CN109981343A - Microblogging spread and epidemic degree prediction technique and device based on propagating accelerated degree - Google Patents
Microblogging spread and epidemic degree prediction technique and device based on propagating accelerated degree Download PDFInfo
- Publication number
- CN109981343A CN109981343A CN201910119940.5A CN201910119940A CN109981343A CN 109981343 A CN109981343 A CN 109981343A CN 201910119940 A CN201910119940 A CN 201910119940A CN 109981343 A CN109981343 A CN 109981343A
- Authority
- CN
- China
- Prior art keywords
- microblogging
- degree
- popularity
- model
- acceleration
- 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
- 230000001902 propagating effect Effects 0.000 title claims abstract description 40
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000001133 acceleration Effects 0.000 claims abstract description 30
- 238000006467 substitution reaction Methods 0.000 claims abstract description 4
- 208000025099 Absence of the pulmonary artery Diseases 0.000 claims abstract 7
- 238000012546 transfer Methods 0.000 claims description 18
- 238000012549 training Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 2
- 238000000691 measurement method Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 description 17
- 230000000644 propagated effect Effects 0.000 description 7
- 238000012417 linear regression Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 244000097202 Rathbunia alamosensis Species 0.000 description 2
- 235000009776 Rathbunia alamosensis Nutrition 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000006854 communication Effects 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 241000287196 Asthenes Species 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 230000001568 sexual effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 210000003813 thumb Anatomy 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L51/00—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
- H04L51/04—Real-time or near real-time messaging, e.g. instant messaging [IM]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L51/00—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
- H04L51/52—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail for supporting social networking services
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- Computing Systems (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of microblogging spread and epidemic degree prediction technique and device based on propagating accelerated degree, which comprises determine the object time T of Twitter message mt;By the acceleration A of each timesliceiWeighted arithmetic mean value as t0~TrPropagating accelerated degree;Increase the feature of the propagating accelerated degree on the basis of SH model, constructs Regress Forecast model;It determines that microblogging enlivens intensity relatively, obtains relative popularity N*(Tr) and counterpropagate acceleration A*(Tr);By relative popularity N*(Tr) and counterpropagate acceleration A*(Tr) N (T in the substitution Regress Forecast modelr) and A (Tr), new prediction model UAPA is obtained, and predict microblogging spread and epidemic degree based on the prediction model UAPA.
Description
Technical field
The present invention relates to social networks big data analysis field more particularly to a kind of microblogging propagation based on propagating accelerated degree
Popularity prediction method and device.
Background technique
In recent years, it incorporates rapidly in people's life by the part social networking application of representative of microblogging, and profoundly changes
The mode of user-to-user information transmitting.In the traditional media using newspaper, broadcast, TV as representative, information is propagated main
Mode is a small number of authoritative node release informations, and most users read, listen to or watch message, and information spread speed is opposite
Relatively slow, depth is shallower, and the difference of the spread scope in identical platform between different information is not very big.And in microblog
On, each user can oneself publication and forwarding information, the propagation of information be with publication, reading, forwarding, again read ... it is this
What the mode of continuous iteration duplication was propagated.Relative to traditional media, Twitter message spread speed faster, the road that information is propagated
Diameter and process randomness are stronger, and " popularity " propagated between different information is also unbalanced in the extreme, and most message are propagated
Range very little, and only a few hot spot message can be propagated in tremendous range.
Lot of domestic and international scholar has carried out a large amount of research work for Popularity prediction problem.HP Lab in 2008
Szabo and Huberman have studied the popularity variation tendency of model in Digg and YouTube, find note in the two websites
The early stage popularity of son and following popularity existing linear relationship after carrying out logarithmic transformation, and mentioned based on this discovery
The linear regression SH model of online information Popularity prediction is gone out.Pinto in 2013 et al. is improved on SH model, will
YouTube video propagation period early stage is divided into multiple timeslices, and future is predicted according to the spread and epidemic degree of each timeslice
Spread and epidemic degree, and a kind of improved multiple linear regression ML model is proposed based on this thought.
Although using SH model and ML model etc. can the spread and epidemic degree to the video model in YouTube carry out it is pre-
It surveys, but to Twitter message prediction effect and unsatisfactory.This is primarily due to for Digg and YouTube, micro-
The rich transmission of news period is shorter, speed faster, mechanism of transmission it is also more complicated, it is therefore necessary to the popularity to Twitter message is pre-
Survey problem further progress research.Existing method is mainly to predict the following prevalence according to the popularity value of message early stage propagation
Degree, does not account for the variation tendency of early stage popularity.Analysis is carried out by the spread and epidemic degree changing trend to microblogging to find, is passed
Broadcasting acceleration and the following popularity has certain correlativity.In addition, the popularity of Twitter message and the initial issuing time of message
It is closely related, the influence of this factor should be fully considered when being predicted.
Summary of the invention
The embodiment of the present invention provides a kind of microblogging spread and epidemic degree prediction technique and device based on propagating accelerated degree, to
Solve the problems of the prior art.
The embodiment of the present invention provides a kind of microblogging spread and epidemic degree prediction technique based on propagating accelerated degree, comprising:
Determine the object time T of Twitter message mt, wherein the object time TtPopularity for the Twitter message becomes
At the time of stablizing no longer growth;
By Twitter message m from t0To TrIt is equally divided into i period this period, the message m of each period Mo is accumulative to be turned
Hair number is N1..., Ni, by t0When forwarding number be set as N0=0, the transfer amount increased newly in each timeslice is Ni-Ni-1, when with i-th
Between the total transfer amount of message of section when starting approximately indicate the spread scope of moment m, be denoted as Ni-1, then the i-th period plus
Speed isTake the acceleration A of each timesliceiWeighted arithmetic mean value as t0~TrPropagating accelerated degree;
Increase the feature of the propagating accelerated degree on the basis of SH model, constructs Regress Forecast model;
It determines that microblogging enlivens intensity relatively, intensity amendment microblogging is enlivened in T according to the microblogging relativelyrPopularity N (Tr)
With propagating accelerated degree A (Tr), and respectively divided by TrThe microblogging at moment enlivens intensity relatively, obtains relative popularity N*(Tr) and phase
To propagating accelerated degree A*(Tr);
By relative popularity N*(Tr) and counterpropagate acceleration A*(Tr) substitute in the Regress Forecast model
N (Tr) and A (Tr), new prediction model UAPA is obtained, and carry out to microblogging spread and epidemic degree based on the prediction model UAPA
Prediction.
Preferably, the object time TtIt is 24 hours.
Preferably, the acceleration A of each timeslice is takeniWeighted arithmetic mean value as t0~TrPropagating accelerated degree tool
Body includes:
T is calculated according to formula 10~TrPropagating accelerated degree:
Wherein, TrTo start to carry out the reference time of prediction task after Twitter message publication a period of time.
Preferably, increase the feature of the propagating accelerated degree on the basis of SH model, construct Regress Forecast
Model specifically includes:
Regress Forecast model is constructed according to formula 2 and formula 3:
The final following Popularity prediction model are as follows:
Wherein,It is Twitter message m in TtThe forwarding number at moment, N (Tt) it is Twitter message m in TtThe reality at moment turns
Send out number, α0, α1, α2For model parameter, obtained by passing through least-squares estimation on training dataset.
Preferably, determine that microblogging enlivens intensity relatively and specifically includes:
It is a vector that microblogging enlivens intensity relatively, represents in daily 24 hours the intensity of enlivening of n-th hour platform, first
The average microblog number u forwarded per hour in microblog is first calculated, then calculating transfer amount average in each hour again is V
[n], wherein then enliven intensity relatively according to what formula 4 determined each hour platform in 1≤n≤24:
6, the method as described in claim 1, which is characterized in that by relative popularity N*(Tr) and counterpropagate acceleration
A*(Tr) N (T in the substitution Regress Forecast modelr) and A (Tr), new prediction model UAPA is obtained, and be based on
The prediction model UAPA carries out prediction to microblogging spread and epidemic degree and specifically includes:
Microblogging is corrected in TrPopularity N (Tr) and propagating accelerated degree A (Tr), it is relatively living divided by the microblogging at Tr moment respectively
Jump intensity, obtains relative popularity N according to formula 5 and formula 6*(Tr) and counterpropagate acceleration A*(Tr):
By relative popularity N*(Tr) and counterpropagate acceleration A*(Tr) substitute in the Regress Forecast model
N (Tr) and A (Tr), new prediction model UAPA is obtained according to formula 7:
Wherein, β0, β1, β2For model parameter, obtained by the Least Square Method on training set.
The embodiment of the present invention also provides a kind of microblogging spread and epidemic degree prediction meanss based on propagating accelerated degree, comprising: deposits
Reservoir, processor and it is stored in the computer program that can be run on the memory and on the processor, the computer
The step of above method is realized when program is executed by the processor.
The embodiment of the present invention is used, it was proved that new prediction model is to the Popularity prediction of Twitter message multiple
Better performance is all had in index, so as to accurately predict microblogging spread and epidemic degree.
The above description is only an overview of the technical scheme of the present invention, in order to better understand the technical means of the present invention,
And it can be implemented in accordance with the contents of the specification, and in order to allow above and other objects of the present invention, feature and advantage can
It is clearer and more comprehensible, the followings are specific embodiments of the present invention.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the present invention
Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Fig. 1 is the flow chart of the microblogging spread and epidemic degree prediction technique in the embodiment of the present invention based on propagating accelerated degree;
Fig. 2 is propagating accelerated degree and the following popularity figure in the embodiment of the present invention;
Fig. 3 is the microblog users liveness figure of different moments in one day in the embodiment of the present invention;
Fig. 4 is the early stage of two microbloggings in the embodiment of the present invention propagating accelerated degree, popularity and the following popularity exemplary diagram.
Specific embodiment
The embodiment of the present invention proposes a kind of new Twitter message Popularity prediction method, after this method is issued by microblogging
Propagating accelerated degree and popularity in a short time, the liveness of microblog users adds early stage popularity and propagation when in conjunction with microblogging publication
Speed is modified, and obtains UAPA (User Activity Propagation Acceleration) model.That is, this
Inventive embodiments are by the multifactor Twitter message Popularity prediction of being included in such as early stage popularity, propagating accelerated degree, user activity
In model, it is intended to propose more accurate and meet actual microblogging Popularity prediction propagation model.The embodiment of the present invention is by UAPA
Model and in the industry representational SH, ML and RPP model exist to the Popularity prediction of Twitter message into comparison, new prediction model
Better performance is all had in multiple indexs.
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
It is fully disclosed to those skilled in the art.
The embodiment of the invention provides a kind of Twitter message Popularity prediction methods, propose propagating accelerated this popularity of degree
The new feature of prediction, and the microblogging Popularity prediction model based on propagating accelerated degree, early stage popularity and user's activity periods
UAPA;The comparison for analyzing the different forecast result of model such as UAPA model and SH, ML and RPP and relative parameters setting are for pre-
The influence of effect is surveyed, the invention belongs to social networks big data analysis fields.
The embodiment of the present invention is main including the following three aspects:
First, the embodiment of the present invention proposes the propagating accelerated new feature for spending this Popularity prediction.Relative to social network
The models such as relatively broad SH [2] and ML [3] of video messaging Popularity prediction, propagate and add in the application of network information Popularity prediction
Velocity characteristic is more satisfactory to the prediction effect of microblogging popularity, can more characterize variation tendency and change big microblogging popularity.
Second, the embodiment of the present invention proposes the microblogging based on propagating accelerated degree, early stage popularity and user's activity periods
Popularity prediction model UAPA model.The model spends the spy newly defined according to microblogging early stage popularity situation, by propagating accelerated
Sign considers the time cycle and redefines the variables such as relative popularity simultaneously, proposes new prediction model UAPA (User
Activity Propagation Acceleration), i.e., increase propagating accelerated this feature of degree on the basis of SH model,
A two variable linear regression is constructed, realizes prediction.
Third, the embodiment of the present invention pass through the experimental analysis comparison and relative parameters setting of different forecast result of model
Influence for prediction effect.
The above description is only an overview of the technical scheme of the present invention, in order to more clearly understand technology hand of the invention
Section cooperates attached drawing to be illustrated below.
The research object of the embodiment of the present invention be Sina weibo message, relative to read number, comment number, thumb up number etc. refer to
Mark, the forwarding number of message can more portray the popularity of message propagation, while being also easier to acquisition and obtaining and carry out quantum chemical method, because
The forwarding number of this Twitter message of the invention portrays the popularity of Twitter message.
For each microblogging sample, the time that can be learnt the issuing time of source microblogging and it is forwarded every time, thus
Transfer sequence chain according to time sequence can be constructed.For a given Twitter message m, it is t that we, which define its issuing time,0
(submission time) uses t to m i-th forwarding timeiIt indicates, then the transfer sequence process of message m can use { t0,
t1,…,ti,…tfinalIndicate, wherein tfinalIndicate the time of message m last time forwarding.
During establishing prediction model, while defining reference time Tr(reference time) and object time
Tt(target time).Wherein reference time Tr is that news release starts to carry out the time of prediction task, T for a period of time afterwardsrGeneration
Time required for table observation message m early stage communication process feature.Object time TtRefer to that the popularity of message tends towards stability
At the time of no longer growth, it is clear that t0<Tr<Tt.N (t) is denoted as message m in the practical forwarding number of t moment by we, and m turns in t moment
The predicted value of hair number is denoted asSo prediction task can indicate are as follows: according to message m from t0To TrThis period of time message
Communication process { t0,t1,…,ti, forwarding number of the prediction m at the Tt momentWherein tiFor TrBefore last time forwarding when
Between.
The specific processing of Popularity prediction is as shown in Figure 1:
(1) object time is arranged
The goal in research of the embodiment of the present invention is prediction Sina weibo in object time TtThe popularity at moment, for each
For microblogging, life cycle is different.But for prediction task of the invention, microblogging life is not known in advance
Period can last long, it is therefore desirable to object time T be set in advancetOccurrence.On the one hand object time is wanted to cover
To most forwarding times of source microblogging, it could sufficiently reflect the following popularity of microblogging in this way;On the other hand because of microblogging
Life cycle is shorter relative to other social networks, and in order to make prediction work have certain timeliness, the object time is also unsuitable
It is arranged too long.
It is found by statistical analysis, microblogging forwards quantity, and there are unbalanced situations on Annual distribution.Relative to microblogging
For even tens days several days life cycles of message, most forwarding behaviors are propagated in initial several hours in microblogging
It completes, we only need to analyze and handle the relatively short time approximately obtain the final popularity of message.We
It has counted each Twitter message on data set to complete 85%, 90%, 95% in life cycle, being averaged the time required to transfer amount
Value, as shown in table 1.
Table 1
From we have observed that the forwarding 90% all occurs after microblogging issues within 24 hours, and completing 95% in table 1
Transfer amount then averagely needs 43 hours, and transfer amount increases very slow after microblogging issues 24 hours.In the subsequent work of the present invention
We are object time T in worktIt is set as 24 hours, is on the one hand connecing for the total transfer amount of transfer amount Zhan occurred within 24 hours
Nearly 90%, the final popularity of microblogging can be preferably embodied afterwards, be on the other hand to cover within 24 hours source microblogging to issue the latter
In the complete User Activity period, microblogging can be more comprehensively embodied in the propagation condition of different periods.
(2) propagating accelerated degree (Propagation Acceleration)
It is SH model and ML model that the Popularity prediction field of social network information is widely used at present, but to micro-
Rich popularity progress prediction effect is unsatisfactory, because video messaging popularity variation tendency is relatively stable, and microblogging
Variation tendency change greatly.The forwarding number in microblogging future is not only related with its absolute quantity of forwarding quantity in Tr, also
With it in t0~TrThe variation tendency of forwarding number is related in this period.
In order to further study the relationship of Twitter message early stage popularity variation tendency and the following popularity, we mention first
Gone out the concept of the propagating accelerated degree of microblogging: we are Twitter message m from t0To TrIt is equally divided into i period this period, each
It is N that the message m of period Mo, which adds up forwarding number,1... Ni, particularly, we are t0When forwarding number be set as N0=0, then each
The transfer amount increased newly in timeslice is Ni-Ni-1.Due to disappearing when the transfer amount increased newly in i-th of timeslice starts with the timeslice
The spread scope of breath is related, and the total transfer amount of message when we are started with the i-th period approximately indicates the biography of moment m
Range is broadcast, N is denoted asi-1.The acceleration of so the i-th period is denoted asThen we take the acceleration of each timeslice
The weighted arithmetic mean value of Ai is as t0~TrPropagating accelerated degree:
Obtain A (Tr) after, the relationship of propagating accelerated degree with the following popularity is further analyzed, is depicted in t0~Tr
Section transmission of news acceleration A (Tr) with the message in TtPopularity N (the T at momentt) relationship scatter plot (T in figurer=4 is small
When, Tt=24 hours, i 4), as shown in Fig. 2.It can be seen from the figure that generally propagating accelerated degree is in the following popularity
Now certain positive correlation, in most cases, the propagating accelerated biggish microblogging future popularity of degree are also higher.Therefore,
Think an important factor for propagating accelerated degree of microblogging can be used as a prediction future popularity, we increase on the basis of SH model
Add propagating accelerated this feature of degree, construct a two variable linear regression:
Wherein α0, α1, α2For model parameter, obtained by passing through least-squares estimation on training dataset.It is final not
Carry out popularity prediction model are as follows:
(3) user activity (User Activity)
The forwarding activity of user has apparent periodical on microblogging.For 24 hours one day, we have counted data
The curve graph for concentrating each period user to issue/forward microblogging quantity is as shown in Fig. 3.User was at one day 24 in microblog
Hour, liveness in different time periods differed greatly, minimum to early 8 this periods of liveness at 2 points of midnight, at the morning 10
The active state of comparison is in when to evening 21, and 22 when next day 1 be then microblog users most active period in one day, than under
The transfer amount in noon and evening has more 50% or so.
On the other hand, single microblogging is in reference time TrPopularity and in TrPropagating accelerated degree in addition to in microblogging
Other than appearance, participating user and social networks are related, and also and t0~TrThis period, the activity of the user was related.Fig. 4 illustrates two
Microblogging is in reference time Tr=4 hours popularities, it is propagating accelerated to spend and in Tt=24 hours popularities.Microblogging maIts is starting
5 points in the morning of time, it is in the microblog users most sluggish period, therefore it is in reference time TrPopularity it is not high,
Same reason, maIn t0~TrThe propagating accelerated degree of this period is not also high, but it have been found that it is in TtThe popularity at moment
Relatively high, this may be that ma itself has stronger popular sexual factor, therefore higher in the popularity of object time.And for
Twitter message mb, the start time is microblog users more active stage, T at 10 points of the morningrThe popularity and biography at moment
It is relatively high to broadcast acceleration, but it is in object time TtPopularity be not but it is very high, this may be because of mbPopularity itself is not
Especially strong, for no other reason than that its start time is the popular period, so its popularity and acceleration for propagating early stage all compared with
Height, but the popularity that itself relatively weak popularity causes it following is not high.
Therefore, only consider TrThe popularity and acceleration at moment come predict the following popularity be it is incomplete, should also fill
Divide the liveness for considering microblog when Twitter message is starting.The present invention proposes that microblogging enlivens the concept of intensity relatively, it is one
A vector, represent in daily 24 hours n-th hour platform enlivens intensity.It is defined as follows: first in calculating microblog
Then the average microblog number u forwarded per hour is V [n] (1≤n≤24) calculating transfer amount average in each hour.Each
Hour platform enlivens intensity relatively are as follows:
It reflects in microblog each period the activity of the user in a time cycle.On this basis we
Microblogging is corrected in TrPopularity N (Tr) and propagating accelerated degree A (Tr), respectively divided by TrThe microblogging at moment enlivens intensity relatively, obtains
To relative popularity N*(Tr) and counterpropagate acceleration A*(Tr),
And substitute the N (T in formula (3) prediction modelr) and A (Tr), obtain a new prediction model UAPA (User
Activity Propagation Acceleration), concrete form are as follows:
Wherein β0, β1, β2For model parameter, obtained by the Least Square Method on training set.
In conclusion representational SH, ML and RPP model is into comparison by UAPA model and in the industry, by means of this hair
The technical solution of bright embodiment, it was proved that new prediction model is to the Popularity prediction of Twitter message in multiple indexs
Better performance is all had, so as to accurately predict microblogging spread and epidemic degree.
Obviously, those skilled in the art should be understood that each module of the above invention or each step can be with general
Computing device realize that they can be concentrated on a single computing device, or be distributed in multiple computing devices and formed
Network on, optionally, they can be realized with the program code that computing device can perform, it is thus possible to which they are stored
It is performed by computing device in the storage device, and in some cases, it can be to be different from shown in sequence execution herein
Out or description the step of, perhaps they are fabricated to each integrated circuit modules or by them multiple modules or
Step is fabricated to single integrated circuit module to realize.In this way, the present invention is not limited to any specific hardware and softwares to combine.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (7)
1. a kind of microblogging spread and epidemic degree prediction technique based on propagating accelerated degree characterized by comprising
Determine the object time T of Twitter message mt, wherein the object time TtTend to be steady for the popularity of the Twitter message
At the time of fixed no longer growth;
By Twitter message m from t0To TrIt is equally divided into i period this period, the message m of each period Mo adds up forwarding number
For N1..., Ni, by t0When forwarding number be set as N0=0, the transfer amount increased newly in each timeslice is Ni-Ni-1, with the i-th period
The total transfer amount of message when beginning approximately indicates the spread scope of moment m, is denoted as Ni-1, then the acceleration of the i-th period
ForTake the acceleration A of each timesliceiWeighted arithmetic mean value as t0~TrPropagating accelerated degree;
Increase the feature of the propagating accelerated degree on the basis of SH model, constructs Regress Forecast model;
It determines that microblogging enlivens intensity relatively, intensity amendment microblogging is enlivened in T according to the microblogging relativelyrPopularity N (Tr) and pass
Broadcast acceleration A (Tr), and respectively divided by TrThe microblogging at moment enlivens intensity relatively, obtains relative popularity N*(Tr) and opposite biography
Broadcast acceleration A*(Tr);
By relative popularity N*(Tr) and counterpropagate acceleration A*(Tr) N in the substitution Regress Forecast model
(Tr) and A (Tr), new prediction model UAPA is obtained, and carry out in advance to microblogging spread and epidemic degree based on the prediction model UAPA
It surveys.
2. the method as described in claim 1, which is characterized in that the object time TtIt is 24 hours.
3. the method as described in claim 1, which is characterized in that take the acceleration A of each timesliceiWeighted arithmetic mean value
As t0~TrPropagating accelerated degree specifically include:
T is calculated according to formula 10~TrPropagating accelerated degree:
Wherein, TrTo start to carry out the reference time of prediction task after Twitter message publication a period of time.
4. the method as described in claim 1, which is characterized in that increase the spy of the propagating accelerated degree on the basis of SH model
Sign, building Regress Forecast model specifically include:
Regress Forecast model is constructed according to formula 2 and formula 3:
The final following Popularity prediction model are as follows:
Wherein,It is Twitter message m in TtThe forwarding number at moment, N (Tt) it is Twitter message m in TtThe practical forwarding number at moment,
α0, α1, α2For model parameter, obtained by passing through least-squares estimation on training dataset.
5. the method as described in claim 1, which is characterized in that determine that microblogging enlivens intensity relatively and specifically includes:
It is a vector that microblogging enlivens intensity relatively, and represent in daily 24 hours n-th hour platform enlivens intensity, is counted first
The average microblog number u forwarded per hour in microblog is calculated, then calculating transfer amount average in each hour again is V [n],
In, intensity is then enlivened according to what formula 4 determined each hour platform in 1≤n≤24 relatively:
。
6. the method as described in claim 1, which is characterized in that by relative popularity N*(Tr) and counterpropagate acceleration A*(Tr)
Substitute the N (T in the Regress Forecast modelr) and A (Tr), new prediction model UAPA is obtained, and based on described pre-
It surveys model UAPA and microblogging spread and epidemic degree is carried out predicting to specifically include:
Microblogging is corrected in TrPopularity N (Tr) and propagating accelerated degree A (Tr), it is relatively active strong divided by the microblogging at Tr moment respectively
Degree, obtains relative popularity N according to formula 5 and formula 6*(Tr) and counterpropagate acceleration A*(Tr):
By relative popularity N*(Tr) and counterpropagate acceleration A*(Tr) N in the substitution Regress Forecast model
(Tr) and A (Tr), new prediction model UAPA is obtained according to formula 7:
Wherein, β0, β1, β2For model parameter, obtained by the Least Square Method on training set.
7. a kind of microblogging spread and epidemic degree prediction meanss based on propagating accelerated degree characterized by comprising memory, processing
Device and it is stored in the computer program that can be run on the memory and on the processor, the computer program is described
It realizes when processor executes such as the step of signal measurement method described in any one of claims 1 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910119940.5A CN109981343A (en) | 2019-02-18 | 2019-02-18 | Microblogging spread and epidemic degree prediction technique and device based on propagating accelerated degree |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910119940.5A CN109981343A (en) | 2019-02-18 | 2019-02-18 | Microblogging spread and epidemic degree prediction technique and device based on propagating accelerated degree |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109981343A true CN109981343A (en) | 2019-07-05 |
Family
ID=67077071
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910119940.5A Pending CN109981343A (en) | 2019-02-18 | 2019-02-18 | Microblogging spread and epidemic degree prediction technique and device based on propagating accelerated degree |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109981343A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8983889B1 (en) * | 1996-03-25 | 2015-03-17 | Martin L. Stoneman | Autonomous humanoid cognitive systems |
CN104915397A (en) * | 2015-05-28 | 2015-09-16 | 国家计算机网络与信息安全管理中心 | Method and device for predicting microblog propagation tendencies |
CN107784387A (en) * | 2017-09-18 | 2018-03-09 | 国家计算机网络与信息安全管理中心 | The continuous dynamic prediction method that a kind of microblogging event information is propagated |
CN108304867A (en) * | 2018-01-24 | 2018-07-20 | 重庆邮电大学 | Information popularity prediction technique towards social networks and system |
-
2019
- 2019-02-18 CN CN201910119940.5A patent/CN109981343A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8983889B1 (en) * | 1996-03-25 | 2015-03-17 | Martin L. Stoneman | Autonomous humanoid cognitive systems |
CN104915397A (en) * | 2015-05-28 | 2015-09-16 | 国家计算机网络与信息安全管理中心 | Method and device for predicting microblog propagation tendencies |
CN107784387A (en) * | 2017-09-18 | 2018-03-09 | 国家计算机网络与信息安全管理中心 | The continuous dynamic prediction method that a kind of microblogging event information is propagated |
CN108304867A (en) * | 2018-01-24 | 2018-07-20 | 重庆邮电大学 | Information popularity prediction technique towards social networks and system |
Non-Patent Citations (1)
Title |
---|
朱海龙 等: "基于传播加速度的微博流行度预测方法", 《计算机研究与发展》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Lindauer et al. | Best practices for scientific research on neural architecture search | |
Vinyals et al. | Grandmaster level in StarCraft II using multi-agent reinforcement learning | |
Shaman et al. | Forecasting seasonal outbreaks of influenza | |
Silver et al. | Mastering the game of Go with deep neural networks and tree search | |
Silvestro et al. | PyRate: a new program to estimate speciation and extinction rates from incomplete fossil data | |
CN110326004A (en) | Use consistency of path learning training strategy neural network | |
Manceau et al. | Phylogenies support out‐of‐equilibrium models of biodiversity | |
Das et al. | The effects of feedback on human behavior in social media: An inverse reinforcement learning model | |
Zhao et al. | Mahrl: Multi-goals abstraction based deep hierarchical reinforcement learning for recommendations | |
Peng | Assortative mixing, preferential attachment, and triadic closure: A longitudinal study of tie-generative mechanisms in journal citation networks | |
Roh et al. | State-dependent doubly weighted stochastic simulation algorithm for automatic characterization of stochastic biochemical rare events | |
CN110235149A (en) | Neural plot control | |
Liu et al. | Link prediction in a user–object network based on time-weighted resource allocation | |
Wejnert et al. | Respondent-driven sampling: operational procedures, evolution of estimators, and topics for future research | |
Pfeiffer et al. | Temporal patterns of genes in scientific publications | |
CN109816544A (en) | Information Propagation Model implementation method and device based on contact probability | |
Rechavi et al. | Not all is gold that glitters: Response time & satisfaction rates in yahoo! answers | |
Shi et al. | Dares: an asynchronous distributed recommender system using deep reinforcement learning | |
CN108520337B (en) | Riadry risk assessment method based on network risk entropy difference | |
Meiss et al. | Agents, bookmarks and clicks: a topical model of web navigation | |
CN109981343A (en) | Microblogging spread and epidemic degree prediction technique and device based on propagating accelerated degree | |
WO2020076679A1 (en) | Distributed digital currency mining to perform network tasks | |
Severiukhina et al. | Parallel data-driven modeling of information spread in social networks | |
Earl et al. | Optimal allocation of replicas to processors in parallel tempering simulations | |
Millar et al. | Consistency and fairness in real-time distributed virtual environments: Paradigms and relationships |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190705 |
|
RJ01 | Rejection of invention patent application after publication |