JP4489668B2 - Information disclosure timing calculation device, program, and method - Google Patents

Description

  The present invention relates to an apparatus for publishing information such as advertisements on a network, and more specifically, an apparatus for calculating timing at which information such as advertisements can be effectively disclosed in a situation where various types of information are disclosed on the network. About.

  In recent years, information distribution on networks has become popular. In particular, the Internet has been recognized as a basis for information disclosure, and various information such as advertisements, news, and stock information has been released. In addition to advertisements, various information such as apology advertisements and IR information have been released via a network, not a paper medium such as a newspaper.

  And the situation that information is complicated by such a flood of information, and precious advertisements are buried in other information, or conversely, there is no noticeable news, and apology information that you do not want to stand out is handled largely There is.

In recent years, in danger of flooding information, attribute information by XML or the like is given to information to be disclosed, and classification and distribution efficiency for each information is being improved.
JP 11-296579 A JP 2003-263561 A

  However, in recent years, the flood of information is so intense that the information on the network is complicated, so it is not possible to objectively grasp what information is currently attracting attention. Therefore, there is a problem that it is not possible to know an accurate publication timing for publishing information.

  An object of the present invention is to solve the above problems and provide a technique for calculating timing for effectively publishing user information based on various information disclosed on the network. .

  The above-mentioned problems are public information storage means for storing public information that is information to be disclosed on the network, and public information that acquires public information that is information disclosed from a predetermined server on the network. An acquisition unit, a disclosure timing value calculating unit that calculates a disclosure timing value of the disclosure target information based on the published information and the disclosure target information, and publishing the disclosure target information on the network based on the disclosure timing value This is solved by an information disclosure timing calculation device characterized by comprising a disclosure determination means for determining whether or not to do so.

  The disclosure timing value calculation means groups the published information acquired by the published information acquisition means for each related information, and based on the number of published information in each group and the publication time of each published information Calculating the degree of attention of the published information for each group, and calculating the degree of similarity between the disclosure target information and the group for each group based on the feature information of the disclosure target information and the feature information of the group. The degree of influence of the published information on the disclosure target information is calculated for each group from the degree and the similarity, and the disclosure timing value of the disclosure target information can be calculated by integrating them. . Here, the publication timing value calculating means may calculate the publication timing value by selecting only a predetermined number of groups from the higher attention degree among the attention degrees calculated for each group. It is also possible to calculate the disclosure timing value by selecting only the group having the attention level equal to or higher than a certain value from the attention levels calculated for each. The feature information is, for example, attribute information or type information.

  Further, the disclosure determining means determines whether or not to disclose the disclosure target information on the network by comparing the disclosure timing value calculated by the disclosure timing value calculating means with a predetermined threshold value. The timing calculation device may further include a public information output unit that uploads the public target information to a predetermined server when the public determination unit determines to disclose the public target information on a network.

  According to the present invention, various types of information disclosed on the network are collected, the degree of attention to the information in the world is calculated, and the timing at which information to be disclosed can be disclosed more effectively according to the degree of attention is calculated. be able to.

  Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings.

(System configuration overview)
FIG. 1 shows a system configuration diagram showing an outline of an embodiment of the present invention. As shown in FIG. 1, this system has a configuration in which an information disclosure timing calculation server 1, a news site 2, a client terminal 3, and a disclosure server 4 are connected to the Internet 5.
The information disclosure timing calculation server 1 is a server that calculates the disclosure timing of information to be disclosed provided from the client terminal 3. The news site 2 is a server that distributes news and various types of information on the Internet, and has a function of placing news such as incidents and accidents that occur from time to time in a state in which transmission is possible.

  In the present embodiment, an example is shown in which the news of news site 2 is used as public information. However, a server that provides a blog or a general homepage in addition to or instead of news You may make it use the content of the blog acquired from, and the information published on the homepage.

  The client terminal 3 is a terminal that transmits advertisements and news (hereinafter referred to as upload target information) that the user thinks should be disclosed to the information disclosure timing calculation server 1. A general PC terminal can be used as the client terminal 3, but any terminal having a communication function such as a mobile phone or a PDA may be used. The public server 4 is a server for publishing the upload target information, and the information disclosure timing calculation server 1 transmits the upload target information to the public server 4 when it is determined that the upload target information may be disclosed. To do. In addition, it replaces with providing the public server 4, and the information disclosure timing calculation server 1 itself may be provided with the function of a public server.

(System operation overview)
In the system shown in FIG. 1, the information disclosure timing calculation server 1 regularly acquires news article information from the news site 2 and classifies and accumulates articles. Then, in response to receiving the upload target information from the client terminal 3, the information disclosure timing calculation server 1 performs a predetermined calculation using the upload target information and the accumulated news article information, and the current time is the upload target information. It is determined whether or not the target information is to be disclosed. If the timing is to be disclosed, the upload target information is uploaded to the disclosure server 4. If the present time is not the timing to be disclosed, it is determined again whether the upload target information is at the timing to be disclosed after a preset time. This determination is repeated until the upload target information is released. Alternatively, it is repeated a predetermined number of times.

  Note that the upload target information is stored in a storage medium such as a floppy (registered trademark) disk, delivered to the information disclosure timing calculation server 1 by physical means such as mail, and stored in the information disclosure timing calculation server 1. May be.

(Configuration outline of the information disclosure timing calculation server 1)
FIG. 2 shows a functional configuration diagram of the information disclosure timing calculation server 1. As shown in FIG. 2, the information disclosure timing calculation server 1 includes an information acquisition unit 11, a calculation unit 12, an input / output IF unit 13, an information classification unit 14, an article DB 15, an upload target information storage unit 16, and a dictionary DB 17. ing. The function outline of each part is as follows.

  The information acquisition unit 11 acquires information from the news site 2 and the client terminal 3. The information classification unit 14 classifies the article information acquired from the news site 2 and stores it in the article DB 15. The upload target information storage unit 16 stores the upload target information acquired from the client terminal 3. The dictionary DB 17 stores the similarity between article information as a matrix.

  The calculation unit 12 uses the information stored in the article DB 15 and the dictionary DB 17 to perform a calculation for determining whether the upload target information stored in the upload target information storage unit 16 is at an appropriate publication timing. . The input / output IF unit 13 is used by an administrator of the information disclosure timing calculation server 1 to obtain an output result, input initial setting information to the calculation unit 12, input / output and rewrite of information in the article DB 15 and the dictionary DB 17, and the like. It has a user interface such as a display and a keyboard to perform.

  The information disclosure timing calculation server 1 is a general computer including a CPU, a memory, a hard disk, and the like. The information acquisition unit 11, the calculation unit 12, and the information classification unit 14 execute the program according to the present invention on the computer. It is realized by doing. Various DBs and storage units can be realized by a storage device such as a hard disk.

(Details of each functional part)
Hereinafter, each functional unit of the information disclosure timing calculation server 1 will be described in detail. Further, the calculation method of the information disclosure timing will be described in detail in the detailed description of each functional unit.

[Information acquisition unit 11]
The information acquisition unit 11 has a communication function such as TCP / IP and communicates with the news site 2 and the client terminal 3 to receive various news article information and advertisement information to be posted.

  Here, the news article information sent from the news site 2 to the information disclosure timing calculation server 1 will be described with reference to FIG. FIG. 3 shows an example of electronic information of news article information distributed from the news site 2 via the Internet.

  As shown in FIG. 3, the news article information distributed from the news site 2 includes information such as article text information (<zokusei>) and date and time information (<time>) in addition to the news body article (<content>). Various information necessary for characterizing and classifying the news is included.

  In addition, in the example of FIG. 3, in addition to the attribute information, the type information (<species>) is defined and distributed from the news site 2, but when such information is not added at the news site 2, The type information may be extracted from the content information (<content>) of the news article information received by the information disclosure timing calculation server 1. As a technique for extracting content type information from content information, for example, there are techniques described in JP-A-6-215049 and JP-A-7-219953.

  Similarly, the news site 2 may add the subject information (<shutai>) corresponding to the subject of the news article information, or the information disclosure timing calculation server 1 that has received the information may use the content information (<content>). It does not matter if it is extracted. As a method for extracting subject information from content information, for example, a proper noun is extracted by a hidden Markov model, and an algorithm that regards a character string in which the proper noun is followed by “ga” or “ha” as the subject is information disclosure timing calculation A method for incorporating the server 1 into a program for operating the server 1 is conceivable.

  In order for the information acquisition unit 11 of the information disclosure timing calculation server 1 to obtain such news article information and the like, the news site 2 periodically distributes the news article information to the information disclosure timing calculation server 1. In this case, the IP address of the information disclosure timing calculation server 1 is stored in the address storage unit of the server constituting the news site 2.

  As another method, an RSS reader is embedded in the information acquisition unit 11 of the information disclosure timing calculation server 1, the address of each news site 2 is stored in the RSS reader, and the information disclosure timing calculation server 1 is stored. May regularly access the news site 2 to obtain news article information.

  Next, upload target information acquired from the client terminal 3 will be described with reference to FIG. FIG. 4 is a diagram illustrating an example of the configuration of the upload target information.

  As shown in FIG. 4, the upload target information is tagged for each information unit in the same way as news article information, and in addition to the body article (<content>), the article attribute information (<zokusei>) Various information necessary for classification is included.

  These tag information may be given in advance by the client terminal 3, or the content information of the received article information (like the news article acquired from the news site 2 by the information disclosure timing calculation server 1) ( Attribute information and type information may be extracted from <content>).

  The upload target information includes a <flag> tag and a <koukaichi> tag. <Flag> Kugu is flag information for identifying whether the user who uses the client terminal 3 wants to “publish the upload target information positively” or “not so”. The user of the client terminal 3 inputs 1 to the flag when he / she wants to actively disclose the upload target information, and inputs 0 otherwise. This information is used when the calculation unit 12 calculates the information disclosure timing value (φ). The <koukaichi> tag is input with a threshold value for determining how many times the information disclosure timing value (φ) is to be uploaded to the public server. Details of these will be described later. The <URL> tag is a tag that describes the address of the public server 4 that uploads the upload target information.

  Now, the information acquisition part 11 which acquired news article information from the news site 2 confirms whether the information classification | category part 14 has sufficient information for classifying news article information. Whether or not all the information necessary for classification is included in the news article information is defined in advance in the information acquisition unit 11 tag information necessary for classification, all the tag information is aligned in the news article information Judging by whether or not. When it is determined that sufficient information for classification is not prepared in the acquired news article information, necessary information is extracted from the content information in the news article information by the method described above.

  For example, when the information necessary for classifying information is <title>, <zokusei>, <species>, <shutai>, the information acquisition unit 11 has the same tag information, that is, <title>. , <Zokusei>, <species>, <shutai> are stored. For example, when the information shown in FIG. 3 is acquired as news article information, the information acquisition unit 11 determines that <shutai>, that is, the subject information is insufficient, and extracts subject information by the above-described hidden Markov model method. , Add to news article information.

  When the acquired information is information to be uploaded from the client terminal 3, the information acquisition unit 11 accumulates the information in the upload target information storage unit 16 and is normal news article information obtained from the news site 2. After confirming necessary information, it is passed to the information classification unit 14.

[Information classification unit 14]
Next, the information classification unit 14 will be described. The information classification unit 14 extracts information such as attribute information, type information, and subject information from the acquired news article information and the like, classifies the news article information using these, and stores the classified article information in the article DB 15.

  FIG. 5 shows an example of the contents of the article DB 15. As shown in FIG. 5, the case ID is assigned to the news article information and stored in the article DB 15 together with various tag information such as time information and content information included in the news article information.

  The incident ID of the newly received news article information is assigned according to the past news article information. That is, if there is news article information similar to the newly received news article information in the news article information received in the past, the same event ID as the incident ID of the past news article information is assigned to the new news article information To do. Further, when similar news article information has not been received in the past, a new case ID is assigned to the newly received news article information. In this way, the information classifying unit 14 groups news article information relating to similar cases, and associates one group with one case ID.

  The information classification unit 14 determines whether or not the received news article information and the past news article information are similar news article information based on whether or not some of the tag information is the same. For example, when the attribute information (zokusei), type information (species), and subject information (shutai) match, it may be determined to be the same article, or the attribute information and type information match. If the type information and the subject information match, it may be determined that the article is similar.

  Next, an example of the processing operation of the information classification unit 14 when new news article information is received will be described. When the news article information shown in FIG. 5 is already stored in the article DB 15, it is assumed that new news article information 151 is input to the information classification unit 14. Here, it is determined that the case is the same when the attribute information, the type information, and the subject information match.

  The contents of the new article 151 are time information (<time> 2005.6.5), attribute information (<zokusei> company), type information (<species> information leakage), subject information (<shutai> ABC Inc.), content information (<Content> ... omitted). That is, the attribute information of the new news article information 151 is a company, the type information is information leakage, the subject information is ABC Corporation, and whether or not news article information having the same attribute information, type information, and subject information has existed in the past. When the article DB 15 is searched, news article information matching the condition with the case ID: 1002 is found. Therefore, the information classification unit 14 assigns the case ID 1002 to the new news article information 151 and then stores the news article information 151 in the article DB 15.

  By storing the case ID in this way and storing the information, the same case ID, that is, an article describing the same case can be extracted instantaneously. For example, as described later, the attention degree of the case is calculated. Thus, the number of articles for each case ID can be calculated immediately.

[Calculation unit 12: Timing calculation]
Next, a procedure for calculating the information disclosure timing value, which is a value used by the calculation unit 12 to determine the upload information disclosure timing, will be described. The calculation unit 12 compares the information disclosure timing value with a predetermined threshold value to determine whether the current time is at the information disclosure timing.

  As for the information to be disclosed, there are information (A) that is desired to be actively disclosed so as to be noticed by the public, and information (B) that is information that should be disclosed but is not desired to be noticed so much. The information of (B) is information such as an internal scandal that should not be publicized as much as the social responsibility. From this point of view, this embodiment uses different information disclosure timing value calculation methods in the case of (A) and the case of (B). Note that (A) or (B) is designated by the user who provides the upload target information.

  In calculating the information disclosure timing value, an article attention degree that is an index value indicating the degree of attention of news article information and an article indicating the degree of similarity between the upload target information and the news article information stored in the article DB 15 Although the similarity is used, a specific method of obtaining the case attention degree and the article similarity will be described later. Hereinafter, the calculation method of the information disclosure timing value will be described for each of the cases (A) and (B).

<In the case of (A)>
In the case of information that is desired to be publicly disclosed in order to attract attention to the public, y (i) is the case attention degree of the case whose case ID is i, and δ (i) is the article similarity degree between the case and the upload target information. In this case, the calculation unit 12 calculates an information disclosure timing value based on the following calculation formula.

Information disclosure timing value (φ) = Σ (incident attention degree (y (i)) × article similarity (δ (i))) (Σ is the sum of i) Equation 1
Equation 1 is the product of the product of the case attention degree for a case i, the degree of similarity between the upload target information acquired from the client terminal 3 and the case i, and the sum of each case i is the information disclosure. It means a timing value (φ).

  In other words, as it is obvious from the product of the case attention degree and the article similarity, articles that have high similarity and highly correlated articles are highly evaluated and articles that are not very relevant Is underestimated.

  Therefore, when the information disclosure timing value (φ) calculated in this way is high, it is a time when similar incidents are attracting attention. This indicates that it is not suitable for the timing of information disclosure for users who think that. Conversely, when the value of the information disclosure timing (φ) is low, there is no case that has attracted attention in similar articles. Therefore, as a timing for information disclosure for users who want to pay attention to their information. Demonstrate what is appropriate.

<In the case of (B)>
In the case of information that should be disclosed but is not desired to be noticed, the calculation unit 12 calculates an information disclosure timing value based on the following calculation formula.

Information disclosure timing value (φ) = Σ (incident attention degree (y (i)) / article similarity degree (δ (i))) (Σ is the sum of i) Equation 2
Formula 2 divides the attention degree of an incident for an incident i by the degree of similarity between the upload target information acquired from the client terminal 3 and the incident i, and summed up each incident i to obtain the information disclosure timing value (Φ).

  In other words, as apparent from the fact that the incident attention is divided by the article similarity, articles with high similarity and articles with high correlation are evaluated low, and articles with little relevance are highly evaluated.

  Therefore, when the value of the information disclosure timing value (φ) calculated in this way is high, it means that an incident that is not related to the upload target information to be disclosed is attracting attention. For users who think that it is not, it is appropriate for the timing of information disclosure. Conversely, when the value of the information disclosure timing (φ) is low, there is no case that has been noticed, or there is a possibility that a similar case has been noticed. Indicates that it is not appropriate for the timing of information disclosure.

  In addition, in the case of bad news that you do not want to publish so much, if there is a similar attention case, you may get noticed as if you are caught in that case, so there is another case that is very much attention However, if it is a case similar to the information disclosed, it is not appropriate for the information disclosure timing. In Formula 2, similar incidents are underestimated from such intentions.

  Note that Formula 2 evaluates a similar article low by dividing the case attention degree (y (i)) by the article similarity (δ (i)), but can handle an article with a high article similarity low. Any case may be used, for example, the case attention degree (y (i)) × (1-article similarity degree (δ (i))) may be calculated.

<About timing judgment>
Whether to apply Formula 1 or Formula 2, that is, in the case of (A) or (B), whether the calculation unit 12 calculates the information disclosure timing value or not is the upload target information Judgment is made based on the information of the <flag> tag in the upload target information stored in the storage unit 16.

  Further, by comparing the information disclosure timing value (φ) obtained as described above with the value of the <koukaichi> tag in the upload target information stored in the upload target information storage unit 16, It is determined whether or not it is the disclosure timing of the information.

  In addition, when the calculation part 12 calculates Formula 1 or Formula 2, you may take the sum total of the case regarding all the case IDs stored in article DB15, For example, the value of case attention degree y (i) It is preferable to take a predetermined number (for example, about 20) from the largest and take the sum, or extract only the case attention degree y (i) greater than a certain value and take the sum.

  For example, if only information having a high value is selected from the degree of attention y (i) and the information disclosure timing value (φ) is calculated, a larger information disclosure timing value (φ is obtained when the attention bias of the world is large. ) And more accurately reflect the concentration and dispersion of public interest. In addition, what is necessary is just to set to the calculation part 12 previously receiving the input from the input-output interface 18 how to take a sum total.

[Calculation unit 12: Attention level calculation]
Next, an example of a method for calculating the case attention degree y (i) for each case ID by the calculation unit 12 will be described. In the present embodiment, with regard to news article information of a certain case ID, attention is calculated by paying attention to the newness of the news article information and the large number of news article information. That is, if a large number of news article information has been issued in the past near the present time for a certain case ID, it is determined that the degree of attention of the news article information of that case ID is high. Specifically, the calculation is performed as follows.

If there is N pieces of news article information with case ID: i = 1001, and the attention degree of the j-th article is x (j), the attention degree of the news article information will fade with time. Is taken into consideration, the degree of attention is expressed by an attenuation function relating to t (j), where t (j) is the difference between the current time and the news article information announcement time. Any function can be used as long as it is an attenuation function. For example, if e is used for this attenuation function,
x (j) = a * e- ^{(t (j) -b)} (a and b are constants)...
It can be expressed as. In this embodiment, the sum of x (j) for articles with the same case ID: i is set as the case attention degree y (i). That means
y (i) = Σx (j) = Σa × e− ^{(t (j) −b)} (a and b are constants, Σ is the sum of j).
It becomes.

This will be described more specifically with reference to FIG. FIG. 6 shows a situation (N = 3) in which there are three news article information with the case ID: i = 1001, and shows the case where the news article information announcement times are different.
Since the j-2nd article shown in (a) has passed a long time since its publication, the value of the attention degree x (j-2) of the j-2nd article is very small. On the other hand, the j-th article shown in (c) has a large value of x (j) since not much time has passed since publication. In this case, the attention degree y (i) of the case with the case ID: i = 1001 is the sum of x (j−2), x (j−1), and x (j) from Equation 4.

  According to the above equation 4, the value of y (i) increases as the number N of articles increases, and the value of y (i) increases as the number of articles has not passed much. In other words, the newer and more incidents are the ones with higher attention.

[Calculation unit 12: Similarity calculation]
Next, a method for calculating the article similarity δ (i) by the calculation unit 12 will be described. The article similarity δ (i) is a numerical value representing the degree of similarity, that is, the correlation between the news article whose case ID is i and the upload target information. In the present embodiment, the degree of similarity between certain news article information and upload target information is obtained using matrix information representing the correlation between news article information and upload target information stored in the dictionary DB 17. A specific example of the similarity calculation method will be described below.

  FIG. 7 shows an example of matrix information stored in the dictionary DB 17. When calculating the similarity between the news article information with the case ID i and the upload target information, the calculation unit 12 acquires the attribute information of the news article information with the case ID i from the storage DB 15. For example, the attribute information of news article information with case ID: i = 1002 in FIG. 5 is “company”. Next, the calculation unit 12 acquires the attribute information of the upload target information from the upload target information storage unit 16. For example, “company” is acquired as attribute information in the case of the upload target information shown in FIG.

  Next, the calculation unit 12 refers to the matrix information in the dictionary DB 17 using these two pieces of information as keys, and extracts the similarity between the news article information and the upload target information. In the above example, since the attribute information is “company” for both news article information and upload target information, the calculation unit 12 acquires 1.0 as the similarity between the two. In this way, the similarity between the news article information of a certain case ID: i and the upload target information is calculated.

  Here, the similarity is calculated based on the correlation of only the attribute information. However, the dictionary DB 17 stores each matrix of type information and subject information, and calculates the similarity including the type information or subject information. Also good. As a method of calculating the similarity based on a plurality of tag information, for example, it is conceivable to take the product of each similarity such as similarity of attribute information × similarity of type information × similarity of subject information.

[Flow of operation]
Next, the operation of the information disclosure timing calculation server 1 having the functional units described so far will be described with reference to the flowchart of FIG.

  The information acquisition unit 11 periodically acquires news article information from the news site 2 in advance, and the information classification unit 14 assigns an incident ID to the acquired article and stores it in the storage DB 15 (step 100).

  When the information acquisition unit 11 receives new upload target information from the client terminal 3, the information acquisition unit 11 stores the received upload target information in the upload target information storage unit 16 (step 101). When the calculation unit 12 recognizes that there is some information in the upload target information storage unit 16, the calculation unit 12 clears a timer for periodically calculating the information disclosure timing value (step 102). For example, a counter that increments at regular intervals can be used as the timer.

  Next, the calculation unit 12 acquires attribute information of news article information of a certain case ID from the article DB 15 and further acquires attribute information of upload target information from the upload target information storage unit 16. And the calculation part 12 calculates the similarity degree of upload object information and the news article information of the said case ID by referring dictionary DB17 using those attribute information (step 103).

  Next, based on the time data of each news article information of the case ID in the article DB 15 and the current time information extracted from the clock of the information disclosure timing calculation server 1, the calculation unit 12 creates a group of the case ID. Is calculated (step 104). In this way, the similarity and attention level for each case ID are calculated for all case IDs (step 105).

  Next, the calculation unit 12 acquires information on the <flag> tag from the target upload target information in order to determine whether the information is to be actively disclosed, and based on the tag information, the formula 1 or the formula 2 And information disclosure timing value is calculated (step 106).

  Next, the calculation unit 12 acquires <koukaichi> tag information as information disclosure threshold information from the upload target information, and compares the tag information with the information disclosure timing value calculated in Step 106. When the <flag> tag is 1, that is, when the upload target information is information that is to be actively disclosed, it is determined whether the information disclosure timing value calculated using Equation 1 is equal to or less than a threshold value, and <flag If the tag is 0, that is, if the upload target information is not information to be actively disclosed, it is determined whether or not the information disclosure timing value calculated using Expression 2 is equal to or greater than a threshold value (step 107).

  When the calculation unit 12 determines that this condition is satisfied (Yes in step 107), the upload target information and the upload destination address (<URL>) are extracted from the upload target information storage unit 16 and uploaded via the information acquisition unit 11. Send the target information to the specified upload destination address (public server).

  When the calculation unit 12 determines in step 107 that the above condition is not satisfied (No in step 107), it is determined whether or not the timer has reached a predetermined time (step 108). If the predetermined time has not been reached, wait until the predetermined time is reached. If the predetermined time has been reached, the processing from step 102 is repeated. That is, the information disclosure timing is determined again while the time has passed and the situation has changed, for example, new news article information has been accumulated. In this way, the processing is repeated until the information is disclosed. Alternatively, if the information disclosure condition is not satisfied even after repeating the process a predetermined number of times, the client terminal 3 may be notified to that effect, and an operation according to an instruction from the user may be performed.

  Thus, the public information timing calculation server 1 collects various current affairs information published on the Internet, and can measure the degree of interest of the case from the number of articles related to the same case. Then, it is possible to calculate the bias of interest in the world from the degree of interest and upload the information at the optimal timing to disclose its own information.

  Also, by calculating the degree of similarity between the information to be uploaded and each case, it is possible to select when the similar case is attracting attention or when it is not attracting attention and disclose the upload target information of the user.

  In particular, when the information to be uploaded is information that you want to actively disclose, you can choose when there is no information of interest in the world. In addition, when the upload target information is information that you do not want to actively spread, you can choose to release information when there is no similar interest information in the world.

  The present invention is not limited to the above-described embodiment, and various modifications and applications can be made within the scope of the claims.

1 is a system configuration diagram showing an outline of an embodiment of the present invention. 3 is a functional configuration diagram of an information disclosure timing calculation server 1. FIG. It is a figure which shows the example of the news article information sent with respect to the information disclosure timing calculation server 1 from the news site 2. It is a figure which shows the example of the upload object information acquired from the client terminal. It is a figure which shows an example of the content of article DB15. It is a figure for demonstrating the calculation method of attention degree. It is a figure which shows an example of the matrix information memorize | stored in dictionary DB17. 4 is a flowchart for explaining the operation of the information disclosure timing calculation server 1.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Information disclosure timing calculation server 2 News site 3 Client terminal 4 Public server 5 Internet 11 Information acquisition part 12 Calculation part 13 Input / output IF part 14 Information classification part 15 Article DB
16 Upload target information storage unit 17 Dictionary DB
151 News article information

Claims (8)

A disclosure target information storage means for storing disclosure target information which is information to be disclosed on the network;
Published information acquisition means for acquiring published information that is information released from a predetermined server on the network;
A disclosure timing value calculating means for calculating a disclosure timing value of the disclosure target information based on the published information and the disclosure target information;
An information disclosure timing calculation device comprising: a disclosure determination unit that determines whether to disclose the disclosure target information on a network based on the disclosure timing value;
The publication timing value calculating means includes
The published information acquired by the published information acquisition means is grouped for each related information, and the degree of attention of the published information is determined based on the number of published information in each group and the publication time of each published information. Calculate for each group,
Based on the feature information of the disclosure target information and the feature information of the group, the similarity between the disclosure target information and the group is calculated for each group,
The degree of influence of the published information on the disclosure target information is calculated for each group from the attention level and the similarity, and the disclosure timing value of the disclosure target information is calculated by integrating them. Information disclosure timing calculation device. The publication timing value calculating means includes
The information disclosure timing calculation device according to claim 1 , wherein the disclosure timing value is calculated by selecting only a predetermined number of groups from the higher attention degree among the attention degrees calculated for each group. The publication timing value calculating means includes
2. The information disclosure timing calculation apparatus according to claim 1 , wherein, among the attention degrees calculated for each group, only the group having the attention degree of a certain value or more is selected to calculate the disclosure timing value. The disclosure determination unit determines whether to disclose the disclosure target information on the network by comparing the disclosure timing value calculated by the disclosure timing value calculation unit with a predetermined threshold,
If the public interest information is determined to be published on the network by the public judgment means, we claim 1, characterized in that further comprising a public information output means for uploading the public interest information to a predetermined server 3 The information disclosure timing calculation apparatus according to any one of the above. Computer
Disclosure target information acquisition means for acquiring the disclosure target information from the storage device storing the disclosure target information that is information to be disclosed on the network;
Published information acquisition means for acquiring published information that is information released from a predetermined server on the network;
A disclosure timing value calculating means for calculating a disclosure timing value of the disclosure target information based on the published information and the disclosure target information;
A program for functioning as a disclosure determining unit that determines whether to disclose the disclosure target information on a network based on the disclosure timing value,
The publication timing value calculating means includes
The published information acquired by the published information acquisition means is grouped for each related information, and the degree of attention of the published information is determined based on the number of published information in each group and the publication time of each published information. Calculate for each group,
Based on the feature information of the disclosure target information and the feature information of the group, the similarity between the disclosure target information and the group is calculated for each group,
The degree of influence of the published information on the disclosure target information is calculated for each group from the attention level and the similarity, and the disclosure timing value of the disclosure target information is calculated by integrating them. Program to do. The disclosure determination unit determines whether to disclose the disclosure target information on the network by comparing the disclosure timing value calculated by the disclosure timing value calculation unit with a predetermined threshold,
The computer according to claim 5, further causing the computer to function as public information output means for uploading the public object information to a predetermined server when it is determined by the public determination means that the public object information is to be disclosed on a network. Program . An information disclosure timing calculation device calculates information disclosure timing,
The disclosure target information acquisition means provided in the information disclosure timing calculation device acquires the disclosure target information from a storage device that stores the disclosure target information that is information to be disclosed on the network;
The published information acquisition means provided in the information disclosure timing calculation device acquires published information that is information published from a predetermined server on the network;
A step of calculating a disclosure timing value of the disclosure target information based on the published information and the disclosure target information, by a disclosure timing value calculating means provided in the information disclosure timing calculation device ;
A method for determining whether or not to publish the information to be disclosed on a network based on the disclosure timing value, wherein the disclosure determination means provided in the information disclosure timing calculation device comprises:
The publication timing value calculating means includes
The published information acquired by the published information acquisition means is grouped for each related information, and the degree of attention of the published information is determined based on the number of published information in each group and the publication time of each published information. Calculate for each group,
Based on the feature information of the disclosure target information and the feature information of the group, the similarity between the disclosure target information and the group is calculated for each group,
The degree of influence of the published information on the disclosure target information is calculated for each group from the attention level and the similarity, and the disclosure timing value of the disclosure target information is calculated by integrating them.
A method characterized by that. The disclosure determination unit determines whether to disclose the disclosure target information on the network by comparing the disclosure timing value calculated by the disclosure timing value calculation unit with a predetermined threshold,
When the disclosure determination unit determines that the disclosure target information is disclosed on the network, the disclosure information output unit included in the information disclosure timing calculation device uploads the disclosure target information to a predetermined server.
The method according to claim 7 .

Images