JP2000172699A - Device and method for supporting hypertext structure change and storage medium with hypertext structure change support program recorded therein - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a hypertext structure change supporting device for comprehending the trend of a user from the access tendency of the user and the similarity degree between the distribution tendencies of contents and supporting the change of contents on a web site, corresponding to the trend of the user. SOLUTION: A log clustering part 2 acquires a link structure of nodes and a log as the access history of the user from a web server 1 and forms a log cluster by clustering the nodes based on the log. Moreover, a content clustering part 3 acquires contents from the web server 1 and forms a content cluster by clustering based on the similarity of contents. An out-of-intention detecting part 4 makes the log cluster correspond to the contents cluster and detects the contents cluster, which cannot be made to correspond to cluster as one with possibility for being out-of-intention. Then, an out-of-intention presenting part 5 presents them to the manager or the like of the web server 1.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a content (hypertext information) published on a Web site for providing information on the Internet.
TECHNICAL FIELD The present invention relates to an apparatus and a method for providing support for changing a hypertext structure based on the similarity between a user's access tendency and a content distribution tendency with respect to, and a storage medium on which a support program for implementing the support apparatus by a computer is recorded.

[0002]

2. Description of the Related Art In recent years, the World Wide Web
(Hereinafter abbreviated as Web), there is an increasing demand for providing information. Companies also have a Web site for providing information on the Web, and advertising activities for their products or buying and selling of goods and services by means such as electronic commerce are being actively performed on the Web site. Under such circumstances, there is an increasing demand for technology for knowing the trends of users who have accessed the Web and quickly changing the structure of information to be provided according to the user trends.

[0003] Information provided on a Web site is called content, and is referred to as hypertext (for example, HTML).
Is represented by Each piece of hypertext information (hereinafter referred to as a node) is connected by a link. A user who wants to access a Web site can access a node in the Web site by following a link to obtain various information. Further, the Web server can record the history of the user accessing the Web server. The recorded content is hereinafter referred to as a log. For example, in the log, the IP address of the computer accessing the information (the computer used by the user to access the Web server or the computer relaying the user's usage information), the time of access,
The command sent to the Web server to access the information is recorded. The instruction includes an identifier (for example, URL) on the server of the node accessed by the user.

[0004] As a conventional technique for knowing a user's trend using an access log, for example, Ming-Syan et al., "Data Mining for Path T".
Raversal Patterns on the
Web ", IEEE Trans. On Knowl
edge and Data Engineerin
g, 1998 and the like. The technique described in this document is based on MF (Maximal F).
This is a technique for extracting a series called “old reference”. FIG. 15 is an explanatory diagram of an example of a Web site structure and an access route of a user. In the example shown in FIG. 15, it is assumed that the nodes A to H are linked as shown. At this time, it is assumed that a certain user browses the nodes in the order of A → B → C → D → E → F → G as shown by the thick line in the figure. In this case, [ABC
D], [ABE] and [AFG] are obtained.
MF is calculated for all users who have access, and M
The access pattern of the user can be obtained by counting the number of appearances of the subsequence of F.

[0005] According to this technique, a linear series of nodes that a user has followed a link can be obtained. However, a series obtained by tracing the user starting from a node returning from a certain node is obtained as a different series. Therefore, it is difficult to obtain a group of nodes accessed by the user as a lump.

Another prior art is disclosed in, for example, Bor
ges et al., “Mining Association
Rules in Hypertext Datab
as ", In Proc. of KDD-98. This technique extracts the number of transitions of a user between nodes connected by a link from a log, and calculates a correlation rule of a node sequence accessed by the user. In this technique, since only the number of transitions of the user between the nodes is observed, the same user does not necessarily follow the sequence. It is explanatory drawing of an example of the transition number of a user.
The direction in which the link was followed is indicated by an arrow, and the number of times followed is indicated by a numerical value. According to this conventional technique, in the example shown in FIG. 16, for example, a correlation is obtained in which a user who follows a link from B to A tends to follow a link from A to F. However, this does not indicate that the same user has visited all of the nodes B, A, and F, and cannot be said to accurately reflect the access trend of the user.

In addition, in the above two prior arts, a frequently accessed route can be known by obtaining a user's access pattern from a log. However, since the content is not integrated and processed in an integrated manner, the access pattern cannot be determined. It is difficult to provide support for maintaining and managing a Web site, such as which content should be changed based on the content.

[0008]

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned circumstances, and shows the tendency of access by users,
A hypertext structure change support apparatus and method that can grasp the user's trend by comparing the similarity between the distribution trends of the content and support the change of the content of the Web site according to the user's trend It is another object of the present invention to provide a storage medium storing a support program for realizing the support device by a computer.

[0009]

According to the present invention, a log cluster is generated by clustering hypertext information groups based on access history information for the hypertext information groups managed by the information system. The information group is clustered based on the similarity between individual hypertexts to generate a content cluster, the content cluster is associated with the log cluster, and the hypertext that is an element in the content cluster that could not be associated is identified. It is presented. When the content cluster cannot be associated with the log cluster, for example, the frequency of reference from the user to hypertexts having similar contents is clearly different, and the website structure is inappropriate, or
There is a possibility that the link is not properly provided.
By presenting such hypertext,
The change of the structure of the hypertext can be promoted, and the structure change can be supported.

[0010]

FIG. 1 is a block diagram showing an embodiment of a hypertext structure change support device according to the present invention. In the figure, 1 is a Web server, 2 is a log clustering unit, 3 is a content clustering unit, 4 is an intention shift detection unit, 5 is an intention shift presentation unit, 11 is a log recording unit, 12
Is a content providing unit.

The Web server 1 stores information (contents) to be provided to a user, and transmits the information on a network. The Web server 1 has a log recording unit 11 and a content providing unit 12. The content providing unit 12 provides a content according to a user's access. In addition, the log recording unit 11 sets a user identifier (I) for identifying a user every time there is an access from the user.
P address), access time, and an address (URL) of the hypertext (hereinafter, referred to as a node) accessed by the user. The IP address recorded in the Web server 1 may be an address of a client computer actually used by a user or an address of a proxy server acting for access from the client computer. In the latter case, a plurality of users use the same IP address. In the following description, W
It is assumed that the user is identified based on the IP address and the date of the computer accessing the web server 1. That is, if the access dates are different even from accesses from the same IP address, they are identified as different users.

The log used in the present invention is not limited to the one described above. For example, as described in Japanese Patent Application Laid-Open No. H10-224349, a log that identifies a user by using a log of a proxy server together is also used. Good. Also,
For example, even in a log recorded by a mechanism having an access notification mechanism on the client side such as a Java applet described in Japanese Patent Application Laid-Open No. 10-207838, the user can be identified, and the location of accessed information and the order of access can be known. Anything should do.

The log clustering unit 2 acquires a link structure formed by links with nodes from the Web server 1 and logs, and forms a log cluster by clustering the nodes based on the logs. This log clustering unit 2
Will be described later.

The content clustering unit 3 is a Web server
Content is acquired from the server 1 and clustered based on, for example, the similarity of the content to form a content cluster. This content cluster is formed without referring to the link structure.

The intention shift detecting unit 4 associates the content cluster obtained by the content clustering unit 3 with the log cluster obtained by the log clustering unit 2. At this time, the content cluster that could not be correlated is detected as possibly having an unintended intention.

The intention shift presenting unit 5 sends the hypertext, which is an element in the content cluster detected as having a possibility that the intention is shifted due to the inability to associate with the intention shift detecting unit 4, to the Web server 1. Present it to the administrator.

FIG. 2 is a block diagram showing an example of a log clustering unit in an embodiment of the hypertext structure change support device of the present invention. In the figure, 21 is an access information storage unit, 22 is an access number counting unit, 23 is a cluster configuration unit, 24 is a cluster generation acceleration unit, and 25 is a cluster generation restriction unit.

The access information storage unit 21 stores, from the log information, a set of an identifier and information in which nodes accessed by a user having the identifier are recorded in chronological order. The identifier here is an identifier identified and given as a different user when the access date is different even if the access is from the same IP address as described above.

The number-of-accesses counting section 22 counts the number of identifiers having a history of accessing the given hypertext set by satisfying the given order constraint by using the information stored in the access information storage section 21. calculate.

The cluster composing unit 23 has a cluster composed of only a predetermined starting node as a vertex, and can be reached by tracing one link from the starting node and one by one from a set of nodes from which a log has been acquired. A cluster set including a set of the extracted node and the originating node is generated as a new cluster candidate in the next lower hierarchy. When the generated cluster candidate is calculated by the access number counting unit 22 to have an access number equal to or greater than a threshold value given in advance, a new cluster is generated as a child cluster of the origin node as a new cluster. Regarding the generated child cluster,
For each of the nodes, the nodes included in the set of nodes reachable by following a link from the set of nodes included in the child cluster and whose logs have been acquired and not included in the child cluster are extracted one by one, and the child By adding to the node set included in the cluster, a cluster candidate of the next lower hierarchy is generated. Further, when it is calculated that the cluster candidate has the number of accesses equal to or greater than the predetermined threshold, the process of generating a new child cluster with the child cluster as a parent cluster is performed until a new child cluster is not generated. repeat. In this way, a hierarchical cluster is generated.

When generating a hierarchical cluster in the cluster composing section 23, the cluster generation accelerating section 24 selects an upper-level cluster having a subset of a node set included in a newly generated cluster candidate. When the number of accesses from a user to a hierarchical cluster is equal to or less than a predetermined threshold, the cluster candidate is not generated. As a result, generation of clusters after a cluster candidate not to be generated can be stopped to shorten the processing time.

When generating a new cluster candidate for a predetermined node, the cluster generation restriction unit 25 generates a candidate for a set of nodes that can be reached by following a link from only that node. Do not use it. Thus, the number of child cluster candidates to be generated can be limited.

Next, an example of the operation of the hypertext structure change support apparatus according to the embodiment of the present invention will be described. FIG. 3 is a flowchart illustrating an example of the operation of the log clustering unit 2. Also, FIG.
FIG. 3 is an explanatory diagram of a specific example of a structure of a content stored in an eb server 1. In the following description, it is assumed that the content as shown in FIG. 4 is associated with the link as shown, and the description will be made with reference to FIG. Note that the main processing shown in FIG. 3 is performed in the cluster configuration unit 23.

First, in step S31, a threshold value of the number of accesses relating to the number of users who have visited all nodes included in the cluster is read. In S32, a non-inherited node set describing nodes that do not follow a link when generating a cluster is read. In this example, the threshold for the number of accesses is 2
0, a non-inherited node set is given in advance as [site entrance]. Note that the set is indicated by being surrounded by '[' and ']'.

In step S33, a link table from which link relations of nodes on the Web site have been extracted in advance is read from the Web server 1. When the link table is generated, a link to a node of another site from which a log cannot be acquired at the Web site is removed.

In step S34, a predetermined starting node is read. In this example, a top page "site entrance" which is a starting node in a Web site having the structure shown in FIG. 4 is used. From the top page, it can be accessed by following links to all nodes published on the website.

Next, in S35, a cluster consisting of only the originating node is generated as an initial cluster and added to the cluster set. In the example shown in FIG. 4, the cluster set is [the entrance of the site].

In S36, it is determined whether or not the cluster set is empty. If the cluster set is empty, it indicates that the processing has been completed, so the process proceeds to S39, where the cluster set generated so far is output as a log cluster, and the processing of the log clustering unit 2 ends. If the cluster set is not empty, the process proceeds to S37 to generate a cluster candidate. In the initial state, since the initial cluster has been generated, the process proceeds to cluster candidate generation processing in S37.

In S37, a cluster is extracted from the cluster set, and a cluster candidate is generated by a combination of the cluster and a link destination node from a node included in the cluster. In the example shown in FIG. 4, the node “site entrance” is linked to the nodes “printer” and “copy”, so that as a new cluster candidate, the node set includes a child cluster consisting of [site entrance, printer]. , [Site entrance, copy] are generated as cluster candidates.

Next, in S38, it is checked whether or not the generated cluster candidate satisfies a criterion that can be a cluster.
Select those that meet the cluster criteria. FIG. 5 is a flowchart illustrating an example of a process of determining whether a cluster candidate satisfies a cluster criterion. In S41, it is determined whether there is a cluster having the same node set, and if there is a cluster having the same node set, it is excluded from the candidates.

Next, in S42, it is determined whether or not a cluster having a subset of cluster candidates has been removed in the past as not satisfying the cluster criterion, and a cluster having a subset of cluster candidates has been removed in the past. in case of,
The cluster candidate is excluded from the candidates. This processing is performed in the cluster generation speed-up unit 24.

Next, in S43, the number of users (the number of user identifiers) accessing all the node sets included in the cluster and accessing the origin node first.
Is counted using the access number counting unit 22. If the number is less than the access number threshold read in S31, it is excluded from cluster candidates. As a method of counting the number of users, for example, a method of counting only those that match the order in which each node is accessed may be used. In this case, the node set included in the cluster also has order information, and is considered to be the same cluster only when the order information matches.

In S38, the cluster candidates that pass the determination criteria shown in FIG. 5 are added to the cluster set. Then, the process returns to S36.

Returning to S36, if the cluster set is not empty, the processes of S37 and S38 are repeated. At this time, if there are a plurality of clusters in the newly generated cluster set, the cluster with the smallest number of clusters is extracted, and cluster candidates are generated similarly. Hereinafter, when generating a cluster candidate, a node that can be accessed by following one link from a node included in the cluster is used. This is because the user may return to the node using the back function of the browser and access another page when accessing.

However, for the nodes which are indexed in the contents of the Web server, the link destination is restricted by the cluster generation restriction unit 25 so as not to be used for generating the child cluster candidate, and the number of combinations is prevented from increasing. be able to. In the example illustrated in FIG. 4, if the node “entrance of the site” is an index, the cluster [entrance of the site] is generated when a subsequent cluster is generated.
Is not expanded to the cluster [copy, printer]. For this index node, it is possible to prevent a cluster from being generated by an explicit input from the user. Here, since the association with the clustering based on the similarity of the contents is performed, if the similarity between the contents is low, the corresponding intention cluster is not assigned by the intention shift detecting unit 4, and the cluster including those nodes is generated. It does not have to be done. Therefore,
It is possible to calculate the similarity between nodes connected by a link from a certain node in advance, detect nodes whose average similarity is less than or equal to a threshold, and limit cluster generation using the detected nodes as index nodes. It is possible.

When generating a cluster candidate in S37, if a child cluster having the same node set has already been generated from another parent cluster, the already generated child cluster is replaced with the child cluster. It is possible to generate a relationship as a child cluster of the parent cluster to be generated.

FIG. 6 is an explanatory diagram of an example of a log cluster. By the processing in the log clustering unit 2 as described above, a log cluster as shown in FIG. 6 is output from the content structure shown in FIG. That is, first, cluster 1 [site entrance] is generated as an initial cluster, and then from node “printer” and “copy” that can follow one link from node [site entrance], cluster 2 [site entrance, printer] Then, the cluster 3 [site entrance, copy] is generated. Similarly, cluster 4 and cluster 5 are generated from cluster 2, and cluster 6 is generated from cluster 3. Cluster candidate 1 [site entrance, copy, model D], which has become a cluster candidate based on cluster 3, does not become a cluster because the number of accesses is less than 20.

From the cluster 4, clusters 8 and 9 are generated. At this time, the same cluster as cluster 8 is generated from cluster 5 as a child cluster candidate. In this case, as described above, a relationship as a child cluster of the cluster 5 is also generated for the already generated cluster 8. Similarly, from cluster 8 to cluster 1
Although 0 is generated, a relationship as a child class of the cluster 9 is also generated.

From the cluster 6, a cluster candidate 2 having a node set [entrance of site, copy, model C, model D] can be generated, but a cluster candidate 1 having a subset of this node set previously satisfies the criterion. Because there is no
It is removed by the condition of 42 and is excluded from cluster candidates.

The clusters generated in this way have an effective acyclic graph (DAG) structure as shown in FIG. 6, and each cluster has which cluster as a child cluster and which cluster as a parent cluster. A relationship has been created. Since the generated log cluster also has a tree structure having a parent-child relationship, it has a structure similar to that of FIG. 4 in the example shown in FIG. 6, but a log cluster is formed as a structure different from the content structure. Note that

Next, a process of generating a content cluster by the content clustering unit 3 will be described. The content clustering unit 3 generates a profile of each node with respect to each node using, for example, the method disclosed in Japanese Patent Application No. 9-153387, calculates the similarity for all combinations of two nodes, Remember. Then, clustering is performed by a known method based on the calculated similarity. As a method of clustering, for example, Ellen M. et al. Voorhees, "I
MPLEMENTING AGGLOMERATIVE
HIERARCHIC CLUSTERING AL
GORITHMS FOR USE INDOCUME
NT RETRIEVAL ”, Information
Processing & Management,
Vol. 22, no. 6, pp. 465-476,19
86 and the like.
Clustering Algorithm Comp
Clustering can be performed using the let link.

FIG. 7 is an explanatory diagram of an example of a content cluster. When the contents as shown in FIG. 4 are stored in the Web server 1, by hierarchically clustering these contents, the structure of the content cluster as shown in FIG. 7 is obtained. That is, since the contents of the node “model A” and the node “function of the model A” are similar, a cluster [model A, function A of the model A] is obtained. Further, since these nodes are similar to the node “model B”, a cluster [model A, function A of model A, model B] is obtained. On the other hand, since the contents of the node “model C” and the node “model D” are similar, a cluster [model C, model D] is obtained. Finally, since these nodes are similar, the cluster [model A, model A
A, model B, model C, and model D] are obtained. Thus, [model A, function A of model A], [model A,
Function A, Model B of Model A], [Model C, Model D], [Model A, Function A of Model A, Model B, Model C, Model D]
One cluster is obtained.

The contents clustering unit 3
Does not consider the links between the nodes, but performs clustering based on, for example, the similarity of the contents. For this reason, in the example shown in FIG. 4, content similar in content is connected by a link, but distant nodes in a link relationship may be combined into one cluster by content clustering. Alternatively, a node having no link for some reason can be included in the content cluster.

FIG. 8 is a flowchart showing an example of the operation of the intention shift detecting section 4 and the intention shift presenting section 5. First, in S51 and S52, the log cluster and the content cluster generated as described above are read. Here, it is assumed that the log cluster shown in FIG. 6 and the content cluster shown in FIG. 7 have been read.

Next, in S53, a leaf cluster of the log cluster (a cluster in which a new node is not added to the node set of the cluster and a cluster has not been generated)
Of content clusters included in. The leaf clusters of the log cluster shown in FIG. As content clusters included in these log clusters, the content clusters [model A, function A of model A], [model A, model A]
Function A, model B] is detected. Hereinafter, the content cluster [model A, function A of model A, model B] is referred to as a content cluster 1, and the content cluster [model A, function A of model A] is referred to as a content cluster 2. Note that the content cluster included in the cluster 6 is not detected.

In S54, the leaf cluster of the log cluster including the content cluster detected in S53 is set 1
Add to The content cluster 1 and the content cluster 2 are included in the leaf cluster 10 of the log cluster. Therefore, the cluster 10 of the log clusters is added to the set 1.

In S55, it is determined whether or not the set 1 is empty. If the set 1 is empty, the process ends. Set 1
If is not empty, the following processing is performed.

At S56, one of the log clusters is extracted from the set 1. Then, in S57, the content cluster is propagated from the extracted cluster to the parent cluster. FIG. 9 is an explanatory diagram of the correspondence between log clusters and content clusters. As described above, since the set 1 includes the cluster 10 among the log clusters, the cluster 10 is extracted from the set 1. Here, the content cluster included in the cluster 10 is shown in FIG. 9 as being the content cluster 1. Then, as shown by an arrow in FIG. 9, the content cluster 1 is transferred from the cluster 10 to the parent clusters 8 and 9.
To propagate.

At this time, in S58, it is determined whether or not a node not included in the parent cluster exists in the content cluster. If a node that is not included in the parent cluster exists in the content cluster, the process proceeds to the determination in S59. FIG. 10 is an explanatory diagram of an example of how a content cluster and a log cluster overlap. In this example, when the content cluster 1 is propagated from the cluster 10 to the parent cluster 9, the node “model B” existing in the content cluster 1 is not included in the cluster 9 as shown in FIG. Therefore, the process proceeds to S59. In the determination in S58, a shift between the log cluster and the content cluster can be detected. That is, it is possible to detect the presence of a portion where the similarity between the content access and the user's access to the node does not match.

In S59, it is determined whether or not the relationship between the content cluster and the parent cluster is in a predetermined format. FIG. 11 and FIG. 12 are explanatory diagrams of an example of a shift between a content cluster and a log cluster. As the predetermined format, for example, the formats shown in FIGS. 11 and 12 can be defined. The format shown in FIG. 11 is an example in which a part of the contents of the node 2 is described in detail in the node 3, and the site creator expects further access from the node 2 to the node 3. Link format. However, it indicates that the access to the node 3 is not actually performed. The format shown in FIG.
Nodes 2 and 3 explain the contents of node 1 in detail, and the site creator expects the same number of accesses to node 1 and node 2 or which user has access It is a link format that is important for knowing the user's selection point depending on whether it is done. The format shown in FIG. 12 indicates that the node 2 is being accessed, but the node 3 is hardly accessed. Here, since the example shown in FIG. 10 has the link format shown in FIG. 12, it is determined that the parent cluster and the content cluster are in a predetermined format.

If it is determined in S59 that the format is a predetermined format, it is determined that an intentional deviation has occurred, and in S61 an evaluation value is calculated and presented to a manager or the like. The evaluation value can be represented by, for example, the number of accesses classified into the parent class divided by the number of accesses classified into the child cluster. The larger the evaluation value, the higher the possibility that the number of accesses of the user can be increased by changing the content, or the clearer the selection of the user. As shown in FIG. 6 or FIG. 9, the number of accesses in which the access tendency is classified into the cluster 9 is 50, and the number of accesses in which the access tendency is classified into the class 10 is 30, so the evaluation value is 1.7. Becomes

FIG. 13 is an explanatory diagram of an example of the presented evaluation result. Using the evaluation value calculated as described above,
For example, it can be presented to an administrator or the like in a format as shown in FIG. Sites that need to be changed include nodes in the content cluster that have left the log cluster,
A node that is a link source to the node is highlighted to show the relationship between the structure of the node included in the log cluster and the node that is no longer included in the log cluster. At this time, each node presents the contents of the node in the same format as the Web browser, or presents only the title of the page when the display size is limited. In FIG. 13,
For convenience of illustration, the highlighting is shown with hatching. Further, the display method is arbitrary, and the display can be performed in an easy-to-understand manner without being limited to the form shown in FIG.

In the above description, the case where the content cluster is propagated to the cluster 9 as the parent cluster of the cluster 10 in S57 has been described. A case will be described in which a content cluster is propagated to cluster 8 which is another parent cluster as a parent cluster of cluster 10. In this case, since the node “function A of model A” included in the content cluster 1 is not included in the cluster 8, the process proceeds from S58 to S59. Since the relationship between the cluster 8 and the content cluster 1 corresponds to the link format shown in FIG. 11, an evaluation value is calculated in S61, and the evaluation result is presented. FIG. 14 is an explanatory diagram of another example of the presented evaluation result. In this case, the node “model A” and the node “function A of model A” are highlighted, and the link between them displays the evaluation value. Thereby, the point of change of the site can be indicated to the administrator or the like. Also in this case, the display form is arbitrary.

In the case where all the nodes of the content cluster are included in the parent cluster in S58, or in the case where the relationship between the parent cluster and the content cluster is not in a predetermined format in S59, calculation of the evaluation value and display of the evaluation result are performed. Proceed to S60 without performing. After the calculation of the evaluation value and the display of the evaluation result in S61, the process proceeds to S60.

In S60, it is determined whether or not the parent cluster still includes a content cluster.
At 2, the parent cluster is added to the set 1. Since the cluster 9 which is the parent cluster of the cluster 10 includes the content cluster 2 [function A of model A, model A], the cluster 9 is added to the set 1. Note that the cluster 8 which is the parent cluster of the other cluster 10 does not include the content cluster, and thus is not added to the set 1.

Then, the flow returns to S55. In this example, cluster 9 still exists in set 1. Therefore, the processing after S56 is repeated. The content cluster 2 is propagated to the cluster 4 that is the parent cluster of the cluster 9, but the cluster 4 does not include the node “function A of model A”. Since the relationship between the parent cluster and the content cluster in this case has the format shown in FIG. 11, the evaluation value is calculated in S61, and the evaluation result is displayed.

Since this cluster 4 does not include a content cluster, it is not added to the set 1, and the process returns to S55, where the set 1 becomes empty, and the processing is terminated.

In this way, it is possible to detect and display a shift between a node intended for similar access as a site that created a site from the contents of the content and a node actually accessed. For example, by referring to the display as shown in FIG. 13, the administrator can know that the model A is popular. Alternatively, it is possible to take measures to increase the access to the model B. Further, by referring to the display as shown in FIG. 14, for example, it is possible to take measures to have more access to the function A of the model A.

The above embodiment can be realized also by a computer program. In that case,
The program and the data used by the program can be stored in a computer-readable storage medium. A storage medium is a type of signal corresponding to a change state of energy such as magnetism, light, electricity, etc., caused to a reading device provided in a hardware resource of a computer in accordance with a description content of a program. Thus, the program description can be transmitted to the reading device. For example, magnetic disk, optical disk, C
A D-ROM, a memory built in a computer, and the like.

[0060]

As is apparent from the above description, according to the present invention, the difference between the access tendency of the user and the intention of the content creator of the Web site is detected, and the change of the content of the Web site is performed. Support can be provided, and there is an effect that the maintenance and management of the Web site can be promptly performed according to the trend of the user.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a hypertext structure change support device of the present invention.

FIG. 2 is a block diagram illustrating an example of a log clustering unit in the embodiment of the hypertext structure change support device of the present invention.

FIG. 3 is a flowchart illustrating an example of an operation in the log clustering unit 2.

FIG. 4 is an explanatory diagram of a specific example of the structure of content stored in the Web server 1.

FIG. 5 is a flowchart illustrating an example of a process of determining whether a cluster candidate satisfies a cluster criterion;

FIG. 6 is an explanatory diagram of an example of a log cluster.

FIG. 7 is an explanatory diagram of an example of a content cluster.

FIG. 8 is a flowchart illustrating an example of an operation in an intention shift detecting unit 4 and an intention shift presenting unit 5;

FIG. 9 is an explanatory diagram of the correspondence between log clusters and content clusters.

FIG. 10 is an explanatory diagram of an example of how a content cluster and a log cluster overlap.

FIG. 11 is an explanatory diagram of an example of a shift between a content cluster and a log cluster.

FIG. 12 is an explanatory diagram of another example of a shift between a content cluster and a log cluster.

FIG. 13 is an explanatory diagram of an example of a presented evaluation result.

FIG. 14 is an explanatory diagram of another example of the presented evaluation result.

FIG. 15 is an explanatory diagram of an example of a Web site structure and an access route of a user.

FIG. 16 is an explanatory diagram of an example of a Web site structure and the number of transitions of a user between nodes.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Web server, 2 ... log clustering part, 3 ... content clustering part, 4 ... intention shift detection part, 5 ...
Intention deviation presentation unit, 11 log recording unit, 12 content providing unit, 21 access information storage unit, 22 access count unit, 23 cluster configuration unit, 24 cluster generation acceleration unit, 25 cluster generation restriction Department.

Continuation of the front page F term (reference) 5B075 ND04 ND36 NK02 NK32 NK44 NR03 NR06 NR12 PR04 QM08 UU40 5B082 FA11 GA02 GA15

Claims (13)

[Claims] 1. A log clustering means for clustering hypertext information groups based on access history information on hypertext information groups managed by an information system, and converting the hypertext information groups into similarities between individual hypertexts. Content clustering means for performing clustering based on the content clustering means, and an intention shift detecting means for associating the content clustering result obtained by the content clustering means with the log clustering result obtained by the log clustering means; A hypertext structure change support device, characterized in that it has a mis-intention presentation means for presenting a hypertext which is an element in a content cluster that could not be attached. 2. The log clustering means according to claim 2, wherein said user identifier accessing said hypertext information group, the time, and the same user identifier accessed during a predetermined period from an access history in which the location of the accessed information are recorded. Access information storage means for giving a single identifier to a user having the identifier and storing information accessed by the user having the identifier in chronological order without duplication and storing the identifier in combination with the given hypertext set Access number counting means for calculating, using the access information storage means, the number of identifiers having a history of accesses satisfying the order constraint given to, and the number of identifiers calculated by the access number counting means. Cluster Construction Method for Generating Hierarchical Cluster from Hypertext Set Based on The cluster constructing means comprises a cluster composed of only a predetermined starting hypertext as a vertex, and a hyperlink that can be reached by following one link from the starting hypertext and whose access history is acquired. A cluster set consisting of a pair of a hypertext and a starting hypertext extracted one by one from the text set is generated as a new cluster candidate of the next lower hierarchy, and the cluster candidate is a threshold given in advance by the access number counting means. If it is calculated to have the above number of accesses, a new cluster is generated as a child cluster of the origin hypertext,
Each of the child clusters is included in a hypertext set that can be reached by following a link from the hypertext set included in the child cluster, an access history is acquired, and the child cluster is included in the child cluster. Unextracted hypertexts are extracted one by one, and are added to the hypertext set included in the child cluster to generate a cluster candidate of the next lower hierarchy. When it is calculated that the child cluster is generated, a hierarchical cluster obtained by repeating a process of generating a new child cluster with the child cluster as a parent cluster until no new child cluster is generated is performed. The hypertext structure change support device according to claim 1, wherein: 3. The log clustering means further comprises, in a higher-level cluster having a subset of a hypertext set included in a cluster candidate newly generated when forming a hierarchical cluster, A cluster generation speed-up unit that does not generate a cluster candidate when the number of accesses from a user to a cluster is equal to or less than a predetermined threshold value; A cluster generation restricting means for restricting the number of child cluster candidates to be generated by not using a hypertext set that can be reached by following a link from text alone when generating a candidate. Item 3. A hypertext structure change support device according to item 2. 4. When a candidate for a child cluster generated from a parent cluster has already been generated as a child cluster of another parent cluster, the cluster composing means sets a new parent cluster to generate a new child cluster. 3. The hypertext structure change support device according to claim 2, wherein a parent-child relationship is generated between the cluster and the already generated child cluster. 5. In the cluster having a hierarchical structure obtained by the log clustering means, the intention shift detection means determines whether or not a log cluster set having no child cluster has a difference between hypertexts obtained by the content clustering means. A content cluster included in each cluster of the log cluster set is selected from the content cluster set based on the similarity, and a hypertext in the content cluster is determined with respect to a parent log cluster of the log cluster associated with the selected content cluster. Is calculated to be included in the parent log cluster, there is a hypertext that is not included, and the hypertext has a predetermined link relationship with a hypertext set included in the upper log cluster, and To the top log cluster The hypertext is detected when the number of accesses to be performed and the change in the number of accesses to the corresponding log cluster are in a predetermined relationship. The related hypertext is presented to the user, and the intention shift detecting means and the intention shift presenting means repeat the above processing when there is a content cluster included in the parent log cluster, thereby detecting the intention shift. 2. The hypertext structure change support device according to claim 1, wherein the detection / presentation is repeated until the content cluster included in the parent log cluster no longer exists or a predetermined stop condition is satisfied. 6. The intention shift detecting means uses a cluster having a hierarchical structure based on the degree of similarity between hypertexts. The hypertext structure change support device according to claim 5, wherein a cluster of a higher hierarchy including the cluster of (a) is excluded. 7. The cluster generation restricting means calculates a similarity between hypertexts linked by a link from the same link source hypertext, and an average value of the similarities is equal to or less than a predetermined threshold. 4. The hypertext structure change support device according to claim 3, wherein the link source hypertext is detected, and the detected hypertext is used as the predetermined hypertext. 8. The hypertext information group is clustered into log clusters based on access history information for the hypertext information group managed by the information system, and the hypertext information group is classified based on the similarity between individual hypertexts. Clustering the content clusters into content clusters, associating the content clusters with the log clusters, and presenting hypertexts as elements in the content clusters that could not be associated with the log clusters. 9. The process of clustering into a log cluster is performed by the same user who has accessed for a predetermined period from an access history in which a user identifier, a time, and a location of the accessed information are recorded. One identifier is given to the user having the identifier, and the information accessed by the user having the identifier is stored as access information as a set of the information recorded in the chronological order without allowing duplication and the identifier. The number of identifiers having a history of accessing the text set satisfying the given order constraint is calculated as the number of accesses using the access information, and a cluster consisting of only a predetermined starting hypertext is defined as a vertex,
A cluster set consisting of a set of a hypertext and a starting hypertext that can be reached by following a link from the starting hypertext and that is extracted one by one from a hypertext set for which an access history has been acquired is one level lower. It is generated as a new cluster candidate of the hierarchy, and when it is calculated that the cluster candidate has the access number equal to or greater than a predetermined threshold, it is generated as a child cluster of the origin hypertext as a new cluster, Each of the child clusters is included in a hypertext set that can be reached by following a link from the hypertext set included in the child cluster, an access history is acquired, and the child cluster is included in the child cluster. Take out the missing hypertext one by one, Is added to the set of hypertexts included in the group, a cluster candidate of the next lower hierarchy is generated, and when it is calculated that the cluster candidate has the number of accesses equal to or more than the predetermined threshold, the child 9. The hypertext structure according to claim 8, wherein a hierarchical cluster is obtained by repeating a process of generating a new child cluster having a cluster as a parent cluster until no new child cluster is generated. Change support method. 10. The process of associating the content cluster with the log cluster and presenting a hypertext which is an element in the content cluster that could not be associated with the cluster having a hierarchical structure obtained by the log clustering process. In the log cluster set having no child clusters, a content cluster included in each cluster of the log cluster set is selected from the content cluster set obtained by the content clustering process based on the similarity between hypertexts Calculating, for the parent log cluster of the log cluster associated with the selected content cluster, whether or not the hypertext in the content cluster is included in the parent log cluster; And the hyper -The text has a predetermined link relationship with the hypertext set included in the upper log cluster, and the number of accesses to the upper log cluster and the change in the number of accesses to the corresponding log cluster have a predetermined relationship. Sometimes, the hypertext is detected as having an unintended intention, and the hypertext having a predetermined relationship with the hypertext is presented to the user, and there is a content cluster included in the parent log cluster. 9. The method according to claim 8, wherein, in the case, by repeating the above processing, the detection and presentation of the intention shift are repeated until the content cluster included in the parent log cluster disappears or a predetermined stop condition is satisfied. Hypertext structure change support method. 11. A log clustering process for clustering hypertext information groups based on access history information for the hypertext information groups managed by the information system, and converting the hypertext information groups into similarities between individual hypertexts. A content clustering process for performing clustering based on the content clustering process; an intention shift detection process for associating the content clustering result obtained by the content clustering process with the log clustering result obtained by the log clustering process; A storage medium storing a hypertext structure change support program for causing a computer to execute an unintended presentation process of presenting a hypertext which is an element in the hypertext. 12. The log clustering process has a user identifier that has accessed a hypertext information group, a time, and the same user identifier that has been accessed for a predetermined period from an access history in which the location of the accessed information is recorded. A single identifier is given to the user, and the information accessed by the user having the identifier is stored in the form of a set of the information recorded in a chronological order without allowing duplication and the identifier as access information, and all the given hypertext sets are stored. , The number of identifiers having a history of accesses satisfying the given order constraint is calculated as the number of accesses using the access information, and a cluster consisting only of a predetermined starting hypertext is set as a vertex, and the starting hyper Reachable and accessible by following a single link from the text A cluster set consisting of a pair of a hypertext and a starting hypertext extracted one by one from the hypertext set whose history is acquired is generated as a new cluster candidate of the next lower hierarchy, and the cluster candidate is given in advance. If it is calculated that the number of accesses is equal to or greater than a threshold value, a new cluster is generated as a child cluster of the originating hypertext, and each of the child clusters is linked from a hypertext set included in the child cluster. Is extracted from the hypertexts that are included in the set of hypertexts that can be reached by tracing and the access histories of which have been acquired, and that are not included in the child cluster. Cluster candidates in the next lower hierarchy Generating, and when it is calculated that the cluster candidate has the number of accesses equal to or greater than the predetermined threshold, the process of generating a new child cluster with the child cluster as a parent cluster is performed by a new child cluster. The storage medium according to claim 11, wherein a hierarchical cluster obtained by repeating the process until the cluster is no longer generated is generated. 13. The content intention clustering based on the similarity between hypertexts in a cluster having a hierarchical structure obtained by the log clustering process with respect to a log cluster set having no child clusters. A content cluster included in each cluster of the log cluster set is selected from the content cluster set obtained by the processing, and a hypertext in the content cluster is determined with respect to a parent log cluster of the log cluster associated with the selected content cluster. Is calculated whether or not is included in the parent log cluster,
There is a hypertext that is not included, and the hypertext has a predetermined link relationship with the hypertext set included in the upper log cluster, and the number of accesses to the upper log cluster and the number of accesses to the corresponding log cluster When the change in the number of accesses is in a predetermined relationship, the hypertext is detected, and the intention shift presentation process presents a hypertext in a predetermined relationship with the hypertext to a user, and In the shift detection process and the intention shift presentation process, if there is a content cluster included in the parent log cluster, the above process is repeated to detect the intention shift and the content cluster included in the parent log cluster. Repeat until it disappears or the predetermined stop condition is met. A storage medium storing hypertext structure changing support program according to claim 11, wherein.