JP5959308B2 - ID assigning apparatus, method and program - Google Patents

Description

  The present invention relates to an apparatus, a method, and a program for assigning an ID to a website.

Conventionally, among websites published on the Internet, there are personal websites managed by one or more online individuals set by offline individuals.
Here, the offline individual means each real user who uses the network (Internet), and manages the online individual on the network. An online individual refers to a virtual user who is provided with a predetermined service group through a network, and an offline individual and an online individual have a one-to-one or one-to-many relationship.
In recent years, especially among junior high school students and high school students, each person manipulates multiple online individuals and creates multiple personal websites with each online individual, not only between their own sites, but also between sites with others. In many cases, hyperlinks are provided and information and messages are disclosed and exchanged.

  By the way, a technique for analyzing the link structure of a website in which links are provided in this way has also been proposed. For example, Patent Document 1 discloses that a boundary of a community is determined by analyzing a link structure.

JP 2006-331070 A

  Since the technique of Patent Document 1 extracts website links and recursively collects linked websites until there are no more links, all websites linked by links are the same. Considered a community.

  However, the above-mentioned online individuals manage a plurality of personal websites linked to each other, but these personal websites do not contain information (ID) that identifies the same online individual. . Further, as described above, the personal website is linked with a personal website managed by another online individual. Therefore, it was not possible to specify a personal website managed by the same online individual only by the presence or absence of the link.

Therefore, a method of classifying a plurality of personal websites for each online individual who is an administrator by assuming a certain condition based on an example of a method for managing a plurality of personal websites by online individuals can be considered.
However, when a case that does not meet the assumed condition appears, there is a problem that the accuracy of classification is lowered.

  An object of the present invention is to provide an ID assignment apparatus, method, and program capable of classifying a plurality of personal websites for each online person who is an administrator.

  The present invention provides the following solutions.

  (1) A storage unit that stores a plurality of personal websites and link information indicating adjacency relationships by hyperlinks between the plurality of personal websites, and two personal websites are common based on the link information A calculation unit that calculates a first index indicating the degree of having an adjacent personal website; a first generation unit that generates a set of clusters of one or more personal websites based on the first index; An ID assigning device comprising: an assigning unit that assigns the same administrator ID to each of the plurality of personal websites, which is different for each cluster and within the cluster.

According to such a configuration, the ID allocation device generates a cluster set from a plurality of personal websites based on an index indicating the degree to which two personal websites have a common adjacent personal website. The ID assigning device can classify a plurality of personal websites for each online individual who is an administrator by assigning different administrator IDs to these clusters. Therefore, the ID allocation apparatus can estimate the personal website of the same administrator more accurately by using a universal index.
As a result, the ID allocation device can easily obtain information on personal websites managed by the same online individual, so teachers and guardians can create personal IDs created by children (especially junior and senior high school students). Useful for website monitoring.

  (2) The ID allocation device according to (1), wherein the calculation unit calculates the first index including itself as the adjacent personal website.

  According to such a configuration, the ID allocation device includes itself as an adjacent personal website. Since the same administrator's personal website is often adjacent to each other, the ID assigning device assigns the same administrator's personal website by weighting the pair of adjacent personal websites. Can be estimated more accurately.

  (3) The ID assignment device according to (1) or (2), wherein the first generation unit includes a combination of personal websites in which the first index is larger than a predetermined value in the same cluster.

  According to such a configuration, since the ID allocation device includes a combination of personal websites whose first index is greater than or equal to a predetermined value in the same cluster, the administrator of the personal website having an adjacency similar to the predetermined value or more. Can be easily generated by assuming that they are the same.

  (4) The first generation unit uses the first index relating to the combination of the personal website belonging to the cluster and the personal website outside the cluster as a cluster and a cluster or a cluster and a personal web according to a predetermined rule. Cluster integrated with a second index relating to a combination of sites, and combining a personal website and a personal website, a personal website and a cluster, or a cluster and a cluster based on the first index or the second index The ID assigning device according to (1) or (2), wherein the process of generating is repeated until the first index and the second index are less than a predetermined value.

  According to such a configuration, the ID allocation device integrates the first index into the second index related to the combination of the cluster and the cluster, or the cluster and the personal website, and based on the second index, The cluster which combined the combination of can be generated. Therefore, the ID assigning device can assign the manager ID by sequentially generating and expanding the same cluster by the manager based on the first index.

  (5) The first generation unit sets the maximum value of the first index relating to a combination of a personal website belonging to the cluster and a personal website outside the cluster as the second index (4). The ID assigning device described.

  According to such a configuration, when a cluster is generated, the ID allocation device can easily set the second index related to the combination with the cluster as the maximum value of the index related to the combination with the personal website belonging to the cluster. Can be sought.

  (6) Predetermined predetermined input / output based on a predetermined number of types representing usage forms classified based on the URLs of each of the plurality of personal websites and the structure of the input / output relationship of the hyperlink The ID allocation device according to any one of (1) to (5), further including a second generation unit configured to combine the combinations of predetermined types of personal websites having the relationship (1) to generate the cluster.

  According to such a configuration, the ID allocation device generates a cluster by combining combinations of predetermined types of personal websites that have a predetermined input / output relationship set in advance. Therefore, the ID assigning device can assign the manager ID more accurately in consideration of a predetermined condition based on the case.

  (7) Based on a service provider account that can be identified by the URL of each of the plurality of personal websites, the account includes a third generation unit that combines the same personal websites to generate the cluster (1) ) To (6).

  According to such a configuration, the ID assigning apparatus generates a cluster by combining personal websites with the same service provider account, and therefore can assign an administrator ID more accurately.

  (8) The calculation unit calculates the first index by excluding the personal website satisfying a predetermined condition of the incoming order and the outgoing order of the hyperlink and the link information related to the personal website (1 ) To (7).

  According to such a configuration, the ID allocation device calculates an index by excluding a specific type of personal website and an adjacent hyperlink according to a predetermined condition of the incoming and outgoing orders of the hyperlink. Therefore, the ID assigning apparatus can remove sites that are not subject to the assignment of the administrator ID as noise, and therefore can assign the administrator ID more accurately.

  (9) The incoming order or the outgoing order is the number of hyperlinks in a case where the adjacent personal website is a predetermined type representing a usage form classified based on the URL of the personal website (8) The ID allocation device described in 1.

  According to such a configuration, the ID assigning apparatus can more reliably remove sites that are not subject to assignment of administrator IDs as noise based on the type of site.

  (10) The ID calculation unit according to any one of (1) to (9), wherein the calculation unit calculates the first index by excluding a hyperlink that has occurred outside a predetermined period of the link information. apparatus.

  According to such a configuration, since the ID allocation device calculates the index except for the hyperlink that occurs outside the predetermined period, the accuracy is limited to recent information or a specific period, and the like. Processing load can be reduced.

  (11) When the number of hyperlinks generated between the same personal websites in a predetermined period is less than a predetermined number in the link information, the calculation unit excludes the hyperlinks and the first information The ID assignment device according to any one of (1) to (10), which calculates an index.

  According to such a configuration, when the number of hyperlinks generated between the same personal websites during a predetermined period is less than the predetermined number, the ID allocation device calculates an index excluding these hyperlinks. Therefore, since the ID allocation device removes noise for hyperlinks linked with a predetermined strength or more, it is possible to determine the identity of the manager more reliably.

  (12) The ID allocation device according to (11), wherein the predetermined number is set for each type representing a usage form classified based on a URL of a personal website that is a link source of the hyperlink.

  According to such a configuration, the ID allocation device sets the hyperlink strength threshold value for each type of link source, and thus more reliably generates noise corresponding to the tendency of hyperlinks to vary depending on the usage mode. Can be removed.

  (13) An ID assignment method in which a computer assigns an administrator ID to each of a plurality of personal websites, wherein the computer is adjacent to the plurality of personal websites and hyperlinks between the plurality of personal websites. A calculation step of storing link information indicating a relationship, and calculating a first index indicating a degree to which two personal websites have a common adjacent personal website based on the link information; and Based on the index, a generation step for generating a set of clusters including one or more personal websites, and for each of the plurality of personal websites, an administrator ID that is different for each cluster and is the same in the cluster And assigning step, and how to execute.

  According to such a configuration, the same effect as in (1) can be expected when the computer executes the ID allocation method.

  (14) An ID assignment program for causing a computer to assign an administrator ID to each of a plurality of personal websites, wherein the computer is connected between the plurality of personal websites and the plurality of personal websites. A calculation step of storing link information indicating an adjacency relationship by a hyperlink, and calculating a first index indicating a degree that two personal websites have a common adjacent personal website based on the link information; A generation step of generating a set of clusters composed of one or more personal websites based on the first index, and for each of the plurality of personal websites, different for each cluster and the same in the cluster ID assigning program for executing the assigning step of assigning the administrator ID.

  According to such a configuration, the same effect as in (1) can be expected by causing the computer to execute the ID assignment program.

  According to the present invention, a plurality of personal websites can be classified for each online individual who is an administrator.

It is a figure which shows the relationship between the personal website which concerns on embodiment, and its administrator. It is a schematic diagram which shows the result by which OnID which concerns on embodiment was allocated. It is a block diagram which shows the function structure of the ID allocation apparatus which concerns on embodiment. It is a figure which shows the collection log | history management table which concerns on embodiment. It is a figure which shows the relationship table between sites which concerns on embodiment. It is a figure which shows the example of calculation of Simality between the personal websites concerning embodiment. It is a figure which shows the example of calculation of Similarity between the clusters which concerns on embodiment. It is a flowchart which shows the process in the control part which concerns on embodiment. It is a 1st figure which shows the procedure of the ID allocation process which concerns on embodiment. It is a 2nd figure which shows the procedure of the ID allocation process which concerns on embodiment. It is a 3rd figure which shows the procedure of the ID allocation process which concerns on embodiment. It is a 4th figure which shows the procedure of the ID allocation process which concerns on embodiment. It is a 5th figure which shows the procedure of the ID allocation process which concerns on embodiment.

Hereinafter, an example of an embodiment of the present invention will be described.
The ID assignment device 1 according to the present embodiment is a device that assigns an online ID for identifying an online individual to a personal website managed by the online individual. The ID assignment device 1 may be various information processing devices (computers) such as a server device and a PC (Personal Computer).

FIG. 1 is a diagram showing a relationship between a personal website according to the present embodiment and an administrator thereof.
An offline individual who is a real person manages one or a plurality of online individuals on a network (Internet). Also, online individuals manage one or more personal websites.

  Each online individual, for example, is a student at the same school, is part of a group with the same hobby, or has a certain relationship with others online. Therefore, hyperlinks are often referenced between personal websites managed by a plurality of online individuals.

  Here, the personal website refers to a website for online individuals to publish information about themselves and to exchange messages with others online. For example, the following types of personal websites are each provided by a plurality of service providers.

Profile (profile): A site where you can publish your personal profile.
Gesbu (Guest Book): A site where visitors can post comments as a history.
Real (real time): A site where you can post your current situation in short sentences.
Blog: A site where you can publish a daily updated diary.
My Link: A site where you can post a link to another person's personal website.
Hompe (homepage)-personal site.

  Since online individuals create the above-mentioned multiple types of personal websites with different accounts for each service provider, they are often not linked (identified) with the same ID. For example, an online individual who has “Online ID (OnID) = 1” manages Prof, Gesbu, and Real. These personal websites do not have the information “OnID = 1” and are managed with different account IDs (1 and 2).

  The ID assigning apparatus 1 assigns the same online ID (OnID) to the personal website that is the same as the online individual who is the administrator by the process described later, and assigns a plurality of personal websites to each online individual. Classify.

  FIG. 2 is a schematic diagram illustrating a result of online ID assignment by the ID assignment apparatus 1 according to the present embodiment.

  Hereinafter, in this embodiment, personal websites to which IDs are assigned are classified into the following three types (class A, class B, and class C) that indicate the usage mode of the site. These classes and the above types can be identified from the URL of the personal website.

Class A (prof, hompe): Personal website created by online individuals to distinguish them from others.
Class B (Gesbu, Mylink): A personal website created by an online individual attached to a Class A site.
Class C (real, blog): A personal website created by an online individual alone or in association with a Class A site to distinguish them from others.

  In some cases, the same personal website includes a class A and has a plurality of types (for example, profile and my link). In this case, the class determined from the URL is class A.

FIG. 3 is a block diagram showing a functional configuration of the ID assignment device 1 according to the present embodiment.
The ID assignment device 1 includes a control unit 10, a storage unit 20, a communication unit 30, an input unit 40, and an output unit 50.

  The control unit 10 is a part that controls the entire ID allocation device 1, and by reading and executing various programs stored in the storage unit 20 as appropriate, the control unit 10 cooperates with the hardware described above, and performs various types in the present embodiment. The function is realized. The control unit 10 may be a CPU (Central Processing Unit). In addition, the function of each part with which the control part 10 is provided is mentioned later.

  The storage unit 20 is a storage area for various programs and various data for causing the hardware group to function as the ID assignment device 1, and may be a hard disk (HDD). Specifically, the storage unit 20 stores a program (ID assignment program) that is executed by the control unit 10 in order to realize various functions of the present embodiment.

  Furthermore, the storage unit 20 includes a site storage DB 21 and a site management DB 22. The site storage DB 21 stores page data (HTML file group) of a personal website acquired by the program. The site management DB 22 also stores a collection history management table and an inter-site relation table, which will be described later, created or edited by a program.

  The communication unit 30 is a network adapter in the case where the ID allocation device 1 transmits and receives information to and from other devices, and accesses the server 100 that manages the personal website via the network (Internet). The page data is acquired and provided to the control unit 10.

  The input unit 40 is an interface device that receives an instruction input from the user to the ID assignment device 1. The input unit 40 is configured by, for example, a keyboard, a mouse, a touch panel, and the like.

  The output unit 50 includes a display device that displays a screen for accepting data input to the user and displays a screen of a processing result by the ID assignment device 1. Further, the output unit 50 may include various output devices such as a printer in addition to a display device such as a cathode ray tube display device (CRT) or a liquid crystal display device (LCD).

Next, the function of the control unit 10 will be described in detail.
First, symbols used in the present embodiment will be described.
-Personal website network (directed graph): G = (V, E)
Personal website (node): v _i ∈V, (i = 1, 2,..., | V |)
Number of personal websites (graph size): | V |
• Hyperlinks from v _i to v _j : (v _i , v _j ) εE, (v _i , v _j εV)
Number of hyperlinks: | E |
- attribute · URL of personal web site: URL _(v i)
Type: TYPE _(v i)
Class: C _(v i)
Service Provider: SP _(v i)
Account ID: ACID _(v i)
-Administrator ID (OnID): ONID (v _i )
- attributes hyperlink hyperlink _{strength: | (v i, v j} ) |
Specifically, representing the number of hyperlinks from v _i to v _j.
・ Hyperlink strength between times t0 and t1: | (v _i , v _j ) | _{t0 → t1}
This represents the number of hyperlinks within a specified period (from time t0 to t1) among hyperlinks having time information, such as hyperlinks generated when posting comments on blogs, reals, and gebs. However, hyperlinks that do not have time information, such as hyperlinks generated by Prof, Hompe, and Mylink, are considered to have occurred during the entire period.

  The control unit 10 includes a site collection unit 11 (collection unit), a metric calculation unit 12 (calculation unit), an account cluster generation unit 13a (third generation unit), and a type cluster generation unit 13b (second generation unit). ), A metric cluster generation unit 13c (first generation unit), an ID allocation unit 14 (allocation unit), and an application unit 15. Each unit is a functional block realized by executing an ID assignment program.

  The site collection unit 11 extracts a hyperlink included in the personal website, and further acquires page data of another personal website that is a reference destination of the hyperlink. And the site collection part 11 memorize | stores page data in site preservation | save DB21, and the adjacent relationship and input / output relationship by the hyperlink between several personal website lists (collection history management table), and several personal websites Is added to the list of link information (inter-site relation table) indicating the site information, and the site management DB 22 is updated.

  For example, the site collection unit 11 firstly, the URL of the personal website of the route that is the source of the collection process, and how many hyperlinks (<a> link in HTML) are traced from the personal website of this route (number of link hops). ) Etc. are accepted.

  Next, the site collection unit 11 accesses the Internet and stores the page data of the root personal website in the site storage DB 21. Further, the URL of the personal website of this route is added to the collection history management table.

  Further, the site collection unit 11 acquires a hyperlink from the personal website, and adds a combination of the link source URL and the link destination URL to the inter-site relation table except for a link destination that is not a personal website. .

  Further, the site collection unit 11 stores the page data of the personal website acquired from the link destination URL in the site storage DB 21 and this URL in the collection history management table. Then, the site collection unit 11 ends the collection process when all URLs are stored in the collection history management table up to the designated collection range. Further, when the URL has not been collected to the designated collection range, the site collection unit 11 repeats the acquisition of the hyperlink, the update of the inter-site relation table, the storage of the page data, and the update of the collection history management table.

  FIG. 4 is a diagram showing a collection history management table stored in the site management DB 22 according to the present embodiment.

  The collection history management table stores a collection ID, route URL, link source URL, personal website URL, personal website class, storage destination, number of site hops, administrator ID (OnID), and collection date and time.

  Here, the collection ID is an identification number assigned to each collection process. The root URL is the URL of the personal website designated for performing the collection process. The storage destination is a URL indicating the storage location of the target personal website in the site storage DB 21. OnID is an ID for identifying an online individual assigned by an ID assigning unit 14 described later.

  The number of site hops is the number of hops from the root node when a designated personal website is a root node and a hyperlink between adjacent linked personal websites is 1 hop.

  FIG. 5 is a diagram showing an inter-site relation table stored in the site management DB 22 according to the present embodiment.

  The inter-site relation table includes the collection ID, the URL of each link source personal website and the link destination personal website, type, class, service provider identification data (SP), and service provider account ID (ACID). The occurrence date and time of the hyperlink and the collection date and time are stored.

  The SP and ACID are identified by the URL of the personal website. Here, if the ACID is the same in the same SP, the administrator (online individual) can be regarded as the same, but if the SP is different, the administrator is not necessarily the same even if the ACID is the same. Absent.

  Further, the occurrence date and time of the hyperlink is the date and time when the hyperlink was written by posting a comment on a personal website. In the case of hyperlinks that are not linked to time information, such as hyperlinks that exist since the creation of personal websites, the date and time of occurrence are not specified, but they are considered to have occurred for the entire period and are used in the processing described below. To do.

  The metric calculation unit 12 calculates a metric (first index) indicating the degree to which two personal websites have a common adjacent personal website based on the link information in the inter-site relation table.

Specifically, one of the following two types of similarity (Similarity) is used as a metric for v _i and v _j .
(1) Similarity = | Γ (v _i ) ∩Γ (v _j ) | / | V |
However,
Γ (v _i ) = {w | (v _i , w) εE or (w, v _i ) εE}
Is a set of adjacent personal websites (adjacent nodes) and does not depend on the direction of the hyperlink. Also,
| Γ (v _i ) ∩Γ (v _j ) |
Is the number of common adjacent nodes.

(2) Similarity = | Γ ⁺ (v _i ) ∩Γ ⁺ (v _j ) | / | V |
However,
Γ ⁺ (v _i ) = {w | (v _i , w) ∈ E or (w, v _i ) ∈ E} ∪ {v _i }
Is a set of adjacent nodes, like Γ (v _i ), but includes its own node (v _i ). Also,
| Γ ⁺ (v _i ) ∩Γ ⁺ (v _j ) |
Is the number of common adjacent nodes.

FIG. 6 is a diagram illustrating an example of calculating Simality between personal websites according to the present embodiment.
Seven nodes are adjacent by hyperlinks as shown. Specifically, in the calculation method (1), the adjacent nodes of the node v ₁ and the node v ₂ are
Γ (v ₁ ) = {v ₂ , v ₃ , v ₄ , v ₅ , v ₆ }
Γ (v ₂ ) = {v ₁ , v ₃ , v ₄ , v ₅ }
It is. At this time, the common adjacent node is
Γ (v ₁ ) ∩Γ (v ₂ ) = {v ₃ , v ₄ , v ₅ }
It is. Therefore,
Similarity = 3/7
It becomes.

In the case of the calculation method (2), adjacent nodes of the node v ₁ and the node v ₂ are
Γ ⁺ (v ₁ ) = {v ₁ , v ₂ , v ₃ , v ₄ , v ₅ , v ₆ }
Γ ⁺ (v ₂ ) = {v ₁ , v ₂ , v ₃ , v ₄ , v ₅ }
It is. At this time, the common adjacent node is
Γ ⁺ (v ₁ ) ∩Γ ⁺ (v ₂ ) = {v ₁ , v ₂ , v ₃ , v ₄ , v ₅ }
It is. Therefore,
Similarity = 5/7
It becomes.

Both calculation methods are based on the fact that personal websites of the same manager tend to have many common adjacent nodes.
Hereinafter, description will be made using the calculation method (2) including the own node.

In addition, the metric calculation unit 12 calculates Similarity except for the personal website in which the incoming order and outgoing order of the hyperlink satisfy a predetermined condition and link information related to the personal website.
Here, the incoming order or outgoing order may be limited to the number of hyperlinks when the adjacent personal website is of a predetermined class or type.

In addition, the metric calculation unit 12 calculates Similarity by excluding hyperlinks that have occurred outside the predetermined period (for example, the past one year) from the link information.
Further, when the strength of the hyperlink generated between the same personal websites in the predetermined period is less than the predetermined value in the link information, the metric calculation unit 12 calculates Simularity except for the hyperlink. . The predetermined value is set for each class or type of the hyperlink link source.

  The account cluster generation unit 13a generates a cluster by combining the personal websites having the same ACID based on the account ID (ACID) of the service provider that can be identified by the URL of each of the plurality of personal websites.

  The type cluster generation unit 13b is configured to combine a predetermined class of personal websites having a predetermined input / output relationship based on a class of each of a plurality of personal websites and a hyperlink input / output relationship structure. To create a cluster.

For example, if there is a hyperlink from class A to class B based on the following hypotheses (A) and (B) based on empirical rules, the type cluster generation unit 13b links the link source and link destination of this hyperlink. Create a cluster by combining your personal websites.
(A) The administrator does not create a hyperlink from a class A personal website to another person's personal website.
(B) The administrator does not create a hyperlink from his / her own personal website to another person's class B personal website.

  The metric cluster generation unit 13c generates a set of clusters including one or more personal websites by including, in the same cluster, a combination of personal websites having a similarity greater than a predetermined value based on the metric.

  Specifically, the metric cluster generation unit 13c determines the similarity (first index) relating to the combination of the personal website belonging to the cluster and the personal website outside the cluster according to a predetermined rule. Integrate with Similarity (Second Indicator) on the combination of personal websites. Then, the metric cluster generation unit 13c selects a personal website and a personal website, a personal website and a cluster, or a cluster obtained by combining clusters and clusters in descending order of the first index or the second index. The generation process is repeated until the first index and the second index are less than a predetermined value.

  In the present embodiment, the second index is the maximum value of the first index related to the combination of the personal website belonging to the cluster and the personal website outside the cluster.

FIG. 7 is a diagram illustrating an example of calculating similarity between clusters according to the present embodiment.
Similarity (second index) regarding the combination of the cluster CL (v ₁ , v ₂ ) composed of the nodes v ₁ and v ₂ and the cluster CL (v ₃ , v ₄ ) composed of the nodes v ₃ and v ₄ is It is obtained from the similarity (first index) between the nodes.

Here, the similarity between the nodes is
| Γ ⁺ (v ₁ ) ∩Γ ⁺ (v ₃ ) | / | V | = 4/7
| Γ ⁺ (v ₂ ) ∩Γ ⁺ (v ₃ ) | / | V | = 4/7
| Γ ⁺ (v ₁ ) ∩Γ ⁺ (v ₄ ) | / | V | = 5/7
| Γ ⁺ (v ₂ ) ∩Γ ⁺ (v ₄ ) | / | V | = 5/7
It is. Therefore, using these maximum values, the similarity between clusters is
Similarity = 5/7
It becomes.

  The ID assigning unit 14 assigns the same administrator ID (OnID) that is different for each cluster to the plurality of personal websites and updates the collection history management table.

  The application unit 15 receives an instruction input from the user via the input unit 40, and outputs the information stored in the site management DB 22 to the output unit 50 and provides it to the user. Specifically, the application unit 15 visualizes and outputs hyperlinks between personal websites or hyperlinks between clusters and further the strength based on the collection history management table and the site relation table. As a result, an online personal network as an administrator is visualized.

Next, the procedure of the process for assigning OnID will be described in detail.
FIG. 8 is a flowchart showing processing in the control unit 10 according to the present embodiment.
It is assumed that the collection of personal websites within the range to be processed has been completed and stored in the site management DB 22. However, OnID in the collection history management table is blank.

  In step S <b> 1, the control unit 10 acquires G (V, E) data from the site management DB 22.

  In step S2, the control unit 10 (metric calculation unit 12) determines that the input order is AD_Deg_ {in} or more and the output order is AD_Deg_ {out} or less from G (V, E) acquired in step S1. Remove nodes and tangent links.

  Here, in order to remove a site such as an advertisement site where many hyperlinks are linked to an unspecified partner, but hardly generate a reverse hyperlink, for example, “AD_Deg_ {in} = Values such as “1000” and “AD_Deg_ {out} = 5” are used. At this time, the link source may be limited to class B or class C.

  In addition, since a site that publishes poetry, lyrics, etc. receives many references from an unspecified partner's class A personal website, for example, “AD_Deg_ {in} = 500”, “AD_Deg_ {out} = A value such as “10” is used. At this time, the link source may be limited to class A.

  In step S <b> 3, the control unit 10 (metric calculation unit 12) removes a node whose outgoing order is equal to or greater than CELEB_Deg_ {out} and a link that contacts the node from G (V, E) acquired in step S <b> 1. In addition, a condition that the incoming order is equal to or less than CELEB_Deg_ {in} may be added.

  Here, since the personal website of a celebrity receives a lot of comments from an unspecified partner, especially on a personal website of class B or class C, a large number of hyperlinks that are linked to themselves are generated. Therefore, for example, values such as “CELEB_Deg_ {out} = 10000” and “CELEB_Deg_ {in} = 100” are used. At this time, the link source may be limited to class B or class C.

  In step S4, the control unit 10 (metric calculation unit 12) determines that the strength of the hyperlink in the specified period (for example, the past one year) from the G (V, E) acquired in step S1 is not more than Strength. Remove the link. This Strength is preset for each link source class.

  In step S <b> 5, the control unit 10 (metric calculation unit 12) calculates Similarity for all node pairs in G (V, E).

  In step S6, when there is a node having a common account ID (ACID) identified from the URL of the personal website, the control unit 10 (account cluster generation unit 13a) generates a new cluster including these nodes. At this time, if the node already belongs to the cluster, the control unit 10 (account cluster generation unit 13a) generates a new cluster including all the nodes belonging to this cluster.

  In step S7, when there is a hyperlink from class A to class B, the control unit 10 (type cluster generation unit 13b) generates a new cluster including nodes at both ends of the hyperlink. At this time, if the node already belongs to the cluster, the control unit 10 (type cluster generation unit 13b) generates a new cluster including all the nodes belonging to this cluster.

  In step S8, the control unit 10 (metric cluster generation unit 13c) combines a pair having the highest similarity among all node and cluster pairs to generate a new cluster.

  In step S <b> 9, the control unit 10 (metric cluster generation unit 13 c) determines whether the similarity of all node and cluster pairs is less than the threshold Th. If this determination is YES, the process proceeds to step S10, and if the determination is NO, the process returns to step S8 to repeat cluster generation.

  In step S10, the control unit 10 (ID allocation unit 14) allocates a unique OnID to each node and cluster.

  Note that Th, AD_Deg_ {in}, AD_Deg_ {out}, CELEB_Deg_ {out}, CELEB_Deg_ {in}, period, and Strength are received from the user via the input unit 40 in advance.

9 to 13 are diagrams showing the procedure of OnID allocation processing according to the present embodiment.
Assume that seven personal websites (nodes) are linked by hyperlinks as shown in FIG.
V = {v ₁ , v ₂ , v ₃ , v ₄ , v ₅ , v ₆ , v ₇ }
_{However, v 1,} v _7: Class _{A, v 2,} v _6: Class _{B, v 3, v 4,} v 5: Class C
E = {(v ₁ , v ₂ ), (v ₂ , v ₃ ), (v ₂ , v ₄ ), (v ₂ , v ₅ ), (v ₃ , v ₁ ), (v ₃ , v ₄ ) _{, (v 4, v 1)} , (v 4, v 5), (v 5, v 1), (v 6, v 1), (v 6, v 7), (v 7, v 6)}

At this time, 21 combinations of each node are calculated in the range of “0/7” to “5/7” as shown in the table.
Hereinafter, it is assumed that the threshold at the time of cluster generation is “Th = 5/7”.

First, as shown in FIG. 10, for a hyperlink (v ₁ , v ₂ ) from class A to class B, a cluster CL (v ₁ , v ₂ ) that combines nodes at both ends is generated.
At this time, in the Similarity table, the nodes v ₁ and v ₂ are deleted, and these are coupled to the cluster CL (v ₁ , v ₂ ).

Further, the maximum value of the similarity regarding v ₁ and v ₂ is adopted as the similarity regarding CL (v ₁ , v ₂ ). For example, for a pair of CL (v ₁ , v ₂ ) and v ₃ , “4/7”, and for a pair of CL (v ₁ , v ₂ ) and v ₆ , “2 / The maximum value of “7” and “1/7” is “2/7”.

Furthermore, as shown in FIG. 11, a cluster CL (v ₆ , v ₇ ) in which nodes at both ends are combined is generated for the hyperlink (v ₇ , v ₆ ) from class A to class B.
At this time, in the Similarity table, nodes v ₆ and v ₇ are deleted, and these are coupled to the cluster CL (v ₆ , v ₇ ).

Next, in FIG. 11, the combination of CL (v ₁ , v ₂ ) and v ₄ , which is the pair with the highest similarity, is combined, and a new cluster CL (v ₁ , v ₂ , v, as shown in FIG. 12). ₄ ) is generated.

As a result, in FIG. 12, all of the Similarity falls below the threshold Th = 5, and thus the generation of the cluster ends.
As shown in FIG. 13, four OnIDs are assigned to each node and cluster.

As described above, according to the present embodiment, the ID assigning apparatus 1 uses a plurality of individuals based on Similarity, which is an index (metric) indicating the degree that two personal websites have a common adjacent personal website. Create a set of clusters from a website. The ID assigning apparatus 1 can classify a plurality of personal websites for each online individual who is an administrator by assigning different administrator IDs to these clusters. Therefore, the ID assigning apparatus 1 can estimate the personal website of the same administrator more accurately by using a universal metric.
As a result, since the ID allocation device 1 can easily acquire information on a personal website managed by the same online individual, teachers, guardians, etc. create this ID allocation device 1 by children (especially junior and senior high school students). Can be used to monitor personal websites.

  Further, the ID assignment device 1 includes itself as an adjacent personal website. Since personal websites of the same administrator are often adjacent to each other, the ID assigning apparatus 1 can be assigned to the personal website of the same administrator by weighting the pairs of adjacent personal websites. Can be estimated more accurately.

  In addition, since the ID allocation device 1 includes a combination of personal websites whose first index is greater than or equal to a predetermined value in the same cluster, it is assumed that the administrators of personal websites having adjacency relationships similar to or greater than the predetermined value are the same. Can easily generate clusters.

Further, the ID assigning apparatus 1 integrates the first index into the second index related to the combination of the cluster and the cluster or the cluster and the personal website, and combines these combinations based on the second index. A cluster can be created. Therefore, the ID assigning apparatus 1 can assign an OnID by sequentially generating and expanding the same cluster based on the first index.
Furthermore, when the cluster is generated, the ID allocation device 1 can easily obtain the second index related to the combination with the cluster as the maximum value of the index related to the combination with the personal website belonging to the cluster.

  Further, the ID assigning apparatus 1 generates a cluster by combining combinations of a predetermined class of personal websites having a predetermined input / output relationship set in advance. Therefore, the ID assignment device 1 can assign an OnID more accurately in consideration of a predetermined condition based on the case.

  In addition, since the ID assignment apparatus 1 generates a cluster by combining personal websites with the same service provider account, it can assign an OnID more accurately.

Further, the ID assigning apparatus 1 calculates Similarity by excluding specific types of personal websites and adjacent hyperlinks according to predetermined conditions of the incoming and outgoing orders of hyperlinks. Therefore, the ID assigning apparatus 1 can remove sites that are not subject to OnID assignment as noise, and therefore can assign OnID more accurately.
Furthermore, the ID allocation device 1 can more reliably remove sites that are not subject to OnID allocation as noise based on the site class.

  In addition, since the ID allocation device 1 calculates Similarity excluding hyperlinks generated outside a predetermined period, the accuracy can be improved only for recent information or a specific period, and the processing load can be reduced. .

Further, when the strength of hyperlinks generated between the same personal websites during a predetermined period is less than the predetermined value Strength, the ID allocation device 1 calculates Simularity excluding these hyperlinks. Therefore, the ID allocation device 1 removes noise for hyperlinks linked with a strength of a predetermined level or more, so that it is possible to determine the identity of the manager more reliably.
Furthermore, since the ID allocation device 1 sets Strength, which is a threshold of hyperlink strength, for each type of link source, noise can be more reliably removed in response to the occurrence tendency of hyperlinks that differ depending on the usage mode. .

  As mentioned above, although embodiment of this invention was described, this invention is not restricted to embodiment mentioned above. Further, the effects described in the present embodiment are merely a list of the most preferable effects resulting from the present invention, and the effects of the present invention are not limited to those described in the present embodiment.

In the above-described embodiment, one of the two types of similarity is used as the metric regarding the nodes v _i and v _j , but the metric is not limited to these. For example, using the Jaccard coefficient,
(3) Similarity = | Γ (v _i ) ∩Γ (v _j ) | / | Γ (v _i ) ∪Γ (v _j ) |
Or
(4) Similarity = | Γ ⁺ (v _i ) ∩Γ ⁺ (v _j ) | / | Γ ⁺ (v _i ) ∪Γ ⁺ (v _j ) |
It is good. Thus, by setting the denominator to the total number of nodes adjacent to v _i or v _j , the ID allocation device 1 can set a threshold value that does not depend on the number of collected nodes | V |.

1 ID allocation device 10 Control unit 11 Site collection unit (collection unit)
12 Metric calculation part (calculation part)
13a Account cluster generation unit (third generation unit)
13b Type cluster generation unit (second generation unit)
13c Metric cluster generation unit (first generation unit)
14 ID allocation unit (allocation unit)
15 Application section 20 Storage section 21 Site storage DB
22 Site management DB
30 communication unit 40 input unit 50 output unit

Claims (14)

A storage unit for storing link information indicating a plurality of personal websites and an adjacency relationship by hyperlink between the plurality of personal websites;
A calculation unit that calculates a first index indicating the degree to which two personal websites have a common adjacent personal website based on the link information;
A first generation unit configured to generate a set of clusters including one or more personal websites based on the first index;
An ID assigning device comprising: an assigning unit that assigns the same administrator ID to each of the plurality of personal websites, which is different for each cluster and within the cluster.   The ID allocation device according to claim 1, wherein the calculation unit calculates the first index including itself as the adjacent personal website.   The ID allocation device according to claim 1, wherein the first generation unit includes a combination of personal websites in which the first index is greater than a predetermined value in the same cluster.   The first generation unit uses the first index relating to a combination of a personal website belonging to the cluster and a personal website outside the cluster as a combination of a cluster and a cluster, or a combination of a cluster and a personal website according to a predetermined rule. And a personal website and a personal website, a personal website and a cluster, or a cluster that combines clusters and clusters is generated based on the first index or the second index. The ID allocation device according to claim 1 or 2, wherein the process is repeated until the first index and the second index are less than a predetermined value.   5. The ID according to claim 4, wherein the first generation unit uses the maximum value of the first index relating to a combination of a personal website belonging to the cluster and a personal website outside the cluster as the second index. Allocation device.   Based on a predetermined number of types representing usage forms classified based on URLs of each of the plurality of personal websites and a structure of the input / output relationship of the hyperlink, a predetermined input / output relationship is established. The ID allocation device according to claim 1, further comprising a second generation unit that generates a cluster by combining a combination of a certain type of personal website.   The third generation unit according to claim 1, further comprising: a third generation unit configured to generate the cluster by combining the same personal websites based on service provider accounts identifiable by URLs of the plurality of personal websites. Item 7. The ID assignment device according to any one of Items 6 to 7.   2. The calculation unit according to claim 1, wherein the calculation unit calculates the first index by excluding a personal website satisfying a predetermined condition of an incoming order and an outgoing order of the hyperlink and the link information related to the personal website. Item 8. The ID assignment device according to any one of Items 7 to 8.   9. The number of hyperlinks when the incoming order or the outgoing order is a predetermined type representing a usage form in which adjacent personal websites are classified based on URLs of the personal websites. ID assignment device.   The ID allocation device according to any one of claims 1 to 9, wherein the calculation unit calculates the first index by excluding a hyperlink that has occurred outside a predetermined period in the link information.   When the number of hyperlinks generated between the same personal websites in a predetermined period of the link information is less than a predetermined number, the calculation unit calculates the first index except for the hyperlinks The ID assignment device according to any one of claims 1 to 10.   The ID assignment device according to claim 11, wherein the predetermined number is set for each type representing a usage form classified based on a URL of a personal website that is a link source of the hyperlink. An ID assignment method in which a computer assigns an administrator ID to each of a plurality of personal websites,
The computer stores link information indicating adjacency relationships by hyperlinks between the plurality of personal websites and the plurality of personal websites;
A calculation step of calculating a first index indicating the degree to which two personal websites have a common adjacent personal website based on the link information;
Generating a cluster set of one or more personal websites based on the first indicator;
A method of performing, for each of the plurality of personal websites, an assigning step of assigning the same administrator ID that is different for each cluster and within the cluster. An ID assignment program for causing a computer to assign an administrator ID to each of a plurality of personal websites,
The computer stores link information indicating adjacency relationships by hyperlinks between the plurality of personal websites and the plurality of personal websites,
A calculation step of calculating a first index indicating the degree to which two personal websites have a common adjacent personal website based on the link information;
Generating a cluster set of one or more personal websites based on the first indicator;
An ID allocation program for executing, for the plurality of personal websites, an allocation step that is different for each cluster and allocates the same administrator ID in the cluster.

Images