JP2006228141A - Information recommendation system, management apparatus, information recommendation method and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an information recommendation system, a management apparatus, an information recommendation method and a program for appropriately managing keywords. <P>SOLUTION: An information terminal 1 extracts a first keyword from transmitted and received e-mail, and transmits it to a service provider system 2, which refers to a synonym dictionary to produce list information including a synonym of the first keyword received from the information terminal, determines a group that the list information belongs to, detects a second keyword effective in a group, and transmits information including the keyword to the information terminal 1. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an information recommendation system, a management device, an information recommendation method, and a program using personal attribute information.

  Conventionally, in order to obtain effective information from a large number of received mails or to process them preferentially, it has been necessary to check one e-mail at a time by human eyes.

  Here, a keyword is extracted by a URL keyword extraction unit from the contents of a home page that is regularly browsed by an Internet user, and a new keyword is derived by learning the keyword by a keyword learning unit, and the home page matches the new keyword. After the URL is searched by the URL search unit, the auto pilot execution unit downloads the home page required by the user automatically, so that the home page required by the user can be easily downloaded. A technique that enables downloading is proposed (see, for example, Patent Document 1).

In addition, a user profile P representing the user's preference feature is provided, and the user profile is automatically updated from the document vector H (i) representing the home page feature and the user's browsing time t (i). The document vector H (i) is obtained by combining a static document vector Hs (i) registered in advance with a dynamic document vector Hd (i) created from the homepage text. In addition, by dynamically adjusting the parameters α, β, and γ related to the synthesis of the document vector and the update of the user profile from the user's homepage browsing history, without burdening the user on thinking and operation, There has been proposed a technique that makes it possible to easily select information desired by a user from information on a plurality of home pages (see, for example, Patent Document 2).
JP 2000-172710 A JP 2002-342366 A

  However, the conventional example described above has the following problems.

  Conventional services using personal attribute information have a problem that information provision and information processing do not match actual personal attribute information as a fundamental problem. The background of this problem is as follows.

  In the prior art, it was possible to recommend information based on personal attributes, which are internal preference information such as personal interests. However, it is necessary to register these personal attributes in advance through questionnaires, etc. The process has a first problem that an individual has to input.

  In addition, in the registration work, considering the ease of registration and reducing complexity, the category of category and category becomes representative type and abstract expression, and is organized as information different from the original preference information, Even when attribute information is used, there is a second problem that information that does not match the actual interest of the individual is provided.

  In addition, since preference information such as personal interest changes in time series, registration processing is performed at the time of initial introduction, and after that, it may not be changed or cannot be changed, so it is processed with attribute information registered in the past. There was a third problem that it did not match the current preferences.

  In addition, when advertising distribution using personal attributes is performed, there is a fourth problem that the interest in advertising is low and the advertising effect is reduced because the types and categories of personal attributes are abstract. .

  Therefore, an object of the present invention is to provide an information recommendation system, a management device, an information recommendation method, and a program that can appropriately manage keywords.

  The invention according to claim 1 is an information recommendation system including an information terminal connected via a network and a management device, wherein the information terminal extracts a first keyword from the transmitted / received e-mail. Means for transmitting to the management device, the management device refers to the synonym dictionary, generates a list information including the synonym of the first keyword received from the information terminal, and to which group the list information belongs An information recommendation system comprising: means for determining; means for detecting a second keyword effective in a group; and means for providing information including the second keyword to an information terminal.

  According to a second aspect of the present invention, in the information recommendation system according to the first aspect, in the case of a received mail, the information terminal is provided with means for determining whether or not a reference is made.

  According to a third aspect of the present invention, in the information recommendation system according to the first or second aspect, the management device is based on at least one of the number of extracted word frequent occurrences, the word reference time count, the total word number, and the total reference time. Means for calculating a word reference rate, and keyword information up to a predetermined number of words is used in descending order of the calculated reference rate for each word.

  According to a fourth aspect of the present invention, in the information recommendation system according to any one of the first to third aspects, the management device includes means for adding a new keyword and updating a reference rate of an existing keyword. The rate update is characterized by weighting the current keyword over the past keyword by multiplying the past reference rate by the weathering coefficient.

  According to a fifth aspect of the present invention, in the information recommendation system according to any one of the first to fourth aspects, the management device is configured such that the group information includes a group ID indicating a group, a division for each group, and information. , Keywords that include keywords for each interest or anomaly indicated in the category, and keyword candidates that may become keywords are recorded as reserved keywords, and the keywords and groups included in the list information A group determination is performed by comparing keywords included in a keyword group of information.

  According to a sixth aspect of the present invention, in the information recommendation system according to the fifth aspect, when the keyword matches, the management device updates and updates the hit rate calculated by using the number of times of matching so far as a ratio of the number of times of matching. It is determined whether or not the hit rate exceeds the threshold value. If the hit rate exceeds the threshold value, the hit rate is deleted from the reserved keyword group and additionally updated to the keyword group.

  A seventh aspect of the present invention is the information recommendation system according to any one of the first to sixth aspects, wherein a search is performed for a distribution target keyword of the stored advertising content and an advertising content that matches the keyword of the search keyword. And means for transmitting advertisement content information having a large number of matches to the information terminal.

  According to an eighth aspect of the present invention, in the information recommendation system according to any one of the first to seventh aspects, an object handled by the information terminal and the management device is an e-mail and / or a home page.

  According to a ninth aspect of the invention, there is provided the information recommendation system according to the eighth aspect, further comprising means for receiving a search keyword, searching for an article on a homepage that matches the keyword of the search keyword, and transmitting the search information to the information terminal. It is characterized by that.

  The invention according to claim 10 includes means for generating list information including the synonym of the first keyword received from the information terminal with reference to the synonym dictionary, means for determining which group the list information belongs to, And a means for detecting an effective second keyword and a means for providing information including the second keyword to the information terminal.

  According to an eleventh aspect of the present invention, in the management device according to the tenth aspect, the word reference rate is calculated based on at least one of the extracted word frequent appearances, the word reference time aggregation, the total word number, and the total reference time. And the keyword information is determined up to a predetermined number of words in descending order of the calculated reference rate for each word.

  According to a twelfth aspect of the present invention, the management device according to the tenth or eleventh aspect includes means for adding a new keyword and updating a reference rate of an existing keyword, and updating the reference rate is weathered to a past reference rate. The present keyword is weighted from the past keyword by multiplying by a coefficient.

  According to a thirteenth aspect of the present invention, in the management device according to any one of the tenth to twelfth aspects, the group information is composed of a group ID indicating a group and a section and information for each group, and is indicated by the section. A keyword group that includes keywords for each interest or anaerobic and keyword candidates that may become keywords are recorded as reserved keyword groups. The keywords included in the list information and the keyword groups of the group information The group determination is performed by comparing the included keywords.

  According to a fourteenth aspect of the present invention, in the management device of the thirteenth aspect, when the keywords match, the hit rate calculated by using the number of times of matching up to now as a ratio of the number of times of matching is updated, and the updated hit rate has a threshold value. It is determined whether it has been exceeded. If the hit rate exceeds a threshold value, it is deleted from the reserved keyword group and additionally updated to the keyword group.

  The invention according to claim 15 is the management device according to any one of claims 10 to 14, wherein the distribution target keyword of the accumulated advertisement content and the advertisement content matching the keyword of the search keyword are searched, A means for transmitting advertisement content information having a large number of matches to an information terminal is provided.

  A sixteenth aspect of the present invention is the management apparatus according to any one of the tenth to fifteenth aspects, wherein the object to be handled is an e-mail and / or a home page.

  According to a seventeenth aspect of the present invention, the management device according to the sixteenth aspect includes means for receiving a search keyword, searching for an article on a homepage that matches the keyword of the search keyword, and transmitting the search information to the information terminal. It is characterized by.

  The invention according to claim 18 is an information recommendation method using an information terminal connected via a network and a management device, wherein the information terminal extracts a first keyword from transmitted / received e-mails And a step of generating list information including the synonym of the first keyword received from the information terminal with reference to the synonym dictionary, and a group to which the list information belongs An information recommendation method comprising the steps of: determining a second keyword effective in a group; and providing an information terminal with information including the second keyword.

  According to a nineteenth aspect of the present invention, in the information recommendation method according to the eighteenth aspect, in the case where the information terminal is a received mail, the information terminal includes a step of determining whether or not the information terminal has referred.

  The invention according to claim 20 is the information recommendation method according to claim 18 or 19, wherein the management device is based on at least one of the number of extraction word frequent occurrences, word reference time count, total word count, and total reference time. And calculating a word reference rate, and keyword information is obtained up to a predetermined number of words in descending order of the calculated reference rate for each word.

  The invention according to claim 21 is the information recommendation method according to any one of claims 18 to 20, wherein the management device includes a step of adding a new keyword and updating a reference rate of the existing keyword. The update of the reference rate is characterized in that the current keyword is weighted over the past keyword by multiplying the past reference rate by a weathering coefficient.

  According to a twenty-second aspect of the present invention, in the information recommendation method according to any one of the eighteenth to twenty-first aspects, the management device includes group information, which includes a group ID indicating a group, a division for each group, and information. , Keywords that include keywords for each interest or anomaly indicated in the category, and keyword candidates that may become keywords are recorded as reserved keywords, and the keywords and groups included in the list information A group determination is performed by comparing keywords included in a keyword group of information.

  The invention according to claim 23 is the information recommendation method according to claim 22, wherein, when the keywords match, the management device updates the hit rate calculated as the ratio of the number of matches until the keyword matches, and updates It is determined whether or not the hit rate exceeds the threshold value. If the hit rate exceeds the threshold value, the hit rate is deleted from the reserved keyword group and additionally updated to the keyword group.

  The invention according to claim 24 is the information recommendation method according to any one of claims 18 to 23, wherein the advertisement target keyword of the stored advertisement content and the advertisement content matching the keyword of the search keyword are searched. And a step of transmitting the advertisement content information having a large number of matches to the information terminal.

  According to a twenty-fifth aspect of the present invention, in the information recommendation method according to any one of the eighteenth to twenty-fourth aspects, an object handled by the information terminal and the management apparatus is an e-mail and / or a home page.

  According to a twenty-sixth aspect of the present invention, in the information recommendation method according to the twenty-fifth aspect, the method includes a step of receiving a search keyword, searching for an article on a homepage that matches the keyword of the search keyword, and transmitting the search information to the information terminal. It is characterized by that.

  A twenty-seventh aspect of the invention is a program that causes an information terminal and a management device to execute processing for realizing the method according to the eighteenth to twenty-sixth aspects.

  According to the information recommendation system, management device, information recommendation method, and program of the present invention, keywords can be managed appropriately.

  Next, an embodiment of the present invention will be described in detail with reference to the drawings.

  Referring to FIG. 1, the information recommendation system in the present embodiment includes an information terminal 1 connected to the Internet 3 and a service provider system 2. The Internet 3 may be replaced with an arbitrary network, and may be, for example, an optical fiber, a public line, a LAN (Local Area Network), an ADSL (Asymmetric Digital Subscriber Line), or the like. The communication method may be wired or wireless.

  The service provider system 2 includes a preference analysis device 21, a mail filter device 22, a recommendation information search device 23, a recommendation information distribution device 24, an advertisement distribution device 25, and an advertisement content storage device 26.

  The information terminal 1 transmits the keyword information S1001 generated inside the information terminal 1 to the preference analysis device 21 of the service provider system 2 via the Internet 3.

  In the service provider system 2, the preference analysis device 21 notifies the mail filter device 22, the recommended information search device 23, and the advertisement distribution device 25 of the search keyword S 1005 derived based on the keyword information S 1001.

  Further, the preference analysis device 21 transmits list information S1006 generated from the keyword information S1001 to the mail filter device 22.

  The mail filter device 22 transmits, to the information terminal 1, mail information S1002 that has been subjected to the analysis and classification processing of the electronic mail from the search keyword S1005 and the list information S1006 that are personal attribute information.

  The recommended information search device 23 searches the search information S1003 from the Internet 3 based on the search keyword S1005, and transmits it as recommended content information S1007 to the recommended information distribution device 24.

  The advertisement distribution device 25 reads out the advertisement content S1009 that matches the search keyword S1005 from the advertisement content storage device 26, and transmits it to the recommended information distribution device 24 as the advertisement content information S1008.

  The recommended information distribution device 24 transmits recommended information S1004 obtained by combining the recommended content information S1007 and the advertising content S1009 to the information terminal 1.

  Next, the flow of processing in the present embodiment will be described in detail with reference to FIGS.

  Referring to FIG. 1, the information terminal 1 transmits keyword information S1001 that is personal attribute information generated inside the information terminal 1 to the preference analysis device 21 of the service provider system 2 via the Internet 3.

  The processing in the information terminal 1 will be described in detail with reference to other drawings.

  FIG. 2 is a block diagram showing a detailed internal configuration of the information terminal 1. FIG. 4 is a flowchart showing processing inside the information terminal 1.

  Referring to FIG. 2, the first network connection device 101 is connected to a mail display device 102, a mail generation device 104, a homepage display device 106, and a keyword transmission device 111.

  When the received mail S1101 received from the first network connection apparatus 101 is displayed on the mail display device 102, the mail reference state detection device 103 monitors the reference state of the received mail S1101, and the reference state and received mail The text information that is the content is transmitted to the keyword detection device 108 as the mail reference information S1103.

  In the mail generation device 104, when a transmission mail is generated and the transmission mail S1104 is transmitted to the first network connection device 101, the mail transmission state detection device 105 monitors the generation and transmission state of the transmission mail and transmits the transmission mail. The text information which is the state and the content of the transmitted mail is transmitted to the keyword detecting device 108 as mail transmission information S1106.

  When the homepage display device 106 displays the homepage information S1107 received from the first network connection device 101, the homepage reference state detection device 107 monitors the reference state of the homepage, and the homepage information S1107 displayed as the reference state. Is sent to the keyword detection device 108 as homepage reference information S1109.

  The keyword detection device 108 receives the mail reference information S1103, the mail transmission information S1106, and the homepage reference information S1109, and generates keyword information S1001.

  In the generation of the keyword information S1001, the keyword detecting device 108 first determines whether or not the mail reference indicates that the mail reception process is being performed (step 101). In the case of the mail reference, the mail is immediately displayed without displaying any mail. It is determined whether or not it is discarded (step 102). If it is immediately discarded, a rejection flag is set to ON (step 103). This rejection flag is set to OFF when not set to ON.

  Next, when it is not the mail reference, it is determined whether or not the mail is transmitted (step 104). If it is not mail transmission, it is further determined whether or not the home page is referenced (step 105).

  Based on the result of each determination process, the medium classification information update process is performed for keyword information S1001 referring to mail, mail transmission, or home page reference (step 106).

  Next, syntax and word extraction analysis is performed from the text information included in the mail reference information S1103, the mail transmission information S1106, and the homepage reference information S1109 (step 107).

  Next, the word reference rate is calculated in consideration of the number of extracted word frequent occurrences, word reference time tabulation, total word count, total reference time, and the like (step 108).

  Up to a predetermined number of words is generated as keyword information S1001 in descending order of the calculated reference rate for each word (step 109).

  FIG. 6 is a conceptual diagram showing a data structure of the keyword information S1001. In the keyword information S1001, a personal ID for identifying an individual, a word to be used as a keyword, a reference rate, a medium classification for determining whether an e-mail is sent / received, or a home page reference, and whether it is immediately discarded when receiving an e-mail It consists of a refusal flag.

  In the keyword information temporary storage device 110, the keyword information 110 generated by the keyword detection device 118 is temporarily stored.

  The keyword transmission device 111 periodically reads the keyword information S1113 from the keyword temporary storage device 110, and transmits the keyword information S1113 to the preference analysis device 21 of the service provider system 2 via the first network connection device 101. The transmission interval for performing regular transmission is the update timing of the personal attribute, and is determined in consideration of personal preference, preference change interval and detection accuracy.

  With reference to FIG. 1 again, the operation in the overall configuration will be described.

  The preference analysis device 21 generates list information S1006 as personal attribute information including synonyms from the received keyword information S1001, and determines from this list information S1006 what kind of group the individual belongs to. Search keyword S1005 which is effective attribute information in the group is searched.

  The preference analysis device 21 transmits the list information S1006 and the search keyword S1005 to the mail filter device 1, and transmits the search keyword S1005 to the recommended information search device 23 and the advertisement distribution device 25.

  The operation of the preference analysis device 21 when the keyword information S1001 is received will be described in detail with reference to other drawings.

  FIG. 3 is a block diagram showing a detailed configuration of the preference analyzer 21. FIG. 5 is a flowchart showing processing of the preference analysis device 21.

  The preference analysis device 21 includes a second network connection device 201, a keyword reception device 202, a list management device 204, a group attribute management device 206, a storage device 210, and synonyms recorded in the storage device 210. It comprises a dictionary 203, a list information database 205, and group information 207.

  The keyword receiving device 202 receives the keyword information S1001 transmitted from the keyword transmitting device 111 of the information terminal 1 as the keyword information S2001 via the second network connection device 201 (step 201).

  The keyword receiving device 202 refers to the synonym dictionary 203 of the storage device 210 from the keyword of the received keyword information S2001 (step 202), and generates list information including the keyword synonym (step 203). The keyword receiving device 202 transmits the generated list information S2003 to the list management device 204.

  The list management device 204 receives the list information S2003, reads the list information corresponding to the personal ID from the list information database 205 of the storage device 210, and updates the list information. The update is performed by adding new keywords and updating the reference rate of existing keywords. The reference rate is updated by multiplying the past reference rate by the weathering coefficient, thereby weighting the current keyword from the past keyword and adapting to the time-series change of the personal attribute (step 204).

  The list management apparatus 204 records the update result of the list information again in the list information database 205 and transmits the list information S2006 to the group attribute management apparatus 206. The list management device 204 also transmits list information S1006 to the mail filter device 22.

  FIG. 7 is a conceptual diagram showing a data structure of list information recorded in the list information database 205. The list information has a data structure including an update time in addition to the information included in the keyword information, and a synonym data record is added as a keyword. The keyword information is managed for each individual, but the list information is composed of a plurality of personal information.

  The keyword “xxxyy” in FIG. 7 is information given as a synonym for the keyword “xxx”, and other information items “reference rate”, “update time”, “medium classification”, and “rejection flag” are: The data of the keyword “xxx” is copied as it is.

  The group attribute management device 206 receives the list information S2006 from the list management device 204, reads the group information 207 from the storage device 210, and searches the received group information S2006 for the group information 207 from which the group belongs. A group determination process is performed. Further, a keyword effective in the corresponding group is read and notified as a search keyword S1005. The group information 207 is managed by grouping similar attributes among personal attributes.

  FIG. 8 is a conceptual diagram showing the data structure of group information.

  The group information is composed of a group ID indicating a group, a classification and information for each group, a keyword group including keywords for each of interests and anaerobic indicated by the classification, and keywords that may become keywords Candidates are recorded as a reserved keyword group. As the group determination process, the group attribute management device 206 compares and determines the keyword included in the list information S2006 and the keyword included in the keyword group of the group information recorded in the group information 207 of the storage device 210. The group that almost matches is searched (step 205).

  If there is no matching group, the list information is added and updated as new group information (step 206).

  If it is determined that the specified group matches, the keyword included in the list information S2006 that does not match the keyword group is compared with the keyword included in the reserved keyword group of the corresponding group (step 207).

  When the keywords match, the hit rate calculated by using the number of times matched so far as the ratio of the number of matches is updated, and it is determined whether the updated hit rate exceeds the threshold (step 208).

  If the hit rate exceeds the threshold value, it is deleted from the reserved keyword group and is additionally updated to the keyword group (step 209).

  When the keyword included in the list information S2006 does not match the reserved keyword group, as a next process, it is determined whether or not the keyword included in the list information S2006 is included in the keyword group and the reserved keyword group of the group information. To determine whether a new keyword exists (step 210).

  If a new keyword exists, the new keyword is added to the reserved keyword group, and the group information is updated (step 211). In addition, keywords whose refusal flag is set to ON in the list information are managed with the classification as anaerobic.

  By these processes, even when the group attribute changes with the change of the personal attribute, the existing attribute information can be updated.

  When the update processing of the group information so far is completed, the keyword is read from the keyword group of the corresponding group, and the mail filtering device 22, the recommended search device 24, and the advertisement distribution device 25 are used as the inspection keyword S1005 of the corresponding individual. Notification is made (step 212).

  With reference to FIG. 1 again, the overall operation will be described.

  The mail filter device 1 receives the list information S1006 and the search keyword S1005, and searches for the keywords included in the list information S1006 and the search keyword S1005 in the contents of the mail addressed to the information terminal 1. If there is an email that matches the keyword, classification processing of importance is performed based on the reference rate of the keyword, and after marking the email (a mark that can be used to determine the category information is given to the email), the information terminal 1 The mail information S1002 is transmitted. In addition, for a keyword for which the rejection flag is ON, the mail information S1002 is transmitted after marking as a mail to be deleted, or is deleted from the viewpoint of individual desire or service (step 213).

  The recommended information search device 23 receives the search keyword S1005, searches the Internet 3 for articles such as a home page that matches the keyword of the search keyword S1005, and uses the search information S1003 as recommended content information S1007. The data is transmitted to the distribution device 24 (step 214).

  The advertisement distribution device 25 receives the search keyword S1005, searches the advertisement content distribution target keyword stored in the advertisement content storage device 26, and the advertisement content S1009 that matches the keyword of the search keyword S1005, and has a large number of matches. The advertising content information S1008 is transmitted to the recommended information distribution device 24 (step 215).

  FIG. 9 is a conceptual diagram showing a data structure of advertising content recorded in the advertising content storage device 26. In the advertising content, keywords related to the advertisement are recorded as distribution target keywords.

  The recommended information distribution device 24 receives the recommended content information S1007 and the advertising content information S1008, combines these pieces of information, and transmits the information to the information terminal 1 as the recommended information S1004 (step 216).

  FIG. 10 is a conceptual diagram showing a data structure of recommendation information. The recommendation information includes recommendation content information and advertisement content information.

  As described above, according to the present embodiment, the process of registering personal attribute information is not necessary. This is because the information terminal automatically monitors the mail reference state, the mail transmission state, and the home page reference state, extracts keywords from the text information, and further analyzes them in the service provider system.

  In addition, the accuracy of the personal attribute information is improved, and the degree of coincidence with the personal preference is greatly improved as compared with the conventional abstract attribute management. This is because the personal attribute information is managed at the word level extracted from the text information.

  Also, even if the personal interests and preferences change, the personal attribute information can be updated by capturing the time-series changes. This is because the information terminal is automatically monitored and the personal attribute information is updated.

  In addition, the advertisement reach using the highly accurate personal attribute information further increases the reach rate of the advertisement. The reason for this is that, in the past, it has been difficult to obtain consent for provision of personal attribute information in advertisement distribution using personal attributes, so processing was performed with low-precision attribute information. This is because personal attribute information can be provided because the purpose is to improve processing efficiency and early discovery of effective information.

  Further, it is possible to obtain effective information from a large amount of received mails and to process it efficiently. This is because personal attribute information is automatically detected, and the mail is classified and filtered using the personal attribute information.

  In addition, it is possible to formalize information that the individual is not aware of. The reason is that collection of personal attribute information is conventionally registered by individuals, such as a questionnaire, and cannot be registered as information unless the individual is conscious of it, but personal attribute information is automatically derived from text information. This is because even if an individual is not conscious, it can be detected and managed as information of interest and anaerobic.

  It can also be applied to users who want to extract effective information from a large amount of information and improve processing efficiency, as well as service providers and corporate systems that provide such information, and advertisements that use personal attribute information. It can be applied to service providers that handle distribution services.

  Each of the above-described embodiments is a preferred embodiment of the present invention, and various modifications can be made without departing from the scope of the present invention. For example, to realize the functions of the information terminal 1, the preference analysis device 21, the mail filter device 22, the recommendation information search device 23, the recommendation information distribution device 24, the advertisement distribution device 25, and the advertisement content storage device 26 in the above embodiment. The processing for realizing the functions of the present system may be performed by causing each device or the like to read and execute the program. Further, the program is transmitted to another computer system by a transmission wave via a CD-ROM or a magneto-optical disk that is a computer-readable recording medium, or via the Internet or a telephone line that is a transmission medium. Also good.

  In each of the above-described embodiments, the information terminal 1, the preference analysis device 21, the mail filter device 22, the recommendation information search device 23, the recommendation information distribution device 24, the advertisement distribution device 25, and the advertisement content storage device 26 are separately connected. However, the present invention can be applied to a configuration in which each function is realized as one computer system and a configuration in which a plurality of server devices are added for each function.

It is a schematic block diagram of the information recommendation system in embodiment of this invention. It is a block diagram which shows the internal structure of an information terminal. It is a block diagram which shows the internal structure of a preference analyzer. It is a flowchart of the internal processing of an information terminal. It is a flowchart of the internal processing of a preference analyzer. It is a conceptual diagram which shows the data structure of keyword information. It is a conceptual diagram which shows the data structure of list information. It is a conceptual diagram which shows the data structure of group information. It is a conceptual diagram which shows the data structure of advertisement content. It is a conceptual diagram which shows the data structure of recommendation information.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Information terminal 2 Service provider system 3 Internet 21 Preference analysis apparatus 22 Mail filter apparatus 23 Recommended information search apparatus 24 Recommended information distribution apparatus 25 Advertisement distribution apparatus 26 Advertisement content storage apparatus

Claims (27)

An information recommendation system including an information terminal connected via a network and a management device,
The information terminal comprises means for extracting a first keyword from the transmitted and received e-mail and transmitting it to the management device;
Means for generating list information including synonyms of the first keyword received from the information terminal with reference to a synonym dictionary;
Means for determining which group the list information belongs to;
Means for detecting a second keyword valid in the group;
Means for providing the information terminal with information including the second keyword.   The information recommendation system according to claim 1, wherein the information terminal includes means for determining whether or not the received mail is referred to.   The management device includes means for calculating a word reference rate based on at least one of the number of extracted word frequent occurrences, total word reference time, the total word number, and the total reference time. 3. The information recommendation system according to claim 1, wherein up to a predetermined number of words is used as keyword information in descending order of rate.   The management device includes means for adding a new keyword and updating a reference rate of an existing keyword. The reference rate is updated by multiplying a past reference rate by a weathering coefficient, so The information recommendation system according to any one of claims 1 to 3, wherein the keyword is weighted.   In the management apparatus, the group information is composed of a group ID indicating a group, a classification and information for each group, and a keyword group including keywords for each of interests and anaerobies indicated by the classification, and a keyword. The keyword candidate which has a possibility is recorded as a pending | holding keyword group, The group determination is performed by comparing the keyword contained in list information with the keyword contained in the keyword group of group information. 5. The information recommendation system according to any one of 4 above.   When the keywords match, the management device updates the hit rate calculated by using the number of matches so far as a ratio of the number of matches, determines whether the updated hit rate exceeds a threshold, However, if the threshold value is exceeded, the information recommendation system is deleted from the reserved keyword group and additionally updated to the keyword group.   2. The information processing apparatus according to claim 1, further comprising: means for searching for advertisement content that matches the keyword to be distributed with the distribution target keyword of the stored advertisement content and the keyword of the search keyword, and transmitting advertisement content information having a large number of matches to the information terminal. 7. The information recommendation system according to any one of items 1 to 6.   The information recommendation system according to any one of claims 1 to 7, wherein the information terminal and the management apparatus handle the electronic mail and / or a homepage.   9. The information recommendation system according to claim 8, further comprising means for receiving a search keyword, searching for an article on a homepage that matches the keyword of the search keyword, and transmitting the search information to the information terminal. Means for generating list information including synonyms of the first keyword received from the information terminal with reference to the synonym dictionary;
Means for determining which group the list information belongs to;
Means for detecting a second keyword valid in the group;
Means for providing information including the second keyword to the information terminal.   A means for calculating a word reference rate based on at least one of the number of extracted word frequent occurrences, word reference time tabulation, the total word count and the total reference time, and in descending order of the calculated reference rate for each word, The management apparatus according to claim 10, wherein up to a predetermined number of words is used as keyword information.   A means for adding a new keyword and updating a reference rate of an existing keyword is provided, and the reference rate is updated by multiplying the past reference rate by a weathering coefficient, thereby weighting the current keyword over the past keyword. The management apparatus according to claim 10 or 11, wherein   The group information is composed of a group ID indicating a group, a classification and information for each group, a keyword group including keywords for each of interests and anaerobic indicated by the classification, and keywords that may become keywords The candidate is recorded as a reserved keyword group, and the group determination is performed by comparing the keyword included in the list information with the keyword included in the keyword group of the group information. The management device according to item.   If the keywords match, update the hit rate calculated using the number of matches so far as the ratio of the number of matches, determine whether the updated hit rate exceeds the threshold, and the hit rate exceeds the threshold In such a case, the management apparatus according to claim 13, wherein the management apparatus deletes the keyword from the reserved keyword group and additionally updates the keyword group.   11. The information processing apparatus according to claim 10, further comprising: means for searching for an advertisement content that matches the keyword to be distributed with the stored advertisement content keyword and the keyword of the search keyword, and transmitting advertisement content information having a large number of matches to the information terminal. 15. The management apparatus according to any one of 1 to 14.   The management apparatus according to any one of claims 10 to 15, wherein an object to be handled is the electronic mail and / or a homepage.   17. The management apparatus according to claim 16, further comprising means for receiving a search keyword, searching for an article on a homepage that matches the keyword of the search keyword, and transmitting the search information to the information terminal. An information recommendation method using an information terminal connected via a network and a management device,
The information terminal has a step of extracting a first keyword from the transmitted and received e-mail and transmitting it to the management device;
The management device generates list information including synonyms of the first keyword received from the information terminal with reference to a synonym dictionary;
Determining which group the list information belongs to;
Detecting a second keyword that is valid in the group;
Providing information including the second keyword to the information terminal.   19. The information recommendation method according to claim 18, further comprising a step of determining whether or not the information terminal has referred to the received mail.   The management device includes a step of calculating a word reference rate based on at least one of the number of extracted word frequent occurrences, word reference time tabulation, the total word count, and the total reference time. 20. The information recommendation method according to claim 18, wherein up to a predetermined number of words is used as keyword information in descending order of reference rate.   The management device includes a step of adding a new keyword and updating a reference rate of an existing keyword, and the update of the reference rate is performed by multiplying a past reference rate by a weathering coefficient, so 21. The information recommendation method according to any one of claims 18 to 20, wherein the keyword is weighted.   In the management apparatus, the group information is composed of a group ID indicating a group, a classification and information for each group, and a keyword group including keywords for each of interest or anaerobic indicated by the classification, and a keyword. 19. A keyword candidate having a possibility is recorded as a reserved keyword group, and the group determination is performed by comparing the keyword included in the list information with the keyword included in the keyword group of the group information. 21. The information recommendation method according to any one of 21.   If the keywords match, the management device updates the hit rate calculated by using the number of matches so far as a ratio of the number of matches, determines whether the updated hit rate exceeds a threshold, 23. The information recommendation method according to claim 22, wherein if the threshold value is exceeded, the information is deleted from the reserved keyword group and additionally updated to the keyword group.   19. The method includes a step of searching for a distribution target keyword of the stored advertising content and an advertising content that matches the keyword of the search keyword, and transmitting advertising content information having a large number of matches to the information terminal. 24. The information recommendation method according to any one of items 1 to 23.   25. The information recommendation method according to claim 18, wherein the information terminal and the management apparatus handle the electronic mail and / or a homepage.   26. The information recommendation method according to claim 25, further comprising a step of receiving a search keyword, searching for an article on a homepage that matches the keyword of the search keyword, and transmitting the search information to the information terminal.   27. A program for causing the information terminal and the management device to execute processing for realizing the method according to claim 18.

Images