JP4940399B2 - Advertisement distribution apparatus and program - Google Patents

Description

  The present invention relates to a technique for distributing an advertisement related to the content of a web page to a web page provided on a network.

In recent years, advertisements are widely displayed on web pages browsed via the Internet. Conventionally, a content matching technique is known as a technique for distributing advertisements on the Internet. In this technique, an advertiser designates a keyword for his / her advertisement, and the advertisement is distributed to a website that matches the keyword (see, for example, Patent Document 1).
JP 2003-187145 A

  However, in the content matching advertisement distribution, only the keyword of the web page is referred to, so the advertisement of the product is distributed to a web page that makes a negative comment about a certain product, or the word as a keyword May be delivered to a web page that describes a completely different article or service to be advertised.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide a technique that enables distribution of an appropriate advertisement in accordance with the content and context of a web page.

  An aspect of the present invention determines a category in which a media web page should be classified from a plurality of categories set in advance based on text information of the media web page provided on a network, Advertising information that matches the category of the medium web page from the advertisement information that specifies the category of the web page to be displayed and displays the advertisement information together with the medium web page to the terminal that browses the medium web page Is an advertisement distribution device that transmits.

  Here, “medium web page” refers to a web page to which advertisement information is distributed. “Category” refers to grouping for classifying web pages according to specific criteria. “Advertisement information” includes simply textual information about ads for specific products and services, link information to sites that offer specific products and services, images, videos, or sounds of specific products and services, and these Any combination of a plurality of these is included.

  According to this aspect, since the advertisement information in which the category that matches the category of the media web page is specified by the advertiser is selected and displayed together with the media web page, the content in accordance with the content and context of the media web page is displayed. Advertising can be delivered.

  Another aspect of the present invention is also an advertisement distribution device. This device displays an advertisement on a medium web page provided on a network, and holds advertisement information in association with category designation information that designates a category condition of the medium web page to which the advertisement is distributed. A media web page is acquired via an information holding unit and a network, and a category in which the media web page is to be classified is determined from a plurality of categories set in advance based on text information of the media web page. A page analysis unit, a matching unit that selects from the advertisement information holding unit advertisement information whose category of the media web page determined by the page analysis unit satisfies the category condition, and a media web page on a terminal connected to the network. When viewed, the advertisement information selected by the matching unit for the terminal Comprising an advertisement transmitting section for transmitting a form to be displayed with the page, the.

  According to this aspect, the category to be classified is determined based on the text information of the medium web page, and the advertisement information that satisfies the category condition set in advance in association with the advertisement information is selected and distributed. Therefore, it is possible to distribute advertisements with contents that match the contents and context of the medium web page without having to report and classify the category of the medium web page in advance.

  The advertisement information holding unit holds advertisement information in association with category designation information for designating whether or not to display an advertisement on a medium web page classified into the category for each of a plurality of preset categories. Good. According to this, it is possible to individually set whether or not the advertiser displays his / her advertisement information for each category of the medium web page.

  The page analysis unit may further include a sentence classification unit that calculates an attribution probability representing a probability that the medium web page belongs to a plurality of preset categories. The advertisement information holding unit holds the advertisement information in association with the category designation information for each category specifying the lower limit of the attribution probability of the medium web page to which the advertisement is distributed, and the matching unit When the attribution probability of the medium web page is equal to or higher than the lower limit value by comparing with the lower limit value of the attribution probability related to the advertisement information, the corresponding advertisement information may be selected from the advertisement information holding unit.

  “Attribution probability” is a probability that a web page should be classified into a certain category, and can be calculated for each category. According to this, by specifying the lower limit value of the attribution probability for each category, it becomes possible to select medium web pages having contents that span a plurality of categories as advertisement distribution destinations.

  The advertisement information holding unit holds the advertisement information in association with the category designation information for designating the non-display category that should avoid the transmission of the advertisement. May be refused. According to this, the advertiser can individually specify the category of the medium web page that the advertisement is not desired to be displayed regardless of the attribution probability calculated by the sentence classification unit.

  A page management unit for transmitting the script program to the administrator's terminal of the medium web page in order to describe the script program for displaying the advertisement information transmitted from the advertisement transmission unit on the medium web page. You may prepare. The script program may have a function of transmitting information on the medium web page to the page analysis unit when the medium web page is viewed on a terminal connected to the network. The page analysis unit may acquire the medium web page according to information transmitted from the script program. According to this, for the first time when the medium web page is browsed, the web page is acquired by the advertisement distribution device, and then the category analysis of the medium web page is executed by the page analysis unit. Therefore, the advertisement distribution apparatus does not need to crawl the medium web page in advance, and the processing load on the advertisement distribution apparatus can be suppressed.

  The advertisement information holding unit holds the advertisement information in association with the distribution destination keyword that specifies the keyword of the medium web page that is the distribution destination of the advertisement. The page analysis unit includes a keyword extraction unit that extracts a keyword set from the medium web page. Further, it may be provided. The matching unit selects, from the advertisement information holding unit, the advertisement information in which the delivery destination keyword included in the keyword set of the medium web page is specified, and then the category of the web page satisfies the category condition from the selected advertisement information. You may select the advertising information you are doing. According to this, both the keyword and the category are specified as the display destination of the advertisement information, and the advertisement information is displayed only for the medium web page that satisfies both conditions. The contents of the page can be specified in more detail.

  Yet another embodiment of the present invention is also an advertisement distribution device. This device displays an advertisement on a medium web page provided on a network, and holds advertisement information in association with category designation information that designates a category condition of the medium web page to which the advertisement is distributed. Connected to the network, an information holding unit, a matching unit that receives a category into which the media web page to which the advertisement information is distributed is classified, and selects the advertisement information for which the category satisfies the category condition from the advertisement information holding unit An advertisement transmission unit that transmits the advertisement information selected by the matching unit together with the medium web page when the medium web page is browsed on the terminal.

  It should be noted that any combination of the above-described components and a representation of the present invention by a method, apparatus, system, recording medium, and computer program are also effective as an aspect of the present invention.

  According to the present invention, the category of the media web page is determined, and the advertisement information that matches the category condition is selected and displayed together with the media web page. Can be delivered.

  One embodiment of the present invention is an advertisement distribution apparatus that distributes advertisements to web pages such as blogs provided on a network.

  FIG. 1 shows an example of a usage pattern of an advertisement distribution apparatus 210 according to the present embodiment. The advertisement distribution apparatus 210 is connected to the general user terminal 206, the advertiser terminal 202, and the page manager terminal 204 via the network 208. The advertiser terminal 202 is a terminal operated by an advertiser who wants to distribute an advertisement to a web page, and the page manager terminal 204 is a web page management that permits distribution of advertisement information to a web page managed by the advertiser terminal 202. Is a terminal operated by a person. In this specification, a web page to which advertisement information is distributed is particularly referred to as a “medium web page”. The general user terminal 206 is a terminal used by a user who browses a medium web page other than the advertiser terminal 202 and the page manager terminal 204.

  The “advertising information” in the present embodiment is displayed on the medium web page in a state associated with the URL (Uniform Resource Locator) of the advertiser site, and when the general user selects the advertisement, it is generally displayed on the advertiser site. It has a role to invite users. Preferably, the advertisement information is text, an image, or a moving image displayed in a clickable form. However, the method of inviting general users to the advertiser's site is not limited to these. Even if the advertiser's URL is displayed together with the advertisement information, or a mailer addressed to the advertiser's email address is activated. Good. Further, when the advertisement information is clicked, it may be connected to the advertiser's site via a proxy server.

  The page manager operates his / her terminal 204 to store the created web page 214 in the database of the provider server 212. When the provider server 212 specifies a URL from the general user terminal 206 that accesses the medium web page and receives a request for the web page 214, the provider server 212 selects the web page 214 corresponding to the URL and transmits it to the general user terminal 206. To do. The browser of the general user terminal 206 interprets the web page 214 and displays it on the display.

  The advertisement distribution device 210 determines to which medium web page the pre-registered advertisement information is distributed. Further, the advertisement distribution apparatus 210 may count the number of times each piece of advertisement information is displayed on the medium page, the number of times the advertisement is clicked, and the advertiser's site is displayed.

  The advertisement distribution device 210 includes a sentence classification unit 100. The sentence classification unit 100 acquires a medium web page specified by a URL and classifies the medium web page into any of a plurality of preset categories.

  The advertiser designates a delivery destination keyword that specifies a keyword of a medium web page that is desired as a delivery destination of advertisement information, and a category condition that designates a category of the media web page that is desired as a delivery destination. The delivery destination keyword designated by the advertiser may be one word or a plurality of words. The advertisement distribution apparatus 210 determines a set of keywords included in the medium web page and a category of the medium web page for the medium web page that has been registered in advance for distribution of the advertisement. The advertisement distribution device 210 compares the keyword set and category of the medium web page determined by the sentence classification unit 100 with the distribution destination keyword and category condition specified for the advertisement information. Then, the advertisement distribution apparatus 210 responds when the keyword set included in the medium web page includes the distribution destination keyword specified by the advertiser, and the category of the medium web page satisfies the category condition specified by the advertiser. The advertisement information is transmitted to the general user terminal 206.

  Hereinafter, the function and configuration of the sentence classification unit 100 that determines a category in which the medium web page should be classified will be described in detail. Thereafter, the advertisement distribution apparatus 210 according to the present embodiment will be described.

1. Sentence Classification Unit FIG. 2 is a functional block diagram of the sentence classification unit 100. The sentence classification unit 100 includes a sentence example storage unit 12, a sentence decomposition unit 18, a sorting unit 20, a classification dictionary creation unit 22, a classification dictionary holding unit 24, a category attribution probability calculation unit 26, a determination unit 28, an element narrowing reference provision unit 30, The determination result storage unit 32 and the text reception unit 36 are provided. These configurations can be realized in terms of hardware by a CPU, memory, or other LSI of any computer, and in terms of software, they are realized by programs loaded into the memory. It depicts the functional blocks that are realized. Accordingly, those skilled in the art will understand that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof.

  The functional blocks included in the sentence classification unit 100 are used in a learning stage for creating dictionary data, and in a classification stage for classifying unclassified sentence data using the dictionary data after the dictionary data is created. Can be divided into functional blocks. First, functional blocks used in the learning stage will be described.

  The sentence example storage unit 12 stores, for each of a plurality of predetermined categories, a category sentence example 14 including sentence data to be classified into the category and a non-category sentence example 16 including sentence data not classified into the category. Here, “category” refers to grouping for classifying text data according to a specific standard. This category can be variously set according to the purpose of use of the sentence classification result. For example, when the classification result by the sentence classification unit 100 is used on a news article distribution site, types of categories such as “politics”, “economy”, “society”, and “sports” are conceivable. When the classification result by the sentence classification unit 100 is provided on a directory-type search site, categories such as “shopping”, “travel”, “movie”, and “music” can be considered. When the questionnaire etc. transmitted from the client terminal are classified by the sentence classification unit 100, categories such as “female” and “young people” can be considered.

  “Text data” refers to data including text information such as text data, HTML, XML, and XHTML files, but the data format is not limited.

  Assuming that M categories (category 1,..., Category M) are defined, a category sentence example 14 and a non-category sentence example 16 are prepared in the sentence example storage unit 12 for each category. The category sentence example is a file in which one or a plurality of sentence data to be classified into a certain category is accumulated. Non-category sentence example data is a file in which one or a plurality of sentence data that are not classified into a certain category are accumulated. The determination of which category sentence example or non-category sentence example the sentence data is included is made manually. A pair of category sentences and non-category sentence examples are prepared for each category.

  Text data included in a non-category sentence example for a certain category may be different from or identical to sentence data included in a non-category sentence example for another category. That is, the sentence data in the “non-category 1 sentence example” and the sentence data in the “non-category 2 sentence example” may overlap. The same applies to category sentence examples. For example, sentence data classified into both “category 1 sentence example” and “category 2 sentence example” may exist. However, it is desirable that there is no sentence data classified into both category sentence examples and non-category sentence examples within the same category.

  The sentence decomposition unit 18 analyzes sentence data included in the category sentence example 14 and the non-category sentence example 16 described above according to a predetermined rule, and extracts constituent elements of the sentence. The sentence decomposition unit 18 can use any known sentence decomposition algorithm. The sentence decomposition unit 18 extracts words, combinations of words and parts of speech, phrases, simple sentences, and the like from the sentence data as constituent elements according to the sentence decomposition algorithm. An example of the sentence decomposition algorithm used in the sentence decomposition unit 18 will be described later.

  The sorting unit 20 rearranges the components extracted from the sentence data of the category sentence example 14 and the non-category sentence example 16 by the sentence decomposition unit 18 according to a predetermined rule. As an example, the readings of the constituent elements may be rearranged in the order of 50 sounds. Or you may rearrange in order of the ASCII code of the first character of a component. When the sentence decomposition unit 18 extracts combinations of words and parts of speech as constituent elements, they may be arranged in the part of speech order of the constituent elements. By rearranging the constituent elements by the sorting unit 20, it becomes easy to search dictionary data using the constituent elements described later as keys, so that the time required for creating the classification dictionary can be shortened.

The classification dictionary creation unit 22 calculates the frequency at which each component extracted by the sentence decomposition unit 18 appears in the sentence data in the category sentence example 14 and the non-category sentence example 16 for each category, and for each component and each category Create dictionary data in association with the appearance frequency of. “Frequency” here may be the number of appearances or the ratio of appearance to the total number of words. Alternatively, it may be an appearance ratio with respect to the number of sentence data included in the category sentence example or the non-category sentence example. Hereinafter, these are collectively referred to as “appearance frequency”. In any case, any numerical value can be used as long as it is a numerical value representing the degree of appearance of a certain component in a category sentence example and a non-category sentence example for one category.
The dictionary data created by the classification dictionary creation unit 22 is stored in the classification dictionary holding unit 24. A more detailed configuration and function of the classification dictionary creation unit 22 will be described later with reference to FIG.

  The element narrowing criteria providing unit 30 stores selection criteria for excluding some components from the components extracted by the text decomposition unit 18. This selection criterion is, for example, a condition that specifies a specific part of speech (for example, a noun + verb, only a noun, only a particle, etc.), specifies the upper limit of the number of characters, specifies a component consisting of only hiragana or only kanji. I mean. A combination of a plurality of conditions may be used as a selection criterion.

The classification dictionary creating unit 22 uses the selection criteria provided from the element narrowing criteria providing unit 30 to narrow down the number of components for which dictionary data is to be created, thereby storing the dictionary data stored in the classification dictionary holding unit 24. It is possible to create dictionary data effective for classification while suppressing the amount of data.
The sentence classification unit 100 of the present embodiment does not limit the language of sentence data to be classified, but the element narrowing reference providing unit 30 is particularly useful when processing Japanese sentence data.

  Subsequently, functional blocks used in the classification stage will be described.

  The text receiving unit 36 receives text data 34 to be classified (hereinafter referred to as “unclassified text data 34”) from an external device (not shown). The data format of the uncategorized text data 34 is not limited as described above. The external device is, for example, a client terminal or server connected to the network, but is not limited to these.

  The sentence decomposing unit 18 receives the uncategorized sentence data 34 from the sentence receiving unit 36 and extracts the constituent elements of the sentence in the same manner as described above. The extracted components are rearranged by the sorting unit 20 according to a predetermined rule. This rule is preferably the same rule as when the components extracted from the category sentence example and the non-category sentence example are rearranged. This rearrangement facilitates the search of the dictionary data in the classification dictionary holding unit 24 using the constituent elements as keys, thereby speeding up the processing in the category attribution probability calculating unit 26 described later.

  The category attribution probability calculation unit 26 refers to the dictionary data stored in the classification dictionary holding unit 24, and calculates the probability that uncategorized sentence data including some components should be classified into each category. calculate. Hereinafter, this probability is referred to as “attribute probability”.

Similar to the classification dictionary creation unit 22, the category attribution probability calculation unit 26 may limit the number of constituent elements serving as a basis for calculating the attribution probability according to the selection criterion provided from the element narrowing criterion provision unit 30.
A more detailed configuration and function of the category attribution probability calculation unit 26 will be described later with reference to FIG.

  The determination unit 28 acquires the belonging probability calculated for each category by the category belonging probability calculating unit 26 and determines whether to classify the unclassified sentence data into any category based on the value of the belonging probability. More specifically, the determination unit 28 classifies the unclassified sentence data into the category having the maximum attribution probability. Alternatively, uncategorized sentence data may be classified into all categories for which an attribution probability equal to or higher than a preset threshold is obtained. By doing so, one uncategorized sentence data can be classified into two or more categories by a series of operations. When there is no category having the attribution probability equal to or higher than the threshold, the determination unit 28 may determine that the uncategorized sentence data is a sentence not classified into any category, or the category having the maximum attribution probability. May be classified. The determination result of the uncategorized sentence data by the determination unit 28 is stored in the determination result storage unit 32 or output to an external device (not shown).

  Next, an outline of the sentence decomposition algorithm used in the sentence decomposition unit 18 will be described.

(1) Morphological Analysis FIG. 3 shows an example in which sentence data is decomposed into constituent elements by morphological analysis. The text data used is the text "The Japan Meteorological Agency announced that the first spring in the Kanto region blew on the 23rd." As shown in FIG. 3, this sentence is “Meteorological Agency / Ha / 2/3 / Day /, / Kanto / Region / De / Spring Ichiban / Ga / Blow / Ta / To / Announcement / Shi / Ta /.” Thus, it is broken down into 18 elements. In the morphological analysis, the utilization form 50, the original form 52, and the part of speech 54 can be determined from the target sentence. Of these utilization forms, original forms, and parts of speech, (original form + part of speech) may be a constituent element, or only the original form may be an element. The utilization form may be used as an element instead of the original form.

  Since components extracted using morphological analysis have high sentence resolution, classification into categories based on dictionary data using these components is expected to be highly accurate. Since morphological analysis is a well-known technique, further explanation is omitted.

(2) Syntax analysis Next, syntax analysis will be described. Parsing breaks a sentence into phrases. When the same sentence data as in the example of FIG. 3 is decomposed by parsing, it is decomposed into six components as follows: “The Japan Meteorological Agency announced on / 23rd / In the Kanto region / Spring Ichiban / Announced”. Is done.

  When parsing is used, the number of components is significantly reduced compared to morphological analysis, which is suitable for high-speed classification, but the accuracy of classification is reduced. Since parsing is a well-known technique, further explanation is omitted.

(3) Minimum Composition Sentence Next, an example of extracting a minimum construction sentence from a sentence using morphological analysis and syntax analysis will be described. Here, the “minimum component sentence” is a sentence that has a minimum meaning. For details, see “Invention of similarity between documents used for imitation report determination, Takahisa Ota, Shigeru Masuyama, 10th Annual Meeting of the Language Processing Society” Pp.729-732, 2004 ”, published in the next conference.

  Extracting the minimum sentence from the same text data as in the example of FIG. 3, “Ministry of Meteorological Agency announced.” “Announced on 23rd.” “Announced that the first spring in the Kanto region was blown.” A composition sentence is obtained. By creating dictionary data using these minimum constituent sentences as constituent elements, it is possible to capture the meaning of words in the context, so it is possible to classify sentences containing words that can be captured in multiple meanings into appropriate categories. Classification is possible, but the calculation cost is high.

  Note that, by using the part-of-speech information obtained as a result of the morphological analysis, it is possible to extract the minimum constituent sentence consisting only of the noun, the adjective, and the verb form. Using the same example as above, three minimum composition sentences are obtained: “Meteorological Agency to announce”, “23 days to announce”, “Kanto region, spring first, blow, to announce”.

  In this manner, by extracting the constituent elements using different sentence decomposition algorithms in the sentence decomposition unit 18, dictionary data having different tendencies can be created in the classification dictionary creation unit 22. Therefore, it is possible to improve classification accuracy and processing speed by selecting an appropriate sentence decomposition algorithm according to the category type.

  FIG. 4 is a detailed functional block diagram of the classification dictionary creation unit 22. The classification dictionary creation unit 22 includes a component reception unit 102, a narrowing information reception unit 104, a category information provision unit 106, a component selection unit 108, a dictionary data search unit 110, and a dictionary data update unit 112.

  The component reception unit 102 receives the components rearranged according to a predetermined rule from the sorting unit 20 and passes them to the component selection unit 108. The category information providing unit 106 informs the element narrowing-down criterion providing unit 30 of information about the categories to which the category examples and the non-category example sentences from which the component elements received by the component element receiving unit 102 are extracted belong. The refinement information receiving unit 104 receives the selection criteria from the element refinement criteria providing unit 30 and passes them to the component selection unit 108. The component selection unit 108 compares the selection criterion with the component, selects a component that satisfies the selection criterion, and passes it to the dictionary data search unit 110. The dictionary data search unit 110 searches the dictionary data held in the classification dictionary holding unit 24 for dictionary data for the same component as the component that satisfies the selection criteria, and the corresponding dictionary data is If there is, it is passed to the dictionary data update unit 112. The dictionary data updating unit 112 counts the number of each constituent element that satisfies the selection criteria, adds the number to the dictionary data, and stores it in the classification dictionary holding unit 24. When the component is new, new dictionary data is created and stored in the classification dictionary holding unit 24.

FIG. 5 is a data structure diagram of dictionary data stored in the classification dictionary holding unit 24. In the dictionary data 40, the constituent element 42 is associated with the appearance frequency 44 at which the constituent element appears in sentences included in the category sentence examples of category 1 to M and the non-category sentence example. If the component elements are represented as W _n (n = 1 to N) and the frequency of appearance in sentences included in the category sentence example or the non-category sentence example where W _n is the category m (m = 1 to M), respectively, X _nm and Y _nm. The dictionary data 46 for a certain component W _n can be expressed as (W _n , X _n1 , Y _n1 , X _n2 , Y _n2 ,..., X _nm , Y _nm ).

In this embodiment, for each component W _n , two values of (appearance frequency in a sentence included in a category sentence example of category m) and (appearance frequency in a sentence included in a non-category sentence example of category m) are set. Hold in pairs. This is for facilitating rewriting of the frequency information when new sentence data is added to the category sentence example or the non-category sentence example and the dictionary data in the classification dictionary holding unit 24 is to be expanded.
In another embodiment, the appearance frequency of the component W _n may be held as a single value. When the appearance frequency of the component W _n with respect to the category m is expressed as F _nm , F _nm = X _nm / Y _nm may be set, or F _nm = X _nm / (X _nm + Y _nm ) may be set. In this case, the dictionary data 46 for a certain component W _n can be expressed as (W _n , F _n1 , F _n2 ,..., F _nM ).

  FIG. 6 is a data structure diagram of the selection criteria stored in the element narrowing criteria providing unit 30. Selection criteria are prepared corresponding to the category type. In FIG. 6, “theme classification”, “style”, and “age” are included as category types. The element narrowing criteria providing unit 30 receives the category sentence information from which the constituent elements are extracted from the category information providing unit 106, and returns the selection criteria in the right column 58 in the case of a specific category shown in the left column 56 in the figure. . If the category received from the category information providing unit 106 does not exist in the left column 56, the standard “noun” selection criterion is returned.

  For example, if the category type is related to a theme or topic classification, for example, a category such as “travel”, “music”, or “movie”, the criterion “noun” is provided. This is because, for such themes and topic classifications, the presence of a specific noun often determines the category classification. When the category type is related to a style, for example, a category such as “formal”, “careful”, or “random”, a criterion of “adjective or particle” is provided. This is because the style can often be determined by a particle such as “Tenanoha” or emotional expression. Furthermore, for the age and gender of the person who created the text, for example “female” and “young people”, a criterion “noun of hiragana” is provided. In this way, the element refinement criterion providing unit 30 selects different criteria for each category depending on which category sentence example or non-category sentence example includes the constituent elements for which dictionary data is to be created. Can be provided. The classification dictionary creation unit 22 can create dictionary data by narrowing down the components prepared as a dictionary with reference to the selection criteria.

  The element narrowing criteria providing unit 30 may provide the number of characters as a selection criterion instead of providing the part of speech of the component as a selection criterion. As a result, the classification dictionary creation unit 22 can create a dictionary for components having a certain number of characters or less. Alternatively, the element refinement criterion providing unit 30 may provide a specific component (for example, a noun “automobile”) as a selection criterion. The classification dictionary creation unit 22 may exclude components that match the classification dictionary from the dictionary data creation target. For example, it is preferable to exclude common nouns (for example, “I” and “thing”) used in a very large number of sentences because they have little influence on the category classification.

  FIG. 7 is a detailed functional block diagram of the category attribution probability calculator 26. The category attribution probability calculation unit 26 includes a component reception unit 122, a narrowing information reception unit 124, a component selection unit 126, a dictionary data search unit 128, an appearance probability calculation unit 130, and an attribution probability calculation unit 132.

  The component receiving unit 122 receives the components rearranged from the sorting unit 20 according to a predetermined rule. The refinement information receiving unit 124 receives the selection criterion from the element refinement criterion providing unit 30 and passes it to the component selection unit 126. The component selection unit 126 compares the selection criterion with the component, selects a component that satisfies the selection criterion, and passes it to the dictionary data search unit 128. The dictionary data search unit 128 searches the dictionary data held in the classification dictionary holding unit 24 for dictionary data for the same component as the component that satisfies the selection criteria, and the corresponding dictionary data is If there is, it is passed to the appearance probability calculation unit 130.

The appearance probability calculation unit 130 calculates the appearance probability a _nm of each component W _n extracted from the unclassified data for each category m. Here, the appearance probability a _nm is calculated by the following equation using the appearance frequencies X _nm and Y _nm in the sentence included in the above-described category sentence example or non-category sentence example.

FIG. 8 shows the appearance probabilities a _{11 to} a _N1 of the components W _{1 to} W _N with respect to the category 1 calculated by Equation ₁ .
Note that the calculation of the appearance probability is not limited to Equation 1. For example, the above-described F _nm may be used as it is.

The attribution probability calculation unit 132 adds up the calculated appearance probabilities for all the components, and calculates the attribution probability for each category for the unclassified sentence data. Preferably, belonging probability calculation unit 132 uses the Bayesian filter method to calculate a membership probability E _n to category n by the following equation.

  Note that the attribution probability may be calculated using a method other than the Bayesian filter method. For example, the attribution probability may be calculated by simply multiplying the appearance probabilities of all the constituent elements, or the average value of the appearance probabilities may be used as the attribution probability.

FIG. 9 is a flowchart showing a process of creating a classification dictionary.
First, the sentence decomposition unit 18 acquires a pair of category sentence examples or non-category sentence examples from the sentence example storage unit 12 (S10). Next, the sentence decomposition unit 18 decomposes the sentence data in the category sentence example and the non-category sentence example into constituent elements based on a predetermined sentence decomposition algorithm, and the sorting part 20 separates the decomposed constituent elements according to a predetermined rule. Rearrange (S12). The execution of the rearrangement is not essential for the present embodiment, and the data search time from the classification dictionary holding unit becomes long, so that the calculation speed can be reduced, but the accuracy of category classification is not affected.

  Next, the classification dictionary creation unit 22 compares the extracted component with the selection criterion received from the element refinement criterion providing unit 30 to determine whether the component is a component for which dictionary data is to be created. (S14). If it is not a dictionary data creation target (N in S14), the process proceeds to S24. If it is a dictionary data creation target (Y in S14), the classification dictionary creation unit 22 searches the classification dictionary holding unit 24 for dictionary data for the component (S16). If the corresponding dictionary data exists (Y in S18), the number of components existing in the current sentence example data is added to the frequency of the category in the dictionary data (S20). If there is no corresponding dictionary data (N in S18), new dictionary data is created (S22). Then, it is determined whether or not all the components decomposed by the text decomposition unit 18 have been processed (S24). If the processing has not ended (N in S24), the processing from S14 is performed on another component. repeat. When the creation of dictionary data for all the constituent elements is finished (Y in S24), this flow is finished.

FIG. 10 is a flowchart showing a process of classifying uncategorized text data into categories.
The text receiving unit 36 receives unclassified text data (S30). The sentence decomposition unit 18 decomposes the sentences in the uncategorized sentence data into constituent elements, preferably based on the same sentence decomposition algorithm as S12 in FIG. 9, and the sorting unit 20 converts the decomposed constituent elements according to a predetermined rule. Rearrange (S32). Next, the category attribution probability calculating unit 6 compares the extracted one component with the selection criterion received from the element narrowing criterion providing unit 30 to calculate the attribution probability of uncategorized sentence data. It is determined whether or not the appearance probability of the component is to be calculated (S34). If it is an appearance probability calculation target (Y in S34), the category attribution probability calculation unit 26 searches the classification dictionary holding unit 24 for dictionary data for the component, and if there is corresponding dictionary data (in S36). Y) Based on the dictionary data, the appearance probability of the component is calculated for each category (S38). If the component is not the target of appearance probability calculation in S34 (N in S34), or if the corresponding dictionary data does not exist in S36 (N in S36), S38 is skipped.
Subsequently, it is determined whether or not all the components extracted by the text decomposition unit 18 have been processed (S40). If the processing has not ended (N in S40), the processing from S34 is performed on another component. repeat.

  When the processing for all the constituent elements is completed (Y in S40), the category attribution probability calculation unit 26 calculates the attribution probability of uncategorized sentence data for each category according to the above-described procedure (S42), and the determination unit 28 Determines the category to which the uncategorized text data belongs based on the attribution probability (S44).

(Example)
Hereinafter, based on a specific Example, operation | movement of the text classification | category part 100 which concerns on this Embodiment is demonstrated. In this embodiment, it is assumed that two categories of “gambling” and “education” are prepared as categories for easy explanation. The dictionary data that has already been created is used.

  FIG. 11 shows dictionary data used in this embodiment, and corresponds to FIG. 5 in the entire description above. This dictionary data is included in the category sentence example for the category “gambling”, the constituent elements extracted from the 30 sentences included in the category sentence example and the 15 sentences included in the non-category sentence example, and the category “education”. And the components extracted from the 18 sentences included in the non-category sentence examples. The number of sentences used is shown in the column 154 as “number of sentences”.

  As shown in the figure, this dictionary includes words such as “Pachinko”, “Manga ticket”, “Youth”, “Healthy”, and “Nurture” as constituent elements. Each component has appearance frequency information for each category. Taking the component “pachinko” as an example, the appearance frequency in the sentence of the category sentence example is 10 times and the appearance frequency in the sentence of the non-category sentence example is 2 times for the category “gambling”. For the category “education”, the frequency of appearance in the sentence of the category sentence example is once, and the frequency of appearance in the sentence of the non-category sentence example is 20 times. The same applies to other components.

  The “total” column 152 is a number obtained by adding the number of sentences in the category sentence example and the number of sentences in the non-category sentence example for all the constituent elements.

  Assuming that such dictionary data is held in the classification dictionary holding unit 24, the sentence “Let's grow the pachinko industry soundly” as uncategorized sentence data is classified into any of the two categories. Explain how. A component is extracted from the sentence by the sentence decomposition unit 18. In this embodiment, the sentence decomposing unit 18 decomposes the sentence by morphological analysis, and as a result, ten constituent elements such as “Pachinko / industry / to / health / to / nurture / do / make / to /.” Is extracted. Subsequently, the component element selection unit 126 in the category attribution probability calculation unit 26 receives the selection criterion “noun” from the element narrowing criterion provision unit 30 and selects only the noun from the extracted component elements. Therefore, four components of “pachinko”, “industry”, “sound” and “nurturing” are selected.

  The dictionary data retrieval unit 128 retrieves corresponding dictionary data from the classification dictionary holding unit 24. In this case, three dictionary data of “pachinko”, “sound” and “nurturing” are obtained. The appearance probability calculation unit 130 calculates the appearance probability for each category using the above equation 1 from the appearance frequencies of the three components in the dictionary data. The result is shown in FIG. Taking the component “Pachinko” as an example, the appearance probability for the category “gambling” is 0.714, and the appearance probability for the category “education” is 0.043. The same applies to other components.

The attribution probability calculation unit 132 calculates the attribution probability of the uncategorized sentence data for each category using the appearance probability according to the above formula 2.
This will be described using specific numerical values. 13, corresponding to Figure 8 of the entire description, is a table summarizing the probability a _n for the category "gambling", "Education" (1-a _n). For the category “gambling”, the appearance probability a ₁₁ of the element “pachinko” is 0.714, the appearance probability a ₂₁ of the element “sound” is 0.200, and the appearance probability a ₃₁ of the element “nurturing” is 0.273. Therefore, when calculated according to Equation 2, the attribution probability E ₁ of the uncategorized sentence data to the category “gambling” is calculated as follows.
(Equation 3)
E ₁ = (0.714 × 0.200 × 0.273) /
{(0.714 × 0.200 × 0.273) + (1−0.714) × (1−0.200) × (1−0.273)} ≈0.190

Category for the "education", the occurrence probability of _{a 12} element "pachinko" is 0.043, the occurrence probability of _{a 22} element "healthy" is 0.783, the occurrence probability of _{a 32} element "training" is 0.844 Therefore, when calculated according to Equation ₂ , the belonging probability E ₂ of the uncategorized sentence data to the category “education” is calculated as follows.
(Equation 4)
E ₂ = (0.043 × 0.783 × 0.844) /
{(0.043 × 0.783 × 0.844) + (1−0.043) × (1−0.783) × (1−0.844)} ≈0.467

  As a result, the determination unit 28 determines that the sentence “Let's nurture the pachinko industry soundly” is classified into the category “education” having the higher attribution probability. Thus, a series of processes for classifying unclassified sentence data is completed.

  As described above, according to the sentence classification unit 100, uncategorized sentence data can be automatically classified according to a predetermined category. Which category the uncategorized sentence data belongs to is determined by the attribution probability of the unclassified sentence data calculated for each category. Therefore, the sentence classification unit 100 can classify one sentence data into one category, and can also classify into two or more categories. In addition to classifying categories based on one viewpoint, categories based on a plurality of viewpoints can be mixed, and the belonging probability of unclassified sentence data can be calculated collectively.

The classification of blog articles using the sentence classification technique described above is described in detail in the following documents.
Koichi Hirano, Kiya Kobayashi, Junichi Takahashi, “Automatic Classification of Japanese-Speaking Blogs”, Information Processing Society of Japan Report (Natural Language Processing), November 2005, 2005-NL-170, p. 21-26
Koya Kobayashi, Koichi Hirano, Junichi Takahashi, “Attempts to Capture Aspects of Consumer Awareness by Automatic Classification of Blog Articles”, NRI Technology Creation, January 2006, 2006 Vol. 7, p. 38-52

2. Advertisement Distribution Device Subsequently, the advertisement distribution device 210 according to the present embodiment will be described.

  FIG. 14 is a functional block diagram of the advertisement distribution apparatus 210. Also in this figure, each functional block can be realized by hardware only, software only, or a combination thereof.

  The medium page management unit 218 manages registration of a medium web page that is a distribution destination of the advertisement. The medium page management unit 218 includes a medium registration reception unit 220 and a script generation unit 222.

  The medium registration receiving unit 220 receives registration from the administrator of the medium web page. The medium registration receiving unit 220 receives the URL of the page to be registered as the medium web page, the name of the administrator, the e-mail address, and the like from the page manager terminal 204 and stores these pieces of information in the medium information holding unit 230.

  The script generation unit 222 creates a script program for displaying the advertisement information transmitted from the advertisement distribution device 210 on the medium web page. The script generation unit 222 transmits the created script program to the page manager terminal 204 by sending an e-mail or presenting a web page describing the program. The administrator of the web page copies the received script program to a markup language file such as an HTML document or an XML document that describes its web page. Then, the administrator uploads this file to the provider server 212. By doing so, the script program is sent from the provider server 212 to the general user terminal 206 accessing the medium web page.

  The media page analysis unit 232 determines a keyword set and a category of the media web page. The medium page analysis unit 232 includes a page acquisition unit 224, a confirmation unit 226, a sentence classification unit 234, a keyword extraction unit 228, and a medium information holding unit 230.

  The page acquisition unit 224 acquires the medium web page 214 from the provider server 212 via the network 208. The page acquisition unit 224 may acquire the medium web page when the medium web page is browsed on the general user terminal 206 as described later, or the page acquisition unit 224 operates the web crawler. Information on media web pages provided on the network may be collected. The page acquisition unit 224 also acquires the URL of the general user terminal 206 browsing the medium web page and passes it to the advertisement transmission unit 246.

  The confirmation unit 226 refers to the medium information holding unit 230 for the medium web page acquired by the page acquisition unit 224 and determines whether or not a keyword set and a category have already been determined. Alternatively, the confirmation unit 226 refers to the distribution result holding unit 242 and determines whether or not advertisement information has been transmitted previously. When the keyword set and the category are not yet determined, the confirmation unit 226 provides the medium web page to the sentence classification unit 234 and the keyword extraction unit 228. When the advertisement information has been transmitted before, the confirmation unit 226 sends the advertisement information stored in the distribution result holding unit 242 to the advertisement transmission unit 246 to the general user terminal 206 browsing the medium web page. To send to.

  Check if the advertisement information stored in the distribution result holding unit 242 is old because a considerable amount of time has passed since the previous distribution, or if you want to send different advertisement information to the media web page each time. The unit 226 selects and transmits new advertisement information to the matching unit 244 using the keyword set and category information determined for the medium web page held in the medium information holding unit 230. You may instruct.

  The sentence classification unit 234 has the same function as the sentence classification unit 100 described above. The sentence classification unit 234 receives the medium web page file as an unclassified sentence, and calculates an attribution probability for a plurality of categories set in advance. The sentence classification unit 234 may determine some categories to which the medium web page belongs according to the comparison between the attribution probability and a predetermined threshold value, or the calculation of the attribution probability and the subsequent You may make it leave the comparison with the category of advertisement information to the matching part 244 mentioned later. This is determined by an aspect in which the advertisement management unit 240 accepts designation of a category as an advertisement distribution destination from the advertiser. More specific description will be described later with reference to FIGS.

  The keyword extraction unit 228 determines a keyword set of the medium web page. Specifically, the keyword extraction unit 228 decomposes the text information of the medium web page into morphemes by morphological analysis, and determines a keyword set based on the appearance frequency of each morpheme and the relationship between morphemes. The keyword extracting unit 228 desirably extracts a plurality of keywords as a keyword set, but may select only one keyword. Since such a keyword extracting method is publicly known, further detailed description is omitted.

  The medium information holding unit 230 holds, for each medium web page, the attribution probability for each category calculated by the sentence classification unit 234 or the determined category, and the keyword set extracted by the keyword extraction unit 228.

  The advertisement management unit 240 manages registration and distribution of advertisement information to be displayed on the medium web page. The advertisement management unit 240 includes a distribution result holding unit 242, a matching unit 244, an advertisement transmission unit 246, an advertisement receiving unit 250, and an advertisement information holding unit 252.

  The advertisement reception unit 250 receives registration of advertisement information from the advertiser. The advertisement reception unit 250 distributes the text to be displayed as an advertisement from the advertiser terminal 202, the category designation information for specifying the category condition of the medium web page to which the advertisement information is distributed, and the medium web page to which the advertisement information is distributed. Advertisement information such as keyword designation information for designating a destination keyword and URL of a linked advertiser site is received. These pieces of information are held in the advertisement information holding unit 252.

  The advertisement information holding unit 252 holds the advertisement information in association with the category designation information and the keyword designation information. There are multiple ways to specify category conditions. One is a method of designating whether or not an advertisement is displayed on a media web page classified into the category for each of a plurality of categories set in advance. The second is a method of designating the lower limit value of the attribution probability of the medium web page as the advertisement distribution destination for each category. When the second method is adopted, a non-display category that should avoid sending an advertisement may be further specified.

  The matching unit 244 acquires the keyword set and the category of the medium web page that is the current distribution target from the medium information holding unit 230. Subsequently, the matching unit 244 selects, from the advertisement information holding unit 252, the advertisement information in which the delivery destination keyword included in the keyword set of the medium web page is designated. After that, the matching unit 244 displays, from the selected advertisement information, the advertisement information holding unit 252 that has the category of the medium web page that is the distribution target this time satisfies the category condition related to the advertisement information. Select from.

  At this time, when the category condition simply specifies whether or not to display an advertisement on the medium web page, the matching unit 244 may select the advertisement information to be displayed for the category of the medium web page. When the lower limit value of the belonging probability is specified for each category, it is as follows. First, the matching unit 244 acquires from the medium information holding unit 230 the attribution probability for each category of the medium web page that is the distribution target of the advertisement. Subsequently, for each category, the matching unit 244 compares the attribution probability of the medium web page with the lower limit value of the attribution probability related to the advertisement information. When there is a category in which the medium web page attribution probability is equal to or higher than the lower limit value, the matching unit 244 selects the corresponding advertisement information from the advertisement information holding unit 252.

The matching unit 244 sends the selected advertisement information to the advertisement transmission unit 246.
The advertisement transmission unit 246 transmits the advertisement information selected by the matching unit 244 to the general user terminal 206 so as to be displayed together with the medium web page. This transmission is performed for the URL of the general user terminal 206 acquired by the page acquisition unit 224. The result of transmission in the advertisement transmission unit 246 is stored in the distribution result holding unit 242.

  The distribution result holding unit 242 holds a combination of a medium web page and advertisement information transmitted for display on the medium web page. In addition, the distribution result holding unit 242 may store information such as the URL of the general user terminal 206 that transmitted the advertisement information, the date when the advertisement is displayed, and the number of times the advertisement is displayed.

FIG. 15 is a flowchart showing an outline of processing for distributing an advertisement related to a web page by the advertisement distribution apparatus 210.
First, advertisement information is registered in the advertisement management unit 240 from the advertiser terminal 202 (S60). Also, the medium web page is registered from the page manager terminal 204 to the medium page management unit 218 (S62). Thereafter, when the registered media web page is viewed on the general user terminal 206, the media page analysis unit 232 determines the keyword set and category of the media web page. Further, the advertisement management unit 240 selects advertisement information that matches both the keyword and the category from the registered advertisement information, and transmits it to the general user terminal 206 (S64). The transmitted advertisement information is displayed together with the medium web page on the general user terminal 206.

FIG. 16 shows an example of an advertisement reception screen 260 displayed on the display of the advertiser terminal 202 by the advertisement reception unit 250.
When the advertiser uses the advertiser terminal 202 to access a dedicated site for registering advertisement information, the advertisement reception unit 250 transmits a markup language file for displaying the advertisement reception screen 260 to the advertiser terminal 202. To do. In the advertisement reception screen 260, the advertiser inputs the company name of the company that is the advertiser and the mail address of the person in charge in the company name column 262 and the email address column 264. In the advertisement text fields 266 and 268, the text of the advertisement text distributed to the medium web page is input. The advertiser inputs a headline that can be displayed in about one line in the advertisement text field 266, and inputs more detailed contents in the advertisement text field 268.
In the keyword column 270, a distribution destination keyword for designating a medium web page for which the advertiser desires distribution of advertisement information is input. The delivery destination keyword to be input may be one word or a plurality of words. The display category button 272 is a button that the advertiser clicks to set in detail a category for which the advertisement information is desired to be distributed. In the jump destination URL column 274, the URL of the advertiser site that is the link destination when the user clicks on the advertisement text displayed on the medium web page is input.
When the advertiser clicks the registration button 276 after filling in the above fields, the input content is transmitted to the advertisement receiving unit 250. The advertisement receiving unit 250 passes the received information to the advertisement information holding unit 252.

17 to 19 show examples of screens for setting categories displayed when the display category button 272 is clicked on the advertisement reception screen 260.
FIG. 17 shows a setting screen 280 for setting a category in which an advertisement should be displayed or a category to be hidden for each category. On the setting screen 280, for each category 282, a check box 284 that is checked when it is designated as a distribution destination category of advertisement information, a check box 286 that is checked when it is not desired to be designated as a distribution destination category of advertisement information, Is displayed. For the category of the medium web page, the matching unit 244 selects advertisement information in which the check box 284 of the same category as the category is checked.

  FIG. 18 shows, for each category 292, a lower limit column 294 for setting a lower limit value of the attribution probability calculated for the medium web page to which the advertisement information is distributed, and a category of media web pages to which the advertisement information is not distributed (hereinafter, “ A category setting screen 290 having a non-display category column 296 for setting “non-display category”) is shown.

  In the lower limit value column 294, the lower limit value when the attribution probability calculated for each category in the sentence classification unit 234 is displayed as a percentage can be set in increments of 20%. Checking a check box for setting a lower limit value means that advertisement information is distributed for a category whose attribution probability is greater than or equal to the number displayed on the right shoulder of the check box. For example, when looking at the category “entertainment” 297, a check box of “20 to 40” is checked in the lower limit column 294. This means that, for the category “entertainment”, the advertisement information should be distributed to a medium web page having an attribution probability of 20% or more. In other words, the advertising information is distributed to all of the media web pages that have a small number of entertainment elements. When the category “Computer” 298 is viewed, the check box “80 to 100” is checked in the lower limit column 294. This means that for the category “computer”, the advertisement should be distributed to a medium web page having an attribution probability of 80% or more. In other words, advertising information is only delivered to media web pages with very high computer elements. Further, when the category “finance” 299 is viewed, the non-display category column 296 is checked. This means that if the attribution probability for the category “finance” is small, advertisement information is not distributed to the medium web page.

  As shown in the figure, on this category setting screen 290, the lower limit of the category belonging probability of the medium web page as the advertisement distribution destination can be determined for each category. Therefore, the advertiser can specify a medium web page that cannot clearly classify the category as a distribution destination of the advertisement information. For example, since it is possible to narrow down the media web pages that are classified across a plurality of categories by the matching unit 244, the advertiser can narrow down the target audience of the advertisement and improve the advertising effectiveness. Can be connected.

  FIG. 19 shows a category setting screen 300 having a lower limit value column 304 for setting a lower limit value of the attribution probability calculated for the medium web page to which the advertisement information is distributed for each category 302. In the lower limit column 304, the lower limit when the attribution probability calculated for each category in the sentence classification unit 234 is displayed as a percentage can be set as a numerical value. Advertisement information is distributed for categories whose attribution probabilities are greater than or equal to the numbers displayed in the lower limit column 304. For example, when looking at the category “entertainment” 306, a numerical value “20” is entered in the lower limit value column 304. This means that for the category “entertainment”, the advertisement information should be distributed to a medium web page whose attribution probability is 20% or more. The category setting screen 300 may also be provided with a non-display category column for setting a non-display category.

  20 is a table 310 of advertisement information that is input to the advertisement reception unit 250 via the advertisement reception screen 260 of FIG. 16 and the category setting screen of any of FIGS. 17 to 19 and is held in the advertisement information holding unit 252. Indicates.

  The number column 312 indicates a number uniquely given to each piece of advertisement information received by the advertisement receiving unit 250. The advertisement text columns 314 and 316, the keyword column 318, and the URL column 324 indicate the advertisement text 1, the advertisement text 2, the distribution destination keyword, and the jump destination URL input on the advertisement reception screen 260. Items to be entered in the category condition column 319 differ depending on which method is used for category setting. As shown in FIG. 17, when a category for displaying an advertisement and a category for non-display are set, “display” or “non-display” is recorded for each category name 320. On the other hand, as shown in FIGS. 18 and 19, when the lower limit value of the category in which the advertisement is to be displayed is designated, the lower limit value is recorded for each category name 320.

  In the example of FIG. 20, the advertisement designated by the number “1” is an advertisement of a company that sells a muffler of clothing, and “fashion” is designated as a category for displaying the advertisement. The advertisement designated by the number “2” is an advertisement of a company that sells an automobile muffler. The lower limit value of the attribution probability of the category “fashion” is 80%, and the lower limit value of the attribution probability of the category “motor sports” is 20. % Is specified. Note that in this example, “muffler” of the same sound is specified as the distribution destination keyword.

FIG. 21 shows an example of a medium registration acceptance screen 330 displayed on the page manager terminal 204 by the medium registration acceptance unit 220.
When the page manager accesses a dedicated site for registering a media web page using the page manager terminal 204, the media registration acceptance unit 220 displays a media registration acceptance screen 330 on the page manager terminal 204. Send markup language files to be used. On the medium registration acceptance screen 330, the page manager inputs the name, address, and mail address of the manager who provides the medium web page as the advertisement distribution destination in the name field 332, address field 334, and mail address field 338. To do. Further, the page manager inputs the URL of the medium web page in the URL field 336.
When the page manager clicks the registration button 340 after filling in the above fields, the input content is transmitted from the page manager terminal 204 to the medium registration receiving unit 220. The medium registration receiving unit 220 passes the received information to the medium information holding unit 230.

  FIG. 22 shows an example of text information of a page registered as a medium web page. When the page acquisition unit 224 receives the URL of the medium web page being browsed from the general user terminal 206, the page acquisition unit 224 accesses the URL to acquire a file of the medium web page, extracts text information therefrom, and extracts a text information from the file. 234 and the keyword extraction unit 228. 22A shows the text 344 described in the media web page specified by the URL 342, and FIG. 22B shows the text 348 described in the media web page specified by the URL 346. As shown in the figure, the text 344 is a diary-like article about fashion, and the muffler is a hot topic. The text 348 is a diary-like article about a motorcycle, and a muffler is described as a replacement part.

  FIG. 23 shows a medium information table 350 held in the medium information holding unit 230. The code column 352 shows a code given by the medium registration receiving unit 220 for each page manager. The URL column 354 shows the URL of the medium web page. The keyword column 356 shows a keyword set extracted from the text in the medium web page by the keyword extraction unit 228. The classification result column 358 displays the attribution probabilities calculated by the text classification unit 234 for a plurality of predetermined categories.

In the example of FIG. 23, the medium web page specified by the code “0001” corresponds to the page shown in FIG. Therefore, the URL column 354 shows the URL 342 shown in FIG. The keyword column 356 shows a keyword set extracted from the text 344 by the keyword extraction unit 228. In the classification result column 358, an attribution probability of 0.76 is calculated for the category “fashion” as the classification result obtained by inputting the text 344 to the sentence classification unit 234, and an attribution probability of 0.68 for the category “shopping”. Is calculated.
The medium web page specified by the code “0002” corresponds to the page shown in FIG. Therefore, the URL column 354 shows the URL 346 shown in FIG. The keyword column 356 shows a keyword set extracted from the text 348 by the keyword extraction unit 228. The classification result column 358 indicates that the attribution probability 0.88 is calculated for the category “motor sports” as the classification result obtained by inputting the text 348 to the sentence classification unit 234.

FIG. 24 is a flowchart for explaining the registration process of the medium web page by the page manager.
The medium registration receiving unit 220 receives medium web page registration information input from the page manager terminal 204 (S70). The medium registration receiving unit 220 assigns a unique code to the page manager and records information registered on the medium web page in the medium information holding unit 230 (S72). Subsequently, the script generation unit 222 transmits the URL of the medium web page to the advertisement distribution device 210 when the medium web page is browsed by the general user terminal 206, and the information transmitted from the advertisement distribution device 210 is the medium. An advertisement distribution script program having a function of displaying with a web page is created, and this advertisement distribution script program is transmitted to the page manager terminal 204 (S74). The page manager inserts the received script program into his media web page and uploads this file to the provider server 212 (S76). Thus, when the medium web page is read by the general user terminal 206, the advertisement distribution script program is also read together.

  As long as the administrator of the medium web page is the same person, the same advertisement distribution script program can be used for a plurality of medium web pages. Since the keyword set and category for each medium web page are individually extracted or classified for each URL on the advertisement distribution device 210 side and stored, even if the same advertisement distribution script program is used, different advertisement information for each page Will be displayed.

FIG. 25 is a flowchart illustrating in detail the process of distributing advertisements to web pages.
When the general user accesses the medium web page using the general user terminal 206 (S80), the web page 214 is transmitted from the provider server 212 to the general user terminal 206, and the medium web page is displayed on the display of the general user terminal 206. (S82). The script program inserted in the medium web page operates on the general user terminal 206, and transmits the URL of the page being browsed and the page manager code to the advertisement distribution apparatus 210 (S84).

  The page acquisition unit 224 receives the transmitted page URL and page manager code, and the confirmation unit 226 determines whether the category of the media web page specified by the page URL has already been classified (S86). ). When the category of the medium web page is unclassified (N in S86), the advertisement transmission unit 246 transmits advertisement information prepared for display regardless of the category to the general user terminal 206 (S88). . The advertisement distribution script program of the general user terminal 206 displays the transmitted advertisement information on the display.

  The page acquisition unit 224 acquires a corresponding medium web page from the provider server 212 using the page URL. According to the acquired text information of the medium web page, the sentence classification unit 234 extracts a keyword set of the medium web page, and the sentence classification unit 234 calculates the belonging probability for each category (S90). The matching unit 244 includes a distribution destination keyword for each advertisement held in the advertisement information holding unit 252 in the keyword set extracted from the medium web page, and the attribution probability for each category of the medium web page holds the advertisement information. The advertisement information satisfying the category condition for each advertisement held in the unit 252 is determined (S92). This result is held in the distribution result holding unit 242.

  If the category of the medium web page is already classified in S86 (Y in S86), the advertisement transmitting unit 246 records the medium web page recorded in the distribution result holding unit 242 in response to an instruction from the confirmation unit 226. Is sent to the general user terminal 206 (S94). The advertisement distribution script program of the general user terminal 206b receives the transmitted advertisement information and displays it together with the medium web page.

  Thus, in the present embodiment, the content of the web page registered as the medium web page is actually acquired when the medium web page is browsed by the general user terminal for the first time. Therefore, at this point in time, information on the media web page is not accumulated on the advertisement delivery device side. Therefore, the advertisement not targeted by the customer is distributed for the time being, and then the media web page is acquired and the keyword set and category are obtained. decide. This is because, when a medium web page is browsed by a general user terminal, if an advertisement is transmitted after executing keyword extraction and category classification of the medium web page, it takes time for calculation processing and transmission. This is because it is impossible to display an advertisement.

  In addition, the web page is acquired after the medium web page is browsed. All the registered medium web pages are crawled before browsing to acquire the web page, and the keyword set and category are determined. This is because the calculation processing load increases. Further, by acquiring the medium web page at the time of browsing by a general user, an advertisement matched according to the keyword set and category classification before the update of the medium web page can be prevented.

  In addition, in order to cope with the case where the content of the media web page is updated after the category classification, the advertisement distribution apparatus discards the old keyword set and category information at a predetermined cycle and newly acquires the media web page. It is preferable to repeat the keyword extraction and category classification.

  In addition, media web pages are acquired before browsing with general user terminals and keyword extraction and categorization are performed, keyword extraction and categorization are performed each time a media web page is browsed, or determined once It is not intended to exclude an embodiment such as continuing to use the keyword set and category. A suitable method can be adopted as appropriate according to the content of the advertisement to be distributed or the content of the medium web page.

FIG. 26 is a flowchart of a process in which the matching unit 244 performs matching between a medium web page and advertisement information.
The matching unit 244 receives the keyword set and the attribution probability for each category from the medium information holding unit 230 (S100). The matching unit 244 includes a distribution destination keyword in which a keyword set extracted from the current medium web page from the advertisement information held in the advertisement information holding unit 252 is held in association with the advertisement information. Advertisement information is selected (S102). Subsequently, the matching unit 244 satisfies the category condition that the attribution probability calculated for each category for the current medium web page among the advertisement information selected in S102 satisfies the category condition held in relation to the advertisement information. Advertisement information is selected (S104). The matching unit 244 passes the selected advertisement information to the advertisement transmission unit 246 or stores it in the distribution result holding unit 242 (S106).

FIG. 27 is a flowchart when the category of the web page is determined in advance.
The page acquisition unit 224 refers to the medium information holding unit 230 and acquires the medium web page according to the URL of the registered medium web page (S110). The keyword extraction unit 228 extracts a keyword set from the text information of the medium web page, and the sentence classification unit 234 calculates the attribution probability for each category (S112). The matching unit 244 selects advertisement information in which the keyword set included in the medium web page includes the delivery destination keyword specified by the advertiser, and the attribution probability for each category of the medium web page satisfies the category condition specified by the advertiser The result is stored in the distribution result holding unit 242 (S114). Thus, if the category of the medium web page is determined in advance, an appropriate advertisement can be distributed even when the medium web page is displayed for the first time by the general user terminal.

FIG. 28 shows a screen 380 of the general user terminal in which the advertisement text 384 is displayed on the medium web page including the text 344 shown in FIG.
When the medium web page is browsed on the general user terminal 206, the keyword set 356 and the category classification result 358 shown in the table 350 of FIG. 23 are passed to the matching unit 244. First, the matching unit 244 selects two items shown in the table 310 of FIG. 20 as advertisement information having the distribution destination keyword “muffler” included in the keyword set of the medium web page. Subsequently, the matching unit 244 acquires the attribution probability 0.76 of the category “fashion” of the medium web page, and compares the category conditions related to the two selected advertisement information. In this case, since the advertisement of the number “1” is “display” for the category “fashion” in the category condition column 319, the matching unit 244 selects the advertisement information of the number “1”. In the advertisement with the number “2”, the attribution probability of the category “fashion” is 0.80 in the category condition column 319. On the other hand, since the attribution probability with respect to “fashion” of the medium web page is 0.76, the matching unit 244 does not select the advertisement information with the number “2”. Therefore, the matching unit 244 outputs only the advertisement information with the number “1”.

  When the advertisement transmission unit 246 transmits this advertisement information to the general user terminal 206, the script program of the general user terminal 206 receives it and displays the advertisement at a predetermined position 382 of the medium web page. As shown in the drawing, the advertisement text 1 “Cashmere muffler 20% off” and the advertisement text 2 “Frozen a nice lady with a popular brand!” Are displayed in a clickable form. When the general user clicks on the advertisement text 384, the URL jumps to the URL shown in the URL column 324 in FIG.

  As described above, according to the advertisement distribution apparatus of the present embodiment, it is possible to classify web pages into categories set in advance and distribute advertisements according to the categories. By specifying the category of the web page to which the advertisement is distributed, the advertiser can display the advertisement on the web page viewed by the customer assumed as the target of the product or service. Therefore, the advertising effect is expected to increase.

  In addition, since it is possible to select display and non-display for each category as the category condition of the advertisement information, the advertiser can specify in advance the content of the web page to which the advertisement is distributed. Moreover, the most effective advertisement can be distributed even to a web page having contents that span a plurality of categories.

  If categories are set and sentence examples are collected at the beginning, there is no need for manual classification for subsequent web page categories or matching with advertisement information. In the conventional category classification, the classification of the categories is performed manually, resulting in individual differences in the classification results, and appropriate classification cannot always be performed. However, this embodiment does not have such a problem.

  In conventional advertisement distribution on a network, a matching method is often used in which an advertiser designates a keyword of a web page on which an advertisement is to be displayed, and the advertisement is displayed on a web page that matches the keyword. Although this method displays an advertisement for a product related to a keyword, it has the effect of increasing the probability of being clicked or purchased by the advertisement, but has the following disadvantages. For example, because only the keywords of the web page are used as a reference, even if the ad is delivered to a web page that makes a negative comment about a product, or the keyword is the same, In some cases, advertisements are distributed on web pages that describe things or services that are completely different. In the former case, the image of the advertiser's company or the product or service being advertised will be degraded, and in the latter case, displaying the advertisement will not be effective, and the advertisement distribution will be wasted End up.

  On the other hand, in addition to extracting a keyword from a web page, the advertisement distribution apparatus according to the present embodiment also determines the category of the web page, and only the advertisement information that satisfies both the keyword and category conditions is included in the web page. Therefore, the advertisement information can be distributed by selecting a web page close to the desire of the advertiser. Therefore, the probability of leading to clicks and purchases of advertisements further increases, which can lead to an increase in advertising effectiveness.

  The present invention has been described based on some embodiments. It is understood by those skilled in the art that these embodiments are exemplifications, and that various modifications can be made to combinations of the respective constituent elements and processing processes, and such modifications are also within the scope of the present invention. By the way.

  It should also be understood by those skilled in the art that the functions to be fulfilled by the constituent elements described in the claims are realized by the individual functional blocks shown in the present embodiment or their linkage.

  In the embodiment, it is described that the advertiser sets both the delivery destination keyword and the category condition for the advertisement information. However, only the category condition may be set.

Depending on how categories and sentence examples are collected in the sentence classification unit, medium web pages can be classified into various categories other than those shown in the embodiment.
For example, a category of “positive text” can be set by preparing a sentence example in which a positive opinion is stated and a sentence example in which a negative opinion is stated in the sentence example storage unit. Combining destination keywords and other categories with "positive text" categories allows advertisers to display advertising information on media webpages that are likely to contain articles that are positive for their products or services Can be made.
Also, by collecting category sentence examples and non-category sentence examples for each age and gender, a category such as “a woman in her 30s is an author” can be set. Furthermore, by collecting category sentence examples and non-category sentence examples for each dialect, it is possible to set a category such as “Osaka dialect web page”.
The above-mentioned categories are all classified for the author of the web page. For example, “sentence examples read by men in their twenties” and “sentence examples not read by men in their twenties” are prepared in the example storage unit, and the categories are classified. For example, it is possible to classify web pages into categories according to readers, such as “web pages that men in their twenties are likely to read”. In this case, since the distribution destination of the advertisement can be set according to the reader, not the writer, it is expected that the effect of the advertisement displayed on the web page is further improved.

In the embodiment, it is described that the advertiser selects the category of the site where the advertisement should be displayed or hidden. However, the provider of the site should be able to display the advertisement that should be displayed or hidden on the site. A category may be selected.
In this case, a reverse nomination unit (not shown) classifies the category for the advertisement text. At this time, the reverse nomination unit may execute the category classification regardless of the category declared by the advertiser. The matching unit finds a combination in which the category permitted on the page manager side matches the category permitted on the advertiser side, and distributes the advertisement information. By doing so, it is avoided that an advertisement of a kind that the administrator does not like is displayed on the medium web page of the page administrator without knowing it.

  In the embodiment, the advertisement distribution apparatus in which the medium page management unit 218, the medium page analysis unit 232, and the advertisement management unit 240 are combined has been described. However, these blocks may operate on different servers. Further, some of these blocks or the entire advertisement distribution apparatus may operate on the provider's server. In this case, the provider may notify the user that advertisements are automatically distributed by the advertisement distribution device to all web pages managed by the provider.

It is a figure which shows an example of the usage condition of the advertisement delivery apparatus which concerns on this Embodiment. It is a functional block diagram of the text classification | category part of FIG. It is a figure which shows the example which decomposed | disassembled text data into the component by morphological analysis. It is a detailed functional block diagram of a classification dictionary creation unit. It is a data structure figure of the dictionary data stored in the classification dictionary holding part. FIG. 6 is a structural diagram of selection criterion data stored in an element refinement criterion providing unit. It is a detailed functional block diagram of a category attribution probability calculation unit. It is a figure which shows the appearance probability of the component about one category. It is a flowchart which shows the process in which a classification dictionary is produced. It is a flowchart which shows the process in which uncategorized text data is classified into a category. It is a figure which shows the dictionary data used in one Example. It is a figure which shows the result of having calculated the appearance probability for every component using the dictionary data of FIG. It is a figure which shows the result of having calculated the belonging probability about each category using the appearance probability of FIG. It is a functional block diagram of the advertisement delivery apparatus concerning this Embodiment. It is a flowchart which shows the outline of the process which delivers the advertisement relevant to a web page by an advertisement delivery apparatus. It is a figure which shows an example of the advertisement reception screen displayed on a display of an advertiser terminal by an advertisement reception part. It is a figure which shows the setting screen which sets the category which should display an advertisement for every category, or the category which should be hidden. For each category, a category setting screen having a lower limit value field for setting a lower limit value of the attribution probability calculated for the medium web page to which the advertisement is distributed and a non-display category field for setting a category of the medium web page to which the advertisement is not distributed FIG. It is a figure which shows the example of a screen for setting the category displayed when a display category button is clicked on an advertisement reception screen. FIG. 20 is a diagram illustrating a table of advertisement information that is input to the advertisement reception unit via the advertisement reception screen of FIG. 16 and any of the category setting screens of FIGS. 17 to 19 and held in the advertisement information holding unit. It is a figure which shows an example of the medium registration reception screen displayed on a page administrator terminal by a medium registration reception part. It is a figure which shows the example of the text information of the page registered as a medium web page. It is a figure which shows the table of the medium information hold | maintained at a medium information holding part. It is a flowchart explaining the registration process of the medium web page by a page manager. It is a flowchart explaining in detail a process in which an advertisement is distributed to a web page. 5 is a flowchart of a process of matching a medium web page and advertisement information in a matching unit. It is a flowchart in the case of determining the category of a web page beforehand. It is a figure which shows the screen of a general user terminal with which the advertisement was displayed on the medium web page containing the article shown to Fig.22 (a).

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 Text classification part, 202 Advertiser terminal, 204 Page manager terminal, 206 General user terminal, 208 Network, 210 Advertisement delivery apparatus, 212 Provider server, 214 Web page, 218 Medium page management part, 220 Medium registration reception part, 222 Script generation unit, 224 page acquisition unit, 226 confirmation unit, 228 keyword extraction unit, 230 medium information holding unit, 232 medium page analysis unit, 240 advertisement management unit, 242 distribution result holding unit, 244 matching unit, 246 advertisement transmission unit, 250 advertisement reception part, 252 advertisement information holding part.

Claims (7)

A device for displaying an advertisement on a media web page provided on a network,
An advertisement information holding unit for holding advertisement information in association with category specification information for specifying category conditions of a medium web page to which an advertisement is distributed;
A page analysis unit for acquiring a media web page via a network and determining a category into which the media web page should be classified from a plurality of categories set in advance based on text information of the media web page; ,
A matching unit that selects, from the advertisement information holding unit, advertisement information in which the category of the medium web page determined by the page analysis unit satisfies the category condition;
When the medium web page is viewed on a terminal connected to a network, comprising an advertisement transmitting section which transmits the advertisement information selected by the matching unit for the terminal in a form to be displayed along with the medium web page, the ,
The advertisement information holding unit holds the advertisement information in association with a delivery destination keyword that specifies a keyword of a medium web page that is an advertisement delivery destination,
The page analysis unit further includes a keyword extraction unit that extracts a keyword set from the medium web page,
The matching unit selects advertisement information in which a delivery destination keyword included in a keyword set of a medium web page is designated from the advertisement information holding unit, and then the category of the web page is selected from the selected advertisement information. An advertisement distribution apparatus that selects advertisement information that satisfies a category condition . The advertisement information holding unit holds the advertisement information in association with category designation information for designating whether or not to display an advertisement on a medium web page classified into the category for each of a plurality of categories set in advance. The advertisement distribution apparatus according to claim 1 . The page analysis unit further includes a sentence classification unit that calculates an attribution probability representing the probability that the medium web page belongs to a plurality of categories set in advance, for each category,
The advertisement information holding unit holds the advertisement information in association with category designation information for designating a lower limit value of the probability of attribution of the medium web page as a delivery destination of the advertisement for each category,
The matching unit compares the attribution probability of the medium web page with a lower limit value of the attribution probability related to the advertisement information, and when the attribution probability of the medium web page is equal to or more than the lower limit value, the corresponding advertisement information The advertisement distribution apparatus according to claim 1 , wherein the advertisement distribution apparatus is selected from the advertisement information holding unit. The advertisement information holding unit holds the advertisement information in association with category designation information that designates a non-display category that should avoid sending an advertisement,
The advertisement distribution apparatus according to claim 3 , wherein the advertisement transmission unit rejects transmission of advertisement information to a medium web page classified into the non-display category. A page management unit for transmitting the script program to the administrator's terminal of the medium web page in order to describe the script program for displaying the advertisement information transmitted from the advertisement transmission unit on the medium web page. In addition,
The script program has a function of transmitting information on the medium web page to the page analysis unit when the medium web page is viewed on a terminal connected to a network;
The advertisement distribution apparatus according to claim 1 , wherein the page analysis unit acquires the medium web page according to information transmitted from the script program. The page management unit generates a code for specifying the administrator, embeds it in the script program, and transmits it.
The script program when said medium webpage a terminal connected to the network was viewed, claims, characterized in that together with the information of the medium web page has a function of transmitting the code to the page analysis unit 5 The advertisement distribution device described in 1. A program that displays advertisements on media web pages provided on a network,
An advertisement information holding function for storing the advertisement information in a predetermined storage area in association with the category designation information for designating the category condition of the medium web page to which the advertisement is distributed;
An acquisition function for acquiring a media web page via a network;
A page analysis function for determining a category in which the media web page should be classified from a plurality of categories set in advance based on the text information of the obtained media web page;
A matching function for selecting, from the storage area, advertisement information in which the category of the media web page determined by the page analysis function satisfies the category condition;
When the medium web page is viewed on a terminal connected to the network, an advertisement transmission function for transmitting the advertisement information selected by the matching function to the terminal in a form to be displayed together with the medium web page;
To the computer ,
The advertisement information holding function includes a function of storing the advertisement information in a predetermined storage area in association with a delivery destination keyword that specifies a keyword of a medium web page that is an advertisement delivery destination,
The page analysis function includes a keyword extraction function for extracting a keyword set from a medium web page,
The matching function selects advertisement information in which a delivery destination keyword included in a keyword set of a medium web page is specified from the predetermined storage area, and then the category of the web page is selected from the selected advertisement information. An advertisement distribution program including a function of selecting advertisement information that satisfies a category condition .

Images