JP5705772B2 - Website analysis method, apparatus, system, and program - Google Patents

Description

  The present invention relates to a website analysis method, apparatus, system, and program, and more particularly, to a website analysis method, apparatus, system, and program for analyzing a user's flow line in a website.

  For website owners, especially e-commerce site owners, to constantly grasp the quality of experience and usage of users for the website users, and to improve the number of users and usage efficiency through improvements in content and structure Is important.

  For example, as a method of quantitatively measuring the user's quality of experience, the transfer time of the website between the web server and the user terminal is estimated, and the communication quality when the user browses the website is measured. A method has been proposed (see, for example, Patent Document 1).

  In addition, as a method for supporting the discovery of points that should be improved in the website, a method using a server log type access analysis tool that performs analysis based on the log file of the web server, or the contents of the analyzed website When the browser of the browsing user executes the JavaScript (registered trademark) tag embedded in the user, the browsing information of the user can be accessed by giving the browsing information to the browsing log measurement server prepared separately. There is a method using a web beacon type access analysis tool that collects and analyzes data (see, for example, Non-Patent Document 1).

JP 2004-78712 A

Yuichi Watanabe, Satoshi Hattori, Haruo Hayami, "Proposal of Temporal and Spatial Analysis System for Access Logs with IP Geolocation Information", Information Processing Society of Japan, Research Report-Groupware and Network Services (GN), Vol 2009, No. 33, pp. 103-107.

  When a website owner plans to improve the content and structure of his / her own website, the server log type access analysis tool and the web beacon type access analysis tool used in the past can be analyzed. It is possible to grasp the content that is not effectively utilized by the user on the website of the company, that is, the website owned by the user. More specifically, it is possible to grasp content that is not effectively used compared with other content of the website that the user owns. However, since it is not compared with similar content on other websites that are similar to your website being analyzed, the content you choose to improve is actually the same type of content on other websites. There is a possibility that it was used more effectively than content. Similarly, it can happen that content that was considered working effectively on its website was less efficient than similar content on someone else's website.

  In other words, in the prior art, it was possible to identify content that was not functioning effectively in the website owned by the subject of analysis, but it was effective to compare websites owned by others. It was impossible to identify content that was not functioning properly.

  The present invention has been made in view of the above circumstances, and a website capable of analyzing a user's flow line in a website by using an index comparable to the website to be compared with respect to the website to be analyzed. An object is to provide an analysis method, an apparatus, a system, and a program.

In order to achieve the above object, according to the website analysis method of the present invention, the collecting means includes the URL of the first browsing web page in the website browsed by the user and immediately before the first browsing web page. A step of collecting a plurality of browsing logs of the website including the URL of the browsed second browsing web page and acquired at the user terminal ; A regular expression that associates a plurality of functions, each of which has a common definition among a plurality of websites, with a regular expression obtained by extracting a common part in URLs of a plurality of web pages corresponding to each of the functions. A function conversion dictionary, each URL of the first browsing web page included in each of a plurality of browsing logs collected by the collecting means, and the first Determining a function corresponding to each of the first browsing web page and each of the second browsing web pages based on each of the URLs of the browsing web pages, and analyzing means comprising the plurality of browsing Based on the log and the function corresponding to each of the first browsing web page and the second browsing web page determined by the determination means, the browsing transition state of each web page in the website Are analyzed as transition probabilities between functions, which is an index comparable to other websites .

According to the website analysis method of the present invention, the collecting means includes the URL of the first browsing web page in the website browsed by the user and the second browsing browsed immediately before the first browsing web page. A plurality of browsing logs of websites including URLs of web pages and acquired at the user terminal are collected. Then, for each of a plurality of web page groups having different uses possessed by the website, the discrimination means has a plurality of functions assigned a common definition among the plurality of websites, and a plurality of webs corresponding to each of the functions. A regular expression-function conversion dictionary that associates a regular expression obtained by extracting a common part in the URL of the page, and each URL of the first browsing web page included in each of the plurality of browsing logs collected by the collecting unit and The function corresponding to each of the first browsing web page and each of the second browsing web page is determined based on each of the URLs of the two browsing web pages. Then, each web page in the website is analyzed based on a function corresponding to each of the plurality of browsing logs and each of the first browsing web page and the second browsing web page determined by the determining unit. The transition state of browsing is analyzed as the transition probability between each function, which is an index that can be compared with other websites .

In this way, for each of a plurality of web page groups having different usages that the website has, a plurality of functions having a common definition among the plurality of websites, and a plurality of web pages corresponding to each of the functions. In order to obtain a transition probability between functions after determining a function corresponding to a browsing web page using a regular expression-function conversion dictionary that matches a regular expression obtained by extracting a common part in a URL, between websites It is possible to analyze the transition state of browsing of each web page in the website, that is, the flow line of the user, by absorbing the difference and using the index that can be compared with the website to be compared for the website to be analyzed.

  In addition, the analysis unit may include at least one of the number of browsing web pages corresponding to each transition state, the number of users viewing the web page corresponding to each transition state, and the number of sessions to the web page corresponding to each transition state. Based on the above, the transition probability between functions can be analyzed.

In addition, the website analysis apparatus of the present invention includes the URL of the first browsing web page in the website browsed by the user and the second browsing web page browsed immediately before the first browsing web page. Collecting means for collecting a plurality of browsing logs of websites including URLs and acquired at the user terminal, and common to a plurality of websites for each of a plurality of web page groups for a plurality of different uses possessed by the websites A regular expression-function conversion dictionary in which a plurality of functions to which a definition is assigned and a regular expression obtained by extracting a common part in URLs of a plurality of web pages corresponding to each of the functions are associated with each other; Based on each of the URLs of the first browsing web page and each of the URLs of the second browsing web page included in each of the browsing logs A discriminating means for discriminating a function corresponding to each of the first browsing web pages and each of the second browsing web pages, the plurality of browsing logs, and the first browsing web discriminated by the discriminating means. Each function which is an index which can compare the transition state of browsing of each web page in the website with other websites based on the function corresponding to each of the pages and the second browsing web page And analyzing means for analyzing as a transition probability between them.

In addition, the website analysis system of the present invention is incorporated in a user terminal, the URL of the first browsing web page in the website browsed by the user, and the first browsing web page just before the first browsing web page. A browsing log acquisition device for acquiring a browsing log of a website acquired at a user terminal, and a collecting means for collecting a plurality of browsing logs acquired by the browsing log acquisition device; For each of a plurality of web page groups having different uses, a plurality of functions assigned a common definition among a plurality of websites , and a common portion in a URL of a plurality of web pages corresponding to each of the functions Included in each of the regular expression-function conversion dictionary corresponding to the extracted regular expression and the plurality of browsing logs collected by the collecting means Corresponding to each of the first browsing web page and each of the second browsing web page based on each of the URLs of the first browsing web page and each of the URLs of the second browsing web page. Based on the function corresponding to each of the first browsing web page and the second browsing web page determined by the determining means for determining the function, the plurality of browsing logs, and the determining means, A website analysis device including analysis means for analyzing a transition state of browsing of each web page in the website as a transition probability between each function, which is an index comparable to the other websites. ing.

  Moreover, the website analysis program of this invention is a program for making a computer perform each step of said website analysis method.

As described above, according to the website analysis method, apparatus, system, and program of the present invention, for each of a plurality of web page groups for different uses that the website has , common to the plurality of websites. Differences between websites to determine transition probabilities between functions after identifying functions corresponding to browsing web pages using regular expression-function conversion dictionaries that regularly express each of a plurality of functions to which definitions have been added. As a result, the flow of the user in the website can be analyzed by using an index that can be compared with the website to be compared.

It is a block diagram which shows the functional structure of the website analysis system which concerns on this Embodiment. It is a figure which shows an example of a browsing log accumulation | storage database. It is a figure which shows an example of a URL regular expression-function conversion dictionary. It is a flowchart which shows the content of the website analysis processing routine in the website analysis apparatus which concerns on this Embodiment.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the present embodiment, an electronic commerce site is mainly assumed as a website. However, the analysis target of the present invention is not limited to the electronic commerce site.

  As shown in FIG. 1, the website analysis system 10 according to the present embodiment includes a browsing log acquisition device 11 incorporated in a user terminal and a website analysis device 12. The website analysis device 12 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), and a ROM (Read Only Memory) that stores a program for executing a website analysis processing routine to be described later. It consists of As shown in FIG. 1, this computer functionally includes a browsing log collection unit 13, a browsing log accumulation database (DB) 14, a function determination unit 15, a URL regular expression-function conversion dictionary 16, a browsing It can be expressed by a configuration including the log analysis unit 17.

  The browsing log acquisition device 11 is a function incorporated in a user terminal where a user browses a website. For example, it can be incorporated into a web browser as an additional function (also called an add-on). Each time the user browses a web page in the website, the browsing log acquisition device 11 has a site ID that can uniquely identify the website that the user is browsing, a web page that is being browsed (hereinafter referred to as “browsing”). URL of web page), URL of the web page that the user was browsing immediately before the current browsing web page (hereinafter referred to as “referrer”), browsing date and time when browsing each web page, title of each web page, A browsing log including a user ID that can uniquely identify a user browsing each web page is acquired.

  The browsing web page is an example of the first browsing web page of the present invention, and the web page indicated by the referrer is an example of the second browsing web page of the present invention.

  The browsing log collection unit 13 collects browsing logs from each browsing log acquisition device 11 incorporated in a plurality of user terminals via a network. The user who is the target of browsing log collection may be a user extracted by sampling from the public, but in order to analyze the website owned by the user more effectively, the website of the analysis target website It is desirable to collect browsing logs for users who are users.

  The browsing log collection unit 13 stores the collected browsing logs in the browsing log accumulation DB 14. An example of the browsing log accumulation DB 14 is shown in FIG. In the example of FIG. 2, each item (user ID, site ID, browsing date, browsing web page URL, referrer, browsing web page title) is sequentially stored for each browsing log. Note that the referrer may be searched from the browsing log already stored in the browsing log accumulation DB 14 based on the user ID, the site ID, the browsing date and the URL of the browsing web page, without including the referrer in the browsing log. Good.

  The function determination unit 15 uses the URL regular expression-function conversion dictionary 16 to determine a function corresponding to the URL of the browsing web page included in each browsing log accumulated in the browsing log accumulation DB 14. An example of the URL regular expression-function conversion dictionary 16 is shown in FIG. In the example of FIG. 3, a site ID, a function, and a URL regular expression are associated with each other.

  Here, the function refers to a web page group created for the same purpose in the web page group in the website to be analyzed. As an example, in the case of an electronic commerce site, there is generally a web page group that is displayed when a product is put into a cart (sometimes referred to as “shopping basket” depending on the website). Such a web page group can be defined as [function: cart]. Similarly, in the e-commerce site, there are a web page group for ordering products, a web page group for confirming the order when the order procedure is completed, a web page group for displaying detailed product information, etc. [Function: Order], [Function: Order confirmation], [Function: Product details], etc. can be defined. Thereby, web page groups created for the same purpose can be abstracted as “functions”.

  The URL regular expression is a regular expression used as a URL common to a group of web pages corresponding to each function in a website to be analyzed. Here, not only electronic commerce sites but also websites created by companies in recent years, each web page in the website is often generated dynamically by a program, and the generated web page In many cases, a unique ID such as a web page or a product is given as a parameter in the URL so that the URL is unique. As a result, in the web page group corresponding to the same function, when paying attention to the URL, the URL is often the same except for the parameter portion. Therefore, a part obtained by removing the parameter from each URL of the web page corresponding to each function is defined as a URL regular expression of the web page group corresponding to the function. In the example of FIG. 3, the “*” portion of the URL regular expression indicates that a parameter was used in the original URL. As a result, it is possible to efficiently determine the function corresponding to the web page group.

  The site ID is an identifier that can uniquely identify a website. In the following, a website with a site ID “A” is represented as website A.

  In addition, it is necessary to register a plurality of URL regular expressions for a function that cannot represent a web page group corresponding to each function with a single URL regular expression. For example, when the website A has two types of web pages as web pages for ordering products, the website A corresponds to [function: order] as shown in the example of FIG. Thus, two URL regular expressions are registered.

  The function discriminating unit 15 refers to the correspondence between the URL regular expression defined for the website to be analyzed and the function in the URL regular expression-function conversion dictionary 16 defined as described above, and the browsing log accumulation DB 14 The function corresponding to the browsing web page included in each browsing log accumulated in the database is determined.

  Specifically, the URL regular expression-function conversion dictionary 16 compares each URL regular expression of the site ID of the website whose analysis target is the site ID with the URL of the browsing web page included in the browsing log. If any of the URL regular expressions matches the URL of the browsing web page, the function corresponding to the URL regular expression is determined as the function corresponding to the browsing web page.

  Similarly, the function determination unit 15 also determines the function corresponding to the referrer included in each browsing log accumulated in the analysis target browsing log accumulation DB 14. As for the function corresponding to the referrer, the URL of the browsing web page having the same URL as the referer is searched from the browsing log having the site ID “A” stored in the browsing log accumulation DB 14, and the browsing web page searched The function determined as the function corresponding to the URL may be determined as the function corresponding to the referrer.

  The browsing log analysis unit 17 analyzes the user's flow line in the analysis target website based on the browsing web page function determined by the function determination unit 15. Here, the user's flow line refers to a transition state of browsing of each web page in the website to be analyzed by the user. The browsing log analysis part 17 calculates | requires the transition probability between functions as a parameter | index showing this transition state. As described above, since the functions are abstracted from a group of web pages for the same purpose, the difference in characteristic content and site structure existing in each website is analyzed by analyzing the transition probability between functions. It is possible to analyze the user's flow line. That is, the transition probability between functions calculated for the website to be analyzed can be used as an index that can be compared with the transition probability between functions calculated for the website to be compared.

  The browsing log analysis unit 17 calculates transition probabilities between functions based on three types of indicators: the number of PV (Page View), the number of UU (Unique User), or the number of sessions.

For example, in a certain website A, when obtaining the transition probability P _A (order | cart) representing the transition state from [function: cart] to [function: order], when using the PV number, the following (1) ) To calculate the transition probability P _A (order | cart).

  In the formula (1), “the number of URLs corresponding to [function: cart]” has a function corresponding to the URL of the browsing web page by the function determination unit 15 with the site ID “A” in the browsing log accumulation DB 14. It is obtained by counting the lines identified as “cart”. In addition, “the number of cases where the referrer corresponds to [function: cart] and the URL corresponds to [function: order]” is the site ID “A” in the browsing log accumulation DB 14 and is referred to by the function determination unit 15. It is obtained by counting the lines in which the corresponding function is determined as “cart” and the function corresponding to the URL of the browsing web page is determined as “order”.

When calculating the above P _A (order | cart) using the number of UUs, the number of cases of the numerator and denominator of the equation (1) is stored in the viewing log accumulation DB 14 as the user ID that holds the row that meets the condition. The number of For example, in the case of the above example, “the number of URLs corresponding to [function: cart]” indicates that the site ID is “A” in the browsing log accumulation DB 14 and the function corresponding to the URL of the browsing web page is “cart”. It is obtained by counting the user IDs in the rows without duplication. “The number of referrers corresponding to [function: cart] and URL corresponding to [function: order]” indicates that the function corresponding to the referrer is “A” in the browsing log accumulation DB 14 as “ The function corresponding to the URL of the “cart” and the browsing web page is obtained by counting the user ID of the “order” line without duplication.

When calculating the above P _A (order | cart) using the number of sessions, the browsing log acquisition device 11 defines the end of the session as the case where the web page is not browsed for a preset time. The browsing log is acquired for each session in units of user IDs. And the browsing log acquired for every session by the browsing log collection part 13 is collected similarly to the above, and it preserve | saves in browsing log accumulation | storage DB14. When calculating the transition probability, the number of cases corresponding to the numerator and denominator of equation (1) is counted according to the number of sessions by counting the rows corresponding to the condition in the browsing log accumulation DB 14 as described above. The number of hits can be obtained.

  Note that the transition probability may be calculated using a combination of the PV number, the UU number, and the session number.

  The transition probability as described above is calculated for each function. Further, when obtaining the transition probability between functions, it is also possible to obtain a transition probability representing a transition state in a specific period by using information on the browsing date and time stored in the browsing log accumulation DB 14.

  Next, the operation of the website analysis system 10 of the present embodiment will be described. In the user terminal, when the user browses the website and the browsing log acquisition device 11 acquires the browsing log, the CPU executes the website analysis processing routine shown in FIG. 4 from the ROM in the website analysis device 12. The program is read and executed.

  In step 100, the browsing log collection unit 13 transmits a user ID, a site ID, a browsing date, a browsing web page URL, a referrer, and a browsing ID from each browsing log acquisition device 11 incorporated in a plurality of user terminals. A browsing log including information on the title of the browsing web page is collected and stored in the browsing log accumulation DB 14.

  Next, in step 102, the function determination unit 15 causes the URL regular expression-function conversion dictionary 16 to include each of the URL regular expressions of the site IDs of the websites to be analyzed and the browsing web pages included in the browsing log. Is compared with the URL. Then, a URL regular expression included in the URL of the browsing web page is searched, and a function corresponding to the searched URL regular expression is determined as a function corresponding to the browsing web page. Similarly, the function corresponding to the referrer included in each browsing log accumulated in the analysis target browsing log accumulation DB 14 is also determined.

  Next, in step 104, the browsing log analysis unit 17 determines the user based on the browsing log accumulated in the browsing log accumulation DB 14 and the functions corresponding to each browsing web page and referrer determined in step 102. The transition probability between functions representing the transition state of browsing of each web page in the website to be analyzed by is calculated. The transition probability calculated for each of the functions is output as the analysis result of the analysis target website, and the website analysis processing routine is terminated.

  As described above, according to the website analysis system 10 of the present embodiment, a function that abstracts a group of web pages for the same purpose is defined, and a URL regular expression-function in which each function is represented in a URL format. In order to determine the transition probability between functions after identifying the function corresponding to the browsing web page using the conversion dictionary, the characteristic content and site structure existing in each website are absorbed and held by others It is possible to calculate an index that can be easily compared with the website. By calculating and comparing such indicators for the analysis target website and the comparison target website, the analysis target website has a function that is less efficient than the comparison target website, that is, the efficiency is low. It becomes possible to specify a web page group as a candidate for improvement.

  The present invention is not limited to the above-described embodiment, and various modifications and applications are possible without departing from the gist of the present invention.

  In addition, the website analysis apparatus described above has a computer system inside, but the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used. .

  In the present specification, the embodiment has been described in which the program is installed in advance. However, the program can be provided by being stored in a computer-readable recording medium.

DESCRIPTION OF SYMBOLS 10 Website analysis system 11 Browsing log acquisition apparatus 12 Website analysis apparatus 13 Browsing log collection part 14 Browsing log accumulation database 14
15 Function Discriminator 16 URL Regular Expression-Function Conversion Dictionary 17 Browsing Log Analyzer

Claims (5)

The collecting means includes the URL of the first browsing web page in the website browsed by the user and the URL of the second browsing web page browsed immediately before the first browsing web page, and the user terminal Collecting a plurality of website browsing logs acquired in
A plurality of functions in which the determination unit assigns a common definition among a plurality of websites to each of a plurality of web page groups having different uses possessed by the website , and a plurality of web pages corresponding to each of the functions Each of the URLs of the first browsing web pages included in each of the plurality of browsing logs collected by the collecting means, Determining a function corresponding to each of the first browsing web page and each of the second browsing web page based on each of the URLs of the second browsing web page;
Based on the function corresponding to each of the plurality of browsing logs and each of the first browsing web page and the second browsing web page determined by the determining unit, Analyzing the transition state of each web page browsing as a transition probability between functions, which is an index comparable to other websites ;
Website analysis method including.   The analysis means is based on at least one of the number of browsing web pages corresponding to each transition state, the number of users viewing web pages corresponding to each transition state, and the number of sessions to the web page corresponding to each transition state. The website analysis method according to claim 1, wherein the transition probability between the functions is analyzed. The URL of the first browsing web page in the website browsed by the user and the URL of the second browsing web page browsed immediately before the first browsing web page, and acquired at the user terminal A collection means for collecting multiple website browsing logs;
For each of a plurality of web page groups having different uses on a website, a plurality of functions assigned a common definition among the plurality of websites , and a common URL in a plurality of web pages corresponding to each of the functions A regular expression-function conversion dictionary in which a regular expression from which a part is extracted is associated with each other, each of the URLs of the first browsing web pages included in each of the plurality of browsing logs collected by the collecting unit, and the second Discrimination means for discriminating a function corresponding to each of the first browsing web page and each of the second browsing web page based on each of the URLs of the browsing web pages;
Based on the functions corresponding to each of the plurality of browsing logs and each of the first browsing web pages and each of the second browsing web pages determined by the determination unit, each of the web pages in the website An analysis means for analyzing the transition state of browsing as a transition probability between each function, which is an index that can be compared with other websites ;
Including website analysis equipment. A URL of a first browsing web page in a website browsed by a user, which is incorporated in a user terminal, and a URL of a second browsing web page browsed immediately before the first browsing web page; and A browsing log acquisition device for acquiring a browsing log of a website acquired in a user terminal;
Collecting means for collecting a plurality of browsing logs acquired by the browsing log acquisition device, and a plurality of functions provided with a common definition among a plurality of websites for each of a plurality of different web page groups possessed by the website And a regular expression-function conversion dictionary that associates a regular expression obtained by extracting common parts in URLs of a plurality of web pages corresponding to each of the functions, and a plurality of browsing logs collected by the collecting means Corresponding to each of the first browsing web page and each of the second browsing web page based on each of the URLs of the first browsing web page and each of the URLs of the second browsing web page A discriminating means for discriminating a function to be performed, each of the plurality of browsing logs, each of the first browsing web pages determined by the discriminating means, and the second browsing web. Based on the function corresponding to each of the Bupeji, the analysis means for analyzing the transition state of viewing of each web page in the web site, the transition probabilities between each feature in a comparable indicator with other websites Including a website analysis device,
Including website analysis system.   A website analysis program for causing a computer to execute each step of the website analysis method according to claim 1.

Images