JP2004199273A - Information automatic classification device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To remove complication of bookmark management generated when using many bookmarks, to perform accurate classification, and to thereby improve serviceability without damaging simplicity of function of the bookmark. <P>SOLUTION: This information automatic classification device 104 generates learning information of a characteristic of a Web site in each classification item based on directory information provided by a directory service server 103. On the other hand, the information automatic classification device 104 extracts characteristic information of a corresponding Web site based on bookmark information held by an information client 101. The extracted characteristic information is compared respectively with the learning information in each generated classification item, and similarity between the characteristic information and each learning information is calculated respectively. The classification item corresponding to the learning information related to the maximum similarity among each calculated similarity is determined as the classification item to the bookmark information. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an automatic information classification apparatus, and more particularly, to an automatic information classification apparatus connected to a network environment including a directory service server and a client apparatus for providing directory information by classifying a plurality of website information according to a predetermined classification rule. Equipment related.
[0002]
[Prior art]
With the advance of electronic computers and the spread of the Internet, a large amount of information has been transmitted from Web sites. Since the information provided on the Internet is enormous, it is required to efficiently classify and organize the information.
[0003]
Conventionally, there is a directory service as one method for responding to such a request. The directory service manually recognizes information transmitted from a large number of Web sites, classifies the information into appropriate classification categories, organizes the information, and provides the classified information. A general user can easily reach desired information by tracing an appropriate category using a directory service.
[0004]
However, in this directory service, specialized personnel having experience and knowledge examine Web site information and determine an appropriate category based on the content, so that the classification accuracy of the generated classification information is low. Although it has the advantage of being expensive, the task of manually classifying information on a large number of websites is extremely complicated, and it is difficult to maintain appropriate categories for frequently updated website information. There is also a drawback that category determination of newly generated information is difficult to be performed smoothly.
[0005]
As another method for responding to the above request, there is a search service. The search service is a service that automatically and regularly collects information on a Web site using a technique called a Web robot, and enables a search to be performed on the obtained information. This search service automatically collects information on a Web site, so that the problem associated with a high frequency of updating information provided from a Web site as in a directory service can be solved. In order to obtain the information, it is necessary to provide an appropriate search word, and therefore, there is a problem that it is difficult for a general user who is unfamiliar with the search act to easily obtain desired information.
[0006]
In addition, the above two services provided on the Internet have a problem of security risk when using the service, and a problem of stopping the service.
[0007]
By the way, there is a bookmarking method as a simple method that does not provide the service provided on the Internet and that meets the above-mentioned request. This technique is generally included in software for browsing Web sites. According to this bookmark, URLs (Uniform Resource Locators), which are access information to Web sites, can be classified hierarchically, and Web sites can be arbitrarily named and stored.
[0008]
However, in the above-mentioned method of bookmarks, since the creation and organization of bookmarks must all be performed by the user, the user must perform all management such as maintaining an appropriate classification state and deleting unnecessary bookmarks. There is a problem that it is complicated for the person.
[0009]
On the other hand, for example, in the technology described in Patent Document 1, when a user attempts to register a URL, a keyword specifying tag in an HTML source of the URL is searched, and the keyword is specified based on the keyword specified by the keyword specifying tag. One that categorizes and registers URLs has been devised.
[0010]
[Patent Document 1]
JP-A-11-167580
[0011]
[Problems to be solved by the invention]
However, according to the technology described in Patent Document 1, classification is performed according to a keyword specification tag set by the creator of the HTML source on the Web. Therefore, an appropriate keyword may not be specified. There is a problem that is not done.
[0012]
The present invention has been made in view of such a problem, and eliminates the complexity of bookmark management that occurs when a large number of bookmarks are used, and performs accurate classification so that a bookmark function is provided. It is an object of the present invention to provide an automatic information classification device which has improved usefulness without impairing the simplicity.
[0013]
[Means for Solving the Problems]
In order to achieve the above object, according to the present invention, information connected to a network environment including a directory service server and a client device that classify a plurality of website information according to a predetermined classification rule and provide directory information In the automatic classification device, directory information acquisition means for acquiring the directory information provided by the directory service server via a network, and classification item information and a Web site corresponding to each classification item from the acquired directory information. Extracting means for extracting access information, and first site information obtaining means for accessing a Web site corresponding to the access information and obtaining site information provided by the Web site using the extracted access information And obtained by the first site information obtaining means. Analyzing the extracted site information to extract the characteristics of the site information, and integrating the characteristic information of the plurality of site information extracted by the first characteristic extracting unit for each classification item. Learning means for generating learning information through the network, bookmark information obtaining means for obtaining the bookmark information held by the client device via the network, and a Web corresponding to the bookmark information using the obtained bookmark information. A second site information acquisition unit that accesses the site to acquire site information provided by the Web site, and analyzes the site information acquired by the second site information acquisition unit to determine the characteristics of the site information. A second feature extraction unit to be extracted, and feature information extracted by the second feature extraction unit, the feature information extracted by the learning unit. Calculating means for comparing each of the pieces of learning information generated for each classification item with each other and calculating the similarity between the feature information and each piece of learning information; and the largest similarity among the similarities calculated by the calculating means. And a classification item corresponding to the learning information related to the bookmark information related to the site information acquired by the second site information acquisition unit. A classification device is provided.
[0014]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0015]
[First Embodiment]
FIG. 1 is a block diagram showing a basic configuration of a network system including an automatic information classification device according to the first embodiment of the present invention.
[0016]
In FIG. 1, reference numeral 101 denotes an information client. The information client 101 provides functions for receiving a search request for a Web site from a user and performing processing according to an operation performed by the user on search results.
[0017]
102 is an information server. Web site information is provided using a protocol such as HTTP (Hypertext Transfer Protocol).
[0018]
Reference numeral 103 denotes a directory service server that provides a directory service. The directory service server 103 categorizes and sorts information transmitted from a number of Web sites into predetermined classification categories, and provides directory information. This directory information is used when the user of the information client 101 specifies a classification item corresponding to desired site information and searches for an appropriate Web site from a plurality of Web sites.
[0019]
Reference numeral 104 denotes an automatic information classification device according to the present invention. The automatic information classification device 104 performs appropriate management of bookmarks, and will be described later in detail with reference to FIGS. Note that the automatic information classification device 104 in the present embodiment exists independently of the information client 101, but may be provided in the information client 101 instead.
[0020]
The information client 101, the information server 102, the directory service server 103, and the automatic information classification device 104 are each configured by a computer and are connected to each other by a network 105. Although only one information client and one information server are shown in FIG. 1 for the sake of convenience, generally, a plurality of these can exist on the network 103.
[0021]
FIG. 2 is a block diagram showing a functional configuration of the information client 101.
[0022]
In the figure, reference numeral 201 denotes a dialogue unit which receives a search request and an editing operation from a user. The dialogue unit 201 is realized by various input devices such as a keyboard, a pointing device, a touch panel, a joystick, a pen, and a tablet of a computer, a display device such as a bitmap display, and a basic operation system (operating system) on the computer. Then, an instruction from the user is received via the input device, and information is presented to the user via the display device.
[0023]
202 is a control unit. The control unit 202 corresponds to a web browser, and the web browser is a program for interpreting and executing information in a specific format. The Web browser has a bookmark function for storing a URL (bookmark), which is information necessary for accessing a Web site. With this bookmark function, the URL of the Web site currently being browsed by the Web browser can be saved.
[0024]
203 is a communication unit. The control unit 202 performs communication with the information server 102, the directory service server 103, and the automatic information classification device 104 existing on the network 105.
[0025]
FIG. 3 is a diagram showing a basic configuration of the automatic information classification device 104.
[0026]
Reference numeral 302 denotes a server information processing unit which acquires information provided by various servers (including the information server 102 and the directory service server 103) via the network 105.
[0027]
Reference numeral 301 denotes a directory information processing unit. The directory information processing unit 301 uses the server information processing unit 302 to acquire directory information provided by the directory service server 103, and converts the data format of the directory information into a predetermined format. This is analyzed to extract classification identification information (classification category name) and reference information (URL for identifying a Web site on the Internet) corresponding to each classification category. In addition, using the server information processing unit 302, a Web site corresponding to the extracted reference information (URL) is accessed to acquire site information.
[0028]
Reference numeral 306 denotes a feature information extraction processing unit that performs feature extraction processing on given Web site information and generates feature information corresponding to the classification identification information. Specifically, only the document information is extracted from the site information, at least a word contained in the document information is extracted, and a vector having the word as a component is created, and this is used as feature information.
[0029]
A learning processing unit 303 performs a learning process on the plurality of pieces of feature information extracted by the feature information extraction processing unit 306. That is, learning information is generated by integrating a large number of pieces of feature information for each classification item.
[0030]
Reference numeral 304 denotes a reference information processing unit which acquires bookmark information (URL) held by the control unit 202 (Web browser) of the information client 101 via the network 105, requests the server information processing unit 302, and outputs the bookmark information. To access the information server corresponding to the Web server and acquire Web site information.
[0031]
Reference numeral 305 denotes a classification processing unit. The classification processing unit 305 compares the feature information related to the bookmark information with each piece of learning information for each classification item related to the directory information, and determines the learning information having the highest similarity to the feature information. Is calculated. The bookmark information is classified by determining the classification item related to the calculated learning information as the classification item for the bookmark information related to the feature information.
[0032]
FIG. 4 is a flowchart illustrating a procedure of bookmark information classification processing executed by the information automatic classification device 104.
[0033]
Steps S1 to S5 are a learning process. Briefly, directory information is obtained from the directory service server 103 on the network 105, and the obtained directory information is analyzed to obtain classification identification information (classification category name) and reference information (URL). ) Is obtained, and further, the site information of the Web site corresponding to the reference information is obtained, and the content is analyzed to generate learning information for each classification item.
[0034]
First, in step S1, the directory information processing unit 301 uses the server information processing unit 302 to acquire directory information provided by the directory service server 103 via the network 105.
[0035]
Next, in step S2, the directory information processing unit 301 analyzes the directory information acquired from the directory service server 103, and extracts classification identification information (classification category name) and reference information (URL) for each classification category.
[0036]
In step S3, the directory information processing unit 301 uses the server information processing unit 302 to acquire, via the network 105, site information of a Web site corresponding to the reference information (URL) extracted in step S2.
[0037]
In step S4, the feature extraction processing unit 306 performs feature extraction processing based on the website information acquired in step S3, and extracts features of the website. Next, the learning processing unit 303 performs multiplication of many pieces of feature information for each classification item (classification category) to generate learning information corresponding to the classification item.
[0038]
In step S5, it is determined whether or not there is still directory information to be continuously executed in steps S1 to S4, and if it is to be executed continuously, the process returns to step S1.
[0039]
Steps S6 to S9 are a classification process. This time, the reference information (bookmark URL) uniquely stored in the information client 101 is obtained, and the feature extraction process of the corresponding website information is performed. Then, by comparing the obtained feature information with the learning information for each classification item generated in the learning process of steps S1 to S4, a classification item that well corresponds to the reference information (bookmark URL) is automatically determined. It is.
[0040]
First, in step S6, the reference information processing unit 304 acquires bookmark information (reference information) held by the control unit 202 (Web browser) of the information client 101.
[0041]
Next, in step S7, the reference information processing unit 304 uses the server information processing unit 302 to change the site information of the Web site corresponding to the bookmark information (reference information, URL) acquired in step S6, as in step S3. , Via the network 105.
[0042]
In step S8, first, the feature extraction processing unit 306 performs feature extraction processing based on the website information acquired in step S7, and extracts feature information. Next, the classification processing unit 305 compares the characteristic information related to the bookmark information with each piece of learning information for each classification item generated in step S4, and calculates the similarity between the characteristic information and each piece of learning information. . Then, a classification item corresponding to the learning information related to the maximum similarity among the calculated similarities is determined as a classification item for the bookmark information.
[0043]
In step S9, it is determined whether or not the processing of steps S6 to S8 has been executed for all of the bookmark information held by the control unit 202 (Web browser) of the information client 101. Proceeds to step S10, and returns to step S6 if there is any bookmark information that has not been executed.
[0044]
In step S10, the classification item for each bookmark information obtained in step S8 is sent to the information client 101, and the information client 101 uses the classification item for bookmark information management.
[0045]
As described above, since the classification of bookmarks is performed automatically, the complexity of bookmark management that occurs when a large number of bookmarks are used is eliminated, and the usefulness is maintained without impairing the simplicity of the bookmark function. It can be improved.
[0046]
In addition, since the bookmarks stored in the device used by the user are classified using the classification characteristics of the server providing the directory service, appropriate and accurate classification is performed.
[0047]
[Other embodiments]
In the above embodiment, there is only one directory service server that provides a directory service. However, a plurality of directory service servers may exist. In this case, the contents of the plurality of directory services are individually stored in individual learning data. Or all the directory service information may be collectively regarded as one virtual directory service and treated as a single learning data.
[0048]
Further, in the above-described embodiment, the single classification identification information having the highest similarity is calculated in step S8. Instead, a plurality of classification identification information may be calculated and classified into each. .
[0049]
Further, in the above-described embodiment, the single classification identification information having the highest similarity is calculated in step S8. However, instead of this, a certain threshold value is provided, and a similarity value exceeding this value cannot be obtained. In this case, a result that no classification identification information is matched may be calculated.
[0050]
In addition, the present invention may be applied to an apparatus including a plurality of devices (for example, a host computer, an interface device, and the like) or an apparatus including a single device. That is, the information automatic classification device 104 may be configured as the same device as at least one of the information client 101, the information server 102, and the directory service server 103.
[0051]
Further, an object of the present invention is to supply a medium recording a program code of software for realizing the functions of the above-described embodiments to a system or an apparatus, and a computer (or CPU or MPU) of the system or apparatus to store the medium. It is also achieved by reading and executing the program code stored in the medium. In this case, the program code itself read from the storage medium implements the functions of the above-described embodiment, and the storage medium storing the program code constitutes the present invention.
[0052]
Examples of the storage medium for supplying the program code include a floppy (registered trademark) disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD-R, magnetic tape, nonvolatile memory card, and ROM. Can be used. When the computer executes the readout program code, not only the functions of the above-described embodiment are realized, but also an OS (Operating System) running on the computer based on the instruction of the program code. Performs part or all of the actual processing, and the processing realizes the functions of the embodiments.
[0053]
Further, after the program code read from the storage medium is written into a memory provided on a function expansion board inserted into the computer or a function expansion unit connected to the computer, the function expansion is performed based on the support of the program code. This includes the case where the CPU or the like provided in the board or the function expansion unit performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.
[0054]
When the present invention is applied to the storage medium, the storage medium stores program codes for executing processing corresponding to the flowcharts described above.
[0055]
As described above, various embodiments of the present invention have been shown and described. Examples of the embodiments of the present invention are listed below.
[0056]
[Embodiment 1] In an information automatic classification device connected to a network environment including a directory service server and a client device that classify a plurality of website information according to a predetermined classification rule and provide directory information,
Directory information obtaining means for obtaining directory information provided by the directory service server via a network;
Extracting means for extracting classification item information and access information to a Web site corresponding to each classification item from the obtained directory information;
A first site information acquisition unit that accesses a website corresponding to the access information and acquires site information provided by the website using the extracted access information;
A first feature extracting unit that analyzes the site information acquired by the first site information acquiring unit and extracts features of the site information;
A learning unit for generating learning information by integrating feature information of a plurality of site information extracted by the first feature extracting unit for each classification item;
Bookmark information acquisition means for acquiring the bookmark information held by the client device via the network;
A second site information acquisition unit that accesses the website corresponding to the bookmark information using the acquired bookmark information and acquires site information provided by the website;
A second feature extraction unit that analyzes the site information acquired by the second site information acquisition unit and extracts a feature of the site information;
Calculating means for comparing the feature information extracted by the second feature extracting means with the learning information for each classification item generated by the learning means, and calculating the similarity between the feature information and each learning information, respectively; When,
A classification item corresponding to the learning information related to the maximum similarity among the similarities calculated by the calculation unit is defined as a classification item corresponding to the bookmark information related to the site information acquired by the second site information acquisition unit. Classification item determination means to be determined
An information automatic classification device, comprising:
[0057]
[Embodiment 2] The directory information provided by the directory service server is such that the user of the client device specifies a classification item corresponding to desired site information, and selects an appropriate Web site from a plurality of Web sites. Which is used when searching,
The apparatus according to claim 1, wherein the directory information obtaining unit converts a data format of directory information provided by the directory service server into a predetermined format.
[0058]
[Embodiment 3] The first and second feature extracting means respectively extract document information from the site information acquired by the first and second site information acquiring means, and analyze the document information. The information automatic classifying apparatus according to the first embodiment, wherein a feature of the site information is extracted by using the information.
[0059]
[Embodiment 4] Each of the first and second feature extraction means extracts at least a word contained in the extracted document information, and performs the analysis by using a vector having the word as a component. The information automatic classification device according to embodiment 3, characterized in that:
[0060]
[Embodiment 5] An automatic information classification apparatus applied to an information classification apparatus connected to a network environment including a directory service server and a client apparatus for providing directory information by classifying a plurality of Web site information according to a predetermined classification rule. In the classification method,
Directory information obtaining step of obtaining directory information provided by the directory service server via a network;
An extraction step of extracting classification item information and access information to a Web site corresponding to each classification item from the obtained directory information;
A first site information obtaining step of using the extracted access information to access a Web site corresponding to the access information to obtain site information provided by the Web site;
A first feature extraction step of analyzing the site information acquired by the first site information acquisition step and extracting features of the site information;
A learning step of generating learning information by integrating feature information of a plurality of site information extracted by the first feature extraction step for each classification item;
Bookmark information acquisition step of acquiring the bookmark information held by the client device via the network,
A second site information obtaining step of using the obtained bookmark information to access a Web site corresponding to the bookmark information to obtain site information provided by the Web site;
A second feature extraction step of analyzing the site information acquired by the second site information acquisition step and extracting features of the site information;
A calculating step of comparing the feature information extracted in the second feature extracting step with learning information for each classification item generated in the learning step, and calculating a similarity between the feature information and each learning information, respectively; When,
A classification item corresponding to the learning information related to the largest similarity among the similarities calculated in the calculation step is classified into a classification item corresponding to the bookmark information related to the site information acquired in the second site information acquisition step. And the classification item determination step to determine
An automatic information classification method, comprising:
[0061]
[Sixth Embodiment] The directory information provided by the directory service server is such that the user of the client device specifies a classification item corresponding to desired site information, and selects an appropriate Web site from a plurality of Web sites. Which is used when searching,
The information automatic classification method according to embodiment 5, wherein the directory information obtaining step converts a data format of directory information provided by the directory service server into a predetermined format.
[0062]
[Embodiment 7] The first and second feature extraction steps include extracting document information from the site information acquired by the first and second site information acquisition steps, respectively, and analyzing the document information. The method for automatically classifying information according to embodiment 5, wherein the feature of the site information is extracted by the following.
[0063]
[Eighth Embodiment] In the first and second feature extraction steps, at least a word contained in the extracted document information is extracted, and the analysis is performed by using a vector including the word as a component. The information automatic classification method according to embodiment 7, characterized in that:
[0064]
[Embodiment 9] An automatic information classification apparatus applied to an automatic information classification apparatus connected to a network environment including a directory service server and a client apparatus for providing directory information by classifying a plurality of Web site information according to a predetermined classification rule. In a program for causing a computer to execute the classification method,
The information automatic classification method,
Directory information obtaining step of obtaining directory information provided by the directory service server via a network;
An extraction step of extracting classification item information and access information to a Web site corresponding to each classification item from the obtained directory information;
A first site information obtaining step of using the extracted access information to access a Web site corresponding to the access information to obtain site information provided by the Web site;
A first feature extraction step of analyzing the site information acquired by the first site information acquisition step and extracting features of the site information;
A learning step of generating learning information by integrating feature information of a plurality of site information extracted by the first feature extraction step for each classification item;
Bookmark information acquisition step of acquiring the bookmark information held by the client device via the network,
A second site information obtaining step of using the obtained bookmark information to access a Web site corresponding to the bookmark information to obtain site information provided by the Web site;
A second feature extraction step of analyzing the site information acquired by the second site information acquisition step and extracting features of the site information;
A calculating step of comparing the feature information extracted in the second feature extracting step with learning information for each classification item generated in the learning step, and calculating a similarity between the feature information and each learning information, respectively; When,
A classification item corresponding to the learning information related to the largest similarity among the similarities calculated in the calculation step is classified into a classification item corresponding to the bookmark information related to the site information acquired in the second site information acquisition step. And the classification item determination step to determine
A program characterized by having:
[0065]
[Embodiment 10] The directory information provided by the directory service server is such that the user of the client device specifies a classification item corresponding to desired site information, and selects an appropriate Web site from a plurality of Web sites. Which is used when searching,
The program according to embodiment 9, wherein the directory information obtaining step converts a data format of directory information provided by the directory service server into a predetermined format.
[0066]
[Embodiment 11] The first and second feature extraction steps include extracting document information from the site information acquired by the first and second site information acquisition steps, respectively, and analyzing the document information. The program according to the ninth embodiment, wherein the feature of the site information is extracted by the following.
[0067]
[Embodiment 12] The first and second feature extraction steps each include extracting at least a word contained in the extracted document information and performing the analysis by using a vector having the word as a component. 12. The program according to embodiment 11, wherein
[0068]
[Thirteenth embodiment] An automatic information classification apparatus applied to an automatic information classification device connected to a network environment including a directory service server and a client device that classifies a plurality of Web site information in accordance with a predetermined classification rule and provides directory information. In a computer-readable storage medium storing a classification method as a program,
The information automatic classification method,
Directory information obtaining step of obtaining directory information provided by the directory service server via a network;
An extraction step of extracting classification item information and access information to a Web site corresponding to each classification item from the obtained directory information;
A first site information obtaining step of using the extracted access information to access a Web site corresponding to the access information to obtain site information provided by the Web site;
A first feature extraction step of analyzing the site information acquired by the first site information acquisition step and extracting features of the site information;
A learning step of generating learning information by integrating feature information of a plurality of site information extracted by the first feature extraction step for each classification item;
Bookmark information acquisition step of acquiring the bookmark information held by the client device via the network,
A second site information obtaining step of using the obtained bookmark information to access a Web site corresponding to the bookmark information to obtain site information provided by the Web site;
A second feature extraction step of analyzing the site information acquired by the second site information acquisition step and extracting features of the site information;
A calculating step of comparing the feature information extracted in the second feature extracting step with learning information for each classification item generated in the learning step, and calculating a similarity between the feature information and each learning information, respectively; When,
A classification item corresponding to the learning information related to the largest similarity among the similarities calculated in the calculation step is classified into a classification item corresponding to the bookmark information related to the site information acquired in the second site information acquisition step. And the classification item determination step to determine
A storage medium comprising:
[0069]
[Embodiment 14] In the directory information provided by the directory service server, a user of the client device specifies a classification item corresponding to desired site information, and selects an appropriate Web site from a plurality of Web sites. Which is used when searching,
The storage medium according to embodiment 13, wherein the directory information obtaining step converts a data format of directory information provided by the directory service server into a predetermined format.
[0070]
[Embodiment 15] The first and second feature extraction steps include extracting document information from the site information acquired by the first and second site information acquisition steps, respectively, and analyzing the document information. The storage medium according to embodiment 13, wherein the feature of the site information is extracted by using the following.
[0071]
[Embodiment 16] The first and second feature extraction steps each include extracting at least a word contained in the extracted document information, and performing the analysis by using a vector including the word as a component. The storage medium according to embodiment 15, wherein:
[0072]
【The invention's effect】
As described in detail above, according to the present invention, classification item information and access information to a Web site corresponding to each classification item are extracted from the directory information provided by the directory service server, and the Web information is extracted using the access information. The user accesses the site to obtain site information provided by the Web site. The acquired site information is analyzed to extract the features of the site information, and by repeating this, the feature information of each of the plurality of extracted site information is integrated for each classification item to generate learning information. On the other hand, the client accesses the Web site using the bookmark information held by the client device, acquires the site information, analyzes the acquired site information, and extracts the features of the site information. The extracted feature information is compared with the generated learning information for each classification item, and the similarity between the feature information and each learning information is calculated. The classification item corresponding to the learning information related to the maximum similarity among the calculated similarities is determined as the classification item for the bookmark information.
[0073]
Thus, since the bookmarks are automatically classified, the complexity of bookmark management that occurs when a large number of bookmarks are used is eliminated, and accurate classification is performed, so that the simplicity of the bookmark function is not impaired. It is possible to improve the usefulness.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a basic configuration of a network system including an automatic information classification device according to a first embodiment of the present invention.
FIG. 2 is a block diagram illustrating a functional configuration of an information client.
FIG. 3 is a diagram showing a basic configuration of an automatic information classification device.
FIG. 4 is a flowchart illustrating a procedure of bookmark information classification processing executed by the information automatic classification device.
[Explanation of symbols]
101 Information client (client device)
102 Information server (Web site server)
103 Directory service server
104 Automatic Information Classifier
105 Network

Claims (1)

An information automatic classification device connected to a network environment including a directory service server and a client device that classify a plurality of website information according to a predetermined classification rule and provide directory information,
Directory information obtaining means for obtaining directory information provided by the directory service server via a network;
Extracting means for extracting classification item information and access information to a Web site corresponding to each classification item from the obtained directory information;
A first site information acquisition unit that accesses a website corresponding to the access information and acquires site information provided by the website using the extracted access information;
A first feature extracting unit that analyzes the site information acquired by the first site information acquiring unit and extracts features of the site information;
A learning unit for generating learning information by integrating feature information of a plurality of site information extracted by the first feature extracting unit for each classification item;
Bookmark information acquisition means for acquiring the bookmark information held by the client device via the network;
A second site information acquisition unit that accesses the website corresponding to the bookmark information using the acquired bookmark information and acquires site information provided by the website;
A second feature extraction unit that analyzes the site information acquired by the second site information acquisition unit and extracts a feature of the site information;
Calculating means for comparing the feature information extracted by the second feature extracting means with the learning information for each classification item generated by the learning means, and calculating the similarity between the feature information and each learning information, respectively; When,
A classification item corresponding to the learning information related to the maximum similarity among the similarities calculated by the calculation unit is defined as a classification item corresponding to the bookmark information related to the site information acquired by the second site information acquisition unit. An automatic information classification apparatus, comprising: a classification item determining means for determining.

Images