KR20140038834A - Apparatus and method for analyzing web page - Google Patents

Abstract

The present invention relates to a web page analyzing apparatus and a web page analyzing method. According to the present invention, an information block for providing information required by a user is accurately identified from web pages which handle various topics without being constructed in a typical format, for example, from a web page for introducing persons. Further, the web pages are not simply divided into an information block and a non-information block, but divided into various types of blocks based on the content of each information block. [Reference numerals] (100) Web page analyzing apparatus; (101) Division unit; (102) Extraction unit; (103) Sorting unit

Description

[0001] APPARATUS AND METHOD FOR ANALYZING WEB PAGE [0002]

The present invention relates to an apparatus and method for analyzing web pages.

Most of the web pages on the web are providing mixed information of useful information and unnecessary data. It is necessary to identify only useful information from web pages, excluding unnecessary data such as banners, advertisements, navigation bars, menus, copyright marks, and provide them to users in various fields such as web mining, information extraction, It is becoming an important issue. Therefore, it is required to divide a web page into blocks and classify the blocks in the web page into information blocks and non-information blocks according to whether the main contents are included, and to identify only the information blocks in the web pages. When the information block is identified from the web page, the performance of web mining, information extraction and retrieval can be improved, and the accuracy of extracted and retrieved information can be improved.

A conventional information block identification method is a method for identifying a domain constituted by a content having a uniform format such as a news or an e-commerce domain and composed of a single consistent subject and considering the structural characteristics of a web page such as the structure and layout of the web page And identified information blocks containing the main content. For this reason, the conventional information block identification method has a problem that it does not have a formal format such as academic hompages and it is difficult to accurately identify an information block for a web page of a domain providing contents on various subjects.

An embodiment of the present invention aims at extracting information useful for a user accurately and quickly from a personal homepage having a non-standardized format such as a homepage of a web page, especially a researcher and dealing with various topics.

The apparatus for analyzing a web page according to an embodiment of the present invention includes: a division unit for dividing a web page into at least one block; Extracting at least one of words included in the block and word data related to the frequency of the word and part of speech data relating to the part of speech included in the block and the frequency of the part of speech; And a classifier for classifying the block based on at least one of the word data and the part of speech data.

The web page analysis method according to an embodiment of the present invention divides a web page into blocks and classifies the blocks based on the feature values extracted from the blocks. The method of analyzing a web page comprises: dividing the web page into at least one block; Extracting at least one of words included in the block and word data related to the frequency of the words and part of speech data relating to the part of words included in the block and frequency of the part of speech; And classifying the block based on at least one of the word data and the part-of-speech data.

The apparatus for creating a personal profile according to an embodiment of the present invention includes: a receiving unit for receiving a web page introducing a person; A web page analyzer for dividing the web page into blocks and analyzing the web page to classify the blocks; An information providing unit for providing information on a type of a block constituting the web page; And a profile creating unit for creating a profile of the person using the information of the type of the block and the contents of the block. The web page analyzing unit may include a division unit dividing the web page into at least one block, ; Extracting at least one of words included in the block and word data related to the frequency of the word and part of speech data relating to the part of speech included in the block and the frequency of the part of speech; And a classifier for classifying the block based on at least one of the word data and the part of speech data.

The method for analyzing a web page according to an embodiment of the present invention may be implemented by a computer-executable program and recorded on a computer-readable recording medium.

According to the present invention, information blocks can be correctly identified not only in a standard web page having a standardized structure and covering a single topic, but also in an informal web page configured in a free format and handling various subjects.

In addition, the present invention does not merely divide a block in a web page into an information block and a non-information block, but may classify the information block by a predetermined block type according to the content provided by the block.

As described above, the present invention accurately extracts and retrieves information requested by a user from a web page by accurately identifying the information block from the web page and classifying the information block according to the content in various ways Can be obtained.

1 is a block diagram illustrating a web page analyzer according to an embodiment of the present invention.
FIG. 2 shows a news web page constructed in a formatted format and dealing with a single topic divided into blocks.
FIG. 3 shows a homepage of a researcher who is structured in an informal format and deals with an unconventional subject, which is divided into blocks.
4 illustrates a web page divided into a plurality of blocks according to an embodiment of the present invention.
5 shows an example of word data extracted from the block of the web page shown in FIG.
FIG. 6 is a graph showing the parts-of-speech distribution of words extracted from the block of the web page shown in FIG.
7 is a flowchart illustrating a web page analysis method according to an embodiment of the present invention.
8 is a flowchart illustrating a process of extracting word data from a block according to an embodiment of the present invention.
9 is a flowchart illustrating a process of extracting parts of speech data from a block according to an embodiment of the present invention.
10 is a flowchart illustrating a process of classifying blocks by type according to an embodiment of the present invention.
11 is a block diagram illustrating an apparatus for creating a user profile according to an embodiment of the present invention.

Other advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail with reference to the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims.

Unless defined otherwise, all terms (including technical or scientific terms) used herein have the same meaning as commonly accepted by the generic art in the prior art to which this invention belongs. Terms defined by generic dictionaries may be interpreted to have the same meaning as in the related art and / or in the text of this application, and may be conceptualized or overly formalized, even if not expressly defined herein I will not.

The terminology used herein is for the purpose of illustrating embodiments and is not intended to be limiting of the present invention. In the present specification, the singular form includes plural forms unless otherwise specified in the specification. As used herein, the terms' comprise 'and / or various forms of use of the verb include, for example,' including, '' including, '' including, '' including, Steps, operations, and / or elements do not preclude the presence or addition of one or more other compositions, components, components, steps, operations, and / or components. The term 'and / or' as used herein refers to each of the listed configurations or various combinations thereof.

It should be noted that the terms such as '~', '~ period', '~ block', 'module', etc. used in the entire specification may mean a unit for processing at least one function or operation. For example, a hardware component, such as a software, FPGA, or ASIC. However, '~ part', '~ period', '~ block', '~ module' are not meant to be limited to software or hardware. Modules may be configured to be addressable storage media and may be configured to play one or more processors. ≪ RTI ID = 0.0 >

Thus, by way of example, the terms 'to', 'to', 'to block', 'to module' refer to components such as software components, object oriented software components, class components and task components Microcode, circuitry, data, databases, data structures, tables, arrays, and the like, as well as components, Variables. The functions provided in the components and in the sections ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ' , '~', '~', '~', '~', And '~' modules with additional components.

1 is a block diagram illustrating a web page analysis apparatus 100 according to an embodiment of the present invention. The web page analyzing apparatus 100 may divide the web page into blocks and classify the blocks by type based on feature values extracted from the contents of the blocks. The feature value is a value representing a characteristic of the content provided by the block, and the web page analyzing apparatus 100 may classify the block in the web page into a predefined block type based on the feature value.

As shown in FIG. 1, the web page analyzing apparatus 100 may include a dividing unit 101, an extracting unit 102, and a classifying unit 103. The division unit 101 may divide a web page into at least one block. The extraction unit 102 may extract at least one of words included in the block, word data related to the frequency of the word, and part of speech data related to the frequency of the words included in the block. The classifying unit 103 may classify the block based on at least one of the word data and the part-of-speech data.

FIG. 2 illustrates a web page divided into blocks according to an embodiment of the present invention. The news web page 20 shown in Fig. 2 is a web page configured in a formatted format and handling a single topic. The news web page 20 is generally configured in such a format as to display a company logo (CNN) at the top, a menu at the left, a copyright information at the bottom, and a main content such as an article at the center do. Therefore, the information block 201 and the non-information block 202 can be relatively clearly distinguished by using structural features such as the size and position of the block in the web page 20.

FIG. 3 shows a homepage of a researcher divided into blocks according to an embodiment of the present invention. Unlike the news web page 20, the homepages 31, 32, 33, and 34 of the researcher are configured in unformatted formats as shown in FIGS. 3A to 3D, This is a typical example of a non-compliant web page that does not provide it. The researchers 'homepages 31 and 32 shown in FIGS. 3A and 3B display the names of the researchers in the upper blocks 311 and 321 of the web page. However, the researchers' homepages 33 and 32 shown in FIGS. 34 are displayed on the portions 333, 342 other than the upper block of the web page.

In addition, the photograph of the researcher is also posted on the upper right 312 of the researcher's homepage 31 of FIG. 3A and FIG. 3C, but posted on the second block 322 of the researcher's homepage 32 of FIG. And is posted on the left upper end 342 in the researcher homepage 34. [ It is difficult to accurately identify the information block by only the structural characteristics of the web page such as the size and the position of the block of the researchers' homepages (31, 32, 33, 34).

Accordingly, the web page analyzing apparatus 100 according to an embodiment of the present invention divides a web page into blocks, and then analyzes the contents provided by the block to extract feature values indicating the characteristics of the blocks. Then, the web page analyzing apparatus 100 classifies the blocks in the web page by predefined block types based on the feature values.

The division unit 101 divides the web page into at least one block. Tags such as <Table>, <UL>, and <P> may be used for structuring the content in the web page in order to increase the readability of the web page, and the division unit 101 may use the DOM structure based on this tag A web page can be divided into one or more blocks. In addition to the DOM structure, the dividing unit 101 may detect a block by further considering visual cues of a web page such as the font, color, and block size of the contents in the web page .

FIG. 4 illustrates an example of a web page divided into a plurality of blocks according to an embodiment of the present invention. 4, the division unit 101 divides the web page 40 into a first block 401, a second block 402, a third block 403, and a fourth block 403 according to a visual characteristic. (404). According to an embodiment of the present invention, the division unit 101 may divide the web page 40 into blocks using a Vision-based Page Segmentation (VIPS) algorithm. The VIPS algorithm is a page segmentation algorithm using a vision technique for dividing a web page into blocks.

The extracting unit 102 extracts at least one of words included in the block and word data related to the frequency of the word and part of speech data related to the frequency of the part of speech included in the block from the block have. In other words, the extraction unit 102 may extract feature values indicating the characteristics of the block from the contents provided by the blocks 401, 402, 403, and 404, and the feature values may include the word data, &Lt; / RTI &gt; or both.

5 shows an example of word data extracted from the blocks 401, 402, 403, and 404 of the web page 40 shown in FIG. As shown in FIG. 5, the extraction unit 102 may extract the words included in the blocks 401, 402, 403, and 404, and may calculate the occurrence frequency of each word for each block.

According to an embodiment of the present invention, the extracting unit 102 may detect a keyword having a high frequency of appearance among the words included in the block, and use a word vector (term and a word vector generator for generating a vector vector. For example, the word vector generation unit may select a predetermined number of words in order of frequency of occurrence in a block, and may select the word as a keyword. Then, the word vector generation unit may generate a word vector having the occurrence frequency of the keyword as a component.

According to an embodiment of the present invention, the word vector generation unit may perform a stemming operation to remove a stop word among the words included in the block before detecting the keyword. Then, the word vector generator may assign a term frequency-inverted document frequency (TF-IDF) weight to the frequency of each word to form a word vector.

According to an embodiment of the present invention, the extracting unit 102 determines a part of speech included in the block, and calculates a part-of-speech distribution that calculates a ratio of the frequency of each part of speech to the number of words included in the block And a calculation unit. The part-of-speech distribution calculator may determine the part-of-speech word as one of ordinary nouns, proper nouns, pronouns, prepositions, conjunctions, verbs, investigations, adverbs, adjectives, surveys, observations and exclamations, May be variously set according to the embodiment. For example, the part-of-speech distribution calculating unit may determine the part-of-speech distribution calculating unit to be one of plural nouns, singular verbs, foreign language, sign, and possessive endings in addition to the above-mentioned parts-of-speech.

6A shows an example of parts of speech data extracted from the blocks 401, 402, 403, and 404 of the web page 40 shown in FIG. As shown in FIG. 6A, the parts-of-speech distribution calculating unit may perform part-of-speech tagging on words included in each block, and calculate the distribution of parts-of-speech for each block to utilize as a feature value of the block. 6A, NN is a noun, NNP is a proper noun, NNS is a plural noun, IN is a preposition, CC is a conjunction, VBP is a singular verb, FW is a foreign language, CD is an investigation, SYM is a sign, RB is an adverb, , POS stands for the possessive term.

Then, the part-of-speech distribution calculating unit may calculate the ratio of the frequency of each part of speech to the number of words included in each block. FIG. 6B shows a part-of-speech distribution graph of each block calculated from the part-of-speech data of FIG. 6A. 6B, the first block 401 has a high percentage of proper nouns, the third block 403 has a high ratio of proper nouns and rhetoric, and the fourth block 404 has a high ratio of proper nouns and rhetoric And it is composed of various other parts companies.

The classifier 103 may classify the blocks by type based on at least one of the word data and the part-of-speech data extracted by the extractor 102. According to an embodiment of the present invention, the web page analyzed by the web page analyzing apparatus 100 may be a web page for introducing a person, for example, a researcher. In this case, the classifying unit 103 may classify the block into one of the following blocks:

(i) a basic block containing information about a person's position or position;

(ii) a contact block containing information about the contact of the person;

(iii) a photo block containing a photographic image of the person;

(iv) a text block containing information about the article written by the person; And

(v) a blend block that contains more than one of information about a person's position or position, information about a contact of a person, a photographic image of a person, and information about a person's writing.

According to an embodiment of the present invention, the classifying unit 103 may classify the block into a non-information block if the block does not correspond to any of the above-mentioned five blocks.

The types of blocks may be defined in advance by analyzing the content in the plurality of sample homepages prior to the analysis of the web page. For example, the researcher's homepage can provide profile information related to the researcher's personal information and research information related to the research field. Attributes belonging to profile information and research information are shown in Table 1 below.

Information property Profile Information photo, position, affiliation, phone, fax, email, address, bsuniv, bsmajor, bsdate, msuniv, msmajor, msdate, phduniv, phdmajor, phddate Research Information research area / interesting, publication list, research interest (implicit)

Where bsuniv is the undergraduate university, bsmajor is the undergraduate major, bsdate is the bachelor's degree, msuniv is the master's degree university, msmajor is the master's degree major, msdate is the master's degree date, phduniv is the doctoral degree university, phdmajor is the doctoral degree, . In addition, research area / interesting means the field of research / interest explicitly stated on the researcher homepage, and research interesting (implicit) means the research field / field of interest implied on the researcher's homepage.

The information provided by a specific domain can be classified into essential information and optional information. The essential information means information that a web page in a domain must necessarily include, and the selection information means information that is selectively included as needed. According to an embodiment of the present invention, the occurrence frequency of each property corresponding to the profile information and the research information is analyzed in an arbitrary number of homepages, and the attribute found in a predetermined ratio or more (for example, 70% or more) I can decide with my essential information. In addition, the block types corresponding to the above six categories may be defined in consideration of the association between the attributes.

According to an embodiment of the present invention, the classifying unit 103 may classify word data or part-of-speech data of blocks into reference word data or reference part-of-speech data set for each of the basic block, contact block, photo block, Can be compared. If the word data is matched with the reference word data or the part-of-speech data is matched with the reference part-of-speech data, the classifying unit 103 classifies the block into blocks corresponding to the matched reference word data or the reference part-of-speech data can do.

The reference word data or reference word data set for each of the basic block, the contact block, the photo block, the text block, and the mixed block may be stored in a storage unit such as a memory and read out when the classification process is executed. The reference word data and the reference part-of-speech data are data representing the characteristics of the above-described block type. For example, a contact block contains contacts such as a person's phone number, fax number, e-mail address, and work address, so you can select "Tel", "Phone", "Fax", "Mobile" May be included as key words. Therefore, in the case of the contact block, a word vector having the expected frequency of the keyword in the block as a component may be set as the reference word data. Further, since the contact block includes various telephone numbers of the person, the ratio of the investigation can be relatively high as compared with other blocks. Therefore, the reference part data can be set for the contact block so that the occurrence rate of the block is 30% or more so that the contact block is classified as the contact block, and the occurrence ratio threshold for the investigation is 30%.

According to an embodiment of the present invention, the reference word data or reference part data may be constructed using a Naive Bayes classifier or a SVM (Support Vector Machine) using a normalized polynomial kernel.

7 is a flowchart illustrating a web page analysis method according to an embodiment of the present invention. The web page analysis method 70 may divide a web page into blocks, and classify the blocks based on the feature values extracted from the blocks. Here, the feature value may include at least one of words included in the block, word data related to the frequency of the words, and parts of speech data related to the part of words included in the block and the frequency.

As shown in FIG. 7, the web page analysis method 70 according to an embodiment of the present invention includes dividing a web page into at least one block (S71), extracting words included in the block and the frequency (S72) of extracting at least one of word data relating to a word and word data related to a word included in the block, and part of speech data relating to the frequency of the word included in the block, and classifying the block based on at least one of the word data and the parts of speech data Step S73.

According to an embodiment of the present invention, the step S71 of dividing the web page into blocks may include displaying a plurality of web pages reflecting a visual characteristic such as a font, a font color, and a block size together with a DOM structure of the web page Into blocks of &lt; / RTI &gt;

8 is a flowchart illustrating a process of extracting word data from a block according to an embodiment of the present invention. As shown in FIG. 8, the step S72 of extracting word data from the block includes extracting words included in the block S721, removing an insoluble word among the words included in the block S722, A step S723 of detecting a keyword having a high appearance frequency among the words included in the block, and a step S724 of generating a word vector having the frequency of the keyword as a vector component.

For example, the extracting step S72 removes words (for example, articles, prepositions, surveys, connections, etc.) that are not used as search terms in the information retrieval among the words included in the block, A predetermined number of words (for example, the top four words) can be selected as key words in the order of appearance frequency of the words. Then, a word vector can be generated using the frequency of the word selected as the keyword as a vector component.

9 is a flowchart illustrating a process of extracting parts of speech data from a block according to an embodiment of the present invention. 9, the extracting of part-of-speech data (S72) includes extracting words included in a block (S721), determining parts of speech included in the block (S722), and And calculating a ratio of parts-of-speech frequency to the number of words included in the block (S723). The part of the word included in the block may be any one of a noun, proper noun, pronoun, preposition, conjunction, verb, investigation, adverb, adjective, survey, adjective, and exclamation, Can be variously set according to the example.

According to an embodiment of the present invention, the web page may be a homepage for introducing a person, for example, a homepage of a researcher. In this case, the step of classifying the block may include: classifying the block into one of the basic block, the contact block, the photo block, the text block, and the mixed block; And classifying the block into a non-information block if the block does not correspond to any of the five blocks.

10 is a flow diagram illustrating the steps of classifying blocks in accordance with one embodiment of the present invention. As shown in FIG. 10, in the step S73 of classifying the block, the word data or the part-of-speech data of the block may be set to reference word data or reference part-of-speech data set for each basic block, contact block, photo block, text block, and mixed block. (S731), when the word data or the part-of-speech data match the reference word data or the reference part-of-speech data of a specific block (YES in S732), the block corresponding to the matched reference word data or the reference part-of-speech data. And classifying the block by the type of S733.

According to an embodiment of the present invention, the step S73 of classifying the block may include a step of classifying the block if the word data or the part-of-speech data of the block does not match any of the reference word data or the reference parts data (NO in step S732) And sorting the block into non-information blocks (S734).

11 is a block diagram illustrating a person profile creating apparatus according to an embodiment of the present invention. The person profile creating apparatus 200 can extract various information about a person from a homepage of a specific person, create a profile of the person, and provide the profile to the user. 11, a person profile creating apparatus 200 according to an embodiment of the present invention includes a receiving unit 201, a web page analyzing unit 202, an information providing unit 203, and a profile creating unit 204 ).

The receiving unit 201 can receive a web page introducing a person. The web page introducing the person may be the homepage of the researcher as shown in Figs. 3 and 4.

According to an embodiment of the present invention, the receiving unit 201 may receive code data representing the web page. For example, the receiving unit 201 may receive HTML code or XML code of the web page. According to another embodiment of the present invention, the receiving unit 201 may receive the URL address of the web page instead of the code of the web page, and may receive the HTML code or the XML code of the web page by accessing the URL address .

The web page analyzing unit 202 may analyze the web page to divide the web page into blocks and classify the blocks. The web page analyzing unit 202 may include the web page analyzing apparatus 100 described with reference to FIG. That is, the web page analyzing unit 202 includes a dividing unit 2021 for dividing a web page into at least one block, word data included in the block and word data related to the frequency, An extracting unit 2022 for extracting at least one of part of speech data and part-of-speech data about the frequency, and a classifying unit 2023 for classifying a block based on at least one of word data and part-of-speech data.

The information providing unit 203 may provide information on the type of blocks constituting the web page. According to an embodiment of the present invention, the type of the block includes: (i) a basic block including information on the position or position of the person; (ii) a contact block containing information about the contact of the person; (iii) a photo block containing a photographic image of the person; (iv) a text block containing information about the article written by the person; (v) a mixing block including at least two of information on a person's position or position, information on a contact of a person, a photographic image of a person, and information on a manuscript; And (vi) non-information blocks that do not contain any of the above information. However, the type of the block is not limited thereto and may be variously defined according to embodiments.

The profile creator 204 may create a profile of a person based on the information about the type of the block and the contents of the block. The profile of the person may include at least one of the name, affiliation, position, contact, and photograph of the person, and may further include items such as a person's history or an article list depending on the embodiment.

The person profile creating apparatus 200 may further include a storage unit 205. The storage unit 205 may store various data used for web page analysis. For example, reference word data or reference part-of-speech data for a predefined block type may be stored. The classifying unit 2023 reads the reference data from the storage unit 205, compares the word data or the part-of-speech data of the blocks extracted by the extracting unit 2022, and compares the blocks of the web page for each predefined block type. Can be classified.

The web page analysis method 70 according to an embodiment of the present invention may be stored in a computer-readable recording medium that is manufactured as a program to be executed in a computer. The computer-readable recording medium includes all kinds of storage devices in which data that can be read by a computer system is stored. Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like.

In the above, the web page is divided into a plurality of blocks, and feature values representing the characteristics of the blocks, such as word vectors and word parts of speech distribution, are extracted from the contents of each block, and the blocks are classified by type based on the feature values. An apparatus for creating a person profile using a web page analyzing apparatus, a web page analyzing method, and a web page analyzing apparatus according to an embodiment of the present invention has been described. According to the present invention, it is possible to accurately identify an information block that provides information requested by a user from a web page dealing with various topics, for example, a homepage introducing a person, without being configured in a formal format. Furthermore, it is possible to obtain an effect of not only dividing the web page into information blocks and non-information blocks, but also dividing the information block into various types according to its contents.

Further, when analyzing the homepage of a person using the web page analyzing apparatus and method according to the present invention, it is possible to extract information used for profile creation of the person more accurately and quickly.

100: Web page analyzing apparatus 101:
102: extracting unit 103:
200: person profile creating apparatus 201: receiving section
202: Web page analysis unit 203: Information providing unit
204: profile creation unit 205: storage unit

Claims (20)

A division unit for dividing the web page into at least one block;
Extracting at least one of words included in the block and word data related to the frequency of the word and part of speech data relating to the part of speech included in the block and the frequency of the part of speech; And
A classification unit classifying the block based on at least one of the word data and the part-of-speech data;
The web page analyzing apparatus comprising: The method according to claim 1,
Wherein the extracting unit comprises:
And a word vector generator for detecting a keyword having a high occurrence frequency among words included in the block, and generating a word vector having the frequency of the keyword as a component of the vector. 3. The method of claim 2,
And the word vector generator is configured to remove stop words from words included in the block before detecting the key word. The method according to claim 1,
Wherein the extracting unit comprises:
And a part-of-speech distribution calculating unit for determining a part of words included in the block and calculating a ratio of parts-of-speech frequency to the number of words included in the block. 5. The method of claim 4,
Wherein the part-of-speech distribution calculating unit determines the part-of-speech word as one of a normal noun, a proper noun, a pronoun, a preposition, a conjunction, a verb, an investigation, an adverb, an adjective, an investigation, an adjective and an admirable. The method according to claim 1,
Wherein the web page includes a homepage for introducing a person. The method according to claim 6,
Wherein the classifying unit comprises:
A basic block containing information about at least one of a person's position and position;
A contact block containing information about the contact of the person;
A photo block containing a photo image of a person;
A text block containing information about the article written by the person; And
A mixing block including at least two of information on the person's position or position, information on the contact of the person, a photograph image of the person, and information on the article written by the person;
The web page analyzing apparatus comprising: 8. The method of claim 7,
Wherein the classifying unit classifies the block into a non-information block if the block does not correspond to any one of the basic block, the contact block, the photo block, the text block, and the mixed block. 8. The method of claim 7,
The classification unit compares the word data or the part-of-speech data of the block with reference word data or reference part-of-speech data set for each of the basic block, the contact block, the photo block, the writing block, and the mixed block. And classifying the block into the matched reference word data or blocks corresponding to the reference part-of-speech data when the reference word data matches or the part-of-speech data matches the reference part-of-speech data. A web page analysis method comprising dividing a web page into blocks and classifying the blocks based on feature values extracted from the blocks. 11. The method of claim 10,
The web page analysis method is:
Dividing the web page into at least one block;
Extracting at least one of a word included in the block and word data relating to a frequency of the word, and a part of speech included in the block and part of speech data relating to a frequency of the part of speech; And
Classifying the block based on at least one of the word data and the part-of-speech data;
/ RTI &gt; 12. The method of claim 11,
Extracting at least one of the word data and the part of speech data may include:
Removing stop words from words included in the block;
Detecting a key word having a high frequency of appearance among words included in the block; And
Generating a word vector including the frequency of the key word as a component of a vector;
/ RTI &gt; 12. The method of claim 11,
Extracting at least one of the word data and the part of speech data may include:
Determining a part of the word included in the block; And
Calculating a ratio of parts-of-speech frequency to the number of words included in the block;
/ RTI &gt; 12. The method of claim 11,
Wherein the web page includes a homepage for introducing a person. 15. The method of claim 14,
Wherein classifying the blocks comprises:
(i) the basic block including information on the person's position or position; A contact block containing information about the contact of the person; A photo block containing a photo image of a person; A text block containing information about the article written by the person; And a mixing block including at least two of information on the person's position or position, information on the contact of the person, photo image of the person, and information on the article written by the person. And
(ii) classifying the block into a non-information block if the block does not correspond to any one of the basic block, the contact block, the photo block, the text block, and the mixed block;
/ RTI &gt; 16. The method of claim 15,
Wherein step (i) comprises:
Comparing word data or part-of-speech data of the block with reference word data or reference part-of-speech data set for each of the basic block, the contact block, the picture block, the text block, and the mixed block;
Detecting reference word data or reference parts data matching the word data or the part-of-speech data; And
Classifying the block into blocks corresponding to the matched reference word data or reference parts data;
/ RTI &gt; A receiving unit for receiving a web page introducing a person;
A web page analyzer for dividing the web page into blocks and analyzing the web page to classify the blocks;
An information provider for providing information on the type of blocks constituting the web page; And
A profile creation unit which creates a profile of the person based on information on the type of the block and the contents of the block;
Wherein the web page analyzing unit comprises:
A division unit for dividing the web page into at least one block;
Extracting at least one of words included in the block and word data related to the frequency of the word and part of speech data relating to the part of speech included in the block and the frequency of the part of speech; And
A classification unit classifying the block based on at least one of the word data and the part-of-speech data;
And a person profile creation device. The method of claim 17,
And the receiving unit receives the HTML code or the XML code of the web page. The method of claim 17,
Wherein the receiving unit receives the URL address of the web page and accesses the URL address to receive the HTML code or the XML code of the web page. A computer-readable recording medium,
Dividing a web page into at least one block;
Extracting at least one of a word included in the block and word data relating to a frequency of the word, and a part of speech included in the block and part of speech data relating to a frequency of the part of speech; And
Classifying the block based on at least one of the word data and the part-of-speech data;
And a program code for executing the web page analysis process.

Images