TWI575470B - A global relationship model and a relationship search method for internet social networks - Google Patents

Description

Global relationship system and search relationship method for social networks

The invention relates to a relational network search method, in particular to a relational network search method based on integrating heterogeneous social networks.

Today's social networking sites, such as Facebook (Facebook ^TM), LinkedIn (Linkedin ^TM), etc. By providing a simple user liaison, communications, and share a lot of information access platform, and then change today's society. Users can also get in touch with old friends through social networks, and build new social relationships by sharing information such as hobbies, interests, and overlapping relationships. Due to the use of the social network population, and the rapid increase in the types of social networks. For example, active users of Facebook are expected to reach hundreds of millions of people. Although it is difficult to estimate exactly, there are more than a thousand social networks in the world and the different services they provide.

As a result, a single user may sign up for multiple social networking sites to achieve different social networking applications, have multiple social network accounts, and get in touch, post, and access with friends on various social networks. The content and sharing of content across the community of each social network. While providing different services across social networks, it is a key point to provide a way for users to connect with other users on different social networks. However, each social network is an independent network, so users must manage and create profiles on different social networks. And social connections. The posts posted by users on different social networks may be duplicated, so maintaining content on different social networks will place a heavy burden on users.

In summary, it is a technical problem in the art to provide a method and system for integrating a heterogeneous social network.

In order to solve the technical problems of the prior art, one object of the present invention is to provide a method and system for integrating a heterogeneous social network.

In order to achieve the above object, the present invention provides a peer network method for integrating a heterogeneous social network, the method is applied to a server device, and the method further comprises the following steps: First, the server device is connected to a plurality of peer nodes. Each peer node defines a user and is used to access at least one social network. Then, the server device is connected to the peer-to-peer social network according to a social association between the social networks, and is integrated into a peer-to-peer social network (P2P for short). -iSN), wherein the peer community network configures a plurality of social paths between different peer networks of different peer networks.

To achieve the above objects, the present invention further provides a peer network system for integrating a heterogeneous social network. The system includes a communication module and a processing module. The communication module is connected to a plurality of peer nodes, and each peer node defines a user and is used to access at least one social network. The processing module is connected to the communication module and according to a social association between the social networks, to link the equivalent nodes of the association, and to integrate into the same social network, wherein the peer social network system Configure multiple social paths between different peer networks of each peer node.

In summary, the system and method of the present invention integrates heterogeneous social networks and generates social paths to enable users to connect with friends on the same/different social network.

1,na, nb‧‧‧ peer node

1.1‧‧‧FeedRequestListener

1.2‧‧‧SampleAuthListener

1.3‧‧‧CreateFriendListListener

1.4‧‧‧Friend List

1.5‧‧‧Peer Agent

1.6‧‧‧Update_Tvalue()

1.7‧‧‧Update_FriendList()

1.8‧‧‧Relationship_Finding()

1.9‧‧‧Phone Book API

1.10‧‧‧Background Service

2‧‧‧Same index node

2.1‧‧‧IndexPeerAgent

2.2‧‧‧receivePacket.getData()

2.3‧‧‧receiveSocket.receive()

2.4‧‧‧receiveSocket.send()

2.5‧‧‧Global ID list

3‧‧‧Facebook Graph API

3.1‧‧‧SessionEvents.addAuthListener(new SampleAuthListener())

3.2‧‧‧mAsyncRunner.request("me/friends", new reateFriendListListener())

3.3‧‧‧mAsyncRunner.request("me/feed", newFeedRequestListener())

4‧‧‧Twitter REST API

5‧‧‧other social network API

Figure 1 is a system architecture diagram of the peer community network of the present invention.

Figure 2 is a schematic diagram of the format of a friend list of the present invention.

Figure 3 is a software architecture diagram of the peer community network of the present invention.

Figure 4 is a Global ID list of the present invention.

Figure 5 is a message flow diagram of the login procedure of the present invention.

Figure 6A is a social graph of the heterogeneous social network of the present invention.

Figure 6B is an i-Search algorithm of the present invention.

Figures 7 and 8 show the results of the analysis and simulation of the present invention.

The specific embodiments are described below to illustrate the embodiments of the invention, but are not intended to limit the scope of the invention.

FIG 1 has three lines presented aspects of the social network community peer network (e.g., Facebook (Facebook ^TM), Twitter (Twitter ^TM), Columbia Valley (Google+ ^TM)), and two kinds of nodes: Peer node 1 and the same index node 2.

The aforementioned peer node 1 is installed in a terminal device (for example, PDA, smart phone, desk machine, etc.) to allow the user to access the social network and connect to the server device. The main function of the server device is to integrate the heterogeneous social network. . The user of the peer node 1 can register one or more social networks through its terminal device and can simultaneously log in to the aforementioned social network. In this case, the unique user ID (UUID) is used to associate the heterogeneous social network account held by the same user. The aforementioned UUID can be specific authentication information, such as the power of the user holding the phone. The phone number, or the verified email address. The aforementioned peer index node 2 is installed in the servo device and is used to provide communication status (for example, online, offline status) with each peer node 1 and routing information (for example, an IP address). When each peer node 1 is turned on, the online status is reported to the peer index node 2, and the online state includes information such as the ID and IP address of the peer node 1. After receiving the online status, the peer index node 2 updates the related information of each peer node 1. If the user a of the peer node na and the user b of the peer node nb are on the other party's friend list on a social network, and the peer node na and the peer node nb are both turned on, the second line is online. The peer node 1 can query the peer index node 2 for the IP address of the other party to communicate. Accordingly, peer node 1 can establish a social path between different social networks for the user to form a global relationship.

Through the peer to peer network architecture, the P2P-iSM in this case integrates multiple heterogeneous social networks so that users on different social networks need to use any particular community. Group networks can communicate with each other. In other words, the aforementioned integration does not create additional social networks. Through P2P-iSM, the case provides a global relationship mode to access the global association strength between two users located in a heterogeneous social network. Through the global association model, the case also provides a search mechanism called i-Search to find the social path between the users of the heterogeneous social network. The case further provides an analysis model for analyzing the probability of the i-Search search mechanism in finding social paths and its search performance, and conducting extensive simulation studies to verify the analysis results of the case.

This case uses the mobile phone number as the only user identity identified above. The phone book in the handheld device (for example, Jenny's terminal device) serves as the basis for integrating heterogeneous social networks. In the example shown in Figure 2, Jenny has a friend named John, and John's phone number. The code is "0910456".

The case also provides and stores information about users' friends through a database and a list of friends. Figure 2b shows the format of the friend list. The aforementioned friend list contains three types of information: personal information, social network information, and address information.

The personal information field stores the ID of the user's friend (including the ID of the social network), the phone number (phone no), and the email address (Email). In different social networks, users may use different IDs. See the example in Figure 2b. Jenny's friend John uses the ID of John_f on Facebook and the ID of John_t on Twitter. The phone number is used as an entry for the list of friends in the associated phone book. This entry can be further mapped to multiple entries in the friends list.

The social network information field contains four sub-fields: SN type, T value, Timestamp, and Online status. The social network type describes the registered social network. As illustrated by Figure 2b, Jenny's friend John registered Facebook's ID as "John_f". The T value is stored in the global correlation model and the result is calculated by equation (1). The T value is used to indicate how often a user interacts with a friend through a social activity (eg, Jenny posts a post on John's Facebook page, presses "Like", or sends a message, etc.). As illustrated in Figure 2b, when applying equation (1), Jenny ← John has a T value of 0.9 on Facebook. The time stamp is the time at which the T value is calculated. The online status indicates whether a particular friend is on this social network and when he or she recently logged in. If the online status is "ON" / "OFF", the recorded time is the time when John logs in/out of Facebook. Illustrated in Figure 2b, "On_12'0215_1430" means John's time to log in to Facebook February 15, 2012 14:00, and currently hanging on Facebook.

The address information field records the IP address and the connection nickname of the friend terminal device. This information is valid when it is opened on the same side of the friend.

Figure 3 shows the software architecture of the P2P-iSM in this case. The software architecture of peer node 1 contains five classes and one function: PeerAgent 1.5, FeedRequestListener 1.1, SampleAuthListener 1.2, CreateFriendListListener 1.3, BackgroundService 1.10, and PhoneBook API 1.9. The detailed description of each category is as follows: FeedRequestListener 1.1 It is used to get the social activity status of the user on the social network. It is called API mAsyncRunner.request ("me/feed", newFeedRequestListener()) 3.3 provided by the community network, for example: Facebook Graph API 3, Twitter REST API 4, other social network API5...etc.

SampleAuthListener 1.2 is used when the user opens the peer node 1 and logs into the social network. SampleAuthListener 1.2 is implemented by the API provided by the social network (SessionEvents.addAuthListener(new SampleAuthListener())3.1).

CreateFriendListListener 1.3 obtains the user's friend by calling the SessionEvents.addAuthListener(new SampleAuthListener()) 3.1 application (API) mAsyncRunner.request ("me/friends", new CreateFriendListListener()) 3.2 provided by the social network. The ID of the social network and a list of friends who provide users.

BackgroundService class 1.10 provides information exchange between peer node 1, or peer node 1 and peer index node 2. This class provides a communication channel between peer nodes 1 under the i-Search mechanism. Further, the peer node 1 requests through this category. Other peer nodes 1 perform the i-Search mechanism (this section will be detailed later). The peer node 1 further informs the online status of the peer index node 2 through this category.

Peer Agent 1.5 is the main class. This category defines three functions: Update_Tvalue()1.6, Update_FriendList()1.7, and Relationship_Finding()1.8. Update_Tvalue() 1.6 and Update_FriendList() 1.7 are used to update the T value of the friend list and the online status field. Relationship_Finding() 1.8 then performs an i-Search mechanism to identify the directed social path between the two users.

Phone Book API 1.9 is used to retrieve friends information from users' mobile phones. Some of the smart phone operating systems have APIs that provide such an option, such as logging in to the Android (Anddriod) API that executes the program. By using the user's phone number, two or more accounts belonging to the same user can be identified and used as a basis for integrating different social networks.

The peer index node 2 is used to provide a database of Global ID list 2.5, and its data format content is as shown in FIG. Peer index node 2 and establish an item for each online peer node 1 on Global ID list 2.5. Similar to Friend List 1.4, Global ID list 2.5 contains three types of information: Personal Information, Social Network Information, and Address Information.

The personal information field is used to store the ID of the user (eg, the ID used by the user to log in on the social network), the phone number, and the email address. Note that the user should open the peer node 1 while logging into one or more social networks at the same time, so that the same user's different social network accounts can exist in multiple items in the Global ID list 2.5. The plurality of items are linked by the same phone number (or email address) of the user. In this case, you can choose to use One of the IDs (for example, the user's phone number) is stored in the peer index node 2 so that the user's other IDs can remain unknown to the peer index node 2.

The social network information field is the type of social network that stores the current user login (online).

The field of the address information records the IP address and nickname of the device of the peer node 1 opened by the user, and the information is valid when the peer node 1 is turned on.

Figure 3 is a software architecture diagram of the peer index node 2. It contains a main category IndexPeerAgent 2.1 and a database Global ID list 2.5. In the main class IndexPeerAgent 2.1, the receiveSocket.receive() function 2.3 is used to receive the message transmitted by the peer node 1. After receiving the message, call receivePacket.getData()2.2 to get the information in this message. The receiveSocket.send()2.4 is used to send a response message to the peer node 1.

When the user opens the peer node 1 at the terminal device, a login procedure is executed. The flowchart of the login procedure is shown in Figure 5:

Step 1: When the user opens peer node 1, create a SampleAuthListener 1.2 and execute a SessionEvents.addAuthListener(new SampleAuthListener()) function to authorize the user in the social network.

Step 2: If the authorization step is successful, the social network will respond to the user's social network ID by theSessionEvents.addAuthListener() function.

Step 3: The peer node 1 creates a Background Service 1.10 class to transmit a message carrying the user ID, phone number, email address, nickname, and social network type (ie, User_Online_Message message) to the peer index node 2. Then, the same Index node 2 creates an item in the Global ID list.

Steps 4 and 5: The peer node 1 creates CreateFriendListListener 1.3 (ie, the FriendList_Request and FriendList_Response message pairs) to obtain the ID of the user's friend in the social network, and creates an entry for each friend in the friend list. .

Steps 6 and 7: The peer node 1 uses the BackgroundService class to transmit a message (ie, the friends_OnlineStatus_Request and the Friends_OnlineSta-tus_Response message pair) to the peer index node 2 to query the user who has the user online.

Steps 8 and 9: The peer node 1 creates a FeedRequestListener 1.1 category to collect social activity information, and calculates a T value by exchanging the T Value_Parameter_Request and the T Value_Parameter_Response message pairs of the social network.

This case provides a global association model for identifying the global social association of a user on a heterogeneous social network by P2P-iSM. This case first provides a tool for measuring the strength of the global association between any two users located in a heterogeneous network. Next, the present invention further provides an i-Search mechanism for searching for a meaningful directed social path between any two peer nodes 1 in the P2P_iSM.

Before searching for a user's global association, the tool first needs to measure the strength of the association between any two users on the heterogeneous social network. This case modifies the traditional social network association attenuation function to propose a more accurate measurement scheme in the heterogeneous social network.

The associated social link a→b associated with frequency is defined as f(a,b), which is used to capture the frequency with which user a interacts with user b through certain social activities (eg, user a is User b's message board, leave a message, press "Like", send a message, call the other party, etc.). Consider the existence of c social activities, and 1 i C , let λ _{i be} defined as user a interacting with user b through the social activity of the i-th, and defining f ( a , b ) as:

Where w _i is the weight value of the i-th social activity, 0 w _i 1 and 1 i C , and . In equation (1), the weight value w _i is a fine-tuning tool used to reflect the degree of interaction in social associations. For example, pressing "Like" represents a more casual interaction intention, while sending an email indicates a stronger interaction intention. Therefore, we have a large weight value for social activities that send emails. As for λ _i, it means to obtain measurement information from the social network during a specified period (for example, monthly or daily). In the directed social association a→b, the larger the value of f ( a , b ), the more attention is paid to the user b on the user a. For example, suppose there is only one social activity: post ( w ₁ =1), and user a if the average is posted 5 times per day, when the user is on the message board of the user's website (if λ ₁ =5/ day ), then f ( a , b )= w ₁ λ ₁ =5/ day , and use the threshold θ to define f ( a , b ). In other words, if f ( a , b ) θ represents that user a has sufficient attention to user b.

In the case of a→b and b→a association, the interaction factor F ( a , b ) between user a and user b is revealed. The interaction factor is defined as:

When the value of F ( a , b ) is larger, there is more interaction between user a and user b. In equation (2), 0 F ( a , b ) 1 and F ( a , b ) = F ( b , a ).

Consider social graphics formed by heterogeneous social networks. As illustrated in Figure 6A, the social graph contains two social networks. Suppose a directional social path exists between user u ₁ (on community network SNS ₁ ) and user u _{L +1} (on community network SNS ₂ ), and the directed social path Via the user u ₂ , u ₃ ,..., u _L , at least one of the L+1 users is the peer node 1. And defining the directed social path as a set of links " P = { u ₁ → u ₂ , u ₂ → u ₃ , ..., u _{L -1} → u _L , u _L → u _{L +1} }". The directional social path contains L directional links (the distance between u ₁ and u _{L +1} is | P |= L ). To express this directional social path, this case defines that the global association exists between u ₁ and u _{L +1} . The case further provides a Z(P) function for measuring the strength between u ₁ and u _{L +1 in the} global correlation. The Z(P) function is defined as follows:

In equation (2), we have assumed 0 F ( u _i , u _{i +1} ) 1 and at 1 i When L is lower, F ( u _{i +1} , u _i )= F ( u _i , u _{i +1} ). Therefore 0 Z ( P ) 1. Moreover, the reverse social path P' of its social path P ( P '={ u _{L +1} → u _L ,..., u ₃ → u ₂ , u ₂ → u ₁ }) can obtain Z ( P ' ) = Z ( P ). When Z(P) is larger, it means that its global correlation intensity is higher. The correlation strength Z(P) provides more accurate friend recommendations and trust/credit indicators. It can be used as a basis for sharing content across social networks.

The present invention further provides an i-Search mechanism to find a directed social path between two peer nodes 1 in the P2P-iSN. Its i-Search establishes a social path through a link. When a join joins a path, its global correlation strength is calculated by Z(P) of equation (3). If the newly added path has a global correlation strength less than the threshold △, the search will stop. Threshold value Δ is used to ensure that the global association strength of the path is established so that the user has sufficient motivation to use the global community association model for other social networking applications.

In this case, the threshold value △ is set according to the search result of the social network (that is, the interactive factor F(a, b)=0.5) of the a→b link. Consider the path P whose length |P|=4, and then use the Z(P) function of equation (3) to find the strength of this path Z(P)=0.5 ⁴ =0.0625, the result representing a weak social association . Therefore, in this case, the threshold value △=0.5 ³ =0.125 is set in the subsequent performance discussion.

In other words, the social path searched by the i-Search mechanism has a social length of no more than three. As long as the a→b link interaction factor is F ( a , b ) β <1, whose global correlation strength will decay exponentially. Therefore, it is possible to maintain a lower complexity under a large number of searches.

The i-Search mechanism is described in detail as follows: The peer index node 2 provides the uplink state of the uplink peer node 1 (including the ID of the peer node 1 and the IP address). The friend list provided by the peer node 1 stores all friends' online information. To simplify the explanation, this case uses the friend b of the peer node a to indicate the existence of the a→b community connection relationship.

When the peer node 1 is turned on, its online information is reported to the peer index node 2, and the latest friend online state is received from the peer index node 2. With the latest online information, peer node 1 can determine whether his friend is online (in other words, the peer node 1 is also on). In order to let the peer node 1 on the line and communicate directly with his friends. Please refer to FIG. 6B. In this case, a recursive algorithm and i-Search are performed in the same node 1, in which the G set (setG) is a friend list of the peer node 1, and the input parameter s is Store the call algorithm 1 ID of the peer node 1. The parameter r is the ID of the peer node 1 being searched. At the beginning, we set P ← .

Considering the situation in which the peer node a searches for the peer node b, the user a communicates directly with the friend b under the online condition, and requests the friend b to perform the i-search algorithm (that is, executes b.iSearch() in algorithm 1. ). In other words, the directed social path P is established along the peer node 1 of the upper line.

If the i-Search mechanism finds multiple global social associations between any two peer nodes, the peer node 1 that triggers the i-Search mechanism can use the maximum global community association strength as the basis. according to. The system can also speed up the execution of i-Search by caching the search results of peer node 1.

In P2P-iSM, all peer nodes 1 and associated social paths form a social graph. Peer node 1 may be turned on or off when i-Search is performed, and the i-Search request needs to be contacted when a friend goes online. In other words, if the peer node a or the peer node b is off (offline), the social links a, b no longer exist. Therefore, when i-Search is executed, the physical network topology of its P2P-iSM will dynamically change.

In this case, the peer node a performs the i-Search mechanism to find the peer node b, and when the directed social path exists, the probability of path finding is defined as P _f . The state of the line above the peer node 1 will significantly affect P _f . In this paragraph, the case would provide a model for obtaining an approximation of P _f.

In order to simplify the discussion, the case assumes that the peer node 1 in the P2P-iSM has its network behavior, such as online status, interaction, etc., which are independent and have the same distribution (iid). As discussed earlier, in this paragraph, we set the threshold for the i-Search mechanism to be Δ = 0.5 ³ = 0.125. In the analysis model of this case, we use the constraint |P| 3 without using △ 0.125. That is, the i-Search mechanism will determine that the global social path has not been found and exits the search when the path length reaches 3. Assume that the peer node 1 on state (ie, the on-line state) is the x period (the probability density function is f _x (.), the average value is 1 u _x ) is; its off state (ie, the offline state) is the y period (its The probability density function is f _y (.) and the mean is 1 u u _y ). The peer node 1 rotates during the x period and during the y period. It is assumed that the i-Search mechanism requests to reach the peer node 1 to form a Poisson process. Therefore, when the peer node 1 goes online, and the probability of the i-Search request arrives, p _on :

This case provides a social graph for the P2P-iSM by the Watts-Strogatz model. The aforementioned Watts-Strogatz model has three parameters α (the probability of reconnection), n (the total number of peer nodes 1 in P2P-iSM), and m (the average number of friends of peer node 1), and sets: 0< α <1 and n>>m>>ln n>>1 (5)

The Watts-Strogatz model has a small-world property that includes a small average path length and a high cluster for exploring social networks.

In this case, when the peer node a performs i-Search, the expected value of receiving the i-Search request message is defined as N _t . Consider the following scenario: the peer node 1 performs an i-Search mechanism to search for the directed social path to the peer node b. If b belongs to one of the N _t peer nodes 1, the directed social path of the peer nodes a to b is found, and Available:

The N _t derivation is as follows: Two node states are included in the Watts-Strogatz model: far-node and near-node. The far node system defines that peer node 1 has a social path after reconnection, and its probability is α . The near node definition has an initial social link.

In the social graph of P2P-iSM, Nf and Nn are respectively represented on the far node and the near node receives the expected value of the i-Search request when i-Search is executed. According to this: N _t = N _f + N _n

The acquisition of Nf and Nn is as follows: One round represents that when users a, b are online, and i-Search requests to use the directed community link a → b to complete the delivery. In the i-Search mechanism, there are up to three loops to construct a directed community path. In each loop, the peer node 1 It can be a far node or a near node when the trigger loop operates:

Case 1: The peer node 1 triggers the start loop as a far node. In this case, there are mαp _on far nodes _on average, and m (1- α ) p _on near nodes can receive the i-Search request.

Case 2: When node 1 triggers the start loop, it is a near node. Since the near node has a relatively high probability of receiving the i-Search request in advance when transmitting the i-Search request to other near nodes, we consider that only the far node can receive the i-Search request to obtain an approximation. In this case, an average of mαp _on far nodes can receive the i-Search request.

Next, use the following iterative procedure to calculate Nf and Nn:

Procedure 1:

Input parameters: α , m , u _x , u _y

Output measured value: N _f , N _n , N _t

Step 1: Select the initial value, N _f ←1, N _n ←0, and round ←0

Step 2:

Step 3: If (round 3), then go to Step 2.Otherwise,go to the next step.

Step 4: N _t ← N _f + N _n ; return.

This analytical model is validated by a simulation model driven by a discrete event. The simulation model has been widely adopted and used to simulate several research models in mobile communication networks. Preliminary experiment. The simulation model simulates the above-line/offline behavior of peer node 1, and the behavior of the i-Search mechanism.

For the simulation model, the case uses the aforementioned discrete event-driven approach and defines the following five event types:

‧The QUERY_ARRIVAL event: indicates that the peer node 1 on the line opens the i-Search mechanism to find other peer nodes 1.

‧The QUERY_FORWARD event: indicates that the peer node 1 on the line sends an i-Search request to the friend who is online.

‧ The QUERY_RESPONSE event: indicates that an on-line peer node 1 performs an i-search algorithm for the peer node 1 transmitting the i-Search request, and returns a result (ie, finds a path).

‧The ONLINE event: indicates that peer node 1 is on.

‧The OFFLINE event: indicates that peer node 1 is off.

The case further provides a timestamp t _s to indicate when the event occurred. The aforementioned event is further inserted into an event list and deleted/processed according to one of the non-decreasing time stamps in the list. When performing the simulation, an analog clock is provided to illustrate the progress of the simulation. The parameters used to simulate the model are as follows:

‧ N _r : The number of cycles executed for the i-Search request.

‧a: ID of the peer node 1 that triggers the i-Search mechanism.

‧d: ID of the peer node 1 found.

‧l: Indicates whether the social link exists between the two peer nodes 1.

In the above simulation model, the following counter is used to calculate the output measurement:

‧ The C _f counter counts the total number of successful paths found.

‧ The C _q counter counts the total number of times the QUERY_ARRIVAL event has been processed.

This simulation was repeated in this case until C _q reached 100,000 times to ensure the stability of the simulation results and to obtain the following output measurements:

Fig. 7 and Fig. 8 show the comparison results of the analysis values and the simulation values. The parameter settings of the details are as follows. The figure shows that the simulation results are very similar to the analysis results.

Next, the case investigate the influence of the input parameters P _f effectiveness and i-Search Mechanism. In the discussion here, the input parameters of this case are based on the constraints of equation (5), and the total number of peer nodes 1 is set to n=100. The effect of its input parameters is as follows.

In Figures 7 and 8, the case changes u _y / u _x from 0.8 to 8. When the value of u _y / u _x is larger, the time for the peer node 1 to go online is longer. For example, when we know that u _y / u _x =0.5 and u _y / u _x =8 from equation (4), we can know that p _on =1/3 and p _on =8/9, respectively. This two graph shows that the path search probability p _f increases with u _y / u _x . It is worth noting that, please refer to Figure 7. When u _y / u _x =8, α=0.8, m=6, p _f will be 15% larger; please refer to Figure 8, and when u _y / When u _x = 8 and m = 10, p _{f is} about 40 percent.

Please observe Figure 7, when we set m=6 and study the effect of α. When α is larger, the social path representing P2P-iSN is sparse (that is, more distant nodes). Figure 7 shows that p _f increases as α increases, which means that the i-Search mechanism can achieve better chances of finding in a sparse social path.

Please observe Figure 8, we set α = 0.4 and observe the effect of m. When the m value is larger, it means that the peer node 1 has more friends. Figure 8 indicates that when there is more friends, i-Search mechanism can achieve a better performance p _f.

In summary, when located at a relatively sparse and the peer node 1 has more friends, the i-Search mechanism has a 40% chance of finding a global social association for the user, in other words, the social path has a close social relationship. Association

The detailed description of the preferred embodiments of the present invention is intended to be limited to the scope of the invention, and is not intended to limit the scope of the invention. The patent scope of this case.

1,na, nb‧‧‧ peer node

2‧‧‧Same index node

Claims (8)

A peer network method for integrating a heterogeneous social network, the peer network method being applied to a server device, comprising the steps of: connecting a plurality of peer nodes, each peer node defining a user terminal and for accessing At least one social network; and the server device analyzes the social association between the at least one social network, the social activity during the specified period, and the weight value of the social activity to analyze between the two peer nodes Orienting social associations and generating directional and weighted social frequencies, and associating the equivalent nodes with mutual relevance and the social frequency in accordance with the threshold, thereby integrating into a social network, wherein the peers The social network system determines that the equivalent node has a valid social association between the at least one social network; wherein the server analyzes the social strength of the social association between the equivalent nodes to determine the validity a social association, or an associated peer node with an indirect association within a set of the social associations between the two peer nodes The strength of the social, to determine whether the associated social valid, then the servo means to provide interaction between the two peer node; wherein the social frequency is expressed by the following equation of: Where f ( a , b ) represents the social frequency of the social interaction associated with the user a → user b, w _i represents the weight of the i-th activity, and λ _i represents that the user a operates the b The frequency of class i activities. The method of claim 1, wherein the strength of the interactive activity associated with the mutual orientation is represented by the following equation: Where F ( a , b ) represents the intensity of the interaction between user a and user b, and f ( b , a ) represents the social frequency of the directed social association between the user b and the user a, θ represents It is the threshold. The method of claim 2, wherein the social intensity of the interactive activity intensity is represented by the following equation: Where u denotes the social associations, L denotes the associated association value of the social associations, and P denotes one of the social associations. The method of claim 1 further analyzes the added social strength of the social association to determine whether to identify the added social association. The method of claim 1, further comprising: accessing an online state of the peer node of the online connection; providing the peer node with an online status of the corresponding friendship connection to the online node, so that the peer node is in accordance with the online state The other equivalent node communication. A peer network system for integrating a heterogeneous social network, comprising: a communication module connecting a plurality of peer nodes, each peer node defining a user and for accessing at least one social network; and a a processing module, connected to the communication module, the processing module analyzes any two of the peers according to a social association between the at least one social network, a social activity during a specified period, and a weight value of the social activity Directed social associations between nodes and generate directional and weighted social frequencies, and associate the equivalent nodes with mutually related relevance and the social frequency in line with the threshold, thereby integrating into a social network. Wherein the peer community network determines that the equivalent node has a different effective social association between the at least one social network; wherein the processing module analyzes the social strength of the social association between the equivalent nodes, To identify the valid social association, or to analyze the association between the peer nodes with indirect associations within the set of the social associations between the two peer nodes The social strength, to determine whether the associated social effective, then the processing module provides the interaction between two peer node; wherein the social frequency is expressed by the following equation of: Where f ( a , b ) represents the social frequency of the social interaction associated with the user a → user b, w _i represents the weight of the i-th activity, and λ _i represents that the user a operates the b The frequency of class i activities. The system of claim 6, wherein the indirect association further comprises a friend recommendation value. The system of claim 6, wherein the processing module further accesses an online state of the peer node of the online connection, and provides the peer node with an online status corresponding to one of the friendship connections to the peer node, so that the same node is based on the same node. The online state communicates with other such equivalent nodes.