JP2012064132A - Blog body specification device and blog body specification method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a blog body specification device and a blog body specification method that specify a blog body while suppressing a control load.SOLUTION: A first origin specification part (2) of a blog body specification device (1) generates external path information representing a path of a tag, included in a source code, in the source code with respect to each of a plurality of blog pages made to correspond to the same under by analyzing the respective blog pages, and specifies, as an origin path including a blog body, a path which can uniquely specify the tag from the external path information (common path) and is deepest even when being a path including the body (body-including path). Then blog body extraction means of the blog body specification device (1) extracts the body of a blog page of the user based upon the origin path.

Description

  The present invention relates to a blog text specifying device and a blog text specifying method for specifying a text position common to a plurality of blogs of the same user.

  With the spread of the Internet, users can browse a large amount of data on the network, and search systems are being developed every day to extract data according to user needs from such a large amount of data. Yes. As a search system, a robot-type search system is known. In a robot-type search system, important keywords are automatically extracted from a large amount of data (Web pages) published on the Internet and indexed. Thus, desired data is provided to the user.

  By the way, in document data such as blogs, in addition to the main text, a lot of information that is not suitable for extracting keywords such as calendar information, information described in headers and footers, past log information, and advertising information is mixed. Therefore, if keywords are extracted from such information and an index is generated, there is a risk that search accuracy may be reduced.

  As a technique for extracting a text from a blog (text data), a noise removal system disclosed in Patent Document 1 is known. In this noise removal processing system, since the appearance frequency of the character string included in each sentence data exceeds a certain degree, it is determined that the character string is a fixed character string such as a header or footer, and is removed from each sentence data. Then, noise other than the text included in the sentence is removed.

JP 2009-271796 A

  In addition, the blog site supports RSS (RDF Site Summary, Rich Site Summary, Really Simple Syndication) by default, so in the search system, the text summary from the RSS description element (characters in the specified number of sentences at the beginning of the text). Column) can be collected automatically. In the search system, by extracting a character string that matches the summary of the text included in the description element from the blog, at least a part of the text included in the blog can be specified.

  However, there are an enormous amount of blogs on the network, and if the text is extracted from the blogs by the above method, the load on the search system becomes enormous. That is, in the noise removal system of Patent Document 1, it is necessary to individually remove noise for all blogs (documents). For example, when a new blog is created, Noise removal must be performed individually. The same applies to the text specification using RSS, and the character strings must be individually searched for all the blogs. Furthermore, a technique for specifying the scope of the text based on a part of the specified text has not been proposed.

  The present invention has been made in view of such problems, and an object of the present invention is to provide a blog text specifying device and a blog text specifying method that can specify a blog text while suppressing a control load.

  (1) A blog page DB that stores a blog page that includes a blog text created by a user and summary information of the blog text in association with information that identifies the user, and a plurality of blog pages that are associated with the same user. Analyzing and generating, for each of the plurality of blog pages, external path information generation means for generating external path information indicating paths in the source code of tags included in the source code, and the external path information generation means Out of the external path information, an external common path acquisition unit that acquires an external common path common to each of the plurality of blog pages of the same user, and a blog among the external path information generated by the external path information generation unit A text inclusion external path for acquiring a text inclusion external path including a text for each of the plurality of blog pages of the same user. And a column origin specifying means for specifying the deepest path as the origin path including the blog text among the paths that are the external common path and the text inclusion external path in all of the plurality of Web pages; A blog text identification device comprising:

According to the blog text specifying device of (1), the external common path acquisition means is common to each of a plurality of blog pages among the paths indicating the positions of the tags included in the source code of the plurality of blog pages created by the same user. An external common path is acquired, and a text inclusion external path acquisition unit acquires a text inclusion external path including a blog text among tag path information for each of a plurality of blog pages of the same user. Then, the column origin specifying means determines the blog body text common to the blog body of the user's blog page from among the paths that are external common paths and that are the body-inclusive external paths, Specify as the origin path to include.
Thereby, it is possible to specify a starting path indicating a position of a tag including a blog text, which is a path common to a plurality of blog pages for a plurality of blog pages of the same user. In addition, once the starting path is specified, the position of the blog text can be easily determined from the starting path unless the tag structure of the user's blog page changes significantly. can do.

  (2) Internal path information generating means for generating internal path information indicating a path of a tag lower than a tag specified by the starting path specified by the column starting point specifying means for each of a plurality of blog pages of the same user; The internal path information generating means generates an internal common path acquisition means for acquiring an internal common path common to each of the plurality of blog pages of the same user, and the internal path information generation means includes: A text inclusion internal path acquisition means for acquiring a text inclusion internal path including the blog text in the generated internal path information for each of a plurality of blog pages of the same user, the internal common path, and Of the paths that are the text inclusion internal paths in all of a plurality of Web pages, the deepest path is the text path including the blog text. And body position specifying means for specifying, blog body identifying device according to comprise (1) a.

  According to the blog text specifying device of (2), an internal common path and a text inclusion internal path are further acquired for tags lower than the tag specified by the starting path specified by the column starting point specifying means, Specify the text path to include. As a result, the position of the text can be specified in more detail, and since the position of the blog text can be easily specified once specified, the position of the blog text can be specified while suppressing the control load.

  (3) a blog page DB that stores a blog page including a blog text created by a user and summary information of the blog text in association with information for identifying the user, and a plurality of the blog pages associated with the same user For each of the plurality of blog pages, path information generating means for generating path information indicating a path in the source code of the tag included in the source code, and the path information generating means Among the path information, a common path acquisition unit that acquires a common path that is common to each of the plurality of blog pages of the same user, and a text inclusion path that includes a blog text among the path information generated by the path information generation unit. A body inclusion path acquisition unit that acquires each of a plurality of blog pages of the same user, the common path, Among the paths that are the text inclusion paths in all of the Web pages, a column origin specifying unit that identifies the deepest path as a starting path, and a specification that determines whether or not the column origin specifying unit has identified the origin path Passability determining means, and the path information generating means is based on a tag specified by the starting path, provided that the specifying availability determining means determines that the column starting point specifying means has specified the starting path. Further generate path information indicating the path of the lower tag, the common path acquisition means and the text inclusion path acquisition means acquire the common path and the text inclusion path from the path information of the lower tag, and The column starting point specifying unit specifies a new starting point path based on the common path and the text inclusion path, and the specifying possibility determining unit determines that the column starting point specifying unit sets the starting point path. A blog text specifying device further comprising: text position specifying means for specifying the starting path information as a text inclusion path including the blog text on the condition that it is determined that the URL cannot be specified.

  According to the blog text specifying device of (3), the new starting path is specified from the path information of the lower tag until the column starting point specifying unit cannot specify the new starting path. As a result, the position of the text can be specified in more detail, and since the position of the blog text can be easily specified once specified, the position of the blog text can be specified while suppressing the control load.

  (4) A computer including a blog page DB that stores a blog page including a blog text created by a user and summary information of the blog text in association with information for identifying the user identifies the text position of the blog page of the same user This is a method for specifying a blog body, and analyzes a plurality of blog pages associated with the same user, and indicates a path in the source code of a tag included in the source code for each of the plurality of blog pages. A step of generating external path information; a step of acquiring an external common path common to each of a plurality of blog pages of the same user among the generated external path information; and a blog text of the generated external path information A step of acquiring a body-inclusive external path including the URL for each of the plurality of blog pages of the same user. And a step of identifying the deepest path as a starting path including the blog text among the paths that are the external common path and the text inclusion external path in all of a plurality of Web pages. Blog body identification method.

  According to the blog text specifying method of (4), the same effect as that of the blog text specifying device of (1) can be obtained.

  According to the present invention, it is possible to specify a blog text while suppressing a control load.

It is a figure which shows the function structure of the blog text specific | specification apparatus of this embodiment. It is a figure which shows each database memorize | stored in a memory | storage part. It is a figure which shows the example of the production | generation of the external path information by an external path information generation means. It is a figure which shows the example of the produced | generated external path information. It is a figure which shows the example of the production | generation of the origin path | pass by a column origin identification means. It is a figure which shows the example of the production | generation of the origin path | pass by a column origin identification means. It is a figure which shows the example of the production | generation of the origin path | pass by a column origin identification means. It is a figure which shows the example of a production | generation of the internal path information by an internal path information production | generation means. It is a figure which shows the example of the production | generation of the text path | pass by a text position specific | specification means. It is a flowchart which shows the flow of a process of a blog text specific | specification apparatus. It is a figure which shows the function structure of the blog text identification apparatus of another embodiment. It is a figure which shows a blog page and its source code. It is a figure which shows a blog page and its source code.

Embodiments of the present invention will be described below with reference to the drawings.
First, the outline of the present invention will be described with reference to FIGS. FIG. 12 shows a web page 100 that is a blog page of a certain user and its source code 110, and FIG. 13 shows a web page 200 that is a blog page of another day of the same user and its source code 210.

  Referring to FIG. 12, Web page 100 includes a blog title column 101, an advertisement column 102, a calendar column 103, and a body column 104. Referring to FIG. 13, Web page 200 includes a blog title field 201, an advertisement field 202, a calendar field 203, and a body text field 204. Thus, the structure of the blog page of the same user is basically common to a plurality of Web pages. The present invention focuses on the fact that a plurality of blog pages of the same user have a common structure, and specifies the position of the text of the blog page. In particular, the tag included in the source code specifies the position of the text. It is specified by the position information (path).

[First Embodiment]
Then, with reference to FIGS. 1-10, the blog text identification device 1 of 1st Embodiment is demonstrated.

[Configuration of Blog Body Identification Device 1]
With reference to FIG. 1, the blog text identifying device 1 of the present invention will be described. The blog text specifying device 1 includes a first starting point specifying unit 2, a second starting point specifying unit 3, a storage unit 4, and a blog text extracting unit 5.

<Configuration of the first starting point specifying unit 2>
The first starting point specifying unit 2 includes an external path information generating unit 21, an external common path acquiring unit 22, a text inclusion external path acquiring unit 23, and a column starting point specifying unit 24.

The external path information generation means 21 analyzes a plurality of blog pages associated with the same user, and external path information indicating the path in the source code of the tag included in the source code for each of the plurality of blog pages. Is generated. The external path information generation unit 21 reads out HTML data stored in the blog page DB 41 of the storage unit 4 and generates external path information.
Here, as shown in FIGS. 3B and 3C, the external path information generation unit 21 generates the external path information of the tag included in the source code with a hard path and a soft path. “Hard path” refers to the tag path (tag [attribute] [number of appearances]) defined by the tag type, tag attribute, and the number of appearances of the tag, and “soft path” refers to the tag A tag path (tag [attribute]) defined by the type and tag attribute.

  Referring to FIG. 3, FIG. 3 (1) is the source code of the Web page 100, and FIG. 3 (2) is external path information that defines the path of the tag included in the source code with a hard path, FIG. 3B shows external path information that defines a path of a tag included in the source code with a soft path.

The external path information generation unit 21 generates a path (hard path and soft path) in the source code for each tag included in the source code. For example, for the generation of the path of the tag 11 of the line 4 of the source code, since the tag 11 is the “tr” tag having the first appearance count and the attribute is not specified, the external path information generation unit 21 An external path 11A (hard path) “tr [] [0]” is generated, and an external path 11B (soft path) “tr []” is generated.
The path can be specified arbitrarily. In the present embodiment, tags that do not change in position used in many Web pages, such as “html” tags, are omitted.

Further, regarding the generation of the path of the tag 12 of the line 16, the tag 12 is a tag lower than the “div” tag of the line 15 lower than the “td” tag of the line 13 lower than the “tr” tag of the line 9. is there. The “tr” tag of the line 9 is the “tr” tag whose appearance count is the second, and the attribute is not defined. The “td” tag of the line 13 is the “tr” tag of the line 9. The second “td” tag and no attributes are defined, and the “div” tag of the line 15 is the second “div” tag among the “td” tags of the line 13 and the attributes are defined. Absent. Further, the tag 12 itself of the line 16 is the first “center” tag in the “div” tag of the line 15, and the attribute is not defined.
Therefore, the external path information generation unit 21 uses the external “tr [] [1] / td [] [1] / div [] [1] / center [] [0]” as the path of the tag 12 of the line 16. A path 12A (hard path) is generated, and an external path 12B (soft path) of “tr [] / td [] / div [] / center []” is generated.

  FIG. 4 shows external path information of tags included in the source code of the Web page 100 (FIG. 12) and the Web page 200 (FIG. 13). 4 (1) is external path information (hard path) of the Web page 100, FIG. 4 (2) is external path information (soft path) of the Web page 100, and FIG. The external path information (hard path) of the page 200, and FIG. 4 (3) is the external path information (soft path) of the Web page 200. The external path information generated by the external path information generation unit 21 is supplied to the external common path acquisition unit 22 and the text inclusion external path acquisition unit 23.

Returning to FIG. 1, the external common path acquisition unit 22 uniquely identifies a tag corresponding to the path, which is common to each of a plurality of blog pages of the same user among the external path information generated by the external path information generation unit 21. Get external common path that can be identified.
The text inclusion external path acquisition unit 23 acquires the text inclusion external path including the blog text from the external path information generated by the external path information generation unit 21 for each of a plurality of blog pages of the same user.

The acquisition of the external common path and the text inclusion external path will be described with reference to FIGS.
Since the present invention is characterized in that the position of the blog text is specified by the tag path, when specifying the starting path, it is an external path common to each of a plurality of blog pages of the same user, and the corresponding tag It is necessary to identify an external path that can uniquely identify

  Therefore, the external common path acquisition unit 22 counts the number of appearances of each external path (hard path and soft path) included in the external path information for each of the plurality of blog pages. At this time, as shown in the count column of FIG. 5, it is preferable to count the number of appearances of the external path so that the number of digits does not overlap between the Web page 100 and the Web page 200. In this embodiment, the number of appearances of the external path of the web page 200 is multiplied by 100 and counted. That is, 100 in the count column means that the number of appearances is one. Then, the external common path acquisition unit 22 adds the number of appearances of the external path of the Web page 100 and the number of appearances of the external path of the Web page 200. From this addition result, the external common path acquisition unit 22 acquires an external common path, that is, an external path that is common to each of a plurality of blog pages of the same user and can uniquely identify the corresponding tag.

  The count field of FIG. 5 shows the number of appearances of the external path (hard path) of the Web page 100 and the number of appearances of the external path (hard path) of the Web page 200, and the count field of FIG. 6 shows the external path (hard path). The result of adding the number of appearances is shown. Further, the count field in FIG. 7A shows the number of appearances of the external path (soft path) of the Web page 100 and the number of appearances of the external path (soft path) of the Web page 200, and the count field in FIG. Shows the result of adding the number of appearances of external paths (soft paths).

  The addition result “001” indicates an external path that appears once on the Web page 100 but does not appear on the Web page 200, and the addition result “100” does not appear on the Web page 100, but the Web page 200. Shows an external path that appears once. The addition result “101” indicates an external path that appears once in the Web page 100 and the Web page 200, and the addition result “303” indicates an external path that appears three times in the Web page 100 and the Web page 200. Show.

  Here, the addition result “303” is an external path common to the Web page 100 and the Web page 200. Since this external path appears three times, the tag corresponding to this external path can be uniquely specified. Can not. Therefore, in the present embodiment, the external path of the addition result “101” is an external path that is common to each of a plurality of blog pages of the same user, and an external path (external common path) that can uniquely identify the corresponding tag. Determine as. In FIGS. 6 and 7 (2), the external common path is indicated by a dotted line.

  Further, the present invention is characterized in that the position of the blog text is specified by the tag position information (path), and therefore, the tag specified by the starting path needs to include the text.

  Therefore, the text inclusion external path acquisition unit 23 determines whether or not the tag specified by each of the external paths (hard path and soft path) included in the external path information includes a blog text. Whether or not the tag includes a blog text is determined based on RSS data and HTML data stored in the blog page DB 41 (FIG. 2) of the storage unit 4. That is, the text inclusion external path acquisition unit 23 specifies a tag including RSS data (summary of the text included in the description element) from the HTML data, and sets the external path including the tag to the external tag corresponding to the tag including the blog text. It is determined that the path.

  With reference to FIG. 3A, in the Web page 100, the blog text is included in the “center” tag of the line 16. Therefore, the external path including the “center” tag on the line 16 is determined as the external path corresponding to the tag including the blog text. Specifically, the external path (hard path) “tr [] [1] / td [] [1] / div [] [1]” is an external path including the “center” tag of the line 16; An external path corresponding to a tag containing the blog text. On the other hand, the external path (hard path) “tr [] [1] / td [] [0]” does not include the “center” tag of the line 16, and thus is an external path corresponding to a tag that does not include the blog text. . In the present embodiment, an external path including a blog text is set as a text inclusion external path. In FIG. 6 and FIG. 7 (2), the text inclusion external path is indicated by a two-dot chain line. In FIG. 7, there is an external path indicated by “△” in the text inclusion column. This is because some of the external paths that appear multiple times include the text, and the other part contains the text. Is not included.

Returning to FIG. 1, when the external common path and the text inclusion external path are acquired by the external common path acquisition means 22 and the text inclusion external path acquisition means 23, the column origin specifying means 24 is an external common path, and The deepest (ie, the lowest) external path is identified as the starting path among external paths that are text-included external paths in all of the plurality of Web pages.
For example, referring to FIG. 6, the hard path is an external common path (dotted line), and the text inclusion external path of the Web page 100 and the Web page 200 (the Web page 100 and the Web page out of the two-dot chain line) The external paths that are “circle” in 200) are the external paths (hard paths) “tr [] [1]” and “tr [] [1] / td [] [1]”. In these external paths (hard paths), “tr [] [1] / td [] [1]” is deeper by “td [] [1]”. On the other hand, as for the soft path, as shown in FIG. 7, there is no external path that is an external common path and is a text inclusion external path in all of a plurality of Web pages.

Therefore, the column starting point specifying unit 24 specifies the external path (hard path) “tr [] [1] / td [] [1]” as the starting point path. In the case of a soft path, if there is an external path that is an external common path and is a body-inclusive external path in all of a plurality of Web pages (hereinafter referred to as an “external path of a starting path candidate”), The deepest external path is specified as the origin path among the external paths of the path origin path candidates and the soft path origin path candidates. If the path depth is the same for the external path of the hard path starting path candidate and the external path of the soft path starting path candidate, in this embodiment, the external path of the soft path starting path candidate is used as the starting path. As specified. This is because the soft path has a simpler notation than the hard path and is suitable for managing the text positions of many blog pages. Of course, the external path of the hard path starting path candidate may be specified as the starting path.
When the column starting point specifying unit 24 specifies the starting point path, this starting point path is stored in the text position DB 42 (FIG. 2 (2)) of the storage unit 4. Note that “H ′” in FIG. 2 indicates a hard path, and “S ′” indicates a soft path.

  As described above, the first starting point specifying unit 2 specifies the starting point path for specifying the text positions of the Web page 100 and the Web page 200 as “H′tr [] [1] / td [] [1]”. . Here, referring to the source code 110 of the Web page 100 shown in FIG. 12, the tag specified by the starting path is the second of the “td” tags in the lower order of the “tr” tag of the second time (line 9). The “td” tag, ie, the “td” tag on line 13. Therefore, the origin path can specify that the text of the Web page 100 exists between the “td” tag on the line 13 and the “/ td” tag on the line 19.

By the way, from the “td” tag of the line 13 to “/ td” of the line 19, in addition to the body (line 16), an image (line 14) and Footer (line 18) are included. Here, images and Footer (for example, topics, comments (1), trackbacks, etc. in FIG. 12) are information that is not suitable for extracting blog keywords, and keywords including such information are extracted and indexed. If this is generated, the search accuracy may be reduced.
Therefore, in the present embodiment, the tag lower than the tag specified by the starting path specified by the first starting point specifying unit 2 is a path common to each of the plurality of blog pages, and the tag is uniquely identified by the path. The path that can be specified by the user and the path that includes the text is specified again.

<Second origin specifying unit 3>
Returning to FIG. 1, the second origin specifying unit 3 includes an internal path information generating unit 31, an internal common path acquiring unit 32, a text inclusion internal path acquiring unit 33, and a text position specifying unit 34. The

  The internal path information generating unit 31 generates internal path information indicating a path of a tag lower than the tag specified by the starting point specified by the column starting point specifying unit 24 for each of a plurality of blog pages of the same user. The internal path information generation unit 31 reads out the starting path stored in the text position DB 42 of the storage unit 4 and the HTML data stored in the blog page DB 41 of the storage unit 4 to generate internal path information.

  The generation of the internal path information is basically the same as the generation of the external path information, and thus detailed description thereof will be omitted. However, the internal path information generation unit 31 uses the internal path information shown in FIGS. 8 (1) to (4). (Hard path and Soft path) are generated.

The internal common path acquisition unit 32 is an internal common path that is common to each of a plurality of blog pages of the same user among the internal path information generated by the internal path information generation unit 31 and can uniquely identify a tag corresponding to the path. To get.
That is, the internal common path acquisition unit 32 acquires, as an internal common path, an internal path in which the addition result of the number of appearances of the internal path of the Web page 100 and the number of appearances of the internal path of the Web page 200 is “101”.
The text inclusion internal path acquisition means 33 acquires the text inclusion external path including the blog text among the internal path information generated by the internal path information generation means 31 for each of a plurality of blog pages of the same user.
FIG. 9 shows the addition result of the number of appearances of the internal path and the presence / absence of inclusion of the blog text in the internal path.

When the internal common path and the text inclusion internal path are acquired by the internal common path acquisition means 32 and the text inclusion internal path acquisition means 33, the text position specifying means is an internal common path and includes all of the plurality of Web pages. Among the internal paths that are the text inclusion internal paths, the deepest (that is, the lowest) internal path is specified as the text path.
For example, referring to FIG. 9 (2), the soft path is an internal common path (dotted line), and the internal path that is the text-containing internal path of the Web page 100 and the Web page 200 is the internal path (soft Path) “div [] / center []”. On the other hand, with respect to the hard path, there is no internal path that is an internal common path as shown in FIG. 9A and that is a text inclusion internal path in all of the plurality of Web pages.

  Therefore, the text position specifying unit 34 specifies the internal path (soft path) “div [] / center []” as the text path. When the text position specifying unit 34 specifies the text path, the text path is stored in the text position DB 42 (FIG. 2 (2)) of the storage unit 4.

<Blog text extraction means 5>
Returning to FIG. 1, the blog text extracting means 5 extracts the blog text from the user's blog page based on the origin path and the tag specified from the text path stored in the text position DB 42 of the storage unit 4. 4 is stored in the blog text DB 43.
For example, referring to FIG. 12, the blog text extracting means 5 determines the position of the blog text from the starting path “H′tr [] [1] / td [] [1]” from the “td” tag on the line 13. The line 19 is narrowed down to the “/ td” tag, and further narrowed down to the line 16 from the text path “S'div [] / center []”. As a result, the image of the line 14 and the footer of the line 18 included when the origin path is specified can be removed, and the blog text of the blog page of the same user can be extracted more accurately.

<Storage unit 4>
Returning to FIG. 1, the storage unit 4 includes a blog page DB 41, a text position DB 42, and a blog text DB 43.
As illustrated in FIG. 2B, the blog page DB 41 stores a user ID, a URL, and Web page information in association with each other. The Web page information includes HTML data and RSS data. The external path information generation unit 21 or the internal path information generation unit 31 generates external path information or internal path information from the HTML data. The text inclusion external path acquisition means 23 or the text inclusion internal path acquisition means 33 determines whether or not the blog text is included in the tag specified by the external path or the internal path from the RSS data.

As shown in FIG. 2 (3), the text position DB 42 stores the user ID, URL, and blog text position in association with each other. The blog text position includes the starting path specified by the column starting point specifying means 24 and the text path specified by the text position specifying means 34.
Although not shown, the blog text DB 43 stores the blog text extracted by the blog text extraction unit 5 in association with the user ID and URL. Keywords are extracted based on the blog text stored in the blog text DB 43, and an index for the blog page is generated.

[Hardware configuration of blog text identification device 1]
The hardware of the blog text specifying device 1 described above can be configured by a general computer. For example, a general computer includes a central processing unit (CPU) as a control unit, and a memory (RAM, ROM), a hard disk (HDD), and an optical disk (CD, DVD, etc.) as a storage unit, and a network communication device. Various wired and wireless LAN devices as display devices, for example, various displays such as liquid crystal displays and plasma displays, and as input devices, for example, keyboards and pointing devices (mouse, tracking ball, etc.) Are connected by a bus line. In such a general computer, the CPU controls the blog text specifying device 1 in an integrated manner, reads and executes various programs as appropriate, and cooperates with the hardware described above, thereby enabling various functions according to the present invention. Is realized.

[Process of Blog Text Specifying Device 1]
Next, processing of the blog text identification device 1 will be described with reference to FIG.

S1: First, the external path information generation unit 21 analyzes the HTML data stored in the blog page DB 41, and generates external path information indicating a tag path included in the source code by a hard path and a soft path.
S2: Subsequently, the external common path acquisition unit 22 is an external path that is common to each of a plurality of blog pages of the same user in the generated external path information and can uniquely identify the corresponding tag. Get the external common path.
S3: Also, the text inclusion external path acquisition unit 23 acquires the text inclusion external path including the blog text from the generated external path information for each of a plurality of blog pages of the same user.
S4: The column origin specifying means 24 identifies the deepest external path as the origin path among the external paths that are the external common paths acquired in S2 and are the text inclusion external paths acquired in S3. The specified starting path is stored in the text position DB 42.

S5: Subsequently, the internal path information generation unit 31 analyzes the HTML data stored in the blog page DB 41, and obtains internal path information indicating the path of the tag lower than the tag specified by the starting path specified in S4. Generate hard and soft paths.
S6: Next, the internal common path acquisition unit 32 is an internal path that is common to each of a plurality of blog pages of the same user in the generated internal path information, and can uniquely identify the corresponding tag. Get the internal common path.
S7: Also, the text inclusion internal path acquisition unit 33 acquires the text inclusion internal path including the blog text from the generated internal path information for each of the plurality of blog pages of the same user.
S8: Then, the text position specifying unit 34 specifies the deepest internal path as the text path among the internal common paths acquired in S6 and the text inclusion internal path acquired in S7. The identified text path is stored in the text position DB 42.

  S9: Subsequently, the blog text extracting means 5 extracts the blog text from the user's blog page based on the starting path and the tag specified from the text path stored in the text position DB 42, and the blog text in the storage unit 4 is extracted. Store in DB43.

Heretofore, the blog text specifying device 1 of the first embodiment has been described. According to the blog text specifying device 1 of the first embodiment, it is possible to generate a path for designating a tag including a blog text in a plurality of blog pages of the same user. As a result, in the user's blog page, the position of the blog body can be easily specified as long as the tag structure does not change greatly. Therefore, it is not necessary to specify the text of the body every time a new page is generated, and the control load is reduced. The position of the blog text can be specified.
At this time, in the blog text specifying device 1 of the first embodiment, since the second starting point specifying unit 3 further narrows down the text position narrowed down by the first starting point specifying unit 2, the blog text of the blog page of the same user is selected. It can be extracted more accurately.

[Second Embodiment]
Next, the blog text specifying device 1A of the second embodiment will be described with reference to FIG. The blog text specifying device 1 of the first embodiment narrows down the position of the blog text in two steps, whereas the blog text specifying device 1A of the second embodiment narrows down in a plurality of times. It is said.
FIG. 11 (1) shows the configuration of the blog text specifying device 1A, and FIG. 11 (2) shows the configuration of the text position DB 42A. In addition, about the structure same as the blog text identification apparatus 1 of 1st Embodiment, the same code | symbol is attached | subjected and detailed description is abbreviate | omitted.

[Configuration of Blog Text Identification Device 1A]
The blog text specifying device 1A includes a starting point specifying unit 2A, a storage unit 4A, and a blog text extracting unit 5.
The starting point specifying unit 2A includes a path information generating unit 21A, a common path acquiring unit 22A, a text inclusion path acquiring unit 23A, a column starting point specifying unit 24A, a specifiability determining unit 25A, and a text position specifying unit 26A. The storage unit 4A includes a blog page DB 41, a text position DB 42A, and a blog text DB 43.

The path information generation means 21A, the common path acquisition means 22A, the text inclusion path acquisition means 23A, and the column origin specifying means 24A are the external path information generation means 21 (internal path information generation means 31), the external common path acquisition means 22 (internal common The path acquisition means 32), the text inclusion external path acquisition means 23 (text inclusion internal path acquisition means 33), and the column origin specifying means 24 (text position specifying means 34) have the same functions.
That is, the path information generation unit 21A, the common path acquisition unit 22A, the text inclusion path acquisition unit 23A, and the column start point specification unit 24A include a plurality of the same users among the paths included in the path information generated by analyzing the blog page. The tag is uniquely specified (common path) and the deepest path is specified as the starting path even if the path includes the text (text inclusion path).

The identification availability determination unit 25A determines whether the column origin specifying unit 24A has identified the origin path. At this time, if the column starting point specifying unit 24A can specify the starting point path, the path information generating unit 21A further generates path information indicating a path of a tag lower than the tag specified by the starting point path, The common path acquisition means 22A and the text inclusion path acquisition means 23A acquire the common path and the text inclusion path from the path information of the lower tag, and the column origin specifying means 24A generates a new path based on the common path and the text inclusion path. Identify the origin path.
On the other hand, when the column starting point specifying unit 24A cannot specify the starting point path, the body position specifying unit 26A specifies the position of the blog body from the starting point path specified so far.

That is, the blog text specifying device 1A of the second embodiment repeatedly specifies a new starting path until a lower starting path cannot be specified. In other words, the path information generation means is common until a tag can be uniquely specified (common path) common to each of a plurality of blog pages of the same user, and a path including the text (text inclusion path) does not exist. The processing of 21A through the column origin specifying means 24A is repeated.
Therefore, as shown in FIG. 11 (2), the starting position path repeatedly specified as the blog text position is stored in the text position DB 42A.

  According to such a blog text specifying device 1A, a new starting path is specified from path information of lower tags until a new starting path cannot be specified. As a result, the position of the text can be specified in more detail, and since the position of the blog text can be easily specified once specified, the position of the blog text can be specified while suppressing the control load.

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

In the above embodiment, after generating path information (including external path information and internal path information) of the source code of the blog page, a common path (external common path) that is common to multiple pages and can uniquely identify the corresponding tag , Including an internal common path) and a text inclusion path including a blog text (including a text inclusion external path and a text inclusion internal path). Here, the order of acquiring the common path and the text inclusion path from the path information can be arbitrarily designed.
For all paths included in the path information, it is not necessary to determine whether the path is a common path or whether the path is a text inclusion path, and some paths included in the path information. Only this determination may be made. For example, it may be possible to determine whether or not to include the text only for the path with the appearance count addition result “101”, and to calculate the addition result of the appearance count only for the path including the text. It is good.

  Further, since the hard path always appears only once due to its nature, it is possible to calculate only the addition result of the soft path and determine whether the hard path is common to a plurality of blog pages.

In the present embodiment, among the tags included in the source code, tags that do not change the position used in many Web pages, such as the “html” tag, are omitted in the path generation. The tag type to be omitted can be arbitrarily designed. For example, in a blog page, even if the basic configuration is the same, it is possible to change the font to be used every day. Therefore, for example, a tag that does not affect the configuration of the blog page such as a “font” tag may be omitted in the path generation, and can be arbitrarily designed.
Further, the attributes constituting the path may be omitted in the generation of the path, and which attribute is omitted can be arbitrarily designed. For example, a blog page that changes the background of the blog text column every day can be considered. Therefore, for example, an attribute such as “bgcolor” that is an attribute of the “td” tag may be omitted in the path generation, and can be arbitrarily designed.

DESCRIPTION OF SYMBOLS 1 Blog text identification apparatus 2 1st origin specifying part 21 External path generation means 22 External common path acquisition means 23 Body inclusion external path acquisition means 24 Column origin specification means 3 2nd origin specification part 31 Internal path information generation means 32 Internal common path Acquisition means 33 Text inclusion internal path acquisition means 34 Text position specifying means 4 Storage unit 41 Blog page DB
42 Text position DB
43 Blog text DB
5 Blog text extraction means

Claims (4)

A blog page DB for storing a blog page including a blog text created by a user and summary information of the blog text in association with information for identifying the user;
External path information for analyzing a plurality of blog pages associated with the same user and generating external path information indicating a path in the source code of a tag included in the source code for each of the plurality of blog pages Generating means;
Of the external path information generated by the external path information generation means, an external common path acquisition means for acquiring an external common path common to each of the plurality of blog pages of the same user,
A text inclusion external path acquisition means for acquiring a text inclusion external path including a blog text among the external path information generated by the external path information generation means for each of a plurality of blog pages of the same user;
Column origin specifying means for specifying the deepest path as the origin path including the blog text among the paths that are the external common path and are the text inclusion external paths in all of the plurality of Web pages;
A blog text identification device comprising: Internal path information generating means for generating internal path information indicating a path of a tag lower than a tag specified by the starting path specified by the column starting point specifying means for each of a plurality of blog pages of the same user,
Of the internal path information generated by the internal path information generation means, an internal common path acquisition means for acquiring an internal common path common to each of the plurality of blog pages of the same user,
A text inclusion internal path acquisition means for acquiring a text inclusion internal path including the blog text among the internal path information generated by the internal path information generation means for each of a plurality of blog pages of the same user;
Text position specifying means for specifying the deepest path as the text path including the blog text among the internal common paths and the text inclusion internal paths in all of the plurality of Web pages;
The blog text identifying device according to claim 1, comprising: A blog page DB for storing a blog page including a blog text created by a user and summary information of the blog text in association with information for identifying the user;
Path information generation for analyzing a plurality of the blog pages associated with the same user and generating path information indicating paths in the source code of tags included in the source code for each of the plurality of blog pages Means,
Of the path information generated by the path information generation means, a common path acquisition means for acquiring a common path common to each of the plurality of blog pages of the same user;
A text inclusion path acquisition means for acquiring a text inclusion path including a blog text in the path information generated by the path information generation means for each of a plurality of blog pages of the same user;
Column origin specifying means for specifying a deepest path as a starting path among paths that are the common path and the text inclusion path in all of a plurality of Web pages;
Identifiability determining means for determining whether or not the sequence starting point specifying means has specified the starting point path;
With
The path information generation unit indicates a path of a tag lower than the tag specified by the starting path on condition that the specifying availability determination unit determines that the column starting point specifying unit has specified the starting path. Path information is further generated, the common path acquisition means and the text inclusion path acquisition means acquire the common path and the text inclusion path from the path information of the lower tag, and the column origin specifying means is the common origin identification means Based on the path and the text inclusion path, a new origin path is identified,
Text position specifying means for specifying starting path information as a text inclusion path including the blog text, on the condition that the identification availability determining means determines that the column starting point specifying means cannot specify the starting path;
A blog text specifying device further comprising: A blog text in which a computer including a blog page DB that stores a blog page including a blog text created by a user and summary information of the blog text in association with information for identifying the user identifies the text position of the blog page of the same user A specific method,
Analyzing a plurality of blog pages associated with the same user, and generating external path information indicating a path in the source code of a tag included in the source code for each of the plurality of blog pages;
Of the generated external path information, obtaining an external common path common to each of the plurality of blog pages of the same user;
Acquiring a text inclusion external path including a blog text in the generated external path information for each of the plurality of blog pages of the same user;
Specifying the deepest path as a starting path including the blog text among the paths that are the external common paths and are the text inclusion external paths in all of a plurality of Web pages;
Blog body identification method including

Images