CN106503859A - A kind of message propagation prediction method and device based on online social relation network - Google Patents
A kind of message propagation prediction method and device based on online social relation network Download PDFInfo
- Publication number
- CN106503859A CN106503859A CN201610963409.2A CN201610963409A CN106503859A CN 106503859 A CN106503859 A CN 106503859A CN 201610963409 A CN201610963409 A CN 201610963409A CN 106503859 A CN106503859 A CN 106503859A
- Authority
- CN
- China
- Prior art keywords
- message
- time
- forwarding
- popularity
- online social
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 230000006870 function Effects 0.000 claims description 43
- 238000010276 construction Methods 0.000 claims description 10
- 230000006399 behavior Effects 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 238000002372 labelling Methods 0.000 claims description 7
- 238000007476 Maximum Likelihood Methods 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 2
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 abstract description 2
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 abstract description 2
- 238000011160 research Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 5
- 238000009792 diffusion process Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 244000097202 Rathbunia alamosensis Species 0.000 description 2
- 235000009776 Rathbunia alamosensis Nutrition 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 2
- 238000007418 data mining Methods 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003780 insertion Methods 0.000 description 1
- 230000037431 insertion Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000012958 reprocessing Methods 0.000 description 1
- 230000007480 spreading Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The present invention proposes a kind of message propagation prediction method and device based on online social relation network, it is related to social media and big data technical field, the method comprising the steps of 1, for a piece of news d, obtain after which sends [0, T] in the time period, user pays close attention to the time serieses of behavior arrival to which;The time serieses are modeled by step 2, and the model that modeling is generated is learnt, the model parameter of the model is trained, and according to the model parameter, obtain message popularity anticipation function.The present invention cope with data gush discovery as;Contrasted by MAPE, the method accuracy rate is higher;Flexible form, may apply to other application scene.
Description
Technical field
The present invention relates to social media and big data technical field, more particularly to a kind of based on online social relation network
Message propagation prediction method and device.
Background technology
In recent years, the online service of the social network such as social networkies, Social Media, ecommerce is fast-developing, people
Substantial amounts of large-scale graph data has been descended in accumulation, and Sina weibo register user number has formed complexity more than 600,000,000 between these users
Concern relation, transmission microblogging is nearly 100,000,000 daily, and concern relation of the microblogging along between user is propagated, and forms Spreading and diffusion net
The business department of network network-oriented big data and the operation system based on network big data are also collected and have accumulated the figure of magnanimity
Data, the storage, tissue, analysis and process for magnanimity diagram data, have become depth analysis and effectively utilizes magnanimity figure number
According to technical bottleneck and business crucial.
For social medias such as recent fast-developing Twitter, Sina weibos, research worker is pre- around link
The aspects such as survey, power of influence analysis, Information Communication, information recommendation expand corresponding research, and typical work has:Meeder et
al.(S.H.Yang,B.Long,A.Smola,N.Sadagopan,Z.Zheng and H.Zha.Like like alike:
joint friendship and interest propagation in social networks.In Proceedings
of the 20th international conference on World wide web,WWW’11,pages 537-546,
2011.) one kind is proposed under the scene of Twitter effectively according to Exist Network Structure and user's creation time projected relationship
The algorithm of setup time, for the evolution of network has further insight.Yang et al.(S.Wu,J.M.Hofman,
W.A.Mason and D.J.Watts.Who says what to whom on Twitter.In Proceedings of
the 20th international conference on World wide web,WWW’11,pages 705-714,
2011.) high correlation of user interest network and relational network is demonstrated, it is proposed that a kind of integration user interest network and pass
It is the algorithm frame of network, the algorithm is positioned in interest simultaneously and customer relationship is predicted and achieves preferably effect in the two tasks
Really.Wu et al.(D.M.Romero,B.Meeder and J.Kleinberg.Differences in the mechanics
of information diffusion across topics:idioms,political hashtags,and complex
contagion on Twitter.In Proceedings of the 20th international conference on
2011.) user on Twitter is classified by World wide web, WWW ' 11, pages 695-704, is counted and is divided
Analysed different classes of between user relation, message propagate etc. feature.Romero et al.(E.Bakshy,J.M.Hofman,
W.A.Mason and D.J.Watts.Everyone’s an influencer:quantifying influence on
Twitter.In Proceedings of The 4th International Conference on Web Search and
Data Mining, WSDM ' 11, pages 65-74,2011.) to hashtag on Twitter in different topics propagation row
For conducting in-depth research.Bakshy et al.(J.Teevan,D.Ramage and M.R.Morris.#
TwitterSerach:a comparison of microblog search and web search.In Proceedings
of The 4th International Conference on Web Search and Data Mining,WSDM’11,
Pages 35-44,2011.) measure power of influence by the size of event propagation tree, it is proposed that the power of influence based on regression tree model
Prediction algorithm simultaneously analyzes different characteristic for the impact for predicting the outcome.
The emerging in large numbers property (Emergence of information dissemination) of diffusion of information:With micro-blog,
During the online interactions such as social networkies, blog, network forum are for the online social relation network of primary service mode, the propagation of information
The media such as Traditional Newspaper's, broadcast, TV are different from, any individual therein, colony can both make and release news, it is also possible to logical
Cross the operation such as label, comment, reply, modification, forwarding to realize the reprocessing of information and propagate again, strong interactivity and strong evolution properties
It is an obvious characteristic of Information Communication in online social relation network.This strong interactivity and evolutive by force so that online society
Bulk information in relational network is propagated rapidly in a streaming manner and is spread, and the propagation moment of information is in one kind from content
Be distributed to the unstable of power of influence covering and emerge in large numbers state, existing research meanses be difficult to tackle effective modeling of diffusion of information rule and
The objective metric of Information Communication power of influence.
Content of the invention
For the deficiencies in the prior art, the present invention proposes a kind of message propagation forecast side based on online social relation network
Method and device.
The present invention proposes a kind of message propagation prediction method based on online social relation network, including:
Step 1, for a piece of news d, obtains after which sends within [0, the T] time period, and user pays close attention to behavior arrival to which
Time serieses;
The time serieses are modeled by step 2, and the model that modeling is generated is learnt, the model is trained
Model parameter, according to the model parameter, obtain message popularity anticipation function.
Message d includes [message id:message_id].
The step 1 includes:According to the message_id, obtaining from message library all has concern with message d
The massage set of relation, the forwarding massage set that the massage set is designated as message d, is designated as retweet_message_
sequence;
From issuing time value publish_time for forwarding and obtaining every message in massage set, then construction forwarding
Message time arrangement set retweet_time_sequence;
The forwarding message time arrangement set retweet_time_sequence and message_id is together returned
Return.
Traversal message library, judges whether every message in message library is to forward message, if it is, obtain forwarding message
Root_id, judge that whether the root_id is equal to the message_id, if equal to, then forward the message to be described in labelling
The forwarding message of message d;
The forwarding message of all message d after traversal terminates, is returned, retweet_time_sequence is designated as.
The step of model parameter for training the model in the step 2 is:Construction logarithm normal distribution relaxation function:
Wherein, t represents moment, μdFor mathematic expectaion, σd
For standard deviation, d represents message d;
Structural environment probability functionIf the i-th -1 time time concern due in isSo i & lt concern
At the momentThe probability of arrival meets such as minor function:
WhereinRepresent message d in moment t, λdIt is the intrinsic captivation of message, m is intrinsic initial concern number, θdRefer to
μd, σd;
Structural environment probability functionAt the momentThe probability for being not concerned with reaching and moment T between meets
Such as minor function:
Structural learning function, observes the popularity dynamic process of message d in time interval [0, T]Likelihood
For:
Wherein, L (λd, θd) refer to λdAnd θdMaximum-likelihood estimation,Refer at the momentAnd moment T between
It is not concerned with the probability for reaching;
By maximal possibility estimation, learn the parameter lambda for message ddAnd θd;
Return model parameter.
The concrete steps for obtaining message popularity anticipation function in the step 2 include:
Rate function x of the construction for message dd(t)=λdfd(t;θd)id(t), id(t)=m+i-1, wherein id
T () is truly to be concerned about to reach at the i-th -1 time to be truly concerned about the concern number up to period with i & lt;
Make cdT () is message popularity anticipation function, then have equation below according to Poisson process:
Wherein, in above formula, cd(t) be predict message d moment t popularity quantity, λdIt is intrinsic captivation f of messaged
(t;θd) it is normal distribution relaxation function;
The equation is calculated, message popularity anticipation function is obtained as follows:
Return message popularity anticipation function.
The present invention also proposes a kind of message propagation forecast device based on online social relation network, including:
Time serieses module is obtained, for for a piece of news d, obtaining after which sends within [0, the T] time period, user couple
Which pays close attention to the time serieses that behavior is reached;
Message popularity anticipation function is obtained, for being modeled to the time serieses, the model that modeling is generated is entered
Row study, trains the model parameter of the model, according to the model parameter, obtains message popularity anticipation function.
Message d includes [message id:message_id].
The acquisition time serieses module includes:According to the message_id, obtain from message library all with described
Message d has the massage set of concern relation, the forwarding massage set that the massage set is designated as message d, is designated as
retweet_message_sequence;
From issuing time value publish_time for forwarding and obtaining every message in massage set, then construction forwarding
Message time arrangement set retweet_time_sequence;
The forwarding message time arrangement set retweet_time_sequence and message_id is together returned
Return.
Traversal message library, judges whether every message in message library is to forward message, if it is, obtain forwarding message
Root_id, judge that whether the root_id is equal to the message_id, if equal to, then forward the message to be described in labelling
The forwarding message of message d;
The forwarding message of all message d after traversal terminates, is returned, retweet_time_sequence is designated as.
From above scheme, it is an advantage of the current invention that:
1st, cope with data gush discovery as;
2nd, contrasted by MAPE, the method accuracy rate is higher;
3rd, flexible form, may apply to other application scene.
Description of the drawings
Fig. 1 is overview flow chart of the present invention;
Fig. 2 is present invention concern time serieses construction flow chart;
Fig. 3 is that model parameter of the present invention trains flow chart;
Fig. 4 is anticipation function construction flow chart of the present invention.
Specific embodiment
When the research of social relation network is carried out, inventor has found that online social relation network is answering for various dimensions
Miscellaneous system, influences each other between the subsystem of network internal and interacts frequently.Within the system, user group's behavior is general
All there is the characteristics of gradually running up to a sunset and break out, and such accumulation is not linear, not being can be with simple superposition solution
Certainly, a series of often little changes, are each not enough to impact total system, but ought reach certain critical shape
During state, there is critical phase transformation in whole system, we term it emerging in large numbers phenomenon.The multi-source heterogeneous network information in Society information net
Interact in different levels, from being overlapped with the property of subsystem, entirety is the property for being showed whole network
Method description and solve that the property emerged in large numbers of system can not be divided and rule with subsystems, traditional model and analysis method without
System as method research.
Find through the research to emerging in large numbers phenomenon, by operator (AutoRegression) of the investigation based on Time-Series analyses,
Sequential logarithm auto-correlation prediction algorithm (Szabo&Huberman), SpikeM algorithms, popularity classification prediction algorithm, Lerman etc.
The stochastic prediction model of people, self-reinforcing Poisson process prediction algorithm etc., can be from structure diversity, timing dependence, popularity
Model insertion, Selection of kernel function multiple angle Selection high accuracy, enhanced scalability message popularity prediction algorithm right to realize
The prediction of message popularity.
General steps below for the present invention are as follows, as shown in Figure 1:
(1) for a piece of news d, obtain after which sends within [0, the T] time period, user pays close attention to behavior arrival to which
Time serieses;
(2) time serieses to reaching are modeled;
(3) model is learnt, trains the parameter of model;
(4) according to the model parameter for obtaining, final message popularity anticipation function is obtained.
Further, according to message d, obtain after which sends within [0, the T] time period, user to its pay close attention to behavior arrival when
Between sequential manner, as shown in Figure 2:
Input:[message id:message_id];
Output:[the time serieses that concern behavior is reached:publish_time_sequence].
Step 1, verifies the form of input data, if checking is not by directly returning failure information;
Step 2, obtains all other message sets for having concern relation with the message from message library according to message_id
Close, the forwarding massage set that these massage sets are designated as the message is designated as retweet_message_sequence;
Step 3, obtains issuing time value publish_ of every message from step 2 in the forwarding massage set for obtaining
Time, then constructs and forwards message time arrangement set retweet_time_sequence;
Step 4, time serieses retweet_time_sequence for obtaining together are returned with the message_id.
Further, being obtained from message library according to message m essage_id all has with the message other of concern relation to disappear
The mode of breath is as follows:
Whether step 1, travels through message library, be to forward message to every message in message library, if turning next step;
Step 2, obtains the root_id of the message, and root_id refers to the message_id of primary message, judges the message
Whether root_id is equal to message_id, if equal to then the labelling message is the forwarding message for being input into message;
Step 3, after traversal terminates, the forwarding message for returning all message d forwards message, is designated as retweet_
time_sequence.
Further, message library is constructed, and pretreatment is carried out to message, and whether every message of labelling is to forward message, with
And the method for the primary message id for forwarding, as shown in Figure 3:
Whether step 1, message collection device are to forward message to the every information authentication for collecting;
Step 2, if the message is to forward message, the message_id for obtaining the root message that the message is forwarded is designated as
Root_id, what root message issued the message content most start primary message.
Further, according to the time serieses for obtaining, model parameter is trained, the parameter of model is trained:
Input:[retweet_message_sequence];
Output:[model parameter].
Step 1, constructs logarithm normal distribution relaxation function:
Wherein, t represents moment, μdFor mathematic expectaion, σdFor
Standard deviation, d represent message d;
Step 2, structural environment probability functionBetween continuous concern, time interval length obedience is uneven twice
Poisson process.Therefore, if the i-th -1 time concern due in isSo i & lt focuses on the momentThe probability of arrival is full
Foot such as minor function:
WhereinRepresent message d in moment t, λdIt is the intrinsic captivation of message, m is intrinsic initial concern number;
Step 3, makes nd=i-1,Represent message d in moment ti-1, λdIt is the intrinsic captivation of message, m is intrinsic
Initial concern number, θdRefer to μd, σd, wherein μdFor mathematic expectaion, σdFor standard deviation, structural environment probability function?
MomentThe probability for being not concerned with reaching and moment T between meets such as minor function:
Step 4, structural learning function, the length of each time interval is separate.Therefore, time interval [0,
T] the interior popularity dynamic process for observing message dLikelihood be:
In above formula, L (λd, θd) refer to λdAnd θdMaximum-likelihood estimation,Refer at the momentWith moment T it
Between be not concerned with reach probability,Represent the moment is focused in i & ltThe probability of arrival, t represent moment, μdFor
Mathematic expectaion, σdFor standard deviation, m is intrinsic initial concern number,
Step 5, by Maximum-likelihood estimation, learns the parameter lambda of outbound message ddAnd θd;
Step 6, returns model parameter.
Further, according to the model parameter for obtaining, forecast model is created, as shown in Figure 4:
Input:[model parameter:λdAnd θd]
Output:[anticipation function]
Step 1, rate function x of the construction for message dd(t)=λdfd(t;θd)id(t), id(t)=m+i-1, wherein id
T () is truly to be concerned about up to i & lt to be truly concerned about the concern number up to period, λ at the i-th -1 timedIt is the intrinsic attraction of message
Power fd(t;θd) it is normal distribution relaxation function;
Step 2, structure forecast function, makes cdT () is the anticipation function of message popularity, then had according to Poisson process as follows
Equation:
In above formula, cd(t) be predict message d moment t popularity quantity, λdIt is intrinsic captivation f of messaged(t;
θd) it is normal distribution relaxation function;
Step 3, the differential equation shown in solution second step, the forecast model function for obtaining message spread and epidemic degree are as follows:
In above formula, m is intrinsic initial concern number, λdIt is the intrinsic captivation of message,
Return message propagation forecast function.
The present invention also proposes a kind of message propagation forecast device based on online social relation network, including:
Time serieses module is obtained, for for a piece of news d, obtaining after which sends within [0, the T] time period, user couple
Which pays close attention to the time serieses that behavior is reached;
Message popularity anticipation function is obtained, for being modeled to the time serieses, the model that modeling is generated is entered
Row study, trains the model parameter of the model, according to the model parameter, obtains message popularity anticipation function.
Message d includes [message id:message_id].
The acquisition time serieses module includes:According to the message_id, obtain from message library all with described
Message d has the massage set of concern relation, the forwarding massage set that the massage set is designated as message d, is designated as
retweet_message_sequence;
From issuing time value publish_time for forwarding and obtaining every message in massage set, then construction forwarding
Message time arrangement set retweet_time_sequence;
The forwarding message time arrangement set retweet_time_sequence and message_id is together returned
Return.
Traversal message library, judges whether every message in message library is to forward message, if it is, obtain forwarding message
Root_id, judge that whether the root_id is equal to the message_id, if equal to, then forward the message to be described in labelling
The forwarding message of message d;
The forwarding message of all message d after traversal terminates, is returned, retweet_time_sequence is designated as.
Claims (10)
1. a kind of message propagation prediction method based on online social relation network, it is characterised in that include:
Step 1, for a piece of news d, obtains after which sends within [0, the T] time period, user to its pay close attention to behavior arrival when
Between sequence;
The time serieses are modeled by step 2, and the model that modeling is generated is learnt, the mould of the model is trained
Shape parameter, according to the model parameter, obtains message popularity anticipation function.
2. the message propagation prediction method based on online social relation network as claimed in claim 1, it is characterised in that described
Message d includes [message id:message_id].
3. the message propagation prediction method based on online social relation network as claimed in claim 1 or 2, it is characterised in that
The step 1 includes:According to the message_id, obtaining from message library all has disappearing for concern relation with message d
Breath set, the forwarding massage set that the massage set is designated as message d are designated as retweet_message_
sequence;
From issuing time value publish_time for forwarding and every message being obtained in massage set, then construct and forward message
Time serieses set retweet_time_sequence;
The forwarding message time arrangement set retweet_time_sequence and message_id is together returned.
4. the message propagation prediction method based on online social relation network as claimed in claim 3, it is characterised in that traversal
Message library, judges whether every message in message library is to forward message, if it is, obtaining the root_id for forwarding message, sentences
Whether the root_id that breaks is equal to the message_id, if equal to, then it is message d to forward message described in labelling
Forward message;
The forwarding message of all message d after traversal terminates, is returned, retweet_time_sequence is designated as.
5. the message propagation prediction method based on online social relation network as claimed in claim 1, it is characterised in that described
The step of model parameter for training the model in step 2 is:Construction logarithm normal distribution relaxation function:
Wherein, t represents moment, μdFor mathematic expectaion, σdFor mark
Accurate poor, d represents message d;
Structural environment probability functionIf the i-th -1 time time concern due in isSo i & lt focuses on the momentThe probability of arrival meets such as minor function:
WhereinRepresent message d in moment t, λdIt is the intrinsic captivation of message, m is intrinsic initial concern number, θdRefer to μd,
σd;
Structural environment probability functionAt the momentThe probability for being not concerned with reaching and moment T between meets following letter
Number:
Structural learning function, observes the popularity dynamic process of message d in time interval [0, T]Likelihood be:
Wherein, L (λd, θd) refer to λdAnd θdMaximum-likelihood estimation,Refer at the momentDo not have and moment T between
It is concerned about the probability for reaching;
By maximal possibility estimation, learn the parameter lambda for message ddAnd θd;
Return model parameter.
6. the message propagation prediction method based on online social relation network as claimed in claim 1, it is characterised in that described
The concrete steps for obtaining message popularity anticipation function in step 2 include:
Rate function x of the construction for message dd(t)=λdfd(t;θd)id(t), id(t)=m+i-1, wherein id(t) be
Truly be concerned about for the i-th -1 time to reach the concern number up to period is truly concerned about with i & lt;
Make cdT () is message popularity anticipation function, then have equation below according to Poisson process:
Wherein, in above formula, cd(t) be predict message d moment t popularity quantity, λdIt is intrinsic captivation f of messaged(t;
θd) it is normal distribution relaxation function;
The equation is calculated, message popularity anticipation function is obtained as follows:
Return message popularity anticipation function.
7. a kind of message propagation forecast device based on online social relation network, it is characterised in that include:
Time serieses module is obtained, for for a piece of news d, obtaining after which sends within [0, the T] time period, user is closed to which
The time serieses that note behavior is reached;
Message popularity anticipation function is obtained, the model for being modeled, to modeling generation to the time serieses
Practise, train the model parameter of the model, according to the model parameter, obtain message popularity anticipation function.
8. the message propagation forecast device based on online social relation network as claimed in claim 7, it is characterised in that described
Message d includes [message id:message_id].
9. message propagation forecast device as claimed in claim 7 or 8 based on online social relation network, it is characterised in that
The acquisition time serieses module includes:According to the message_id, obtain from message library all relevant with message d
The massage set of note relation, the forwarding massage set that the massage set is designated as message d, is designated as retweet_
message_sequence;
From issuing time value publish_time for forwarding and every message being obtained in massage set, then construct and forward message
Time serieses set retweet_time_sequence;
The forwarding message time arrangement set retweet_time_sequence and message_id is together returned.
10. the message propagation forecast device based on online social relation network as claimed in claim 9, it is characterised in that time
Message library is gone through, judges whether every message in message library is to forward message, if it is, the root_id for forwarding message is obtained,
Judge whether the root_id is equal to the message_id, if equal to, then it is message d to forward message described in labelling
Forwarding message;
The forwarding message of all message d after traversal terminates, is returned, retweet_time_sequence is designated as.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610963409.2A CN106503859A (en) | 2016-10-28 | 2016-10-28 | A kind of message propagation prediction method and device based on online social relation network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610963409.2A CN106503859A (en) | 2016-10-28 | 2016-10-28 | A kind of message propagation prediction method and device based on online social relation network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106503859A true CN106503859A (en) | 2017-03-15 |
Family
ID=58321705
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610963409.2A Pending CN106503859A (en) | 2016-10-28 | 2016-10-28 | A kind of message propagation prediction method and device based on online social relation network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106503859A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109145109A (en) * | 2017-06-19 | 2019-01-04 | 国家计算机网络与信息安全管理中心 | User group's message propagation anomaly analysis method and device based on social networks |
CN109635989A (en) * | 2018-08-30 | 2019-04-16 | 电子科技大学 | A kind of social networks link prediction method based on multi-source heterogeneous data fusion |
CN109961183A (en) * | 2019-03-20 | 2019-07-02 | 重庆邮电大学 | A kind of comment information registers the measure of influence on user |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103699650A (en) * | 2013-12-26 | 2014-04-02 | 清华大学 | Message propagation prediction method and message propagation prediction device |
CN105005583A (en) * | 2015-06-17 | 2015-10-28 | 清华大学 | Method and system for predicting information forwarding increment in social network |
-
2016
- 2016-10-28 CN CN201610963409.2A patent/CN106503859A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103699650A (en) * | 2013-12-26 | 2014-04-02 | 清华大学 | Message propagation prediction method and message propagation prediction device |
CN105005583A (en) * | 2015-06-17 | 2015-10-28 | 清华大学 | Method and system for predicting information forwarding increment in social network |
Non-Patent Citations (2)
Title |
---|
HUA-WEI SHEN等: "Modeling and Predicting Popularly Dynamics via Reinforced Poisson Processes", 《THE TWENTY-EIGHTH AAAI CONFERENCEON ARTIFICIAL INTELLIGENCE》 * |
陈乃月: "微博舆论发展趋势预测方法研究与实现", 《中国硕士学位论文全文数据库-信息科技辑》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109145109A (en) * | 2017-06-19 | 2019-01-04 | 国家计算机网络与信息安全管理中心 | User group's message propagation anomaly analysis method and device based on social networks |
CN109145109B (en) * | 2017-06-19 | 2022-06-03 | 国家计算机网络与信息安全管理中心 | User group message propagation abnormity analysis method and device based on social network |
CN109635989A (en) * | 2018-08-30 | 2019-04-16 | 电子科技大学 | A kind of social networks link prediction method based on multi-source heterogeneous data fusion |
CN109635989B (en) * | 2018-08-30 | 2022-03-29 | 电子科技大学 | Social network link prediction method based on multi-source heterogeneous data fusion |
CN109961183A (en) * | 2019-03-20 | 2019-07-02 | 重庆邮电大学 | A kind of comment information registers the measure of influence on user |
CN109961183B (en) * | 2019-03-20 | 2023-06-23 | 重庆邮电大学 | Method for measuring influence of comment information on user sign-in |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104991956B (en) | Microblogging based on theme probabilistic model is propagated group and is divided and account liveness appraisal procedure | |
US11100411B2 (en) | Predicting influence in social networks | |
Adar et al. | Tracking information epidemics in blogspace | |
Guo | The application of social network analysis in agenda setting research: A methodological exploration | |
CN106980692A (en) | A kind of influence power computational methods based on microblogging particular event | |
Tang et al. | Ranking user influence in healthcare social media | |
CN102945279B (en) | The appraisal procedure of microblog users influence power and device | |
CN106776841A (en) | The acquisition methods and system of a kind of internet public feelings event propagation index | |
Yamaguchi et al. | Tag-based user topic discovery using twitter lists | |
CN103218412B (en) | Public feelings information processing method and device | |
CN104182457A (en) | Poisson-process-model-based method for predicting event popularity in social network | |
CN106503859A (en) | A kind of message propagation prediction method and device based on online social relation network | |
Heredia et al. | Exploring the effectiveness of twitter at polling the united states 2016 presidential election | |
US11269928B2 (en) | Identification and analysis of cohesive and topic-focused groups of user accounts from user-generated content on electronic communication platforms | |
Guille et al. | Predicting the temporal dynamics of information diffusion in social networks | |
Yue et al. | Influences on query reformulation in collaborative web search | |
Schantl et al. | The utility of social and topical factors in anticipating repliers in twitter conversations | |
CN109145109A (en) | User group's message propagation anomaly analysis method and device based on social networks | |
Pálovics et al. | Temporal prediction of retweet count | |
Gün et al. | A hybrid approach for credibility detection in Twitter | |
Singh et al. | Linked data in crowdsourcing purposive social network | |
Java | A framework for modeling influence, opinions and structure in social media | |
Cherichi et al. | Analyzing the behavior and text posted by users to extract knowledge | |
Li et al. | RETRACTED ARTICLE: Capture-removal model sampling estimation based on big data | |
Cui et al. | Burst diffusion of highly retweeted scholarly articles in Social Media. |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170315 |
|
WD01 | Invention patent application deemed withdrawn after publication |