JP5215877B2 - Region characteristic dictionary generation method and apparatus - Google Patents

Description

  The present invention relates to a method and apparatus for generating a regional characteristic dictionary.

  Conventionally, it has been shown that an advertisement distribution system capable of providing an advertisement linked to the content of Web content included in a browsed Web site is provided (for example, Patent Document 1). In the advertisement distribution system described in Patent Literature 1, keywords and topics are extracted from the contents of advertisement insertion target contents on the advertisement insertion website, and are stored in advance in association with the address information of the advertisement insertion target contents. It is shown that when an access request to advertisement insertion target content is made, an advertisement is searched by a keyword and a topic corresponding to the address information of the accessed advertisement insertion target content, and the content with the advertisement is provided to the viewer. .

JP 2007-286833 A

  However, in the method described in Patent Document 1, feature words such as keywords and topics are extracted from the content of the advertisement insertion target content of the advertisement posting website, and advertisements linked to the feature words are only distributed. There is no indication of extracting feature words. For this reason, the advertisement linked to the feature word for each region cannot be distributed.

  The present invention has been proposed in view of such conventional problems, and an object of the present invention is to provide a method and apparatus for extracting a feature word for each region and generating a regional characteristic dictionary related to the feature word. There is.

  The present invention provides the following solutions.

(1) A regional characteristic dictionary generation method for generating a regional characteristic dictionary related to a characteristic word by extracting characteristic words included in a specific regional blog from the bias of appearance frequency of terms,
Extracting terms from a plurality of regional blogs and counting the total number of occurrences of the terms in a specific region;
Counting the number of regional blogs and the number of regional blogs containing the term;
Performing a predetermined calculation based on the total value, the quantity of the regional blog, and the quantity of the regional blog including the term, and calculating a deviation of the appearance frequency of the term in the specific area;
When the calculated deviation is greater than a predetermined threshold, registering the term in the regional characteristic dictionary of the specific region;
A regional characteristic dictionary generation method comprising:

  (1) According to the regional characteristic dictionary generation method described in (1), terms are extracted from a plurality of regional blogs, and the total number of occurrences of the terms in a specific region is counted. Then, the number of regional blogs and the number of regional blogs containing terms are counted. Then, a predetermined calculation is performed based on the total value of the number of appearances of terms, the number of regional blogs, and the number of regional blogs containing the terms, and the deviation of the appearance frequency of terms in a specific region is calculated. When the calculated deviation is larger than a predetermined threshold, the term is registered in the regional characteristic dictionary of the specific region.

  According to such a method, the deviation of the appearance frequency of the term in the specific area can be calculated by performing a predetermined calculation. When the calculated deviation is larger than a predetermined threshold, the term can be registered in the regional characteristic dictionary of the specific region. Therefore, among terms included in regional blogs, terms that appear frequently in regional blogs in a specific region can be used as feature words. Therefore, feature words are extracted for each region, and a regional characteristic dictionary related to feature words. Can be generated.

ただし、Ａ：特定地域
ｔ：用語
Ｄ（Ａ，ｔ）：特定地域Ａにおける用語ｔの出現頻度の偏差
ＴＦ（Ａ，ｔ）：特定地域Ａにおける用語ｔの出現数の合計値
ＤＦ（ｔ）：用語ｔが含まれている地域ブログの数量
Ｎ：地域ブログの数量 (2) The step of calculating a deviation of the appearance frequency of the term performs a predetermined calculation based on the total value, the quantity of the local blog, and the quantity of the local blog including the term by the following formula: The regional characteristic dictionary generation method according to (1), wherein a deviation of the appearance frequency of the term in the specific area is calculated.

However, A: Specific area
t: Term
D (A, t): Deviation in frequency of appearance of term t in specific area A
TF (A, t): Total number of occurrences of the term t in the specific area A
DF (t): Number of local blogs containing the term t
N: Number of regional blogs

  (2) According to the regional characteristic dictionary generation method described in (2), the D (A, t) value increases when TF (A, t) is large and DF (t) is small. That is, when TF (A, t) is large, it is indicated that the term t is frequently used in the specific area A, and when D (A, t) is small, the area where the term t is used. This indicates that the number of blogs is small and the usage frequency of the term t is used in a limited area including the specific area A. Therefore, the deviation D (A, t) of the appearance frequency of the term t in the specific area A can be calculated appropriately.

  (3) The step of counting the total number of occurrences of the terms includes collecting information written by the user within a predetermined period of information written by the user of the regional blog for the plurality of regional blogs. (1) or (2), wherein the term is extracted from the information written by the user within the predetermined period and the total number of occurrences of the term in the specific area is counted. The regional characteristic dictionary generation method described.

  (3) According to the regional characteristic dictionary generating method described in (3), among the information written by a user of the regional blog for a plurality of regional blogs, the information written by the user within a predetermined period is collected, and the predetermined The term is extracted from the information written by the user within the period, and the total number of occurrences of the term in a specific area is counted. By doing in this way, it is possible to monitor terms having a high appearance frequency in a predetermined area within a predetermined period. That is, it is possible to know the fashion of terms used in a specific area.

(4) A family characteristic dictionary generation method for extracting characteristic words included in a family blog composed of a plurality of blogs from bias in appearance frequency of terms and generating a user characteristic dictionary in the family,
Extracting terms from a family blog and counting the total number of occurrences of the terms in the family blog;
Counting the number of blogs of users that make up the family and the number of blogs of the users that contain the term;
A predetermined calculation is performed based on the total value, the number of blogs of the users that constitute the family, and the number of blogs of the users that include the terms, and a deviation in the appearance frequency of the terms in the blogs of the users Calculating steps,
Registering the term in the user's characteristic dictionary if the calculated deviation is greater than a predetermined threshold;
A family characteristic dictionary generation method characterized by comprising:

  (4) According to the family characteristic dictionary generating method described in (4), terms are extracted from a family blog, the total number of occurrences of the term in the family blog is counted, and the number of blogs of users constituting the family and Count the number of user blogs that contain the term. Then, a predetermined calculation is performed based on the total value, the number of users 'blogs constituting the family, and the number of users' blogs including the terms, and the deviation of the appearance frequency of the terms in the user's blog is calculated, If the calculated deviation is greater than a predetermined threshold, the term is registered in the user's characteristic dictionary. According to such a method, it is possible to grasp a feature word used by a specific user in the family.

(5) A regional characteristic dictionary generation device that extracts a characteristic word included in a specific regional blog from a bias in appearance frequency of terms and generates a regional characteristic dictionary related to the characteristic word,
A term appearance counting means for extracting terms from a plurality of regional blogs and counting the total number of occurrences of the terms in a specific area;
Blog number counting means for counting the number of regional blogs and the number of regional blogs containing the term;
Term appearance deviation calculating means for performing a predetermined calculation based on the total value, the number of regional blogs, and the number of regional blogs including the terms, and calculating a deviation of the appearance frequency of the terms in the specific region When,
When the calculated deviation is larger than a predetermined threshold value, a dictionary registration unit that registers the term in an area characteristic dictionary of the specific area;
A regional characteristic dictionary generating device comprising:

  According to such a configuration, the same effect as in (1) can be expected by constructing the device.

  ADVANTAGE OF THE INVENTION According to this invention, the method and apparatus which extract the characteristic word for every area and produce | generate the area characteristic dictionary regarding a characteristic word can be provided.

It is a figure which shows the functional block of the whole area | region structure of the system which concerns on 1st Embodiment, and the regional characteristic dictionary production | generation server 1 used as the core. It is a figure showing tabulation DB21 concerning a 1st embodiment. It is a figure which shows dictionary DB22 which concerns on 1st Embodiment. It is a figure which shows the hardware constitutions of the area characteristic dictionary production | generation server 1 which concerns on 1st Embodiment. It is a flowchart which shows the flow of a process until the characteristic word is registered into an area characteristic dictionary by the area characteristic dictionary production | generation server 1 which concerns on 1st Embodiment. It is a figure which shows the function structure of the advertisement delivery server 4 which concerns on 2nd Embodiment. It is a figure which shows area DB121 which concerns on 2nd Embodiment. It is a figure which shows the term table 122a stored in knowledge DB122 which concerns on 2nd Embodiment. It is a figure showing knowledge table 122b stored in knowledge DB122 concerning a 2nd embodiment. It is a figure which shows advertisement DB123 which concerns on 2nd Embodiment. It is a flowchart which shows the process until the knowledge level for every category is determined by the advertisement delivery server 4 which concerns on 2nd Embodiment. It is a flowchart which shows the process after the browsing terminal 5 which concerns on 2nd Embodiment accesses the advertisement delivery server 4, until an advertisement is delivered. It is a figure which shows the example of a display when an advertisement is delivered with respect to the browsing terminal 5 which concerns on 2nd Embodiment. It is another figure which shows the example of a display when an advertisement is delivered with respect to the browsing terminal 5 which concerns on 2nd Embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
[Entire system configuration]
FIG. 1 is a diagram showing an overall configuration of a system according to the first embodiment and functional blocks of a regional characteristic dictionary generation server 1 that is the core of the system.

  As an overall configuration of this system, a regional characteristic dictionary generation server 1 that generates a regional characteristic dictionary is connected to a plurality of blog servers 2 via a communication network 3. The communication network 3 may be an arbitrary communication network, and may be a LAN (Local Area Network), a WAN (Wide Area Network), or the Internet. The regional characteristic dictionary generation server 1 extracts a term from regional blogs stored in a plurality of blog servers 2 and calculates a deviation of the appearance frequency of the term in a specific region by performing a predetermined calculation. If the threshold is larger than the threshold, the term is registered as a feature word in the regional characteristic dictionary of the specific region.

  The regional blog is a blog that can be classified by region, for example, a blog to which region information is added according to a region category. Also, a blog to which regional information is given as an attribute of a user who uses the blog may be a regional blog because the blog can be classified by region based on the regional information.

[Function configuration]
The regional characteristic dictionary generation server 1 includes a control unit 10 and a storage unit 20. The control unit 10 includes a blog information collection unit 11, a term appearance number counting unit 12, a blog number counting unit 13, a term appearance deviation calculating unit 14, and a dictionary registration unit 15. In addition, the storage unit 20 includes a summary database (hereinafter, the database is referred to as DB) 21 and a dictionary DB 22.

  The blog information collection unit 11 collects regional blog information among the blogs stored in the plurality of blog servers 2. The regional blog information includes the region to which the regional blog belongs, information written on the blog by the user, tags attached to the blog or information written on the blog, and the like.

  In the first embodiment, the information on the local blog is collected without particularly specifying the period, but the present invention is not limited to this. For example, the blog information collection unit 11 may collect information written by the user within a predetermined period from information written by the user of the local blog with respect to the local blog. By doing so, the regional characteristic dictionary generation server 1 can collect information written by the user of the local blog within a predetermined period, so that the term appearance deviation calculating unit 14 and the dictionary registering unit 15 described below can perform predetermined processing. It is possible to monitor terms having a high appearance frequency in a predetermined area within the period. That is, it is possible to know the fashion of terms used in a specific area.

  The term appearance number counting unit 12 extracts terms from the local blog information collected by the blog information collecting unit 11. Then, the number of extracted terms, that is, the total number of occurrences of terms is counted for each region, and the counted result is stored in the tabulation DB 21 (see FIG. 2 described later). More specifically, the term appearance number counting unit 12 extracts a plurality of terms by performing morphological analysis on the information of the local blog collected by the blog information collecting unit 11, and each of the extracted terms is the corresponding region. Count the number contained in the blog information. Further, the counted terms are counted for each region to which the regional blog belongs and stored in the tabulation DB 21.

  In the first embodiment, terms are extracted by performing morphological analysis, but the present invention is not limited to this. For example, a term DB may be provided in the regional characteristic dictionary generation server 1 and the number of terms stored in the regional blog information for the terms stored in the term DB may be counted. Moreover, it is good also as acquiring the dictionary data from a predetermined | prescribed Web server, and using as a target term which counts the noun, verb, etc. which are registered into the said dictionary data.

  The blog number counting unit 13 counts the number of regional blogs including the number of local blogs and terms.

により算出される。
ＴＦ（Ａ，ｔ）は、特定地域Ａにおける用語ｔの出現数の合計値を示す。
ＤＦ（ｔ）は、用語ｔが含まれている地域ブログの数量を示す。
Ｎは、地域ブログの数量を示す。 The term appearance deviation calculation unit 14 performs a predetermined calculation based on the total number of terms appearing in a specific region, the number of regional blogs, and the number of regional blogs containing the terms, and the appearance of terms in the specific region Calculate frequency deviation. When the deviation of the appearance frequency is D (Devation) and the term is t, the deviation D (A, t) of the appearance frequency of the term t in the specific area A is

Is calculated by
TF (A, t) indicates the total number of occurrences of the term t in the specific area A.
DF (t) indicates the number of regional blogs that include the term t.
N indicates the number of regional blogs.

  According to the above formula, the D (A, t) value increases when TF (A, t) is large and DF (t) is small. That is, when TF (A, t) is large, it is indicated that the term t is frequently used in the specific area A, and when D (A, t) is small, the area where the term t is used. It is shown that the number of blogs is small and the usage frequency of the term t is limited to an area including the specific area A. Therefore, the deviation D (A, t) of the appearance frequency of the term t in the specific area A can be calculated appropriately.

  Alternatively, DF (t) may indicate the quantity of the area including the area blog including the term t, and N may indicate the number of all areas (for example, 47 if all areas are prefectures). By doing in this way, the deviation of the appearance frequency of the term for every area can be calculated irrespective of the number of blogs.

により算出される。
ＴＦ（Ｆ，ｔ）は、家族Ｆにおける用語ｔの出現数の合計値を示す。
ＤＦ（ｔ）は、用語ｔが含まれている家族ブログを構成するユーザのブログの数量を示す。
Ｎは、ある家族において家族ブログを構成するユーザのブログの数量を示す。 In the first embodiment, the deviation of the appearance frequency of terms in a region is calculated, but the present invention is not limited to this. For example, for a blog of a family, predetermined based on the total number of occurrences of terms in the family blog, the number of blogs of users who make up the family blog, and the number of blogs of users in the family that include the term It is good also as calculating the deviation of the appearance frequency of the term in a family. In this case, the deviation D (F, t) of the appearance frequency of the term t in the family F is

Is calculated by
TF (F, t) indicates the total number of occurrences of the term t in the family F.
DF (t) indicates the number of blogs of users constituting the family blog including the term t.
N indicates the number of blogs of users who constitute a family blog in a certain family.

  In addition, when calculating the deviation of the appearance frequency of terms in a family, the regional characteristic dictionary generation server 1 needs to provide a family DB indicating the family and the structure of the family in the storage unit 20. Then, the blog information collection unit 11 collects family blog information among the blogs stored in the blog server 2. In addition, the term appearance number counting unit 12 extracts terms from the collected family blog information, counts the total number of terms appearing for each family, and stores the counted result in the tabulation DB 21. In addition, the blog number counting unit 13 counts the number of family blogs and the number of family blogs including terms. Moreover, the dictionary registration part 15 mentioned later registers the said term in a user's characteristic dictionary, when the calculated deviation D (A, t) is larger than a predetermined threshold value. By doing in this way, it can grasp | ascertain about the characteristic word which a specific user uses within a family.

  The dictionary registration unit 15 is a dictionary DB 22 that is a regional characteristic dictionary of a region where the deviation of the appearance frequency of terms is calculated when the deviation of the appearance frequency of terms calculated by the term appearance deviation calculation unit 14 is larger than a predetermined threshold. Register the term. More specifically, it is determined whether or not the appearance frequency D (A, t) of the term t in the specific area A is greater than a predetermined threshold value. If it is larger than the threshold value, the term t is registered in the dictionary DB 22 on the assumption that the term t is a feature word used in a specific area. If it is not larger than the threshold value, the term t is not registered in the dictionary DB 22 because the term t is not a feature word used in a specific area.

  Although the threshold is determined in advance, it may be obtained as follows. That is, when the appearance frequency D (t) of the term t is calculated in all regions, the average value of the appearance frequency D (t) and the standard deviation are obtained. Then, a value obtained by multiplying the standard deviation by a predetermined value and adding it to the average value may be used as the threshold value. For example, when the appearance frequency D (t) follows a normal distribution and the predetermined value is 2, the area having a value larger than the threshold is about 2.3% of the whole. By defining the threshold value in this manner, terms that are used only in a specific region, that is, feature words can be registered in the region characteristic dictionary.

  FIG. 2 is a diagram illustrating the aggregation DB 21 according to the first embodiment. The tabulation DB 21 includes a “term” field indicating the tabulated term, a “region” field indicating the tabulated region, and a “tabulation value” field indicating the tabulated value of the term in the tabulated region. . When the term appearance count counter 12 calculates a term summation value for each region, the summation DB 21 stores the summation value for each region. In addition, the term appearance deviation calculation unit 14 is referred to when the term appearance deviation is calculated.

  FIG. 3 is a diagram illustrating the dictionary DB 22 according to the first embodiment. The dictionary DB 22 is a DB used as a regional characteristic dictionary in which feature words in a region are registered. The dictionary DB 22 includes a “region” field indicating a region where a term determined to be a feature word is used, and a feature. A “feature word” field indicating a word and a “part of speech” field indicating a part of speech of the characteristic word are included. The dictionary DB 22 stores terms that are determined as feature words by the dictionary registration unit 15.

[Hardware configuration of regional characteristic dictionary generation server 1]
FIG. 4 is a diagram illustrating a hardware configuration of the regional characteristic dictionary generation server 1 according to the first embodiment. The regional characteristic dictionary generation server 1 in which the present invention is implemented may be a standard one, and an example of the configuration is shown below.

  The regional characteristic dictionary generation server 1 includes a CPU (Central Processing Unit) 1010 (a plurality of CPUs such as the CPU 1012 may be added in a multiprocessor configuration), a bus line 1005, a communication I / F (I / F). F: Interface) 1040, main memory 1050, BIOS (Basic Input Output System) 1060, display device 1022, I / O controller 1070, input device 1100 such as keyboard and mouse, hard disk 1074, optical disk drive 1076, and semiconductor memory 1078 Prepare. The hard disk 1074, the optical disk drive 1076, and the semiconductor memory 1078 are collectively referred to as the storage unit 20.

  The control unit 10 is a part that controls the regional characteristic dictionary generation server 1 in an integrated manner. By appropriately reading and executing various programs stored in the hard disk 1074, the control unit 10 cooperates with the hardware described above, and Various functions are realized.

  The communication I / F 1040 is a network adapter when the regional characteristic dictionary generation server 1 transmits / receives information to / from the blog server 2 or the like via the communication network 3. The communication I / F 1040 may include a modem, a cable modem, and an Ethernet (registered trademark) adapter.

  The BIOS 1060 records a boot program executed by the CPU 1010 when the regional characteristic dictionary generation server 1 is started, a program depending on the hardware of the regional characteristic dictionary generation server 1, and the like.

  The display device 1022 includes a display device such as a cathode ray tube display device (CRT) or a liquid crystal display device (LCD).

  The I / O controller 1070 can be connected to a storage unit 20 that is a storage device such as a hard disk 1074, an optical disk drive 1076, and a semiconductor memory 1078.

  The input device 1100 is for accepting input by an administrator of the regional characteristic dictionary generation server 1.

  The hard disk 1074 stores various programs for causing the hardware to function as the regional characteristic dictionary generation server 1, a program for executing the functions of the present invention, and the above-described DB. The regional characteristic dictionary generation server 1 can also use an external hard disk (not shown) as an external storage device.

  As the optical disk drive 1076, for example, a DVD-ROM drive, a CD-ROM drive, a DVD-RAM drive, and a Blu-ray Disc (registered trademark) drive can be used. When the optical disk drive 1076 is used, the optical disk 1077 corresponding to the optical disk drive 1076 is used. A program or data may be read from the optical disk 1077 by the optical disk drive 1076 and provided to the main memory 1050 or the hard disk 1074 via the I / O controller 1070.

  The computer in the present invention refers to an information processing apparatus including a storage device, a control unit, and the like, and the regional characteristic dictionary generation server 1 is an information including the control unit 10 and the storage unit 20 as described above. The information processing apparatus is constituted by a processing apparatus and is included in the concept of the computer of the present invention.

  The regional characteristic dictionary generation server 1 is not limited in the number of hardware, and may be configured by one or a plurality of hardware as necessary. In the case of a plurality of hardware, each hardware may be connected via the communication network 3. For example, the functions of the first embodiment may be realized by using separate servers (devices) for each function, and by linking the servers by transmitting and receiving signals between the servers.

[Processing flow]
FIG. 5 is a flowchart showing a flow of processing until a feature word is registered in the regional characteristic dictionary (dictionary DB 22) by the regional characteristic dictionary generation server 1 according to the first embodiment.

  In step S <b> 11, the control unit 10 (blog information collection unit 11) collects regional blog information among the blogs stored in the blog server 2.

  In step S12, the control unit 10 (term appearance number counting unit 12) extracts terms from the information on the local blog collected in step S11.

  In step S13, the control unit 10 (term appearance count counting unit 12) counts the number of extracted terms, that is, the total number of appearances of terms for each area, and counts the terms extracted in step S12. Is stored in the summary DB 21.

  In step S14, the control unit 10 (blog number counting unit 13) counts the number of regional blogs.

  In step S15, the control unit 10 (blog number counting unit 13) counts the number of regional blogs including the term.

  In step S16, the control unit 10 (term appearance deviation calculation unit 14) calculates the above number based on the total value of the number of appearances of terms in the specific area, the number of regional blogs, and the number of regional blogs containing the terms. 2 is calculated, and the deviation of the appearance frequency of the term in the area is calculated.

  In step S17, the control unit 10 (dictionary registration unit 15) determines whether or not the deviation of the appearance frequency of the term calculated in step S16 is larger than a predetermined threshold value. When this determination is YES, the process proceeds to step S18, and when NO, the process of registering the regional characteristic dictionary is ended.

  In step S18, the control unit 10 (dictionary registration unit 15) registers the term in the dictionary DB 22 that is a regional characteristic dictionary of a specific region where the term is a regional feature word and the deviation of the appearance frequency of the term is calculated. .

(Second Embodiment)
In the first embodiment, the regional characteristic dictionary generation server 1 extracts a term from regional blogs stored in a plurality of blog servers 2 and performs a predetermined calculation to thereby change the appearance frequency of the term in a specific region. In the case where it is larger than a predetermined threshold value, the term is registered as a feature word in the regional characteristic dictionary of the specific region. In 2nd Embodiment, the example when the term used as the regional feature word is utilized for the advertisement delivery server of Japanese Patent Application No. 2008-228641 is shown.

  More specifically, the advertisement distribution server 4 according to the second embodiment extracts a category corresponding to the term extracted from the blog server 2 connected via the communication network 3 and used as a characteristic word of a specific area. A category with a high level of knowledge in the region. Then, the advertisement distribution server 4 identifies the area where the browsing terminal 5 is located in response to being accessed from the browsing terminal 5 connected via the communication network 3, and browses the advertisement according to the knowledge level of the area. Deliver to the terminal 5.

[Function configuration]
FIG. 6 is a diagram illustrating a functional configuration of the advertisement distribution server 4 according to the second embodiment.

  The advertisement distribution server 4 includes a control unit 100 and a storage unit 120. The control unit 100 includes a knowledge level determination unit 101, an IP address acquisition unit 102, an access area identification unit 103, a distribution advertisement determination unit 104, and an advertisement distribution unit 105. The storage unit 120 includes a regional DB 121, a knowledge DB 122, and an advertisement DB 123. In addition, although illustration is abbreviate | omitted, suppose that the memory | storage part 120 is equipped with DB of the structure similar to total DB21 and dictionary DB22 with which the regional characteristic dictionary production | generation server 1 of 1st Embodiment is provided.

  The knowledge level determination unit 101 calculates a deviation of the appearance frequency of the term in a specific area by extracting a term from the local blog stored in the plurality of blog servers 2 and performing a predetermined calculation. A term larger than the threshold is a term having a high knowledge level. Then, the category of the term is specified with reference to the term table 122a (see FIG. 8 to be described later), and stored as a category having a high knowledge level in the specific area in the knowledge table 122b (see FIG. 9 to be described later).

  Here, the knowledge level determination unit 101 calculates a deviation of the appearance frequency of the term in a specific region by extracting a term from the regional blog stored in the plurality of blog servers 2 and performing a predetermined calculation. The process of converting a term larger than a predetermined threshold to a term having a high knowledge level is performed. This process includes the blog information collection unit 11, the term appearance count unit 12, and the number of blogs described in the first embodiment. Since it is implemented by the counting unit 13, the term appearance deviation calculating unit 14, and the dictionary registering unit 15, detailed description thereof is omitted. That is, the knowledge level determination unit 101 includes the blog information collection unit 11, the term appearance number counting unit 12, the blog number counting unit 13, the term appearance deviation calculation unit 14, and the dictionary registration unit 15 described in the first embodiment. Is.

  In the second embodiment, the knowledge level determination unit 101 specifies a category having a high knowledge level, but the present invention is not limited to this. For example, each regional feature word stored in the dictionary DB 22 of the region specific dictionary generation server 1 by connecting the advertisement distribution server 4 to the region characteristic dictionary generation server 1 of the first embodiment via the communication network 3. And the term table 122a, a category corresponding to the feature word may be determined as a category having a high knowledge level in each region.

  In the second embodiment, the advertisement distribution server 4 determines a term having a high knowledge level as a term larger than a predetermined threshold and determines a knowledge level of a category corresponding to the term. Absent. For example, a plurality of numerical ranges are defined in the advertisement distribution server 4 instead of a predetermined threshold value. And the advertisement delivery server 4 is good also as determining the knowledge level of the category according to this term according to which numerical range the deviation of the appearance frequency of the term in a region belongs to. By determining the knowledge level in this way, the advertisement distribution server 4 can determine various knowledge levels according to the deviation of the appearance frequency of terms in the region.

  The IP address acquisition unit 102 acquires the IP address of the browsing terminal 5 in response to being accessed from the browsing terminal 5.

  The access area specifying unit 103 specifies the area where the browsing terminal 5 is located based on the IP address acquired by the IP address acquiring unit 102. Specifically, the region according to the IP address acquired by the IP address acquisition unit 102 is specified with reference to the region DB 121 (see FIG. 7 described later).

  The distribution advertisement determination unit 104 determines an advertisement to be distributed to the browsing terminal 5. Specifically, the delivery advertisement determination unit 104 refers to the advertisement DB 123 (see FIG. 10 described later), randomly determines the advertiser of the advertisement to be distributed, and acquires the advertisement category of the advertiser. Here, when the advertiser designates the advertisement distribution for a plurality of categories, the distribution advertisement determination unit 104 randomly acquires any one of the plurality of categories. Then, the distribution advertisement determination unit 104 determines the knowledge level of the category according to the category of the advertisement of the advertiser, which is the category in the region of the browsing terminal 5 specified by the access region specifying unit 103, as shown in the knowledge table 122b (see FIG. 9). In the advertiser's advertisement category, different advertisement data is registered in advance for each knowledge level, so the distribution advertisement determination unit 104 can determine an advertisement to be distributed based on the acquired knowledge level.

  In the second embodiment, the advertiser of the advertisement to be distributed is determined at random. However, the present invention is not limited to this. For example, the adjustment may be performed based on a predetermined rule, such as preferentially selecting an advertisement of an advertiser with a high bid price. In this way, if the advertising bid amount is different among multiple advertisers, the advertiser with the higher advertising bid amount is preferentially selected. Can be delivered.

  Further, the advertisement distribution server 4 may determine the category of the advertisement to be distributed based on the similarity with the content of the web page. That is, the advertisement distribution server 4 receives the content of the Web page displayed on the display unit of the browsing terminal 5, performs morphological analysis on the content of the Web page, and determines the category to which the Web page belongs from the analysis result. It is good as well. In this case, the advertisement distribution server 4 determines whether or not a term belonging to the predetermined category is included in the result of the morphological analysis, and when the word belonging to the predetermined category is included, the Web It is determined that the page belongs to a predetermined category. Thereby, the advertisement distribution server 4 can distribute an advertisement in which the advertisement is in harmony with the content of the Web page and has no sense of incongruity.

  The advertisement distribution unit 105 transmits the advertisement determined to be distributed by the distribution advertisement determination unit 104 to the browsing terminal 5.

  FIG. 7 is a diagram illustrating the regional DB 121 according to the second embodiment. The region DB 121 includes an “IP address” field for storing an IP address and a “region” field indicating the name of the region. The area DB 121 is referred to by the access area specifying unit 103 when acquiring area information based on the IP address.

  FIG. 8 is a diagram illustrating the term table 122a stored in the knowledge DB 122 according to the second embodiment. The term table 122a includes a “category” field classified into a technical field and an academic field, and a term belonging to the category, that is, a “term” field indicating a specific term. The term table is referred to in order to specify a category corresponding to the term in the knowledge level determination unit 101.

  FIG. 9 is a diagram showing the knowledge table 122b stored in the knowledge DB 122 according to the second embodiment. The knowledge table 122b includes a “region” field that indicates the name of the region, a “category” field that is classified into a technical field and an academic field, and a “knowledge level” field that indicates an understanding level of the category. . The knowledge table 122b is referred to when the knowledge level determination unit 101 determines a knowledge level of a category for each region and when the distribution advertisement determination unit 104 determines an advertisement to be distributed.

  FIG. 10 is a diagram illustrating the advertisement DB 123 according to the second embodiment. The advertisement DB 123 includes an “advertisement ID” field that uniquely identifies an advertisement, an “advertiser name” field that indicates the name of the advertiser, a “category” field that is classified into a technical field or an academic field, and an understanding level of the category. And a “display advertisement data” field, which is the content of the advertisement to be distributed. The advertisement DB 123 is referred to when the distribution advertisement determination unit 104 determines an advertisement to be distributed. Here, it is assumed that the advertiser registers in advance different advertisements for each knowledge level and provides the advertisements according to the knowledge level.

[Hardware configuration of advertisement distribution server 4]
Since the advertisement distribution server 4 has the same configuration as the regional characteristic dictionary generation server 1 described in the first embodiment, detailed description thereof is omitted.

[Processing flow]
FIG. 11 is a flowchart illustrating processing until the knowledge level for each category is determined by the advertisement distribution server 4 according to the second embodiment.

  Since the process from step S101 to step S107 is the same as the process from step S10 to step S17 of the process flow of the regional characteristic dictionary generation server according to the first embodiment, the description is omitted. Note that the processing from step S101 to step S107 is performed by the control unit 100 (knowledge level determination unit 101).

  In step S108, the control unit 100 (knowledge level determination unit 101) determines a term having a high knowledge level in the region.

  In step S109, the control unit 100 (knowledge level determination unit 101) refers to the term table 122a and extracts a category corresponding to the term having a high knowledge level determined in step S108.

  In step S110, the control unit 100 (knowledge level determination unit 101) stores the category extracted in step S109 as a category having a high knowledge level in a specific area in the knowledge table 122b.

  FIG. 12 is a flowchart illustrating processing from when the browsing terminal 5 according to the second embodiment accesses the advertisement distribution server 4 until an advertisement is distributed.

  In step S <b> 121, the control unit 100 (IP address acquisition unit 102) acquires the IP address of the browsing terminal 5 in response to being accessed from the browsing terminal 5.

  In step S122, the control unit 100 (access area specifying unit 103) specifies the area where the browsing terminal 5 is located based on the IP address acquired in step S121.

  In step S123, the control unit 100 (distributed advertisement determining unit 104) determines an advertisement to be distributed based on the region of the browsing terminal 5 identified in step S122 and the advertisement DB 123. Specifically, an advertiser is determined at random by referring to the advertisement DB 123. Then, in the area of the browsing terminal 5 identified in step S122, the knowledge level of the category corresponding to the category designated by the advertiser is extracted with reference to the knowledge table 122b of the knowledge DB 122. Then, the advertisement corresponding to the extracted knowledge level is determined as an advertisement to be distributed.

  In step S124, the control unit 100 (advertisement distribution unit 105) distributes the advertisement determined to be distributed in step S123 to the browsing terminal 5.

[Display information display example]
FIG.13 and FIG.14 is a figure which shows a display example when an advertisement is displayed with respect to the display part of the browsing terminal 5 which concerns on 2nd Embodiment.

  The flow until the advertisement shown in FIG. 13 is displayed is shown below. Here, it is assumed that the IP address of the browsing terminal 5 is “210.232. ***. ***.”, The advertiser of the advertisement to be distributed is determined to be “Company A”, and the category is “IT”. .

  When the IP address acquisition unit 102 acquires the IP address of the browsing terminal 5, the access region specifying unit 103 refers to the region DB 121 of FIG. 7 and specifies the region where the browsing terminal 5 is located as “region 1”. Since the advertiser is determined to be “Company A” and the category is “IT” by the distribution advertisement determination unit 104, the knowledge level of “IT” in “Region 1” is determined based on the knowledge table 122b of FIG. It is referred to as “7”, and the advertisement with the advertisement ID “002” is determined to be distributed as the advertisement corresponding to the knowledge level “7”.

  In FIG. 13, the browser 501 is displayed on the display unit 50 of the browsing terminal 5. A screen of the portal site 502 is displayed on the browser 501, and it can be confirmed that an advertisement image 503 a related to a PC (Personal Computer) is displayed on the right side of the portal site 502. The display of the advertisement image 503a is an advertisement corresponding to the advertisement ID “002”, and is an advertisement for a region whose IT knowledge level is “5 to 10”. As described above, when the knowledge level of the specified region is high, an advertisement that is of interest to a user with a high IT knowledge level can be selected and displayed.

  FIG. 14 is another diagram illustrating a display example when an advertisement is distributed to the browsing terminal 5 according to the second embodiment.

  Here, the IP address of the browsing terminal 5 is “210.212. ***. ***”, and the advertiser of the advertisement to be distributed is “Company A” and the category is “IT” as in the description of FIG. ”.

  When the IP address acquisition unit 102 acquires the IP address of the browsing terminal 5, the access region specifying unit 103 refers to the region DB 121 of FIG. 7 and specifies the region where the browsing terminal 5 is located as “region 4”. Then, since the advertiser is determined to be “Company A” and the category is “IT” by the distribution advertisement determination unit 104, the knowledge level of “IT” in “Region 4” is determined by the knowledge table 122b of FIG. Reference is made to “4”, and the advertisement with the advertisement ID “001” is determined as the advertisement to be distributed as the advertisement corresponding to the knowledge level “4”.

  In FIG. 14, the browser 501 is displayed on the display unit 50 of the browsing terminal 5 as in FIG. 13. Then, it is possible to confirm that the screen of the portal site 502 is displayed on the browser 501 and that the advertisement image 503b regarding the PC is displayed on the right side of the portal site 502. The display of the advertisement image 503b is an advertisement corresponding to the advertisement ID “001”, and is an advertisement for a region whose IT knowledge level is “1 to 4”. Thus, when the knowledge level at the specified regional level is low, an advertisement that is of interest to a user with a low IT knowledge level can be selected and displayed.

  The embodiment of the present invention has been described above, but the present invention is not limited to the first embodiment and the second embodiment, and modifications, improvements, etc. within the scope that can achieve the object of the present invention. Is included.

DESCRIPTION OF SYMBOLS 1 Regional characteristic dictionary production | generation server 2 Blog server 3 Communication network 4 Advertisement delivery server 5 Viewing terminal 10 Control part 11 Blog information collection part 12 Term appearance count part 13 Blog number count part 14 Term appearance deviation calculation part 15 Dictionary registration part 20 Storage Part 21 Total DB
22 Dictionary DB
DESCRIPTION OF SYMBOLS 100 Control part 101 Knowledge level determination part 102 IP address acquisition part 103 Access area specific | specification part 104 Distribution advertisement determination part 105 Advertisement distribution part 120 Storage part 121 Area DB
122 Knowledge DB
123 Advertising DB

Claims (5)

An area characteristic dictionary generation method executed by an area characteristic dictionary generation apparatus that extracts a characteristic word included in a specific area blog from a bias in appearance frequency of terms and generates an area characteristic dictionary related to the characteristic word,
The regional characteristic dictionary generation device collects regional information indicating a region to which the regional blog belongs together with terms from a plurality of regional blogs, classifies the regional blogs by region, and calculates the total number of occurrences of the terms for each specific region. Counting step;
The regional characteristic dictionary generating device counting the number of regional blogs and the number of regional blogs containing the term;
The regional characteristic dictionary generation device performs a predetermined calculation based on the total value, the number of regional blogs, and the number of regional blogs including the term, and the frequency of appearance of the term for each specific region. Calculating a deviation;
When the regional characteristic dictionary generation device registers the term as a term having a high degree of understanding in the specific region when the calculated deviation is larger than a predetermined threshold , Registering the knowledge level of the term such that the larger the calculated deviation is, the higher the numerical value is as information used for advertisement delivery ;
A regional characteristic dictionary generation method comprising: The step of calculating the deviation of the appearance frequency of the term performs the predetermined calculation based on the total value, the quantity of the regional blog, and the quantity of the regional blog including the term by the following formula, and The regional characteristic dictionary generation method according to claim 1, wherein a deviation of the appearance frequency of the term in the region is calculated.

However, A: Specific area
t: Term
D (A, t): Deviation in frequency of appearance of term t in specific area A
TF (A, t): Total number of occurrences of the term t in the specific area A
DF (t): Number of local blogs containing the term t
N: Number of regional blogs   The step of counting the total number of occurrences of the term collects information written by the user within a predetermined period from information written by the user of the regional blog for the plurality of regional blogs, The regional characteristic dictionary according to claim 1, wherein the term is extracted from information written by the user within a predetermined period, and a total value of the number of appearances of the term in the specific region is counted. Generation method. A family characteristic dictionary executed by a family characteristic dictionary generation device that extracts characteristic words contained in a family blog composed of a plurality of blogs from the occurrence frequency bias and generates a user characteristic dictionary in the family. A generation method,
Extracting terms from family blogs, collecting information indicating family and family composition, classifying family blogs by family, and counting the total number of occurrences of the terms for each family;
The family characteristic dictionary generating device counting the number of blogs of the users constituting the family and the number of blogs of the users including the term;
The family characteristic dictionary generation device performs a predetermined calculation based on the total value, the number of blogs of users configuring the family, and the number of blogs of the user including the term, and Calculating a deviation of the appearance frequency of the term in the blog;
When the calculated family deviation dictionary has the calculated deviation larger than a predetermined threshold, the term is registered in the user's characteristic dictionary as a term having a high degree of understanding of the user, and the calculated Registering the knowledge level of the term such that the numerical value becomes higher as the deviation is larger as information used for advertisement delivery ;
A family characteristic dictionary generation method characterized by comprising: A regional characteristic dictionary generation device that extracts a characteristic word included in a specific regional blog from a bias in appearance frequency of terms and generates a regional characteristic dictionary related to the characteristic word,
A term appearance number counting unit that collects region information indicating a region to which the regional blog belongs together with terms from a plurality of regional blogs, classifies the regional blog by region, and counts the total number of occurrences of the term for each specific region; ,
Blog number counting means for counting the number of regional blogs and the number of regional blogs containing the term;
Term appearance deviation calculating means for performing a predetermined calculation based on the total value, the number of regional blogs, and the number of regional blogs including the terms, and calculating a deviation of the appearance frequency of the terms in the specific region When,
When the calculated deviation is larger than a predetermined threshold, the term is registered in the regional characteristic dictionary of the specific region as a term having a high degree of understanding in the specific region, and the calculated deviation is larger. A dictionary registration means for registering the knowledge level of the term such that the numerical value is high as information used for advertisement distribution ;
A regional characteristic dictionary generating device comprising:

Images