JP4862553B2 - Update information distribution system, update information distribution method, and update information distribution program - Google Patents

Description

  The present invention relates to an update information distribution system, an update information distribution method, and an update information distribution program, and more particularly to an update information distribution system, an update information distribution method, and an update information distribution program for distributing update information of a web page.

  In recent years, a web server distributes update information called RSS (RDF (Resource Description Framework) Site Summary, Rich Site Summary, or Really Simple Syndication) indicating that information has been updated to the client terminal. Inform the user that the web server information has been updated in a short time.

  FIG. 10 is an explanatory diagram showing a distribution route of update information. As shown in FIG. 10, the update information transmitted by the update information transmission server 100 is distributed to the update information receiving client 200 via the communication network 700.

  FIG. 11 is an explanatory diagram showing the connection between the update information transmission server 100 and the update information receiving client 200. As shown in FIG. 11, the update information is directly distributed from the update information transmission server 100 to the update information receiving client 200 in the same manner as the distribution path of the web page. However, the efficiency of distribution of update information by this basic connection form is low.

  FIG. 12 is an explanatory diagram showing a conventional system using the update information distribution means 300. As shown in FIG. 12, the update information transmitted by the update information transmission server 100 is distributed to the update information receiving client 200 via the update information distribution means 300. Such systems include a system having an aggregator type update information distribution unit and a system having a peer-to-peer type update information distribution unit. Non-Patent Document 1 describes a conventional system having peer-to-peer update information distribution means.

  First, a conventional system having an aggregator type update information distribution means will be described. In a conventional system having an aggregator-type update information distribution unit, the update information distribution unit includes an update information processing unit and an update information retransmitting unit.

  When one or a plurality of update information transmission servers 100 transmit update information, the update information processing unit combines the update information into one. The update information retransmitting means distributes the collected update information to all the update information receiving clients 200 by unicast.

  Next, a conventional system having peer-to-peer update information distribution means will be described. A conventional system having a peer-to-peer type update information distribution unit includes a distribution tree creation unit in addition to the configuration of the system having an aggregator type update information distribution unit. The system having the peer-to-peer type update information distribution means performs the following operation in addition to the operation of the conventional system having the aggregator type update information distribution means.

  The distribution tree creating means creates a distribution tree in which a client group including part or all of the update information receiving clients is configured in a tree structure. FIG. 13 is an explanatory diagram illustrating an example of a distribution tree. The update information retransmitting means distributes the collected update information to the update information receiving client 200 at the apex of the distribution tree. Each update information receiving client 200 constituting the distribution tree distributes update information to the lower update information receiving clients 200 in a chained manner, whereby the update information is distributed to all the update information receiving clients 200.

  Patent Document 1 describes an update notification device that repeatedly accesses one or more websites and notifies other computers that the website has been updated when the website is updated.

  Patent Document 2 describes an information acquisition system that monitors a website and acquires updated information and distributes it to a user when information provided by the website is updated.

  Patent Document 3 describes an update history generation device that generates an update history when a document provided by a website is updated.

  Patent Document 4 describes a content distribution method in which a content distribution tree is formed by a server that distributes content and a client device that receives the content.

  Patent Document 5 describes a client that distributes information to application software according to the format and format of information received from a server.

Japanese Patent Laying-Open No. 2005-284334 (paragraphs 0009 to 0032, FIG. 1) Japanese Unexamined Patent Publication No. 2003-150486 (paragraphs 0007 to 0044, FIG. 1) Japanese Patent Laying-Open No. 2006-99341 (paragraphs 0027 to 0069, FIG. 1) JP 2004-246790 A (paragraphs 0047 to 0056, FIG. 2) JP 2004-348290 A (paragraphs 0021 to 0064, FIGS. 1 and 5) D. Sandler, 3 others, “Sharing Web micronews with peer-to-peer event notification”, Peer-to-Peer Systems 4 4th International Workshop, IPPS 2005, Ithaca, NY, February 24-25, 2005, Peer-to-Peer Systems IV 4th International Workshop, IPTPS 2005, Ithaca, NY, USA, February 24-25, 2005, Revised Secret ed Papers), (the United States), Springer (Springer), 2005, Series: Lecture No Twin Computer Science (Series: Lecture Notes in Computer Science), volume 3640 (Vol.3640), p. 141-151

  The conventional system having the above-described aggregator-type update information distribution means and the conventional system having the peer-to-peer update information distribution means have improved distribution efficiency compared to the connection form shown in FIG. Specifically, assuming that the distribution load C of the update information transmission server 100 is the product of the frequency F of transmission of update information and the number N of distribution destinations, in the most basic connection form shown in FIG. Although the conventional system having F × N and having the above-described aggregator-type update information distribution unit and the conventional system having the peer-to-peer type update information distribution unit has a distribution load C = F × 1.

  However, the conventional system having the above-described aggregator-type update information distribution means and the conventional system having the peer-to-peer type update information distribution means have the following problems. The first problem is that the load on the update information distribution unit 300 increases as the number of update information transmission servers 100 and update information reception clients 200 increases. The reason will be described.

  First, in the update information distribution unit 300 of the conventional system having the aggregator-type update information distribution unit described above, the update information processing unit 301 performs a process of collecting update information distributed from all update information transmission servers 100. Here, assuming that the frequency of receiving update information is F, the load of processing one update information is P, and the number of update information transmission servers 100 is M, the processing load D of the update information processing means 301 is D = F × P × M. Therefore, particularly when the value F of the update information reception server 100 is large and the number M of the update information transmission servers 100 is large, the processing load of the update information processing unit 301 included in the update information distribution unit 300 is heavy. Become.

  Also, assuming that the frequency of retransmitting update information is F ′ and the number of update information receiving clients 200 is N, the distribution load C ′ of the update information retransmitting means 302 is C ′ = F ′ × N. Accordingly, especially when the value F ′ for retransmitting update information is large and the number N of update information receiving clients 200 is large, the retransmit processing of the update information retransmitting unit 302 included in the update information distributing unit 300 The load becomes heavy.

  Next, in the update information distribution unit 300 of the conventional system having the above-described peer-to-peer type update information distribution unit, the distribution tree creation unit 303 creates a distribution tree corresponding to each update information transmission server 100. The number of distribution trees created is the same as the number M of update information transmission servers 100. The distribution tree needs to be reconstructed when the client terminal joins or leaves, and it is necessary to periodically check the existence of the client terminal in order to detect the participation or withdrawal of the client terminal. . This reconstruction and presence confirmation involves communication, and becomes a maintenance load on the distribution tree. If the maintenance load per distribution tree is G, the total maintenance load E of the distribution tree is E = G × M. Therefore, if the value M of the number of update information transmission servers 100 is large, the load of the maintenance process of the distribution tree creation unit 303 becomes heavy.

  The second problem is, in particular, update information of the update information transmission server 100 in which the update information receiving client 200 does not register for subscription (refers to receiving update information) in the peer-to-peer update information distribution means 300. Is not easily distributed to the update information receiving client 200. The reason for this is that, in the case of the peer-to-peer type update information distribution means 300, the update information is distributed by the distribution tree created by the distribution tree creation means 303. It is because it is not delivered.

  The update notification device described in Patent Document 1 has a problem that the load becomes heavy when the number of websites to be accessed and the number of computers that notify the update are large.

  The information acquisition system described in Patent Document 2 has a problem that the load increases when the number of websites to be monitored and the number of users who distribute information are large. The update history generation device described in Patent Document 3 has a problem that the load increases when the number of websites to be monitored is large. The content distribution method described in Patent Document 4 has a problem that the load of the content distribution tree maintenance process becomes heavy when the number of servers that distribute content is large.

  Therefore, the present invention provides an update information distribution system and update information that can distribute the load of a computer that distributes update information even when the number of servers that transmit update information and the number of client terminals that receive update information increases. It is an object to provide a distribution method and an update information distribution program.

  The present invention also provides an update information distribution system, an update information distribution method, and an update information distribution program that can also distribute update information of a server to which the client terminal that receives the update information has not registered a subscription. It is also intended to provide.

An update information distribution system according to the present invention is an update information distribution system connected to a web server and a client terminal, and receives update information indicating that information provided by the web server has been updated from the web server, An update information distribution system for distributing update information to client terminals, based on aggregate information from a plurality of web servers and aggregate information storage means for storing aggregate information indicating the source and distribution destination of the update information. A plurality of update information distribution means for receiving the update information transmitted and distributing the processed information to the client terminal, a client terminal to which the update information is distributed, and a web server from which the update information is transmitted Matrix that replaces each row and column of the matrix from the registration status that shows the relationship between A server client connection aggregating unit that generates aggregated information that is aggregated into a pair of a set of transmission sources and a set of distribution destinations, and stores the generated aggregated information in an aggregated information storage unit. information distribution means, characterized in that it comprises updating information processing means for processing the received update information, the update information recall means updates the processing means delivers to the destination indicated by the aggregated information processing update information .

The server client connection aggregation means selects an arbitrary row of the matrix, extracts a predetermined number of rows close to the selected row, and inserts the extracted rows immediately after the selected row. , Select an arbitrary column of the matrix, extract a predetermined number of columns that are close to the selected column, insert the extracted columns immediately after the selected column, and replace the columns of the matrix Also good.

Update information distribution system, a distribution route of the update information, the distribution tree is a distribution route of the tree structure as components delivery destination update information is created for each set of destination on the basis of the aggregated information delivery The update information retransmitting means may include the tree creation means, and the update information processed by the update information processing means may be distributed to the client terminal at the apex of the distribution tree created by the distribution tree creation means.

Update information distribution system, based on the aggregated information, for each pair of the set of destination and the set of source comprises update information distribution unit allocating means for determining the update information distribution means for the web server sends update information The server client connection aggregating means is configured such that the number of update information distribution means of transmission destinations to which the web server transmits update information is registered in the number of web servers from which the update information is registered in the client terminal and the web server. The update information registration status may be aggregated so that the number is smaller than the number of client terminals to which update information is distributed.

An update information distribution method according to the present invention is connected to a web server and a client terminal, receives update information indicating that information provided by the web server has been updated from the web server, and distributes the update information to the client terminal. An update information distribution method by the update information distribution system, wherein the aggregate information storage means of the update information distribution system stores aggregate information indicating the source and distribution destination of the update information , and the update information distribution system By generating a matrix in which the rows and columns of the matrix are respectively replaced from the registration status indicating the relationship between the client terminal to which the update information is distributed and the web server that is the transmission source of the update information in a matrix format, Aggregated information that is aggregated into a pair of a source set and a destination set is generated, and the generated aggregate information is aggregated information. Stored in憶means, update information distribution system, it receives update information transmitted on the basis of the aggregated information from the plurality of web servers and processing the update information, the delivery destination indicated by aggregating information processed updated information It is characterized by being delivered to.

An update information distribution program according to the present invention causes a computer connected to a web server and a client terminal to receive update information indicating that information provided by the web server has been updated from the web server, and sends the update information to the client terminal. and an update information distribution program to be delivered, the computer receives the update information transmitted on the basis of the aggregated information showing aggregates the source of update information and distribution destination from a plurality of web servers, their Update information distribution processing that distributes information that is processed update information to the client terminal, and the registration status that shows the relationship between the client terminal that is the update information distribution destination and the web server that is the transmission source of the update information in a matrix format, the matrix By generating a matrix with each row and column swapped, Generates aggregated information obtained by aggregating the pair of the original set and the destination set, the generated aggregated information to execute a server client connection aggregation processing to be stored in the aggregation information storage means, the update information distribution process, the received update information processing for processing the update information, characterized in that to execute the update information recall process for delivering processed updated information in the update information processing to the delivery destination indicated by the aggregation information.

  According to the present invention, the update information is processed based on the aggregate information, and the processed update information is distributed to the client terminal based on the aggregate information. Therefore, the web server that transmits the update information or the client that receives the update information Even when the number of terminals increases, the load on the computer that distributes update information can be distributed.

  Also, according to the present invention, update information can be distributed to client terminals that are included in the aggregated registration status and that have not registered for subscription of update information.

  The server client connection aggregation means aggregates the update information registration status shown in matrix format between the update information distribution destination client terminal and the update information distribution source web server by exchanging the matrix rows and columns respectively. In the case where the matrix indicating the aggregate information is generated, the aggregate information can be easily generated.

  When configured to include a distribution tree creation means for creating a distribution tree for each set of distribution destinations based on aggregated information, reduce the number of distribution trees and reduce the maintenance load on the distribution tree Can do.

Embodiment 1 FIG.
A first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration example of a first embodiment of an update information distribution system of the present invention.

  The update information distribution system shown in FIG. 1 includes update information distribution means 300 that receives update information transmitted from an update information transmission server (web server) 100 and distributes the update information to an update information reception client (client terminal) 200. , An aggregate information storage unit 400, a server client connection aggregation unit 500, and an update information distribution unit allocation unit 600. The update information distribution unit 300 is a computer such as a server, and includes an update information processing unit 301 and an update information retransmitting unit 302. Server client connection aggregation means 500 and update information distribution means allocation means 600 are computers such as servers. The update information distribution system includes a plurality of update information distribution means 300.

  The update information processing means 301 receives update information from one or a plurality of update information transmission servers 100, performs processing such as changing the format of the received update information, filtering, and combining a plurality of update information into one, Process update information.

  The update information processing performed by the update information processing unit 301 will be described. The update information format change performed by the update information processing unit 301 is, for example, by parsing an XML (Extensible Markup Language) document in the RSS version 1.0 format and replacing the node name and namespace. This refers to conversion to an XML document in the RSS version 2.0 format. In addition, the filtering of update information means, for example, extracting only an entry including a certain keyword from RSS format update information including a plurality of entries (for example, articles in a web log). Further, combining the plurality of update information into one means generating one update information including information included in the plurality of update information based on the plurality of update information. For example, a plurality of RSS format update information including a plurality of time information-added entries are rearranged based on the time information, and all the entries are grouped into one RSS format update information in time order. That means.

  The update information retransmitting means 302 distributes the update information processed by the update information processing means 301 to one or a plurality of update information receiving clients 200.

  The server client connection aggregation means 500 generates a matrix in which the distribution status is aggregated based on the matrix indicating the distribution status from each update information transmission server 100 to each update information receiving client 200, and the generated matrix is aggregated information. Store in the storage means 400. The update information transmission server 100 and the update information re-transmission means 302 acquire the matrix stored in the aggregate information storage means 400.

  The distribution status will be described. The distribution status refers to a status in which update information is distributed from the update information transmission server 100 to the update information receiving client 200. FIG. 2 is an explanatory diagram illustrating an example of a matrix indicating a distribution status of update information from each update information transmission server 100 to each update information receiving client 200.

  The matrix shown in FIG. 2 indicates whether update information transmitted from each update information transmission server is distributed to each update information reception client with each update information transmission server as a column and each update information reception client as a row. It represents as. Specifically, in the example shown in FIG. 2, since the server A in the first column distributes update information to the client U in the first row, the element in the first row and the first column of the matrix is represented by ◯. . Further, since the server B in the second column does not distribute update information to the client X in the fourth row, the element in the fourth row and the second column of the matrix is represented by x.

  In the example illustrated in FIG. 2, the server A distributes update information to the client U, the client X, and the client Z. Server B distributes update information to client V, client W, and client Y. The server C distributes update information to the client U, the client X, and the client Z. The server D distributes update information to the client W and the client Y. The server E distributes update information to the client V, the client W, the client Y, and the client Z. The server F distributes update information to the client U, the client X, and the client Z.

  In the matrix shown in FIG. 2, the first line is the client U, the second line is the client V, the third line is the client W, the fourth line is the client X, the fifth line is the client Y, and the sixth line is the client Z. It is. In the matrix shown in FIG. 2, the first column is server A, the second column is server B, the third column is server C, the fourth column is server D, the fifth column is server E, and the sixth column is server F. It is.

  The distribution status aggregation will be described. Aggregation of distribution status (also referred to as registration status or connection relationship) refers to a connection relationship between the update information transmission server A and the update information reception client X, and reception of the same update information as another update information transmission server C. When there is a connection relationship with the client X, the connection relationship between the set {A, C} and X is regarded as one, there is a connection relationship between the update information transmission server A and the update information reception client X, and the same When there is a connection relationship between the update information transmission server A and another update information reception client Z, the connection relationship between A and the set {X, Z} is regarded as one, and a composite form thereof, that is, a set The connection relation from {A, C} to the set {X, Z} is regarded as one. However, the size of the set is not limited to 2.

  The aggregate information is information indicating a pair of a set of update information transmission servers and a set of update information receiving clients. Accordingly, the aggregate information collectively indicates the transmission source and the distribution destination of the update information.

  The update information distribution unit allocating unit 600 refers to the matrix stored in the aggregate information storage unit 400 and determines the update information distribution unit 300 to which each update information transmission server 100 transmits update information.

  The computer that is the update information distribution unit 300 receives update information transmitted from a plurality of web servers (update information transmission server 100) to the computer, updates the received update information, and updates information. Update information re-transmission processing for distributing the update information processed by the update information processing processing to the distribution destination (update information receiving client 200) indicated by the aggregate information based on the aggregate information indicating the transmission source and the distribution destination An update information distribution program is installed.

  Next, the operation of the first exemplary embodiment of the present invention will be described with reference to the drawings. FIG. 3 is a flowchart for explaining the operation of the first embodiment of the update information distribution system.

  First, the server client connection aggregating unit 500 displays a matrix indicating which update information transmission server 100 distributes update information to which update information reception client 200, information registered in each update information transmission server 100, and each It is created based on information registered in the update information receiving client 200, or obtained from a database holding the matrix.

  Specifically, for example, update information receiving client 200 stores the update information registration list that indicates beforehand what the update information transmission server 100 of the update information is distributed, server-client connection aggregation means 500, each The matrix illustrated in FIG. 2 is created based on the update information registration list stored in the update information receiving client 200. The creation of the matrix may be performed by the update information transmission server 100 or may be performed by another server (not shown).

  The server client connection aggregation means 500 generates a matrix in which the distribution status is aggregated by changing the order of the rows and columns of the matrix illustrated in FIG. 2, and stores the generated matrix in the aggregation information storage means 400 (step). S101). Therefore, the server client connection aggregation unit 500 generates a matrix indicating the aggregation information. FIG. 4 is an explanatory diagram showing a matrix generated by the server client connection aggregation unit 500.

  In the example illustrated in FIG. 4, the order of the rows and columns of the matrix illustrated in FIG. 2 is changed according to the distribution status of the update information, and the distribution status is aggregated. Specifically, in the matrix shown in FIG. 4, the first line is the client U, the second line is the client Z, the third line is the client X, the fourth line is the client Y, the fifth line is the client V, and the sixth line. Is the client W. In the matrix shown in FIG. 4, the first column is server A, the second column is server F, the third column is server C, the fourth column is server D, the fifth column is server E, and the sixth column is server B. It is.

  Therefore, in the matrix shown in FIG. 4, connection relationships among the server A, the server F, and the server C and the client U, the client Z, and the client X are collected. In the matrix shown in FIG. 4, the connection relationships between the server D, the server E, and the server B and the client Y, the client V, and the client W are aggregated, and the connection relationship between the server B and the client Z is aggregated. Yes. In FIG. 4, the aggregated connection relationships are indicated by broken lines. That is, in the example shown in FIG. 4, the ranges indicated by the broken lines are aggregated units.

  A method of collecting the distribution status of update information will be described. FIG. 5 is a flowchart for explaining a method of aggregating update information distribution statuses.

  The server / client connection aggregation means 500 randomly selects one arbitrary row of the matrix and sets it to P (step S201). The server client connection aggregation means 500 calculates the distance between P and each other row (step S202). The server client connection aggregation unit 500 extracts the top Q rows that are close to P and inserts them immediately after P (step S203). The server client connection aggregation means 500 randomly selects one arbitrary column of the matrix and sets it as R (step S204). The server client connection aggregation unit 500 calculates the distance between R and each of the other columns (step S205). The server / client connection aggregation means 500 extracts the top S rows that are close to R and inserts them immediately after R (step S206). The server client connection aggregation unit 500 repeats step S201 to step S206 N times (step S207).

  Here, Q, S, and N are parameters. Specifically, for example, Q is 1, S is 1, and N is 10,000. Each parameter (particularly N) may be set according to the size of the matrix. Specifically, when the number of rows of the matrix is j and the number of columns is k, for example, Q is set to a value between 1 and j × 0.01, and S is 1 to k × 0.01. And set N to a value between (j + k) × 2 and (j + k) × 100.

  Note that these parameters do not have to be fixed values, and may be changed according to changes in the matrix for each step. Also, a solution of a function f (x1, y1) that satisfies the distance axioms (non-negativeity, identity, objectivity, and triangular inequality) and takes one row and the other row of the matrix as parameters x1, y1, respectively A numerical value (real number) is a distance between one row and the other row. Similarly, the solution of the function f (x2, y2) satisfying the distance axioms (nonnegativeness, identity, subjectivity, and triangle inequality), with one column and the other column of the matrix as parameters x2, y2, respectively. Let (numerical value (real number)) be the distance between one column and the other column.

  In this example, a Hamming distance is used. Specifically, the number of element changes performed to match one row with the other row is the distance between the rows. In other words, when the number of changes of elements performed to match one row with the other row is small, the distance between these rows is short. Similarly, the number of element changes performed to match one column with the other column is defined as the distance between the columns. That is, when the number of element changes performed to match one column with the other column is small, the distance between these columns is short.

  The case where the matrix shown in FIG. 2 is transformed by exchanging the rows using the Hamming distance will be specifically described by taking as an example the case where the server client connection aggregation means 500 selects the first row of the matrix shown in FIG. .

  The server client connection aggregation means 500 that has selected the first row of the matrix shown in FIG. 2 calculates the distance between the first row and other rows. Specifically, the first line is XXXXX, the second line is XXXXX, the third line is XXXXX, and the fourth line is XX. × ○ ×× ○, the fifth row is × ○ × ○○ ×, and the sixth row is ○ × ○ × ○○, so in order to match the second row to the first row, Must be changed 5 times. Therefore, the distance between the first row and the second row is 5.

  In order to make the third line coincide with the first line, the element must be changed six times, so the distance between the first line and the third line is 6. Further, since the elements do not need to be changed in order to make the fourth line coincide with the first line, the distance between the first line and the fourth line is zero. In order to make the fifth line coincide with the first line, the element must be changed six times, so the distance between the first line and the fifth line is 6. In order to make the sixth line coincide with the first line, the element must be changed once, so the distance between the first line and the sixth line is 1.

  Then, the server client connection aggregation unit 500 selects one row that is close to the first row. Specifically, the server client connection aggregation unit 500 selects the fourth row (client X row) that is the closest to the first row (client U row). The server client connection aggregation means 500 inserts the selected client X row below the client U row.

  Next, the server client connection aggregation unit 500 selects any column of the matrix and switches the columns in the same manner as when the rows are switched. Then, the server client connection aggregation unit 500 repeats row replacement and column replacement N times, generates the matrix shown in FIG. 4, and stores the generated matrix in the aggregation information storage unit 400.

  The update information distribution unit allocating unit 600 refers to the matrix stored in the aggregate information storage unit 400 and determines the update information distribution unit 300 to which each update information transmission server 100 transmits update information. That is, the update information distribution unit allocating unit 600 updates the update information distribution unit 300 in which each update information transmission server 100 transmits update information in an aggregated unit based on the aggregate information included in the matrix stored in the aggregate information storage unit 400. To decide. The update information distribution unit allocating unit 600 determines that the update information distribution unit 300 to which each update information transmission server 100 transmits the update information, and the update information distribution unit that transmits the update information to each update information transmission server 100. 300 is also assigned.

  The update information distribution unit allocating unit 600 aggregates, for example, rectangles having a certain size and 90% or more of the circles from the upper left of the matrix stored in the aggregate information storage unit 400. Units are also selected sequentially.

  Note that a rectangle having a certain size is a rectangle having a size within T% (for example, 10%) of a matrix and is as large as possible. Here, a rectangle having a size of T% of the matrix means a rectangle in which the number of rows (row length) and the number of columns (column length) are T% of the number of rows and columns of the matrix, respectively. . Note that a rectangle whose size is T% of the matrix may be a rectangle whose number of elements is T% of the number of elements of the matrix. In this example, the description will be made assuming that a rectangle having a size of T% of the matrix is a rectangle having the number of rows and the number of columns of T% of the number of rows and the number of columns of the matrix, respectively.

  For example, when the number of rows (or columns) of the matrix is T1 and the number of columns (or rows) is T2, the length of the rectangle t1 ≦ T1 × T / 100, A rectangle having a horizontal length t2 ≦ T2 × T / 100 is referred to as a rectangle having a certain size. Specifically, for example, when T is 10, the number of rows of the matrix is 100, and the number of columns is 100, 100 (T1) × 10 (T) / 100 = 10. The vertical length t1 of the rectangle of the size is 10 or less. Further, since 100 (T2) × 10 (T) / 100 = 10, the horizontal length t2 of the rectangle having a certain size is 10 or less.

  When the update information distribution unit allocating unit 600 sequentially selects the aggregation units, the update information distribution unit allocation unit 600 selects the aggregation units so as not to include the elements of the circles already included in the other aggregation units. That is, the update information distribution unit allocating unit 600 sequentially selects the aggregation units so that the circle element is not included in a plurality of aggregation units. Then, the update information distribution unit allocating unit 600 ends the selection of the aggregation unit when all the elements of ○ are included in the aggregation unit.

  One or a plurality of update information transmission servers 100 acquires a matrix from the aggregate information storage unit 400, and based on the acquired matrix, the update information distribution unit allocated to the update information distribution unit allocation unit 600 in an aggregated unit. Update information is transmitted to 300 (step S102). The update information processing unit 301 performs processing such as collecting a plurality of transmitted update information into one update information (step S103).

  The update information retransmitting means 302 acquires the matrix from the aggregated information storage means 400, and distributes (retransmits) the processed update information to each update information receiving client 200 based on the acquired matrix (step S104).

  According to the present embodiment, since the matrix in which the distribution status is aggregated is generated and the generated matrix is configured to be shared (acquired) by the update information transmission server 100 and the update information distribution unit 300, the update information transmission server According to 100, the update information distribution means 300 of the distribution destination of update information can be distributed (assigned). Therefore, even when the number of update information transmission servers 100 increases, it is possible to avoid a load from being concentrated on one update information distribution unit 300.

  Note that the average processing load D of one update information processing unit 301 is aggregated by F, the frequency of receiving update information, P, the load of processing one update information, and M, the number of update information transmission servers 100. If the number of distribution statuses is Q, the average processing load of one update information processing unit 301 is D = F × P × M ÷ Q. That is, the processing of the update information processing unit 301 can be distributed based on the matrix generated in step S101.

  The average distribution load C ′ of one update information retransmitting means 302 is one update information retransmitting means, where F ′ is the frequency of retransmitting update information and N is the number of update information receiving clients 200. 302 average distribution load C ′ = F ′ × N ÷ Q. That is, the processing of the update information retransmitting means 302 can be distributed based on the matrix generated in step S101.

  According to the present embodiment, by using information (aggregated information included in the matrix generated by the server client connection aggregation unit 500) that aggregates the connection relation between the update information transmission server 100 and the update information reception client 200, FIG. As shown in FIG. 5, the update information distribution means 300 can be distributed for each aggregated connection relationship. FIG. 6 is an explanatory diagram showing connections between each update information transmission server 100, update information distribution means 300 assigned for each aggregated connection relationship, and each update information receiving client 200.

  Further, the aggregated connection relationship indicates the relationship between the update information transmission servers 100 and between the update information reception clients 200. Using the aggregated connection relationship, the update information transmission server 100 Update information can be distributed to the update information receiving client 200 having no connection relationship. Specifically, referring to the matrix illustrated in FIG. 2, there is no connection relationship between the server D and the client V, but when the matrix illustrated in FIG. 4 is used, the update information distributed by the server D is the client V Also delivered. Accordingly, the update information receiving client 200 that has not registered the update information subscription to the update information transmission server 100 is distributed to the update information receiving client 200 and the service according to the user's preference of the update information receiving client 200 (Update information) can be provided.

Embodiment 2. FIG.
A first embodiment of the present invention will be described with reference to the drawings. FIG. 7 is a block diagram showing a configuration example of the second embodiment of the update information distribution system of the present invention.

  The update information distribution system according to the second exemplary embodiment of the present invention is that the update information distribution unit 300 includes a distribution tree creation unit 303 in addition to the components of the first exemplary embodiment shown in FIG. Different from the configuration of the first embodiment. Since the other components are the same as those of the first embodiment, the same reference numerals as those in FIG.

  The distribution tree creation unit 303 refers to the aggregate information included in the matrix generated by the server client connection aggregation unit 500 and stored in the aggregate information storage unit 400, and includes a part or all of the update information receiving client 200 Create a distribution tree with groups organized in a tree structure. Accordingly, the distribution tree creation unit 303 creates a distribution tree based on the aggregated connection relationship. Specifically, when the connection relationships are aggregated as in the matrix illustrated in FIG. 4, the distribution tree creation unit 303 includes a distribution tree including the client U, the client Z, and the client X as components, the client Y, A distribution tree having the client V and the client W as components is created.

  FIG. 13 is an explanatory diagram showing a configuration of a distribution tree created by the distribution tree creation unit 303. As shown in FIG. 13, the distribution tree is configured such that each update information receiving client 200 transmits update information received from the upper update information receiving client 200 to a plurality of lower update information receiving clients 200. Yes.

  For example, the distribution tree creating unit 303 receives a predetermined number (see FIG. 13) from the update information receiving client 200 in order from the top of the tree structure in the order received from the update information receiving client 200. In the example shown, the distribution tree is configured such that the predetermined number 2) of update information receiving clients 200 are subordinately connected.

  Also, the distribution tree creation unit 303 receives update information in order from the top of the tree structure, for example, in order of high information processing capability of the update information receiving client 200, in order of wide communication network bandwidth, or lexicographic order of client names. The distribution tree may be configured so that the clients 200 are connected to each other.

  The distribution tree creation unit 303 may be realized by a management server that grasps the structure of the distribution tree and manages the distribution tree. When the management server receives an inquiry to participate in the distribution tree from the update information receiving client 200, for example, the update information receiving client connected to the lower order among the update information receiving clients 200 constituting the distribution tree. The number of 200 is smaller than a predetermined number (in the example shown in FIG. 13, the predetermined number is 2), and a reply is made so that the update information receiving client 200 is connected to a lower level of the apex update information receiving client 200. To do.

  In addition, when the management server receives an inquiry to leave the distribution tree, for example, the update information receiving client 200 connected to a lower level of the leaving update information receiving client 200 causes the update information receiving client 200 to leave. The distribution tree is changed so as to be connected to the upper update information receiving client 200, and the update information receiving client 200 that is higher than the update information receiving client 200 that leaves the message is connected, and the lower update information receiving client 200 that is leaving is connected. The update information receiving client 200 is notified.

  Thereafter, when the management server receives an inquiry to join the distribution tree from the update information receiving client 200, the management server moves the new update information receiving client 200 to the position where the detached update information receiving client 200 is connected in the distribution tree. May participate. In such a configuration, a state in which a predetermined number or more of the update information receiving clients 200 are subordinate to the upper update information receiving client 200 of the detached update information receiving client 200 is indicated as a higher rank of the detached update information receiving client 200. The update information receiving client 200 can be modified to a state in which a predetermined number of update information receiving clients 200 are subordinate.

  The update information retransmitting means 302 distributes update information to the update information receiving client 200 at the top of the distribution tree based on the distribution tree created by the distribution tree creating means 303. Each update information receiving client 200 constituting the distribution tree transmits the update information to each of the lower update information receiving clients 200 in a chain, whereby the update information is distributed to all the update information receiving clients 200.

  Next, the operation of the second exemplary embodiment of the present invention will be described with reference to the drawings. FIG. 8 is a flowchart for explaining the operation of the second embodiment of the update information distribution system.

  The server client connection aggregating unit 500 generates a matrix in which distribution statuses are aggregated based on a matrix indicating which update information transmission server 100 distributes update information to which update information receiving client 200 (step S301). The matrix thus stored is stored in the aggregate information storage means 400.

  The distribution tree creating unit 303 creates a distribution tree with reference to the aggregated connection relationship in the matrix stored in the aggregate information storage unit 400 (step S302).

  The update information distribution unit allocating unit 600 refers to the aggregated connection relation of the matrix stored in the aggregate information storage unit 400 and allocates the update information distribution unit 300 to each update information transmission server 100.

  One or a plurality of update information transmission servers 100 acquires a matrix from the aggregate information storage unit 400, and based on the acquired matrix, the update information distribution unit allocated to the update information distribution unit allocation unit 600 in an aggregated unit. Update information is transmitted to 300 (step S303). The update information processing unit 301 performs processing such as combining a plurality of distributed update information into one update information (step S304).

  The update information retransmitting means 302 acquires the matrix from the aggregate information storage means 400, and based on the acquired matrix, the update information received by the update information receiving client 200 at the vertex in the distribution tree created by the distribution tree creating means 303 is processed. Information is distributed (step S305). Each update information receiving client 200 constituting the distribution tree transmits update information to each lower update information receiving client 200 in a chain.

  According to the present embodiment, a matrix in which the distribution status is aggregated is generated, and the distribution tree creation unit 303 creates a distribution tree based on the matrix. Therefore, the number of distribution trees is reduced, and the maintenance load of the distribution tree is reduced. It can be reduced overall.

  The total maintenance load E of the distribution tree is G = Q × Q, where G is the maintenance load per distribution tree and Q is the number of aggregated distribution statuses. That is, since the number Q of the aggregated distribution statuses can be made smaller than the number M of the update information transmission servers 100 by the matrix generated in step S201, the distribution tree maintenance load by the distribution tree creation unit 303 is reduced. Can be reduced.

  A first embodiment of the present invention will be described. The first embodiment of the present invention corresponds to the update information distribution system of the first embodiment shown in FIG. The update information transmission server 100 is a web server that transmits update information such as RSS. The update information receiving client 200 is a client terminal that receives update information. The update information distribution means 300 is a computer that provides an aggregator service for collecting (processing) the update information transmitted by the plurality of update information transmission servers 100 and distributing the update information to the update information receiving client 200.

  The update information receiving client 200 holds an update information registration list for registering information such as a URL (Uniform Resource Locator) of the update information transmission server 100 in order to receive the update information.

  Specifically, for example, update information is received from the update information transmission server a whose URL is “a”, the update information transmission server b whose URL is “b”, and the update information transmission server c whose URL is “c”. (A, b, c) are registered in the update information registration list of the update information receiving client A.

  An update information receiving client that receives update information from an update information transmission server d whose URL is “d”, an update information transmission server e whose URL is “e”, and an update information transmission server f whose URL is “f”. (D, e, f) is registered in the update information registration list of B.

  An update information receiving client that receives update information from the update information transmission server g having the URL “g”, the update information transmission server h having the URL “h”, and the update information transmission server i having the URL “i”. (G, h, i) is registered in the update information registration list of B.

  The server client connection aggregation unit 500 creates a matrix indicating which update information transmission server distributes update information to which update information reception client based on the update information registration list held by each update information reception client. . FIG. 9 is an explanatory diagram illustrating an example of a matrix created by the server client connection aggregation unit 500.

  In the matrix illustrated in FIG. 9, the element marked with ○ indicates that the URL of the update information transmission server of the other axis is registered in the update information registration list of the update information reception client of one axis, and the update information transmission server transmits It means that the update information receiving client receives the update information.

  When the number of update information transmission servers 100 and the number of update information reception clients 200 are small, the load of distributing update information from any update information transmission server 100 to any update information reception client 200 is not so large. The update information distribution means 300 that provides an aggregator service can be realized by a conventional method, that is, a single server. However, when the number of update information transmitting servers 100 and the number of update information receiving clients 200 are large, that is, when the matrix is large, it is difficult for a single server to provide an aggregator service, and a plurality of servers It is necessary to distribute the load using

  At this time, a server that replaces the rows and columns of the matrix, aggregates the elements of ○ as much as possible, and realizes the update information distribution means 300 is updated to the update information transmission server 100 and the update information reception client 200. By making them as independent as possible, the amount of communication between the update information distribution unit 300 and the update information transmission server 100 and the amount of communication between the update information distribution unit 300 and the update information receiving client 200 are increased. Can be prevented.

  It should be noted that the server that implements the update information distribution means 300 is installed so as to be as independent as possible between the update information transmission server 100 and the update information receiving client 200. That is, the server that realizes each update information distribution unit 300 has a total number of the update information transmission server 100 and the number of update information reception clients 200 that are in charge of all the servers that realize each update information distribution unit 300. Installed to minimize.

  Specifically, in the matrix shown in FIG. 4, the server B updates the transmission destination when the update information is transmitted to the client Z and when the update information is transmitted to the client Y, the client V, and the client W. Servers for realizing the information distribution unit 300 may be different. In such a case, even if the total number of update information transmission servers 100 is k and the total number of update information reception clients 200 is 1, the number of update information transmission servers 100 in charge and the number of update information reception clients 200 are The total number of all servers that realize each update information distribution unit 300 is greater than k + 1. This is because the server that implements the two update information distribution means 300 is in charge of the server B. Therefore, the server B transmits update information to the server that realizes the update information distribution unit 300 that handles the client Z and the server that implements the update information distribution unit 300 that handles the client Y, the client V, and the client W. There must be.

  That is, if the total number of update information transmission servers 100 in charge and the number of update information receiving clients 200 for all the servers realizing each update information distribution unit 300 is large, the update information distribution unit 300 and the update information transmission unit The amount of communication between the server 100 and the amount of communication between the update information distribution unit 300 and the update information receiving client 200 increases. In order to prevent such an increase in communication volume, the server that implements each update information distribution unit 300 determines the number of update information transmission servers 100 and the number of update information reception clients 200 that are in charge of each update information distribution unit 300. It is installed so that the total number of all realized servers is minimized.

  Here, the number of servers that implement the update information distribution means 300 is the same as the number of update information distribution means 300. The servers that implement the update information distribution means 300 are the update information transmission server 100 and the update information reception client 200. In general, the number of servers is limited by physical constraints. When the number of update information distribution means 300 is larger than the number of servers that implement the update information distribution means 300, one server plays the role of a plurality of update information distribution means 300. In this case, the aggregation unit to which the update information distribution unit 300 corresponds (in this example, the set of update information transmission server 100 of the update information transmission source and update information reception client 200 of the distribution destination of update information) is displayed on the matrix. It is preferable to overlap in rows or columns because the connection destination is limited.

  As a method of deforming such a matrix, there is the method described in the first embodiment using the flowchart illustrated in FIG.

  Specifically, as shown in FIG. 5, the server client connection aggregation unit 500 randomly selects one arbitrary row of the matrix and sets it to P (step S201). The server client connection aggregation means 500 calculates the distance between P and each other row (step S202). The server client connection aggregation unit 500 extracts the top Q rows that are close to P and inserts them immediately after P (step S203). The server client connection aggregation means 500 randomly selects one arbitrary column of the matrix and sets it as R (step S204). The server client connection aggregation unit 500 calculates the distance between R and each of the other columns (step S205). The server / client connection aggregation means 500 extracts the top S rows that are close to R and inserts them immediately after R (step S206). The server client connection aggregation unit 500 repeats step S201 to step S206 N times (step S207).

  Here, Q, S, and N are parameters. Specifically, for example, Q is 1, S is 1, and N is 10,000. Each parameter (particularly N) may be set according to the size of the matrix. Specifically, when the number of rows of the matrix is j and the number of columns is k, for example, Q is set to a value between 1 and j × 0.01, and S is 1 to k × 0.01. And set N to a value between (j + k) × 2 and (j + k) × 100.

  Note that these parameters do not have to be fixed values, and may be changed according to changes in the matrix for each step. In the first embodiment, the matrix is deformed by replacing the rows and columns using the Hamming distance, but the rows and columns are replaced using other methods such as the Euclidean distance and the cosine distance. May be.

  A method of exchanging rows using the Euclidean distance will be described. In the matrix illustrated in FIG. 2, if ○ is replaced with 1 and × is replaced with 0, the fourth row is 101001. Similarly, the fifth line is 010110. The sixth line is 101011.

And the Euclidean distance between the 4th and 5th lines is
((1-0) ² + (0-1) ² + (1-0) ² + (0-1) ² + (0-1) ² + (1-0) ² ) ^1/2 = 6 ^{1 / 2} .

The Euclidean distance between the 4th and 6th lines is
((1-1) ² + (0-0) ² + (1-1) ² + (0-0) ² + (0-1 ² ) + (1-1) ² ) ^1/2 = 1 .

  Therefore, since the Euclidean distance between the 4th line and the 6th line is closer than the Euclidean distance between the 4th line and the 5th line, the server-client connection aggregation means 500 has the 4th line (client X line). The sixth line (client Z) is inserted below.

Next, a method for replacing rows using the cosine distance will be described. The cosine distance between the 4th and 5th lines is
(1 × 0 + 0 × 1 + 1 × 0 + 0 × 1 + 0 × 1 + 1 × 0) / ((1 ² +0 ² +1 ² +0 ² +0 ² +1 ² ) ^1/2 × (0 ² +1 ² +0 ² +1 ² +1 ² +0 ² ) ^{1 / 2} ) = 0.

The cosine distance between the 4th and 6th lines is
(1 × 1 + 0 × 0 + 1 × 1 + 0 × 0 + 0 × 1 + 1 × 1) / ((1 ² +0 ² +1 ² +0 ² +0 ² +1 ² ) ^1/2 × (1 ² +0 ² +1 ² +0 ² +1 ² +1 ² ) ^{1 / 2} ) = 3 / (3 ^1/2 × 4 ^1/2 ).

  Therefore, since the cosine distance between the 4th line and the 6th line is closer than the cosine distance between the 4th line and the 5th line, the server client connection aggregating means 500 has the 4th line (the line of the client X). The sixth line (client Z) is inserted below. The cosine distance is closer as the value increases.

  By this method, as shown in FIG. 4, a matrix in which the ◯ elements are aggregated can be obtained. The optimality of the aggregated result is not necessary. This is because the object of the present invention can be achieved even if the aggregation result is not optimal. The optimum aggregation result is a matrix in which the elements of the circles are aggregated so that the number of the parts in which the elements of the circle are aggregated is small and the size is uniform.

  Note that the aggregation unit may include a small number of x elements, such as the 5 × 4 element of the matrix shown in FIG. 4. For example, from the upper left of the matrix, elements in which 90% or more of the elements of the matrix surrounded by a rectangle of a certain size are ◯ are sequentially selected as the aggregation unit.

  Then, when the number of aggregation units is equal to or greater than the number of servers that realize the update information distribution unit 300, the distribution process of update information is distributed to each server that realizes the update information distribution unit 300. Note that when distributing the update information distribution process to each server, the process can be more uniformly distributed by considering the performance of each server and the size of the aggregation unit.

  Further, when the number of aggregation units is smaller than the number of servers that implement the update information distribution unit 300, the number of aggregation units can be increased by dividing a large aggregation unit. Specifically, for example, when the number of servers in the aggregation unit is four and the number of client terminals is three, the number of servers is larger than the number of client terminals. Is divided into two aggregation units each having two client terminals and three client terminals.

  The matrix modified by the above-described method is shared by the update information transmission server 100 and the update information re-transmission means 302. The destination when the update information distribution server 100 distributes update information to the update information distribution unit 300 and the destination when the update information distribution unit 300 distributes update information to the update information receiving client 200 are the minimum necessary. It is preferable to squeeze. By doing so, the load distribution of the update information distribution means 300 using the matrix can be performed.

  By the way, the matrix before transformation changes depending on the number of update information transmitting servers 100, the number of update information receiving clients 200, and the update information registration status. Therefore, the matrix is preferably deformed periodically or according to the degree of change of the matrix, and the update information is preferably distributed using the latest matrix. Note that performing matrix deformation according to the degree of matrix change means, for example, trying to aggregate elements at each step of matrix deformation, and if the number of aggregated elements exceeds a certain value, continue matrix deformation. If the number of aggregated elements is less than a certain value, the transformation of the matrix is terminated.

  A second embodiment of the present invention will be described. The second embodiment of the present invention corresponds to the update information distribution system of the second embodiment shown in FIG.

  The update information distribution system according to the second embodiment is different from the update information distribution system according to the first embodiment in that the update information distribution unit 300 includes a distribution tree creation unit 303. According to the FeedTree system described in Non-Patent Document 1, the distribution tree creation unit 303 creates a distribution tree for each update information identifier such as the URL of the update information transmission server 100, and the update information includes a plurality of update information. It is efficiently distributed to the receiving client 200.

  However, in order to maintain each distribution tree, a maintenance load for performing regular communication or the like is generated. Therefore, in the second embodiment, a matrix is created and modified in the same manner as in the first embodiment, and a delivery tree is created for each aggregation unit, thereby reducing the number of delivery trees.

  In addition to the method described in the first embodiment, the matrix deformation and grouping (element aggregation) method can be performed by repeating only steps S204 to S206 shown in FIG. is there.

  The transformed matrix is shared by the update information transmission server 100 and the distribution tree creation unit 303. Then, it is preferable that the update information transmission server 100 narrows down the destination when the update information is distributed to the update information distribution unit 300 to the minimum necessary. The update information distribution unit 300 receives the update information at the apex of the corresponding distribution tree. It is preferable to distribute update information to the client 200. By doing so, it is possible to prevent an increase in the maintenance load of the distribution tree even when the number of update information transmission servers 100 is increased while taking advantage of FeedTree.

  Further, when the distribution tree of this embodiment is used, even if the URL of the update information transmission server 100 is not registered in the update information registration list of the update information receiving client 200, the update information transmission server 100 is aggregated. If it is included in the unit, that is, in the matrix illustrated in FIG. 4, it is registered in the update information registration list of another update information receiving client 200 if it is an element of × but is included in the aggregation unit. Update information of the update information transmission server 100 is also distributed. Specifically, in the example illustrated in FIG. 4, the update information transmitted by the server D is distributed to the client V. The update information is expected to be highly related to the update information distributed to other update information receiving clients 200 by the matrix configuration method, and the update information receiving client 200 that is not registered for subscription (this example) Then, the client V) may be meaningful.

  The third embodiment of the present invention is different from the first embodiment and the second embodiment in that the update information is transmitted from the update information transmission server 100 that is the transmission side of the update information and the update information distribution means 300 on the reception side. Instead of distributing the update information to a certain update information distribution means 300 and the update information reception client 200 in the push type, the update information distribution means 300 on the reception side and the update information reception client 200 are each the transmission side. A method of requesting transmission of update information to the update information transmission server 100 and the update information distribution means 300 and distributing it in the PULL type is adopted.

  Specifically, in the first embodiment and the second embodiment of the present invention, the update information transmission server 100 distributes update information to the update information distribution means 300 in the PUSH type, and the update information distribution means 300 updates. Update information is distributed to the information receiving client 200 in the push type.

  On the other hand, in the third embodiment of the present invention, the update information distribution means 300 requests the update information transmission server 100 to transmit the update information, the update information is distributed in the PULL type, and the update information receiving client 200 The update information distribution means 300 is requested to transmit the update information, and the update information is distributed in the PULL type.

  Therefore, the update information receiving client 200 does not install the software for receiving the update information distributed by the update information distribution unit 300 in the push type, and the update information is used by using the software of the web browser used for other purposes. Therefore, the update information distribution system according to the present invention can be easily introduced.

  The present invention can be applied to a system that distributes information from a plurality of web servers to a plurality of client terminals via a communication network. Further, the present invention can also be applied to a system that provides a service according to the taste of the user of the client terminal.

It is a block diagram which shows the structural example of 1st Embodiment of an update information delivery system. It is explanatory drawing which shows an example of the matrix which shows the delivery condition of the update information from each update information transmission server to each update information receiving client. It is a flowchart explaining operation | movement of 1st Embodiment of an update information delivery system. It is explanatory drawing which shows the matrix which the server client connection aggregation means produced | generated. It is a flowchart explaining the deformation | transformation method of a matrix. It is explanatory drawing which shows the connection with each update information transmission server, the update information delivery means allocated for every aggregated connection relationship, and each update information receiving client. It is a block diagram which shows the structural example of 2nd Embodiment of an update information delivery system. It is a flowchart explaining operation | movement of 2nd Embodiment of an update information delivery system. It is explanatory drawing which shows an example of the matrix which the server client connection aggregation means produced in the 1st Example. It is explanatory drawing which shows the delivery path | route of update information. It is explanatory drawing which shows the connection of an update information transmission server and an update information receiving client. It is explanatory drawing which shows the conventional system using an update information delivery means. It is explanatory drawing which shows the structure of a delivery tree.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Update information transmission server 200 Update information receiving client 300 Update information delivery means 301 Update information processing means 302 Update information redistribution means 303 Distribution tree creation means 400 Aggregation information storage means 500 Server client connection aggregation means 600 Update information distribution means allocation means 700 Communication network

Claims (6)

An update information distribution system connected to a web server and a client terminal,
Wherein the web server, the information web server provides the received update information indicating that it has been updated, the update information distribution system for distributing the update information to the client terminal,
Aggregated information storage means for storing aggregated information indicating the source and distribution destination of update information in an aggregated manner;
A plurality of update information delivery means for receiving update information transmitted from a plurality of web servers based on the aggregated information, and delivering information obtained by processing the update information to a client terminal;
From the registration status indicating the relationship between the client terminal to which the update information is distributed and the web server that is the transmission source of the update information in a matrix format, by generating a matrix in which the rows and columns of the matrix are respectively replaced, Server client connection aggregation means for generating aggregate information aggregated into a pair of a set of transmission sources and a set of distribution destinations, and storing the generated aggregate information in the aggregate information storage means,
The update information distribution means includes:
Update information processing means for processing the received update information;
An update information distribution system, comprising: update information retransmitting means for distributing update information processed by the update information processing means to a distribution destination indicated by the aggregate information. The server client connection aggregation means selects one arbitrary row of the matrix, extracts a predetermined number of rows that are close to the selected row, and inserts the extracted rows immediately after the selected row. The rows of the matrix are exchanged, an arbitrary column of the matrix is selected, a predetermined number of columns close to the selected column are extracted, and the extracted column is inserted immediately after the selected column. to, update information distribution system according to claim 1, wherein the interchanging columns of the matrix. A distribution tree creation means for creating a distribution tree that is a distribution path of update information and is a distribution path of a tree structure having a distribution destination of update information as a constituent element for each set of distribution destinations based on aggregated information;
The update information distribution system according to claim 1 or 2, wherein the update information retransmitting unit distributes the update information processed by the update information processing unit to a client terminal at the apex of the distribution tree generated by the distribution tree generation unit. . Based on the aggregated information, for each pair of the set of destination and the set of source comprises update information distribution unit allocating means for determining the update information distribution means for the web server to send updates,
The server client connection aggregating means is configured such that the number of update information distribution means of transmission destinations to which the web server transmits update information is registered in the number of web servers of the update information registered in the client terminal and the web server. The update information distribution system according to any one of claims 1 to 3 , wherein the registration status of the update information is aggregated so that the number of update information distribution destination client terminals is smaller than the number. Is connected to the web server and the client terminal information which the web server from the web server provides the received update information indicating that it has been updated, the update information to update information by the update information distribution system for distributing to said client terminal In the delivery method,
Aggregated information storage means of the update information distribution system stores aggregated information indicating the update information transmission source and the distribution destination,
A matrix in which the update information distribution system replaces the rows and columns of the matrix from the registration status indicating the relationship between the client terminal that is the distribution destination of the update information and the web server that is the transmission source of the update information in a matrix format. Generating aggregated information aggregated into a pair of the set of transmission sources and the set of distribution destinations, storing the generated aggregated information in the aggregated information storage unit,
The update information distribution system receives update information transmitted from a plurality of web servers based on the aggregate information, processes the update information, and distributes the processed update information to a distribution destination indicated by the aggregate information An update information delivery method characterized by the above. A computer connected to the Web server and the client terminal, wherein the web server, the information web server provides is to receive the update information indicating that it has been updated, the update information for delivering the update information to the client terminal A distribution program,
In the computer,
Update information distribution that receives update information transmitted from a plurality of web servers based on aggregated information that indicates an aggregate of update information transmission sources and distribution destinations, and distributes the processed information to client terminals Processing,
From the registration status indicating the relationship between the client terminal to which the update information is distributed and the web server that is the transmission source of the update information in a matrix format, by generating a matrix in which the rows and columns of the matrix are respectively replaced, Generating aggregate information aggregated in pairs of a set of transmission sources and a set of distribution destinations, and executing server client connection aggregation processing for storing the generated aggregate information in an aggregate information storage unit;
In the update information distribution process,
Update information processing for processing the received update information;
Update information distribution program for executing the pre-Symbol the processed updated information in the update information processing is distributed to the distribution destination indicated by the aggregated information update information redialing process.

Images