such a keyword for directory mapping is also stored in the search service database 222 as part of information on the directory structure.

  The detailed operation of each stage of the tree structure formation stage (S110) has been described above. The formed tree structure is formed by using information included in the classified document itself. In this case, the tree structure can include unfavorable documents, clusters and their concatenation. Thus, the tree structure including noise will be described based on FIG. 3 as a tree structure before update.

  The document providing step (S120) is a step in which the document providing unit 215 provides each document clustered in the user terminal 230 by directory.

  This stage is characterized in that the user can easily search each document in his / her area of interest by visualizing and providing a predetermined subject, that is, a directory structure belonging to a classification and an inclusive relation of clusters belonging to a directory. is there.

  The user transmits information about the subject he is interested in to the search service providing server 210 through the user terminal 230. This is performed, for example, by clicking a link to a web page that provides a search service for the subject.

  The document providing unit 215 receives a request from the user terminal 230 including information on the subject of interest of the user, and, as a response thereto, an access link for each document classified and clustered at each of the above steps. A web page including can be provided. Through this, a search service is provided to the user.

  On the other hand, in order to provide a document search service, the document providing unit 215 acquires information on each clustered document from the search service database 222.

  The form and manner in which the response of the document providing unit 215 is provided is adjusted using CSS (cascading stylesheets) or the like. It is also possible to use a separate content management system (CMS).

  The step (S130) of obtaining user reaction information is a step in which the reaction information acquisition unit 216 obtains user reaction information.

  In this embodiment, the search service can provide the tree structure visualized by the document providing unit 215 to the user. Such visualization of the tree structure is performed in the classification structure display area 510, the search position display area 520, the cluster display area 530, and the like shown in FIGS.

  In such an area, the user can click on the part associated with the directory, cluster and document that he / she wishes to search in each area. Then, the user's action to determine his / her search position and browse the document, that is, the reaction to the document classified in the user's tree structure for browsing the document is displayed. The reaction information acquisition unit 216 collects information related to the response to the document.

  The obtained reaction information is stored in the reaction information database 223, and the user reaction information can include identification information for the user, reaction time, and information about the reaction target. The user response information collected and stored in this way is used as basic data for updating the tree structure.

  On the other hand, the document providing unit 215 can provide search means for searching for information contained in a tree-structured document. Such search means is provided in the form of a search window included in a web page provided by the document providing unit 215 to the user terminal 230. The question words input in such a search window are also used for updating the tree structure as user reaction information.

The step of updating the tree structure of the document (S140) is a step of updating the tree structure of the document by the tree structure updating unit 217. The tree structure update unit 217 can update the tree structure of the document based on the user reaction information stored in the reaction information database 232.
The tree structure is updated in a form that excludes unnecessary directories, clusters, and documents from the tree.

  For example, if a link to a document belonging to a specific cluster is continuously provided to the user but the frequency of browsing by the user is relatively low compared to other documents, the document and the cluster The tree structure can be updated by disconnecting (releasing) the connection with.

  Similarly, when the user's response frequency to a specific cluster is relatively low, it is possible to exclude (delete) the corresponding cluster from the directory.

  In addition, the upper classification can be changed in consideration of the user's browsing frequency for the included lower classification. As an example, when the frequency of user visits to each cluster included in a specific directory is excessively low, the directory can be excluded (deleted) from the tree structure.

  Such a change in the tree structure is performed by quantifying the degree of connection between the components of the tree structure. As an example, by assigning a predetermined score to a frequently used search route, a high score can be given to each link included in a frequently used search route, and vice versa. , A low score can be given to the corresponding connection. In this case, the tree structure can be updated by a method of selecting each connection having a score of a predetermined value or less as a candidate for disconnection (cancellation).

  The ratio of search paths accessed by users for a specific document is also considered as information for updating the tree structure. When a document is related to two or more superordinate concepts, if the user's inflow route (access route) is biased to a specific route, the tree structure can be updated in a form that leaves only that route. . When the ratio of inflow (access) in a specific route is very low, the route can be excluded.

  As described above, the user's inquiry frequency with respect to the search route and the inflow route related to the specific document are calculated based on the user's reaction information with respect to the information provided by the document classification unit 215.

  On the other hand, the tree structure is updated by further considering the question word information input by the viewers who have viewed the documents belonging to the tree structure. When searching and browsing a document on the web, it is determined that a question word input to input a search result including the document is related to the document. As an example, a document viewed by a user who has input a query word “Bordeaux wine” using a search engine is determined to be a document related to “Bordeaux wine”.

  On the other hand, the tree structure is also updated by creating new directories and new clusters. When new documents are added to the original document database 221, these new documents are newly linked to the existing tree structure.

  As described above, when forming a cluster of documents, the number of documents included in the cluster can also be considered. When the number of documents included in a specific cluster becomes excessively large due to the addition of a new document, it is possible to separate the corresponding cluster into a large number of clusters. In this case, changing the cluster to a directory is also considered.

  The tree structure updated in this way is stored in the search service database 222. Also, the updated tree structure is provided to the user by the document providing unit 215, and through this, the user can search for a tree structure having a more accurate classification result.

  Meanwhile, the search service providing server 210 according to an embodiment of the present invention may further include an advertisement providing unit 218.

  The advertisement providing unit 218 can provide advertisement content to the user terminal 230. The advertisement content is stored in the advertisement database 224 and is called by the advertisement providing unit 218 and transmitted to the user terminal 230.

  Information regarding the user, information regarding a document searched by the user, and the like are considered as factors for determining the advertisement content transmitted to the user terminal 230.

  As an example, when a user performs a login procedure in the course of using a document search service, the user's personal information such as the user's age, occupation, gender, and residential area is considered as an advertising content determination factor.

  On the other hand, each piece of information related to a document viewed through the user terminal 230 is considered as an advertisement content determining factor, and question word information input by the user is also considered.

  In addition, each piece of information acquired in the process of using the document search service according to the embodiment of the present invention by the user is also considered as a content determination factor.

  As described above, by determining the advertisement content to be provided by using the information about the user and the information about the document searched by the user, there is an advantage that the effect of the provided advertisement is maximized.

  FIG. 3 is a diagram illustrating a document tree structure before update according to an embodiment of the present invention, and FIG. 4 is a diagram illustrating a document tree structure after update according to an embodiment of the present invention.

  3 and 4, each document belonging to the original document database 221 is structured by a hierarchical structure in which a classification (theme), a directory, a cluster, and a document are connected in this order.

  Referring to FIGS. 3 and 4, the classification (theme), which is a superordinate concept in which each document is structured, is 'wine'. The document classification unit 211 determines whether a specific document falls under the “wine” classification.

  A classification (subject) can include at least one directory as its subordinate concept. The 'Wine' classification includes directories named 'Wine Origin', 'Wine History' and 'Wine Etiquette'.

  Since the name of the directory can function as the name of the group in which the user tries to search for the document, the name of the document group used in the vertical site and blog that is the source of each document stored in the original document database 221 is used. By using it as a directory name, the document search efficiency of the user can be improved.

  A directory can include at least one cluster below it. The 'Wine District' directory contains clusters named 'Bordeaux', 'Burgundy', 'Campaña' and 'Bordeaux TV'.

  The cluster can also include at least one document. In the case of document 2 illustrated in FIG. 3, the title includes “Bordeaux” and “Burgundy”, so it is included in all of the “Bordeaux” cluster and the “Burgundy” cluster. .

  On the other hand, the names of directories used in vertical sites and blogs and the field information of each structured document contained in such sites can be used to classify and cluster documents by subject. This is as described in the detailed description of FIGS.

  On the other hand, Document 3, which has the title 'Bordeaux TV similar to wine', is a document related to TV produced by a Korean electronics industry that is not a document for wine, and is inappropriate noise to be classified as a document related to wine. It is a document.

  However, since Document 3 includes the strings “Wine” and “Bordeaux” in its title, and includes marketing activities and product development concepts related to wine as document content, it is included in a separate cluster under the “Wine” classification. Results.

  Since these documents cannot attract the attention of visitors searching for documents with an interest in wine, a lower number of queries will be recorded compared to other documents. Such visitor reaction information is reflected in the update of the tree structure.

  On the other hand, as described above, the question word information input by the document viewer is also reflected in the update of the tree structure. If the distribution of questions entered by the user viewing the document 3 is biased by the product producers “Samsung Electronics” and “TV” instead of “Wine” and “Bordeaux”, this is represented in a tree structure. The document 3 can be excluded (deleted) from the “wine” classification.

  On the other hand, as described above, the ratio of paths flowing into the document is reflected in the update of the tree structure. As an example, if the document 3 is linked not only to the classification (theme) 'wine' but also to the classification (theme) 'TV' (not shown), the user who viewed the document 3 was associated with the latter The ratio of users accessing the document 3 through the search path can be relatively large. In this case, if the degree of weighting exceeds a predetermined value, the tree structure can be updated by blocking the path of the 'wine' classification associated with the document 3.

  As a result, the cluster ‘Bordeaux TV’ and the document 3 are removed in FIG. 4 representing the tree structure after the update.

  On the other hand, in the case of the document 4 having the title of “Campagna Mozzarella Cheese-Perfect for Bordeaux Wine”, the content included is for the mozzarella cheese, not the document dealing with the content for Bordeaux as a wine producing area.

  Therefore, the inquiry frequency of the user who searches for the cluster 'Bordeaux' with an interest in Bordeaux as a wine-producing region can be lower than that of the document 1. In this case, in the process of updating the tree structure, the connection between the document 4 and the cluster “Bordeaux” is broken (released), and the result shown in FIG. 4 is obtained.

  On the other hand, as described above, not only the inquiry frequency but also the ratio of the inflow route is considered in this process.

  However, 'Campaña' is also an Italian wine production area, so unlike the case of Document 3, the cluster itself may not be removed. As described above, in addition to excluding the cluster and the document itself from the subject, the tree structure is updated by changing the connection in the tree structure.

  FIG. 5 is a diagram illustrating a document search service provision screen before update according to an embodiment of the present invention, and FIG. 6 is a diagram illustrating a document search service provision screen after update according to an embodiment of the present invention. It is. 5 is a document search service providing screen related to FIG. 3, and FIG. 6 is a document search service providing screen related to FIG.

  5 and 6, the document search service providing screen includes a classification display area 500, a classification structure display area 510, a search position display area 520, a cluster display area 530, and a document link 532.

  As described above, the document providing unit 215 can generate and transmit a web page provided to the user terminal 230 in order to provide the document search service of the present embodiment.

  The web page provided by the document search service can have a screen configuration as illustrated in FIGS. 5 and 6. Such a screen configuration includes a display area for visualizing the structure of the search target document.

  The classification display area 500 is an area in which information related to classification (subject), which is a superordinate concept in which a document is structured, is displayed. In this embodiment, “Wine” as a classification name is displayed as a relatively differentiated attribute.

  The classification structure display area 510 is an area for providing the tree structure illustrated in FIGS. 3 and 4 to the user. In the case of FIG. 4, “the wine region” and the cluster “Bordeaux”, which are directories searched by the user, are displayed as attributes differentiated from other items.

  In addition, a search position display area 520 for displaying the directory and cluster being searched by the user is additionally provided.

  The cluster display area 530 is an area that provides an access means for each document belonging to the cluster that the user is searching for. In this embodiment, the central concept of the cluster that the user is searching for is “Bordeaux”, and the cluster display area 530 is provided with a document link 532 as an access means for a document associated with the central concept “Bordeaux”. The

  The cluster display area 530 is provided with a document link 532 for each document belonging to the cluster searched by the user. The document link 532 can display the title information of the document to be referenced as anchor text.

  The document link 532 is a link to each document. By selecting this link, the user can access the contents of the document including the information he / she wants to search. In this case, the content of the selected document is provided by generating a new browser window on the user terminal 230, and a part of the browser window on which the document search service providing screen as shown in FIGS. 5 and 6 is displayed. Or it can be provided by updating everything.

  The user can change his / her search target by selecting items in each area provided in FIG. 5 and FIG. 6 by a click method or the like. In response to this, the document providing unit 215 can provide information regarding the document visualized in the classification structure display area 510, the search position display area 520, and the cluster display area 530.

  On the other hand, the information regarding the user's document search action is stored in the reaction information database 232 as user reaction information. As described above, the tree structure update unit 217 can update the tree structure based on the user response information.

  By visually transmitting the directory structure and clustering structure to the user through the above-described screen structure, information on the field of interest can be efficiently obtained without visiting the vertical sites and blogs visited by the user individually. The tree structure can be updated more accurately and efficiently by reflecting, in the tree structure, user reaction information on the screen described above.

  FIG. 7 is a diagram illustrating a document search service providing screen including an advertisement display area according to an embodiment of the present invention. Referring to FIG. 7, a document search service providing screen including an advertisement display area according to an embodiment of the present invention includes a classification display area 500, a classification structure display area 510, a search position display area 520, a cluster display area 530, and an advertisement display. Region 710 is included.

  The advertisement display area 710 is an area in which advertisement content provided by the advertisement providing unit 218 to the user terminal 230 is displayed. In the advertisement display area 710, text advertisement content 711 and animation advertisement content 712 are displayed.

  The text advertisement content 711 and the animation advertisement content 712 may additionally include a link that can be connected to a site that includes additional information related to the advertiser.

  The advertisement displayed in the advertisement display area 710 includes a PPC (pay per click) model in which the advertising cost is executed according to the number of clicks, and a PPV in which the advertising cost is executed in accordance with at least one of the number of exposures. Operated based on a (payper view) model.

  On the other hand, as described above, information relating to the user, information relating to the document searched by the user, and the like are considered as factors for determining the advertisement content displayed in the advertisement display area 710.

  As an example, information about the subject (classification), directory, cluster, and document title of the document is considered as an advertising content determinant. Referring to FIG. 5, the subject that the user is searching for is “wine”, the directory is “wine origin”, and the cluster is “Bordeaux”. By providing the text advertisement content 711 having the title 'Bordeaux wine joint purchase application' based on such information, the effect of the advertisement can be maximized.

  On the other hand, the document search service providing method can be created by a computer program. Each code and each code segment constituting the program is easily inferred by a computer programmer in the field. In addition, the program is stored in a computer-readable information storage medium, and is read and executed by a computer, thereby realizing a document search service providing method. The information storage medium includes a magnetic recording medium, an optical recording medium, and a carrier wave medium.

  The terminology used in this application is used to describe specific examples and is not intended to limit the invention. An expression used in the singular encompasses the expression of the plural, unless it has a clearly different meaning in context.

In this application, a term such as “including” or “having” designates the presence of a feature, number, step, action, component, part, or combination thereof as described in the specification. It should also be understood as not excluding the existence or additional possibilities of other features or numbers, steps, actions, components, parts or combinations thereof.
Terms such as “first” and “second” are used to describe various components, but each component should not be limited by each term. Each term is used only to distinguish one component from another.

  Although the embodiments of the present invention have been described above, many embodiments other than the above-described embodiments are within the scope of the claims of the present invention. Those skilled in the art to which the present invention pertains will understand that the present invention can be embodied in a modified form without departing from the essential characteristics thereof. Accordingly, the disclosed embodiments are to be considered in an illustrative, not a limiting sense. The scope of the present invention is shown not in the above description but in the claims, and all differences within the equivalent scope should be analyzed as being included in the present invention.

210 Search service provision server 211 Document classification unit 212 Subject extraction unit 213 Cluster formation unit 214 Directory mapping unit 215 Document provision unit 216 Reaction information acquisition unit 217 Tree structure update unit 218 Advertisement provision unit 221 Original document database 222 Search service database 223 Reaction information Database 224 Advertising database 230 User terminal 240 Tree structure forming unit 500 Classification display area 510 Classification structure display area 520 Search position display area 530 Cluster display area 532 Document link 710 Advertisement display area 711 Text advertisement content 712 Animation advertisement content

Claims (7)

A method for providing a user terminal with a document search service capable of searching a classified document by a document search service providing system,
Receiving information on the subject of interest from the user terminal;
Extracting a document related to information on the subject of interest from the classified documents, and providing the extracted document to the user terminal;
The document classification is:
Classify documents by subject,
Extract the title of the document,
Forming a cluster based on the extracted subjects;
A document search service providing method, characterized in that the cluster is formed by a document structure forming method comprising mapping the cluster to a predetermined directory belonging to the subject. The document includes at least one field;
The method for providing a document search service according to claim 1, wherein the title of the document is extracted in consideration of attributes of fields constituting the document. The formation of the cluster is
The document search service provision according to claim 1, wherein the extracted subject is classified in syllable units, and a portion of the subject that is shared with another document is selected as a central concept candidate of the cluster. Method.   A computer-readable recording medium having recorded thereon a computer program for executing the document search service providing method according to any one of claims 1 to 3. A system for providing a user terminal with a document search service capable of searching for documents classified in a tree structure,
A tree structure forming unit for forming a classification structure of the document;
Of the documents classified by the classification structure, a document related to information on a subject received by the user from the user terminal is extracted, and the extracted document is extracted from the user terminal. Including a document providing unit provided to
The tree structure forming unit includes:
A document classification unit for classifying the document according to subject;
A subject extraction unit for extracting the subject of the document;
A cluster forming unit for forming a cluster based on the extracted title;
A document mapping service providing system comprising: a directory mapping unit that maps the cluster to a predetermined directory belonging to the subject. The document includes at least one field;
The system for providing a document search service according to claim 5, wherein the subject extraction unit extracts the subject based on an attribute of a field constituting the document. The cluster forming part is
6. The document search service provision according to claim 5, wherein the extracted subject is classified in syllable units, and a portion of the subject that is shared with other documents is selected as a central concept candidate of the cluster. system.