JP2011060228A - Webpage correlation evaluation device for detecting information spreading - Google Patents

Abstract

Conventionally, there is no technology for accurately identifying related web pages.
A web page relationship evaluation device 30 includes a creation time detection unit 2 that detects a creation time of each of a plurality of determination target web pages, and a similarity between a determination target web page and another determination target web page. The similarity calculation unit 3 that calculates the two and the two determination target web pages used when calculating the similarity for each of the similarities that are equal to or greater than the reference value among the calculated similarities And a related web page specifying unit 5 for specifying that the web pages are mutually related.
[Selection] Figure 2

Description

  The present invention relates to a technique for detecting information propagation by evaluating the relationship between a plurality of web pages.

  In recent years, companies have transmitted information using web pages, and individuals have also transmitted information using web pages, blogs, and SNS (Social Network Service). Each information may be transmitted under the influence of other information. For example, when a company announces on the web page that a new product will be released, an individual who views the web page may publish an opinion about the new product on the blog along with the information. In this way, certain information may spread through the web page.

JP 2008-130044 A

  By the way, propagation of information published on a web page may be uncomfortable. For example, even if no scandal has occurred, if information that slanders a company spreads over many web pages, a large number of people may know the information and hinder the economic activities of the company. Because there is. In such a case, a person belonging to the company wants to know the propagation path of the information. For example, if you know where the core website or source of information that makes up the propagation path is, you can read the site regularly and quickly notice that unfavorable information is flowing. And measures can be taken early.

  In addition, information may be disclosed on a web page or the like with the aim of advertising effectiveness. For example, when a company announces the fact that a new product has been released by posting a press release on its website, the public will see the fact via a news site or a well-known blog site. In such a case, the person who belongs to the company wants to know what route the information is propagated through. For example, if you can identify news sites and prominent blogs that directly propagate information to more people or indirectly influence other prominent blogs, you can identify those sites. More effective advertising can be performed with focus.

  However, there is currently no technique for accurately identifying the propagation path of information spread by a web page. That is, there is no technique for accurately identifying the relationship between related web pages.

  An object of this invention is to provide the web page relationship evaluation apparatus which pinpoints the relationship between related web pages accurately.

  In order to solve the above problems and achieve the above object, the web page relation evaluation device of the present invention includes an acquisition unit that acquires information constituting each of a plurality of determination target web pages, and information acquired by the acquisition unit Based on the creation time detection unit for detecting the creation time of each of the plurality of determination target web pages, and another determination target for each of the plurality of determination target web pages based on the information acquired by the acquisition unit It is determined that two web pages are related to each other between a similarity calculation unit that calculates a similarity between each of the web pages and a plurality of similarities obtained by the similarity calculation unit For each similarity level that is greater than or equal to the reference value, the two web pages to be judged used for calculating the similarity level are related to each other. And a related web page specific unit that identified as the over-di.

  Moreover, the program for making a computer implement | achieve the function of each component of the web page relation evaluation apparatus of this invention is also 1 aspect of this invention.

  The present invention can provide a web page relationship evaluation apparatus that accurately identifies related web pages.

It is a block diagram of the web page analysis system of this Embodiment. It is a block diagram of the web page analysis apparatus of this Embodiment. It is a figure for demonstrating the outline | summary of operation | movement of the web page analysis system of this Embodiment. It is a flowchart which shows each step of the operation | movement which produces the search keyword which the search keyword creation apparatus of FIG. 2 performs. It is a flowchart which shows each step of the operation | movement which evaluates the relationship between the some web pages which the web page relationship evaluation apparatus of FIG. 2 performs. It is a figure which shows the four web pages of judgment object arranged in order of creation time, and the similarity between each web page of judgment object. FIG. 7 is a network diagram created by connecting web pages related to the four web pages to be judged in FIG. It is a flowchart which shows each step of the operation | movement which evaluates the website which the site evaluation apparatus of FIG. 2 performs. It is a figure for demonstrating a website.

  EMBODIMENT OF THE INVENTION Below, the form for implementing this invention is demonstrated with reference to drawings.

  First, the configuration of the web page analysis system of the present embodiment will be described with reference to FIG. FIG. 1 is a configuration diagram of a web page analysis system according to the present embodiment. The web page analysis system of this embodiment is a system for analyzing a web page. As shown in FIG. 1, the web page analysis device 100, the client terminal device 200, the search device 300, the communication network 400, and the like. And a site access number estimation device 500.

  The web page analysis apparatus 100 is an apparatus that analyzes a web page disclosed in the communication network 400. Details of the configuration of the web page analysis apparatus 100 will be described later with reference to FIG. The client terminal device 200 is a device used by a user, and accepts a search keyword, an instruction to search a web page related to the search keyword (hereinafter referred to as “search instruction”), and the like from the user. The client terminal device 200 transmits the received instruction or the like to the web page analysis device 100 or the search device 300, and receives the results obtained by them from the web page analysis device 100 and the search device 300. As shown in FIG. 1, a display device 250 is connected to the client terminal device 200, and information received by the client terminal device 200 is displayed by the display device 250.

  The search device 300 is a device that searches a web page related to a search keyword from web pages disclosed in the communication network 400. The communication network 400 is a network for communicating data such as the Internet. The site access number estimation apparatus 500 is an apparatus that estimates the number of website accesses. The web page analysis device 100, the client terminal device 200, the search device 300, and the site access number estimation device 500 are connected to the communication network 400 and can communicate with each other.

  Next, the configuration of the web page analysis apparatus 100 will be described with reference to FIG. FIG. 2 is a configuration diagram of the web page analysis apparatus 100. As described above, the web page analyzing apparatus 100 analyzes a web page disclosed in the communication network 400. As shown in FIG. 2, the web page analyzing apparatus 100, the communication apparatus 10, the search keyword creating apparatus 20, and the web A page relation evaluation device 30 and a site evaluation device 40 are included.

  The communication device 10 communicates data. The search keyword creation device 20 creates a more appropriate search keyword that has been subjected to appropriate conversion processing based on the keyword selected by the user.

  The web page relationship evaluation device 30 is a device that evaluates the relationship between a plurality of web pages searched by the search device 300 based on the search keyword created by the search keyword creation device 20. As shown in FIG. 2, the web page relationship evaluation apparatus 30 includes an acquisition unit 1, a creation time detection unit 2, a similarity calculation unit 3, a similarity correction unit 4, a related web page specification unit 5, a network It has a diagram creation unit 6 and an output unit 7.

  The acquisition unit 1 acquires information constituting each of a plurality of web pages searched by the search device 300. Since the web page searched by the search device 300 is a web page evaluated by the web page relation evaluation device 30, the web page searched by the search device 300 is hereinafter referred to as “determination target web page”. The creation time detection unit 2 detects the creation time of each of the plurality of determination target web pages based on the information acquired by the acquisition unit 1.

  Based on the information acquired by the acquisition unit 1, the similarity calculation unit 3 calculates the similarity between each of the plurality of determination target web pages and each of the other determination target web pages. Specifically, when calculating the similarity between two determination target web pages, the similarity calculation unit 3 analyzes the morphemes of information constituting each of the two determination target web pages, and calculates TF. Using the IDF (Term Frequency-Inverse Document Frequency) method and the vector space method, the similarity between the two determination target web pages is calculated. In addition to this method, any of the widely used document similarity calculation methods may be used for calculating the web page similarity. For example, the N-Gram method may be used instead of the TF-IDF method. good.

  The similarity correction unit 4 corrects the similarity calculated by the similarity calculation unit 3. Hereinafter, the value obtained by the similarity correction unit 4 is referred to as “corrected similarity”. Since the similarity correction unit 4 corrects the similarity by a plurality of methods, each method will be described later when the operation of the similarity correction unit 4 is described.

  The related web page specifying unit 5 has a corrected similarity that is equal to or higher than a reference value for determining that two web pages are related to each other among the plurality of corrected similarities obtained by the similarity correcting unit 4 Is detected. And the related web page specific | specification part 5 is a web page which mutually relates the web page of 2 judgment object used when calculating the correction similarity about each detected correction similarity. Is identified.

  For each set of two web pages specified by the related web page specification unit 5, the network diagram creation unit 6 creates two web pages constituting the set by the creation time detection unit 2. Virtually arranged in this order and virtually connected by lines to create a network diagram of related web pages. That is, the network diagram creation unit 6 creates a network diagram by connecting related web pages in a tree shape in the order of creation time.

  The output unit 7 outputs the network diagram created by the network diagram creation unit 6 to the communication device 10.

  The site evaluation device 40 is a device that evaluates a website disclosed in the communication network 400.

  Next, the operation of the web page analysis system of this embodiment will be described. FIG. 3 is a diagram for explaining the outline of the operation of the web page analysis system of the present embodiment. As shown in FIG. 3, the operation of the web page analysis system according to the present embodiment includes an operation for creating a search keyword for performing a refined search (phase F1) and an operation for evaluating the relationship between a plurality of web pages. (Phase F2) and an operation for evaluating a website (Phase F3). Therefore, below, operation | movement of the web page analysis system of this Embodiment is demonstrated for every phase.

(Operation to create a search keyword)
First, an operation (phase F1) for creating a search keyword for performing a refined search will be described.

  The user inputs a search keyword and an instruction (search instruction) for searching a web page related to the search keyword to the client terminal device 200. The client terminal device 200 receives the input search keyword and search instruction, and transmits them to the search device 300 via the communication network 400. The search device 300 receives the search keyword and the search instruction from the client terminal device 200, and searches the web page related to the received search keyword from the web pages disclosed in the communication network 400. For convenience of explanation, it is assumed that the search device 300 has searched a plurality of web pages. The search device 300 transmits the titles of the searched web pages to the client terminal device 200 via the communication network 400.

  The client terminal device 200 receives the titles and text content outlines of the plurality of web pages from the search device 300, and displays them on the display device 250 in the order determined by the search device 300. The user looks at the title of each web page displayed by the display device 250, selects any web page, and instructs the client terminal device 200 to acquire the selected web page.

  The client terminal device 200 acquires the web page selected by the user and causes the display device 250 to display the web page. When the user determines that the web page displayed by the display device 250 is an appropriate web page as a material for creating a more appropriate search keyword, the user inputs the determination result to the client terminal device 200.

  The client terminal device 200 transmits information identifying such a web page to the web page analysis device 100 via the communication network 400. The information specifying the web page is, for example, a URL (Uniform Resource Locator) of the web page. As described above, the user selects a plurality of web pages for creating a more appropriate search keyword, and the client terminal device 200 accepts information for specifying the plurality of web pages selected by the user and receives the web. It transmits to the page analysis apparatus 100. The client terminal device 200 also transmits the search keyword input by the user to the web page analysis device 100.

  In the web page analysis device 100, the communication device 10 receives from the client terminal device 200 information specifying a plurality of web pages for creating a more appropriate search keyword and the search keyword input by the user. .

  The following operation will be described with reference to FIG. FIG. 4 is a flowchart showing steps of an operation for creating a search keyword performed by the search keyword creation device 20.

  The search keyword creation device 20 acquires the information received by the communication device 10 for specifying a plurality of web pages for creating a more appropriate search keyword and the search keyword input by the user (S1). Next, the search keyword creating device 20 acquires information constituting each of a plurality of web pages for creating a more appropriate search keyword via the communication network 400 (S2).

  The search keyword creation device 20 specifies keywords related to a plurality of web pages based on the acquired information. Specifically, the search keyword creating device 20 analyzes the morphemes constituting each of the plurality of web pages based on the acquired information, and uses a predetermined number from the larger value obtained by using the TF-IDF method. Are identified as keywords related to a plurality of web pages (S3). Then, the search keyword creating device 20 adds the identified keyword to the search keyword input by the user, and creates a new search keyword for performing a refined search (S4). This technique is an example of a realization method. For example, a search keyword may be created using the Key-Graph method instead.

  Then, the communication device 10 transmits the new search keyword created by the search keyword creation device 20 to the search device 300 via the communication network 400. The search device 300 receives a new search keyword from the communication device 10 of the web page analysis device 100 and searches a web page related to the new search keyword from web pages disclosed in the communication network 400. . Again, for convenience of explanation, it is assumed that the search device 300 has searched a plurality of web pages. The search device 300 transmits information constituting each of the searched web pages to the web page analysis device 100 via the communication network 400.

(Operation to evaluate the relationship between multiple web pages)
Next, the operation | movement (phase F2) which evaluates the relationship between several web pages is demonstrated using FIG. FIG. 5 is a flowchart showing each step of the operation for evaluating the relationship between a plurality of web pages performed by the web page relationship evaluation apparatus 30.

  The communication device 10 of the web page analysis apparatus 100 receives information constituting a plurality of web pages (determination target web pages) searched by the search device 300 in the last step of the phase F1. In the web page relation evaluation device 30, the acquisition unit 1 acquires information constituting each of the plurality of determination target web pages received by the communication device 10 (S11). The creation time detection unit 2 detects the creation time of each of the plurality of determination target web pages based on the information acquired by the acquisition unit 1 (S12).

  Based on the information acquired by the acquisition unit 1, the similarity calculation unit 3 calculates a similarity between each of the plurality of determination target web pages and each of the other determination target web pages (S13).

  The similarity correction unit 4 corrects the similarity calculated by the similarity calculation unit 3 (S14). The operation of the similarity correction unit 4 will be specifically described by taking two determination target web pages out of a plurality of determination target web pages. For convenience of explanation, it is assumed that two web pages to be determined are called a web page X and a web page Y, and the web page X is created before the web page Y. Further, the similarity between the web page X and the web page Y calculated by the similarity calculation unit 3 is referred to as “similarity S”.

  At this time, the similarity correction unit 4 determines whether the web page Y is linked to the web page X based on the information acquired by the acquisition unit 1. When it is determined that the web page Y is linked to the web page X, the similarity correction unit 4 assigns the first correction coefficient C1 to the similarity S. When it is determined that the web page Y is not linked to the web page X, the similarity correction unit 4 does not assign the first correction coefficient C1 to the similarity S.

  Further, the similarity correction unit 4 determines whether the web page X and the web page Y are linked to another web page α based on the information acquired by the acquisition unit 1. When it is determined that the two web pages are both linked to another web page α, the similarity correction unit 4 assigns the second correction coefficient C2 to the similarity S. When it is determined that at least one of the two web pages is not linked to the other web page α, the similarity correction unit 4 does not assign the second correction coefficient C2 to the similarity S.

  Further, the similarity correction unit 4 determines the difference between the creation time of the web page X and the creation time of the web page Y based on the creation time detected by the creation time detection unit 2, and determines the difference according to the difference. A correction coefficient C3 of 3 is assigned to the similarity S. For example, the similarity correction unit 4 assigns the third correction coefficient C3 specified by the following expression (1) to the similarity S.

C3 = exp (−α × (sp / tp) (1)
α is a constant determined by experiment, sp is the difference between the creation time of the web page X and the creation time of the web page Y, and tp is the first creation of all the web pages to be judged This is the difference between the creation time of the created web page and the creation time of the last created web page.

  In addition, the similarity correction unit 4 generates another determination target web created between the creation time of the web page X and the creation time of the web page Y based on the creation time detected by the creation time detection unit 2. Detect the number of pages. The similarity correction unit 4 gives the similarity S a fourth correction coefficient C4 corresponding to the detected number. For example, the similarity correction unit 4 assigns the fourth correction coefficient C4 specified by the following expression (2) to the similarity S.

C4 = exp (−β × (sn / tn) (2)
β is a constant determined by experiment, and sn is the number of web pages created between the creation time of web page X and the creation time of web page Y among all the web pages to be judged. And tn is the number of all web pages to be determined.

  Further, the similarity correction unit 4 gives a fifth correction coefficient C5 corresponding to the number of accesses per unit period of the web page X to the similarity S calculated by the similarity calculation unit 3. In this case, the similarity correction unit 4 first causes the acquisition unit 1 to acquire the number of accesses per unit period of all the determination target web pages including the web page X and the web page Y. And the similarity correction | amendment part 4 provides the 5th correction coefficient C5 specified by Formula (3) shown below with respect to the similarity S. FIG.

C5 = exp (−γ × (n / N) (3)
γ is a constant determined by experiment, n is the number of accesses per unit period of the web page X, and N is a unit of each of the plurality of web pages to be judged including the web page X and the web page Y This is the maximum number of accesses per period.

  Then, the similarity correction unit 4 adds all the correction coefficients (first correction coefficient C1 to fifth correction coefficient C5) described above, and multiplies the similarity S by the addition correction coefficient obtained by the addition. Thus, the similarity S is corrected.

  So far, the operation in which the similarity correction unit 4 corrects the similarity S between the web page X and the web page Y calculated by the similarity calculation unit 3 has been described. The similarity correction unit 4 performs the same operation as that operation on all the similarities calculated by the similarity calculating unit 3, thereby correcting all the similarities calculated by the similarity calculating unit 3, The corrected similarity is calculated for all the similarities (S14).

  The related web page specifying unit 5 has a corrected similarity that is equal to or higher than a reference value for determining that two web pages are related to each other among the plurality of corrected similarities obtained by the similarity correcting unit 4 Is detected. And the related web page specific | specification part 5 is a web page which mutually relates the web page of 2 judgment object used when calculating the correction similarity about each detected correction similarity. (S15).

  For each set of two web pages specified by the related web page specification unit 5, the network diagram creation unit 6 creates two web pages constituting the set by the creation time detection unit 2. A network diagram of related web pages is created (S16).

  Here, specific examples of operations of the related web page specifying unit 5 and the network diagram creating unit 6 will be described with reference to FIGS. 6 (A) and 6 (B). FIG. 6A is a diagram showing four determination target web pages arranged in the order of creation time and the similarity between each determination target web page. FIG. 6B is a network diagram created by connecting the web pages related to the four web pages to be judged in FIG. 6A with lines. The horizontal axis t in FIGS. 6A and 6B is a time axis.

  For convenience of explanation, it is assumed that the four web pages to be determined shown in FIGS. 6A and 6B are web page A, web page B, web page C, and web page D. The circles in FIGS. 6A and 6B are web pages. 6A and 6B, the web page A is created first, the web page B is created next, and then the web page C is created. Assume that the web page D was created and finally created.

  Further, as shown in FIG. 6A, the corrected similarity between the web page A and the web page B is 0.8, the corrected similarity between the web page A and the web page C is 0.8, The corrected similarity between web page A and web page D is 0.3, the corrected similarity between web page B and web page C is 0.2, and the corrected similarity between web page B and web page D Is 0.3, and the corrected similarity between the web page C and the web page D is 0.6. Further, it is assumed that the reference value for determining that two web pages are related to each other is 0.5.

  At this time, the related web page specifying unit 5 specifies each of the web page A and the web page B, the web page A and the web page C, and the web page C and the web page D as mutually related web pages. . The network diagram creation unit 6 connects the web page A and the web page B, the web page A and the web page C, and the web page C and the web page D with a line, and creates the network diagram shown in FIG.

  The output unit 7 outputs the network diagram of the web page created by the network diagram creation unit 6 to the communication device 10.

  The communication device 10 outputs the network diagram of the web page from the web page relationship evaluation device 30 to the client terminal device 200 via the communication network 400. The client terminal device 200 receives the network diagram of the web page obtained by the web page relationship evaluation device 30 and displays it on the display device 250.

  As a result, the user knows the network associated with the search keyword determined by the user and the web page related to the web page selected by the user among the web pages related to the search keyword in the order of creation time. Can do. That is, the user can know the relevance between a plurality of web pages related to the search keyword determined by the user.

  In the above description, the similarity correction unit 4 adds all the correction coefficients of the first correction coefficient C1 to the fifth correction coefficient C5 and adds the addition correction coefficient obtained by the addition to the similarity S. The similarity S is corrected by multiplication. However, the similarity correction unit 4 adds some correction coefficients of the first correction coefficient C1 to the fifth correction coefficient C5 to obtain an addition correction coefficient, and multiplies the similarity S by the addition correction coefficient. The degree of similarity S may be corrected. Further, the similarity correction unit 4 may correct the similarity S by multiplying the similarity S by one correction coefficient of the first correction coefficient C1 to the fifth correction coefficient C5.

  In the above description, the related web page specifying unit 5 determines the reference value for determining that two web pages of the plurality of corrected similarities obtained by the similarity correcting unit 4 are related to each other. The corrected similarity as described above is detected. However, the related web page specifying unit 5 may detect a similarity that is equal to or higher than the reference value among a plurality of similarities (before correction) obtained by the similarity calculating unit 3. In that case, the related web page specifying unit 5 determines that for each detected similarity, the two web pages to be determined used when calculating the similarity are web pages that are related to each other. Identify.

(Action to evaluate website)
Next, the operation for evaluating the website (phase F3) will be described with reference to FIG. FIG. 7 is a flowchart showing steps of an operation for evaluating a website performed by the site evaluation apparatus 40.

  The site evaluation device 40 classifies all the web pages to be judged received by the communication device 10 for each domain, thereby specifying each website (S21). The website is a set of pages having a common domain among the web pages to be determined. The website will be described with reference to FIG. FIG. 8 is a diagram for explaining a website. In FIG. 8, the web page H, the web page J, the web page K, and the web page M are web pages having the same domain, and a set of these is a website.

  Therefore, in FIG. 8, the web page H, the web page J, the web page K, and the web page M are lower-level web pages of the website G. The ranks of the web page H, the web page J, the web page K, and the web page M are h, j, k, and m, respectively. The rank is a value determined by the search device 300. For example, when a general search engine is used as a search device, the rank is the display order of search results.

  Next, the site evaluation device 40 uses the site access number estimation device 500 via the communication device 10 to obtain the number of accesses to each website (S22 in FIG. 7). In the example of FIG. 8, the access number g of the website G is acquired.

  Next, the site evaluation device 40 acquires, for each website, the rank of a web page having the same domain as that of the website via the communication device 10 (S23 in FIG. 7). In the example of FIG. 8, the site evaluation device 40 acquires the ranks of the web page H, the web page J, the web page K, and the web page M.

  Next, for each website, the site evaluation device 40 uses the number of accesses to the website and the rank of the web page belonging to the website, and calculates the importance of the website using an arbitrary method. (S24 in FIG. 7).

  Finally, the site evaluation device 40 assigns ranks to all websites in descending order of importance (S25 in FIG. 7).

  The communication device 10 transmits the title of each website obtained by the site evaluation device 40 to the client terminal device 200 via the communication network 400 together with the rank and the importance thereof. The client terminal device 200 receives the title, ranking, and importance of each website obtained by the site evaluation device 40, and causes the display device 250 to display the title of each website according to the ranking. At that time, the client terminal device 200 displays the importance of each website on the display device 250 in combination with the title.

  As described above, in the present embodiment, the web page relationship evaluation apparatus 30 virtually arranges a plurality of determination target web pages in the order of creation time, and virtually associates web pages related to them. To create a network diagram of a tree-like web page. In the created network diagram, it means that sites that are higher in the time series order are sensitive to information, and sites that have many related sites in the lower order in the time series order are able to propagate information to other websites. Means high. Further, the site evaluation device 40 calculates the importance of each website. A site at the top of the list of websites ranked based on importance means that there is a high influence on the propagation of information from the viewpoint of the number of accesses and the rank of the search device.

  For example, a website that plays a major role in information propagation in the case by inputting a keyword related to a case where factless information that slanders a company is propagated and obtaining the above analysis result. Can be found. By requesting self-restraint to the site, it can lead to the study of recurrence prevention measures. In addition, for example, when a manufacturer wants to perform an advertisement effect analysis in the announcement of the release of its own product, a plurality of in-house product names are input to the device of the present invention to obtain the above analysis result. Of the analysis results obtained, sites that tend to have high information propagation capabilities in many cases of in-house product analysis will continue to play a major role in information propagation in the announcement of their product launches in the future. It is done. You can get a more effective advertising effect by working on this site.

  Note that the functions of the devices constituting the web page analysis device 100 are realized by the cooperation of a computer program and hardware such as a computer CPU and memory. However, each of the above functions may be realized in any form such as realized by a dedicated circuit.

  The web page relationship evaluation device 30 also has dedicated memories corresponding to the creation time detection unit 2, the similarity calculation unit 3, the similarity correction unit 4, the related web page specification unit 5, and the network diagram creation unit 6, respectively. These components may have a common memory. Each of these components may store data for processing and processing results in a dedicated memory or a common memory.

  100 web page analysis device, 200 client terminal device, 300 search device, 400 communication network, 500 site access number estimation device, 250 display device, 10 communication device, 20 search keyword creation device, 30 web page relation evaluation device, 40 site evaluation Device, 1 acquisition unit, 2 creation time detection unit, 3 similarity calculation unit, 4 similarity correction unit, 5 related web page specification unit, 6 network diagram creation unit, 7 output unit.

Claims (8)

An acquisition unit for acquiring information constituting each of a plurality of web pages to be determined;
Based on the information acquired by the acquisition unit, a creation time detection unit that detects the creation time of each of a plurality of determination target web pages;
Based on the information acquired by the acquisition unit, for each of a plurality of determination target web pages, a similarity calculation unit that calculates a similarity between each of the other determination target web pages;
When calculating the similarity for each of the similarities that are equal to or higher than a reference value for determining that two web pages are related to each other among the plurality of similarities obtained by the similarity calculating unit. A web page relationship evaluation device comprising: a related web page identifying unit that identifies the two web pages to be determined used in the above as web pages that are related to each other. Further, for each set of two web pages specified by the related web page specifying unit, two web pages constituting the set are virtually divided in the order of the creation time detected by the creation time detection unit. The web page relationship evaluation apparatus according to claim 1, further comprising a network diagram creation unit that creates a network diagram of web pages that are connected side by side and virtually connected with a line. Further, based on the information acquired by the acquisition unit, for each similarity obtained by the similarity calculation unit, one of the two determination target web pages used when calculating the similarity When linked to the other, a similarity correction unit that calculates the correction similarity by multiplying the similarity by a correction coefficient,
The related web page specifying unit is used when calculating the corrected similarity for each of the corrected similarities that are equal to or more than the reference value among the plurality of corrected similarities obtained by the similarity correcting unit. The web page relationship evaluation apparatus according to claim 1, wherein the web pages to be determined are identified as web pages that are related to each other. Further, based on the information acquired by the acquisition unit, for each similarity obtained by the similarity calculation unit, both of the two web pages to be judged used for calculating the similarity are different. A similarity correction unit that calculates a corrected similarity by multiplying the similarity by a correction coefficient when linked to the web page of
The related web page specifying unit is used when calculating the corrected similarity for each of the corrected similarities that are equal to or more than the reference value among the plurality of corrected similarities obtained by the similarity correcting unit. The web page relationship evaluation apparatus according to claim 1, wherein the web pages to be determined are identified as web pages that are related to each other. Further, based on the creation time detected by the creation time detection unit, for each similarity obtained by the similarity calculation unit, two web pages to be used for calculating the similarity A similarity correction unit that calculates a correction similarity by multiplying the similarity by a correction coefficient according to the difference in creation time of
The related web page specifying unit is used when calculating the corrected similarity for each of the corrected similarities that are equal to or more than the reference value among the plurality of corrected similarities obtained by the similarity correcting unit. The web page relationship evaluation apparatus according to claim 1, wherein the web pages to be determined are identified as web pages that are related to each other. Further, based on the creation time detected by the creation time detection unit, for each similarity obtained by the similarity calculation unit, two web pages to be used for calculating the similarity A similarity correction unit that calculates a correction similarity by multiplying the similarity by a correction coefficient according to the number of other web pages to be determined created during the creation time of
The related web page specifying unit is used when calculating the corrected similarity for each of the corrected similarities that are equal to or more than the reference value among the plurality of corrected similarities obtained by the similarity correcting unit. The web page relationship evaluation apparatus according to claim 1, wherein the web pages to be determined are identified as web pages that are related to each other. Furthermore, a similarity correction unit that corrects each similarity obtained by the similarity calculation unit,
The acquisition unit acquires the number of accesses per unit period for each of a plurality of web pages to be determined,
The similarity correction unit, for each similarity obtained by the similarity calculation unit based on the number of accesses per unit period of each of the plurality of determination target web pages acquired by the acquisition unit, the similarity Of the two web pages to be judged used for calculating the web page, the similarity is multiplied by a correction coefficient corresponding to the number of accesses per unit period of the previously created web page. Calculate
The related web page specifying unit is used when calculating the corrected similarity for each of the corrected similarities that are equal to or more than the reference value among the plurality of corrected similarities obtained by the similarity correcting unit. The web page relationship evaluation apparatus according to claim 1, wherein the web pages to be determined are identified as web pages that are related to each other. A function of acquiring information constituting each of a plurality of web pages to be judged;
Based on the acquired information, a function that detects the creation time of each of a plurality of judgment target web pages,
Based on the acquired information, for each of a plurality of judgment target web pages, a function for calculating the similarity between each of the judgment target web pages,
Of the plurality of calculated similarities, two of the similarities that are equal to or higher than a reference value for determining that two web pages are related to each other are used for calculating the similarities. A program that causes a computer to realize the function of identifying a web page to be judged as a web page that is related to each other.

Images