CN103258248A - Method, device and system for predicting microblog fashion trend - Google Patents
Method, device and system for predicting microblog fashion trend Download PDFInfo
- Publication number
- CN103258248A CN103258248A CN2013101902253A CN201310190225A CN103258248A CN 103258248 A CN103258248 A CN 103258248A CN 2013101902253 A CN2013101902253 A CN 2013101902253A CN 201310190225 A CN201310190225 A CN 201310190225A CN 103258248 A CN103258248 A CN 103258248A
- Authority
- CN
- China
- Prior art keywords
- microblogging
- user
- behavioural characteristic
- time interval
- forwarding
- 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.)
- Granted
Links
Images
Abstract
The invention provides a method, device and system for predicting a microblog fashion trend. The method for predicting the microblog fashion trend comprises the steps: behavior characteristic description of a microblog user group is obtained; the behavior characteristics for releasing a microblog message of a user and the behavior characteristics of forwarding and commenting the microblog message of the user are represented by the behavior characteristic description of the microblog user group. The method for predicting the microblog fashion trend further comprises the steps: according to the forwarding amount and the commenting amount of the microblog message from the first time interval to the i-1 time interval after the microblog message is released and the behavior characteristic description of the microblog user group, the forwarding amount and the commenting amount of the microblog message at the i time interval are calculated, wherein the i is a positive integer larger than or equal to 1. According to the method for predicting the microblog fashion trend, behavior characteristics of the microblog user group are described from the prospect of time dimension, the accuracy and efficiency of predicting the microblog fashion trend are guaranteed and meanwhile, prediction can be conducted on line in real time.
Description
Technical field
The present invention relates to the network data analysis in the social networks, relate in particular to a kind of microblogging fashion trend Forecasting Methodology, Apparatus and system.
Background technology
Microblogging is a kind of platform of sharing, propagating and obtain based on the customer relationship information of carrying out, and its a kind of new social medium as the Web2.0 epoch occur and the gesture fast development to destroy the decadent force easily.Microblogging comes lastest imformation with the literal about 140 words usually, and realizes sharing immediately.The microblogging system has simplicity, the convenience of issue, information ageing, rich, and a series of characteristics such as the interactivity propagated of information, fissionability.
Simplicity, the convenience of the issue of microblogging system for content, the characteristics of the fission formula of microblogging early warning system propagation are combined with the user group that the microblogging system is huge in addition, make the microblogging platform become a kind of brand-new information channel.For the personal user, the microblogging system is that a good information is shared and the platform that obtains; For the enterprise customer, the microblogging system be a marketing and with the platform of user interaction; And for departments such as governments, the microblogging system is a hot information monitor supervision platform.Though the microblogging system is containing huge value, but excavate and utilize these value, must there be method to find that some interesting or valuable microblogging information attract the notice of platform user, must have method to predict which microblogging content can become popular information in the future, thus save that the user seeks and the time of browsing information or make the user more morning, quickly some problems are handled.The prediction of microblogging fashion trend can overcome the above problems, and can predict in advance by the prediction of microblogging fashion trend becomes popular microblogging information in the future, thereby as required microblogging information is correspondingly processed.
Current, mainly contain two classes at microblogging fashion trend Study on Forecast: a class is some features (such as the bean vermicelli number, pay close attention to number, user whether authenticate etc.) according to user account, and some attributes of microblogging content (as whether comprising picture, content-length etc. in the content) are predicted, but this Forecasting Methodology is unsatisfactory, the popularity effect of predicting certain bar microblogging may be poor, the popularity of all microbloggings of user's issue might predicting out all is the same, thereby causes inaccurate problem; Also having a class is to predict according to the annexation between the forwarding user of certain bar microblogging, but because the data volume that this mode need be obtained and calculate is too big, thereby can only carry out by off-line.In sum, the method that lacks a kind of prediction microblogging fashion trend of can be when guaranteeing forecasting accuracy can also real-time online handling at present.
Summary of the invention
According to one embodiment of present invention, provide a kind of microblogging fashion trend Forecasting Methodology, comprising:
Step 1), acquisition microblogging user group behavioural characteristic are described, user's behavioural characteristic during the behavioural characteristic of described microblogging user group behavioural characteristic description list requisition family issue microblogging and microblogging are transmitted and commented on;
Step 2), issue forwarding and the number of reviews in the 1st to i-1 the time interval of back according to microblogging, and described microblogging user group behavioural characteristic is described, calculate forwarding and the number of reviews in described microblogging i time interval after issue, wherein i is the positive integer greater than 1.
In one embodiment, obtaining microblogging user group behavioural characteristic in the step 1) describes and comprises: obtain that behavioural characteristic that the user issues microblogging is described matrix and microblogging is transmitted and comment in user's behavioural characteristic matrix is described;
Wherein, user's behavioural characteristic of issuing microblogging capable n column element of m of describing matrix user of being characterized in m the time interval in the time cycle issues the ratio that microblogging quantity and the user in n the time interval issue microblogging quantity;
Microblogging transmit and comment in user's the behavioural characteristic capable n column element of m of describing matrix be characterized in user's forwarding in microblogging issue m the time interval of back and the ratio of number of reviews and the user in n time interval forwarding and number of reviews.
In a further embodiment, the described time cycle is 24 hours.
In one embodiment, use following formula to calculate forwarding and the number of reviews in microblogging i time interval after issue step 2):
Wherein, forwarding and the number of reviews in rt (j) expression microblogging j time interval after issue; Rw[i] the described microblogging of [j] expression transmit and comment in user's behavioural characteristic the value of the capable j row of i in the matrix is described; T[i]=(t
0+ i) mod T, t
0In the corresponding time interval in the described time cycle during issue of expression microblogging, T represents the size of described time cycle; Tw[i] the described user of [j] expression behavioural characteristic of issuing microblogging describes the value of the capable j row of i in the matrix.
In one embodiment, also comprise before the step 1):
Step 0), making up microblogging user group behavioural characteristic describes.
In one embodiment, step 0) comprising:
Make up behavioural characteristic that the user issues microblogging matrix is described and make up that microblogging is transmitted and comment in user's behavioural characteristic matrix is described.
In one embodiment, making up behavioural characteristic that the user issues microblogging describes matrix and comprises:
The user in step a), a plurality of time cycles that do not overlap of statistics in interior each time interval issues microblogging quantity;
Step b), calculate each time interval and issue the average of microblogging quantity corresponding to the user in different time cycle, the user who obtains each time interval in the time cycle issues the microblogging number average value;
Step c), structure matrix, the number of its row and column is the number in the time interval in a time cycle, and the corresponding element of the capable n row of its m is that the user in m the time interval issues the ratio that microblogging number average value and the user in n the time interval issue the microblogging number average value in the time cycle.
In a further embodiment, a plurality of time cycles that do not overlap described in the step a) are nearest a plurality of cycles continuous time.
In one embodiment, make up that microblogging is transmitted and comment in user's behavioural characteristic describe matrix and comprise:
Steps A), forwarding and the number of reviews in statistics many microblogging each time interval after issue in a plurality of continuous time cycles;
Step B), calculate that issue each time interval of back transmits corresponding to the user of different microbloggings and the average of number of reviews, the user who obtains microblogging issue each time interval of back transmits and the number of reviews average;
Step C), make up matrix, the corresponding element of the capable n row of its m is that the user in microblogging issue m the time interval of back transmits and the ratio of number of reviews average and the user in n time interval forwarding and number of reviews average.
In a further embodiment, many microbloggings are that forwarding quantity or number of reviews are not 0 many microbloggings steps A).
In one embodiment, step 2) also comprise afterwards:
Step I), positive integer x=i is set;
Step II), with forwarding and the number of reviews in microblogging x time interval after issue of calculating, as forwarding and the number of reviews in microblogging issue x the time interval of back;
Step II I), issue forwarding and the number of reviews in the 1st to x the time interval of back according to microblogging, and described microblogging user group behavioural characteristic is described, calculate forwarding and the number of reviews in described microblogging x+1 time interval after issue, if x+1＜N, then x is set to x+1, returns Step II); Wherein N is the positive integer greater than i.
In one embodiment, described method also comprises:
Total forwarding and the number of reviews in step 3), described microblogging issue the 1st to i the time interval of back of calculating.
According to one embodiment of the invention, a kind of microblogging fashion trend arrangement method is provided, described method comprises:
According to above-mentioned microblogging fashion trend Forecasting Methodology, many microbloggings are calculated issue total forwarding and the number of reviews in the 1st to i the time interval of back;
According to total forwarding and the number of reviews in the 1st to i the time interval after many microblogging issues of calculating, described many microbloggings are carried out rank.
According to one embodiment of the invention, a kind of microblogging fashion trend prediction unit is provided, comprising:
Be used for the module that acquisition microblogging user group behavioural characteristic is described, user's behavioural characteristic during the behavioural characteristic of wherein said microblogging user group behavioural characteristic description list requisition family issue microblogging and microblogging are transmitted and commented on;
Computing module is used for forwarding and number of reviews according to microblogging issue the 1st to i-1 the time interval of back, and the description of described microblogging user group behavioural characteristic, calculates forwarding and the number of reviews in described microblogging i time interval after issue;
Wherein i is the positive integer greater than 1.
In one embodiment, be used for obtaining module that microblogging user group behavioural characteristic describes is used for by obtaining that behavioural characteristic that the user issues microblogging is described matrix and microblogging is transmitted and comment user's behavioural characteristic is described matrix and obtained microblogging user group behavioural characteristic and describe;
Wherein, user's behavioural characteristic of issuing microblogging capable n column element of m of describing matrix user of being characterized in m the time interval in the time cycle issues the ratio that microblogging quantity and the user in n the time interval issue microblogging quantity;
Microblogging transmit and comment in user's the behavioural characteristic capable n column element of m of describing matrix be characterized in user's forwarding in microblogging issue m the time interval of back and the ratio of number of reviews and the user in n time interval forwarding and number of reviews.
In a further embodiment, the described time cycle is 24 hours.
In one embodiment, described computing module uses following formula to calculate forwarding and the number of reviews in microblogging i time interval after issue:
Wherein, forwarding and the number of reviews in rt (j) expression microblogging j time interval after issue; Rw[i] the described microblogging of [j] expression transmit and comment in user's behavioural characteristic the value of the capable j row of i in the matrix is described; T[i]=(t
0+ i) mod T, t
0In the corresponding time interval in the described time cycle during issue of expression microblogging, T represents the size of described time cycle; Tw[i] the described user of [j] expression behavioural characteristic of issuing microblogging describes the value of the capable j row of i in the matrix.
In one embodiment, described device also comprises the description of microblogging user group behavioural characteristic and update module, is used for making up microblogging user group behavioural characteristic and describes.
In one embodiment, described microblogging user group behavioural characteristic is described with update module and is used for describing matrix and making up that microblogging is transmitted and comment user's behavioural characteristic is described matrix and made up microblogging user group behavioural characteristic and describe by making up behavioural characteristic that the user issues microblogging.
In one embodiment, described microblogging user group behavioural characteristic is described and is used for carrying out following function with update module and makes up the behavioural characteristic that the user issues microblogging and describe matrix:
The user who adds up each time interval in a plurality of time cycles that do not overlap issues microblogging quantity; Calculate each time interval and issue the average of microblogging quantity corresponding to the user in different time cycle, the user who obtains each time interval in the time cycle issues the microblogging number average value; Make up matrix, the number of its row and column is the number in the time interval in a time cycle, and the corresponding element of the capable n row of its m is that the user in m the time interval issues the ratio that microblogging number average value and the user in n the time interval issue the microblogging number average value in the time cycle.
In a further embodiment, described a plurality of time cycle that does not overlap is nearest a plurality of cycles continuous time.
In one embodiment, described microblogging user group behavioural characteristic is described with update module and is described matrix for the behavioural characteristic that the following function of execution makes up microblogging forwarding and comment user:
Add up forwarding and the number of reviews in many microblogging each time interval after issue in a plurality of continuous time cycles; Calculate issue each time interval of back corresponding to user's forwarding of different microbloggings and the average of number of reviews, the user who obtains microblogging issue each time interval of back transmits and the number of reviews average; The structure matrix, the corresponding element of the capable n row of its m is that microblogging is issued user's forwarding and the number of reviews average in m the time interval of back and the user in n the time interval transmits and the ratio of number of reviews average.
In a further embodiment, described many microbloggings are that forwarding quantity or number of reviews are not 0 many microbloggings.
In one embodiment, described computing module also is used for:
With forwarding and the number of reviews in microblogging i time interval after issue of calculating, as forwarding and the number of reviews in microblogging issue i the time interval of back; Forwarding and the number of reviews in each time interval of calculating described microblogging successively after i time interval after the issue are up to forwarding and the number of reviews of having calculated described microblogging N time interval after issue; Wherein N is the positive integer greater than i.
In one embodiment, described computing module also is used for calculating total forwarding and the number of reviews in described microblogging issue the 1st to i the time interval of back.
In one embodiment, described device also comprises the renewal of microblogging fashion trend and ranking module, is used for many microbloggings are calculated issue total forwarding and the number of reviews in the 1st to i the time intervals of back; And according to total forwarding and the number of reviews in the 1st to i the time interval after many microblogging issues of calculating, described many microbloggings are carried out rank.
According to one embodiment of the invention, a kind of microblogging fashion trend prognoses system is provided, comprise above-mentioned microblogging fashion trend prediction unit.
In one embodiment, described system also comprises the configuration documentation module, is used for the configuration information of the described microblogging fashion trend prognoses system of storage.
In one embodiment, described system also comprises the data read interface module, is used for reading described user group's behavioural characteristic and describes and update module required data and described computing module required data when the forwarding of prediction microblogging and the number of reviews when making up microblogging user group behavioural characteristic and describe.
In a further embodiment, described system also comprises microblogging fashion trend prediction logic control module, is used for controlling described data read interface module according to described configuration information and reads data; Control forwarding and the number of reviews of described computing module prediction microblogging; Controlling described microblogging user group behavioural characteristic describes and the description of update module structure microblogging user group behavioural characteristic; And control described microblogging fashion trend renewal and with ranking module many microbloggings are carried out rank.
Adopt the present invention can reach following beneficial effect:
The present invention is described user group's behavioural characteristic from time dimension, and utilizes user group's behavioural characteristic to carry out the prediction of microblogging fashion trend.Can dynamically adjust user group's behavioural characteristic as time passes, can online real-time processing when guaranteeing microblogging fashion trend forecasting accuracy and efficient.
Description of drawings
Fig. 1 is the process flow diagram of microblogging fashion trend Forecasting Methodology according to an embodiment of the invention;
Fig. 2 makes up the process flow diagram that microblogging user group behavioural characteristic is described the method for matrix according to an embodiment of the invention;
Fig. 3 upgrades the process flow diagram that microblogging user group behavioural characteristic is described the method for matrix according to an embodiment of the invention;
Fig. 4 is the running environment synoptic diagram of microblogging fashion trend prognoses system according to an embodiment of the invention;
Fig. 5 is the block diagram of microblogging fashion trend prognoses system according to an embodiment of the invention;
Fig. 6 is the flowchart of each module in the microblogging fashion trend prognoses system according to an embodiment of the invention; And
Fig. 7 is predict the outcome statistics synoptic diagram with real microblogging fashion trend of the microblogging fashion trend that adopts the microblogging fashion trend Forecasting Methodology of one embodiment of the invention to predict.
Embodiment
Below in conjunction with the drawings and specific embodiments the present invention is illustrated.
According to one embodiment of present invention, provide a kind of microblogging fashion trend Forecasting Methodology.Fig. 1 shows an embodiment of this Forecasting Methodology, may further comprise the steps:
The present invention can use various identification methods to identify a microblogging uniquely, from the issue microblogging, can the information that this microblogging is relevant correspond to its unique identification, in order to add up these information.The microblogging relevant information for example comprises issuing time, the microblogging content of microblogging, and the forwarding quantity in the issue back different time sections (or claiming the time interval) and number of reviews etc.
In one embodiment, the starting stage refers to microblogging from being published to current time (being known forwarding and number of reviews) during this period of time, and the starting stage can be made up of one or more time intervals.For example, if will be set at 1 hour the time interval, issue back 10 hours at microblogging, can add up this microblogging forwarding and number of reviews in this ten time intervals respectively.
The feature that microblogging user group behavioural characteristic can be divided into two aspects: all users issue the behavioural characteristic of microblogging in the first microblogging platform, and this feature can characterize the active degree of user in the whole microblogging platform; It two is group behavior features of user in the forwarding of every microblogging and the comment, and this feature can be characterized in microblogging user's in life cycle active degree.In one embodiment, above-mentioned microblogging user group behavioural characteristic describe that the matrix form that can adopt respectively as shown in table 1 and table 2 describes respectively that user in the microblogging platform issues the behavioural characteristic of microblogging and microblogging is transmitted and comment in user's group behavior feature.
Table 1
Table 2
By the statistical study of data set being found behavioural characteristic that user group in the microblogging platform issues microblogging has periodically and this cycle is 24 hours, namely (for example per hour) user issues the behavior of microblogging similar (be every day different users constantly issue the variation tendency basically identical of microblogging quantity) the identical period of every day.Therefore, in one embodiment, can set the time interval (step 1004) of user behavior feature (namely issue, transmit and comment on) statistics according to this periodicity.For example in one embodiment, if the time interval is set at 1 hour, then table 1 can be the matrix of 24*24.The user that the user that the element of the capable j of the i of table 1 row can be illustrated in the interior time interval i of one-period (for example 24 hours) issues microblogging quantity and time interval j issues the ratio of microblogging quantity, because time interval mutual relationship between any two is certain, so the behavioural characteristic that the user issues microblogging in the microblogging platform has symmetry in the matrix shown in the table 1, the product of two elements of symmetry is 1 mutually.Should be understood that table 1 described in the time cycle behavioural characteristic that all users in the microblogging platform issue microblogging, can there be multiple expression this time cycle, for example can be 24 hours, 48 hours etc.
The element of the capable j row of i can be represented user's forwarding and the user's forwarding of the time interval j after number of reviews and the issue and the ratio of number of reviews of the time interval i of single microblogging after its issue in the table 2.Table 2 can be a matrix time cycle of table 1 (namely greater than) greater than 24*24, and be lower triangular matrix, this is to have temporal precedence relationship because of forwarding and comment for a microblogging, namely can't learn time interval i+1, i+2 when time interval i ... actual forwarding and number of reviews.
Above described the matrix form that microblogging user group behavioural characteristic is described, described the flow process of matrix and will be described in more detail below and make up this microblogging user group behavioural characteristic.
In one embodiment, total forwarding and the number of reviews of certain bar microblogging of the time interval correspondence after predicting the starting stage, at first to predict forwarding and the number of reviews of microblogging in this time interval, add up this microblogging then from being published to all forwardings and the number of reviews in this time interval.Specifically, can comprise following two steps:
One, forwarding and the number of reviews of predicting microblogging certain time interval in future according to forwarding and the number of reviews of given starting stage.
Specifically, in this step, can predict forwarding and the comment number in i the time interval of microblogging according to the data (transmitting and number of reviews) in i-1 the time interval before the given microblogging.In a further embodiment, if predict forwarding and the number of reviews in the individual time interval of N (N〉i) according to the data in a preceding i-1 time interval, then can at first calculate forwarding and the comment number in i the time interval, calculate successively then, until the forwarding of calculating N the time interval and comment number, the calculating microblogging is as follows at the formula of the forwarding in i the time interval and number of reviews:
Parameter declaration in the formula (1) is as follows:
Rt (i) expression microblogging is in forwarding and the number of reviews in i the time interval; Tw[i] [j] expression user behavioural characteristic of issuing microblogging describes the value of the capable j row of i in the matrix (as shown in table 1); Rw[i] [j] corresponding microblogging transmit and comment in user's behavioural characteristic the value of the capable j row of i in the matrix (as shown in table 2) is described; T[i] represent that microblogging is transmitted and i corresponding user of the time interval of comment issues time interval in time cycle (as 24 hours) of correspondence in the behavioural characteristic of microblogging, it is 1 hour when the time interval for example, rt[i so] just represent microblogging issue back i cycle and just issue forwarding and the number of reviews sum of i hour correspondence in back, t[i] represent the corresponding time interval of behavioural characteristic that i hour corresponding user in microblogging issue back issues microblogging, for example 8 of users issue a certain microbloggings, 2nd hour forwarding comment number is 300, rt[2 so]=300, t[2]=8+2=10, the 10th time cycle in the corresponding time cycle (24 hours).
Therefore, t[i] can be expressed as t
0+ i, t
0A corresponding time interval in a time cycle when representing the microblogging issue.In one embodiment, if the microblogging corresponding time interval in a time cycle in when issue add i greater than this period of time T, then can value t
0+ i carries out modulo operation (t
0+ i) mod T obtains final t[i].
Second step, total forwarding and number of reviews that forwarding and the number of reviews summation in each time interval of microblogging issue back obtained this i time interval correspondence in microblogging issue back, as shown in Equation (2):
In formula (2), rts (i) represent i the time interval correspondence of a microblogging after issue forwarding and the comment total quantity, rt (i) be exactly in the formula (1) in i the time interval forwarding and the number of reviews of microblogging.
Should be understood that above-mentioned four steps only be used for to describe an embodiment of microblogging fashion trend Forecasting Methodology, can also change for example order between the step 1004 and step 1006.
In one embodiment, there has been the microblogging fashion trend to predict the outcome, namely obtain forwarding and the number of reviews of microblogging in certain time interval in future, upgrade the fashion trend of microblogging with this new predicting the outcome then, forwarding and the number of reviews of the microblogging that obtains according to prediction are carried out rank to the fashion trend of each microblogging.
According to one embodiment of present invention, make up process that microblogging user group behavioural characteristic describes matrix in the step 1006 as shown in Figure 2, may further comprise the steps:
As mentioned above, the time interval that microblogging user group behavioural characteristic is described can be set at 1 hour, and the time cycle is set at one day, and then microblogging user group behavioural characteristic is described the microblogging related data that employed data set can be up-to-date 1 middle of the month.This related data comprises: up-to-date 1 the middle of the month all users at the microblogging publish quantities in every day in each time interval, and forwarding and the number of reviews in every microblogging each time interval after issue of 1 issue in the middle of the month.
In one embodiment, 500 microbloggings can only adding up up-to-date 1 middle of the month are in forwarding and the number of reviews in its issue each time interval of back.
In one embodiment, the time cycle that wherein sets is that 1 day, the time interval are 1 hour, and the data set of setting is the set of the microblogging related data in up-to-date month (30 days).Then can add up data concentrated every day of microblogging publish quantities hourly, user hourly issues microblogging quantity to obtain in 30 days every day, then to each hour (being each time interval) corresponding 30 issue microblogging quantity sum up and get average, this average is as the issue microblogging quantity of this time interval correspondence.
In one embodiment, can be at first data centralization find out transmit quantity or number of reviews be not 0 microblogging (in a further embodiment, can in these microbloggings, select, for example 500), then according to the time interval of setting (1 hour) and the related data of data centralization, these microbloggings forwarding hourly and number of reviews after issue are added up, and per hour the forwarding in (each time interval) and number of reviews add and get average as microblogging issue back forwarding and the number of reviews of each hour after issue with these microbloggings.
In one embodiment, can add up forwarding and the number of reviews in microblogging issue one or more time intervals of back.In a further embodiment, can add up forwarding and the number of reviews of issue back 24 hours or more hours.
Wherein, user's behavioural characteristic of issuing microblogging describes that the corresponding value of the capable j row of i is the ratio that the time interval i user corresponding with time interval j issues microblogging quantity in the time cycle in the matrix.In the embodiment that the time cycle was set at 24 hours and the time interval is set at 1 hour, the user of a certain hour relative another hour issues the active degree of microblogging behavior in this matrix representation one day.For example, on the microblogging platform, quantity that 9 users issue microblogging is 500 morning among one day, be 400 and 10 users in morning issue the quantity of microblogging, then user's behavioural characteristic of issuing microblogging is described in the matrix corresponding 9 points of time interval 9() corresponding 10 points with time interval 10(of row at place) the corresponding value of row at place is 500/400=1.25.
Correspondingly, microblogging transmit and comment in user's the behavioural characteristic value of describing the capable j row of i in the matrix be the quantity of the forwarding of i time interval correspondence after the microblogging issue and comment with issue at microblogging after the ratio of quantity of the corresponding forwarding of j time interval and comment.Above-mentioned be among the embodiment of statistics time interval with one hour, each hour after the issue of microblogging of this matrix representation transmits with respect to the users of other hours and the active degree of comment behavior.For example, microblogging issue back the 1st hour forwarding and number of reviews are 1000, and the 2nd hour forwarding and number of reviews are 500, and then the value of the row (as secondary series) of the row of the time interval 1 correspondence (as first row) and the time intervals 2 correspondence is 1000/500=2.
Should be understood that step 2002-2008 has only exemplarily described makes up the process that microblogging user group behavioural characteristic is described matrix, and wherein the Build Order of two kinds of matrixes can be arbitrarily, be not restricted to the described embodiments.
Above-described embodiment has provided the method that initial microblogging user group behavioural characteristic is described that makes up, owing to As time goes on microblogging user group behavioural characteristic can change, for example, the quantity that the user in each time interval issues microblogging in a time cycle can change, and microblogging issue back also can change in forwarding and the number of reviews of different time at interval.Therefore also need be before prediction microblogging trend microblogging user group behavioural characteristic be described and upgrade, according to one embodiment of present invention, the implementation of this renewal mainly may further comprise the steps as shown in Figure 3:
Be similar to and make up the data set that microblogging user group behavioural characteristic is described use, this initiate data set comprises the microblogging related data of nearest a period of time.
Data set and initiate data set when making up microblogging user group behavioural characteristic.
Whether step 3004, judgement reach the maximal value of setting for the data set that makes up the description of microblogging user group behavioural characteristic:
Replace former data centralization time a part of data the earliest if then change step 3006 with initiate data, otherwise forward step 3008 to new data set is joined for the data centralization that makes up the description of microblogging user group behavioural characteristic, thereby obtain new data set.
Comprise that the behavioural characteristic that user in the microblogging platform is issued microblogging adds up, and to microblogging transmit and comment in user's behavioural characteristic add up.
According to one embodiment of present invention, also provide a kind of microblogging fashion trend prognoses system, be used for forwarding number and the comment number of microblogging are predicted.Fig. 4 shows the running environment of this microblogging fashion trend prognoses system, comprises internet, microblogging platform, and such as the terminal of PC, portable computer, mobile phone etc.As shown in Figure 4, microblogging fashion trend prognoses system need be carried out the mutual of data by the internet with existing microblogging platform.By reading of data, can be predicted the fashion trend of microblogging by microblogging fashion trend prognoses system, thereby the user can check microblogging fashion trend prediction result via the internet for example by above-mentioned terminal.
Fig. 5 shows the block diagram of an embodiment of microblogging fashion trend prognoses system.As shown in Figure 5, this system mainly comprises: configuration documentation module 5001, data read interface module 5101, microblogging fashion trend prediction logic control module 5201, user group's behavioural characteristic are described and update module 5301, microblogging fashion trend prediction module 5302, and the microblogging fashion trend is upgraded and ranking module 5303.Below respectively each module or unit in the system are described in detail.
A.
Configuration documentation module 5001
In one embodiment, configuration documentation module 5001 comprises operation some required configuration informations of microblogging fashion forecasting system, and for example: the data layout that reads, microblogging user group behavioural characteristic are described the update strategy that the size of employed data set or time range, microblogging user group behavioural characteristic describe, the time interval (being the time interval of microblogging user group behavior description) that microblogging user group behavioural characteristic statistics is used, the time period of microblogging fashion trend forecast updating etc.
Wherein, the data layout that reads refers to that the data that read comprise the type of which field, field etc.Size or time range that microblogging user group behavioural characteristic is described employed data set refer to: can adopt the data set of fixed size to carry out the description of microblogging user group behavioural characteristic, namely adopt up-to-date some data sets; Perhaps also can adopt the data set in the set time scope to carry out the description of user group's behavioural characteristic, namely adopt the data set of certain nearest time range (for example 1 month).Microblogging user group behavioural characteristic is described the time interval that update strategy comprises that the description of microblogging user group behavioural characteristic is upgraded, and the renewal that is used for the data set of microblogging user group behavioural characteristic description.In a further embodiment, if what user group's behavioural characteristic was described use is the data set of fixed size, then up-to-date data are added microblogging user group behavioural characteristic descriptor data set, replace time data the earliest if data set has reached maximal value; If use the data set of set time scope, then the data set that user group's behavioural characteristic is described is updated to the data set of up-to-date time range.The time interval of microblogging user group behavioural characteristic statistics is namely carried out microblogging user group behavioural characteristic employed time interval when describing.How long the time period of microblogging fashion trend forecast updating refers to every the microblogging that dopes is transmitted number and the comment number upgrades and rank again.
B.
Data read interface module 5101
Data read interface module 5101 is main is responsible for reading of data, according to the data layout of configuration with data read in microblogging fashion trend prognoses system, call for microblogging fashion trend prediction logic control module.In one embodiment, the data owner that relates in the native system will comprise microblogging information and microblogging lastest imformation, i.e. the data of the above-mentioned data centralization of describing for microblogging user group behavioural characteristic.In one embodiment, microblogging information can comprise that microblogging is in the forwarding quantity of the acquisition time of the issuing time of the content of the publisher's of the ID of the microblogging platform of correspondence, microblogging ID, microblogging, microblogging, microblogging, microblogging, number of reviews etc.Table 3 shows an embodiment of microblogging message structure, wherein UID(microblogging publisher unique identification) can be Optional Field.
Table 3
| Field name | Field type | Whether optional | Explanation of field |
| ID | String | Not | The microblogging unique identification |
| Text | String | Not | The microblogging content |
| Created_Time | Date | Not | The microblogging issuing time |
| Crawl_Time | Date | Not | Data acquisition time |
| Repost_Count | Int | Not | The hop count of microblogging during collection |
| Comment_Count | Int | Not | Microblogging comment number of times during collection |
| UID | String | Be | Microblogging publisher unique identification |
In one embodiment, the microblogging lastest imformation can be in information such as the forwarding quantity of different acquisition time microblogging correspondence and number of reviews, mainly comprise microblogging ID, update time, update time correspondence forwarding quantity, update time correspondence number of reviews etc., microblogging ID in the microblogging lastest imformation and microblogging ID corresponding (the microblogging ID that is same microblogging is constant) in the table 1, table 4 shows an embodiment of microblogging lastest imformation structure.
Table 4
| Field name | Field type | Whether optional | Explanation of field |
| ID | String | Not | The microblogging unique identification |
| Repost_Count | Int | Not | The forwarding number of microblogging during renewal |
| Comment_Count | Int | Not | The comment number of microblogging during renewal |
| Update_Time | Date | Not | Update time |
Should be understood that table 3 and table 4 only schematically show the structure of microblogging information and microblogging lastest imformation, and these information can adopt also other different structures to represent, for example can not comprise " whether optional " information.
C.
Microblogging fashion trend prediction logic control module 5201
Microblogging fashion trend prediction logic control module 5201 is used for actuating logic and the flow process of the whole microblogging fashion trend prognoses system of control.In one embodiment, microblogging fashion trend prediction logic control module 5201 is used for providing description and renewal, the microblogging of the reading of data that control describes in conjunction with Fig. 1-3 as mentioned, microblogging user group behavioural characteristic to transmit and comment on the prediction of number and functions such as the renewal of microblogging fashion trend and rank.
D.
User group's behavioural characteristic is described and update module 5301
What user group's behavioural characteristic was described and renewal 5301 is used for describing in conjunction with Fig. 2 as mentioned, configuration information according to configuration documentation module 5001, for example microblogging user group behavioural characteristic is described the time cycle of using, the time interval, the size of data set or time range, and the microblogging information that provides by data read interface module 5101 or microblogging lastest imformation (data set that for example is used for microblogging user group behavioural characteristic), microblogging user group behavioural characteristic is described, comprise make up behavioural characteristic that the user issues microblogging matrix is described and make up that microblogging is transmitted and comment in user's behavioural characteristic matrix is described.Because microblogging user's group behavior feature is dynamic change, As time goes on together with the adding of new data set, this module 5301 also is used for as mentioned describing in conjunction with Fig. 3, finishes the renewal of microblogging user group behavioural characteristic.
E.
Microblogging fashion trend prediction module 5302
Microblogging fashion trend prediction module 5302 is used for carrying out microblogging fashion trend Forecasting Methodology provided by the invention.Described in conjunction with Fig. 1-2 as mentioned, being used for obtaining microblogging user group behavioural characteristic describes, and describe according to microblogging user group behavioural characteristic, and microblogging is in information such as the forwarding of starting stage and number of reviews, the prediction microblogging is in forwarding and the number of reviews of certain time in the future.In one embodiment, this module can comprise for the module that obtains the description of microblogging user group behavioural characteristic, and the prediction microblogging is at the forwarding of certain time in future and the computing module of number of reviews.
F.
The microblogging fashion trend is upgraded and ranking module 5303
The microblogging fashion trend is upgraded and to be used for described in conjunction with Figure 1ly as mentioned with ranking module 5303, according to the update time of setting, microblogging fashion trend prediction result is upgraded and rank again.
According to one embodiment of the invention, above-mentioned user group's behavioural characteristic is described and update module 5301, microblogging fashion trend prediction module 5302, and the microblogging fashion trend is upgraded and ranking module 5303 can constitute microblogging fashion trend prediction unit, being used for obtaining microblogging user group behavioural characteristic describes, and according to forwarding and the number of reviews of microblogging issue back in the starting stage, and this microblogging user group behavioural characteristic is described forwarding and the number of reviews in the calculating time interval of microblogging after the starting stage.
Fig. 6 shows an embodiment of the execution flow process of each module in the above-mentioned microblogging fashion trend prognoses system, mainly may further comprise the steps:
Adopt the present invention for checking and carry out microblogging fashion trend prediction accuracy, the inventor adopts 22 days data set since June 26 in 2011, experimentizes above 234 microbloggings of 500 to transmitting quantity.According to the forwarding of 6 hours given starting stages and the data of number of reviews, predicted 7th hour, the 8th hour until 24th hour forwarding quantity.The inventor adds up the contrast of prediction result and actual value, and the result of statistics as shown in Figure 7.Among Fig. 7, transverse axis represents that the difference of predicted value and actual value accounts for the ratio of actual value, and the longitudinal axis represents that then corresponding microblogging accounts for the ratio of total microblogging, the span of the predicted value of most of microblogging be 1/2 actual value between 3/2 actual value, more satisfactory.As seen, adopt microblogging fashion trend Forecasting Methodology provided by the invention can predict in the future the popularity of microblogging sometime more exactly.
In addition, the inventor will consider Connection Density and consider that a plurality of methods that are connected the degree of depth (comprising annexation between the user) contrast that employed evaluation matrix is respectively shown in formula (3) and formula (4) in microblogging fashion trend Forecasting Methodology provided by the invention and the prior art:
Wherein, Pn and Rn are respectively predict popularity and the true popularities of n bar microblogging.The result of contrast is as shown in table 5:
Table 5
| ? | RMSE | MAE |
| Consider link density | 0.63 | 0.45 |
| Consider to propagate the degree of depth | 0.61 | 0.43 |
| User group's behavioural characteristic | 0.42 | 0.24 |
Wherein, RMSE is root-mean-square error, and MAE is mean absolute error.As seen, compared with prior art, the Forecasting Methodology based on microblogging user group behavioural characteristic provided by the invention has reduced the error of prediction significantly.The density information that uses Forecasting Methodology provided by the invention not need to obtain user's relation data, need not calculate connection simultaneously, only need obtain the forwarding in the starting stage and estimate data such as quantity, arithmetic speed can be improved significantly, and the requirement to the extensive processing of microblogging and real-time online processing can be adapted to.
Should be noted that and understand, under the situation that does not break away from the desired the spirit and scope of the present invention of accompanying Claim, can make various modifications and improvement to the present invention of foregoing detailed description.Therefore, the scope of claimed technical scheme is not subjected to the restriction of given any specific exemplary teachings.
Claims (10)
1. microblogging fashion trend Forecasting Methodology comprises:
Step 1), acquisition microblogging user group behavioural characteristic are described, user's behavioural characteristic during the behavioural characteristic of described microblogging user group behavioural characteristic description list requisition family issue microblogging and microblogging are transmitted and commented on;
Step 2), issue forwarding and the number of reviews in the 1st to i-1 the time interval of back according to microblogging, and described microblogging user group behavioural characteristic is described, calculate forwarding and the number of reviews in described microblogging i time interval after issue, wherein i is the positive integer greater than 1.
2. method according to claim 1 obtains microblogging user group behavioural characteristic and describes and comprise in the step 1): obtain that behavioural characteristic that the user issues microblogging is described matrix and microblogging is transmitted and comment in user's behavioural characteristic matrix is described;
Wherein, user's behavioural characteristic of issuing microblogging capable n column element of m of describing matrix user of being characterized in m the time interval in the time cycle issues the ratio that microblogging quantity and the user in n the time interval issue microblogging quantity;
Microblogging transmit and comment in user's the behavioural characteristic capable n column element of m of describing matrix be characterized in user's forwarding in microblogging issue m the time interval of back and the ratio of number of reviews and the user in n time interval forwarding and number of reviews.
3. use following formula to calculate forwarding and the number of reviews in microblogging i time interval after issue method according to claim 2, step 2):
Wherein, forwarding and the number of reviews in rt (j) expression microblogging j time interval after issue; Rw[i] the described microblogging of [j] expression transmit and comment in user's behavioural characteristic the value of the capable j row of i in the matrix is described; T[i]=(t
0+ i) mod T, t
0In the corresponding time interval in the described time cycle during issue of expression microblogging, T represents the size of described time cycle; Tw[i] the described user of [j] expression behavioural characteristic of issuing microblogging describes the value of the capable j row of i in the matrix.
4. according to any one described method among the claim 1-3, wherein, also comprise before the step 1):
Step 0), making up microblogging user group behavioural characteristic describes.
5. microblogging fashion trend arrangement method, described method comprises:
According to as the described microblogging fashion trend of one of claim 1-4 Forecasting Methodology, many microbloggings are calculated issue total forwarding and the number of reviews in the 1st to i the time interval of back;
According to total forwarding and the number of reviews in the 1st to i the time interval after many microblogging issues of calculating, described many microbloggings are carried out rank.
6. microblogging fashion trend prediction unit comprises:
Be used for the module that acquisition microblogging user group behavioural characteristic is described, user's behavioural characteristic during the behavioural characteristic of wherein said microblogging user group behavioural characteristic description list requisition family issue microblogging and microblogging are transmitted and commented on;
Computing module is used for forwarding and number of reviews according to microblogging issue the 1st to i-1 the time interval of back, and the description of described microblogging user group behavioural characteristic, calculates forwarding and the number of reviews in described microblogging i time interval after issue;
Wherein i is the positive integer greater than 1.
7. device according to claim 6 wherein is used for obtaining module that microblogging user group behavioural characteristic describes and is used for by obtaining that behavioural characteristic that the user issues microblogging is described matrix and microblogging is transmitted and comment user's behavioural characteristic is described matrix and obtained microblogging user group behavioural characteristic and describe;
Wherein, user's behavioural characteristic of issuing microblogging capable n column element of m of describing matrix user of being characterized in m the time interval in the time cycle issues the ratio that microblogging quantity and the user in n the time interval issue microblogging quantity;
Microblogging transmit and comment in user's the behavioural characteristic capable n column element of m of describing matrix be characterized in user's forwarding in microblogging issue m the time interval of back and the ratio of number of reviews and the user in n time interval forwarding and number of reviews.
8. device according to claim 7, wherein said computing module use following formula to calculate forwarding and the number of reviews in microblogging i time interval after issue:
Wherein, forwarding and the number of reviews in rt (j) expression microblogging j time interval after issue; Rw[i] the described microblogging of [j] expression transmit and comment in user's behavioural characteristic the value of the capable j row of i in the matrix is described; T[i]=(t
0+ i) mod T, t
0In the corresponding time interval in the described time cycle during issue of expression microblogging, T represents the size of described time cycle; Tw[i] the described user of [j] expression behavioural characteristic of issuing microblogging describes the value of the capable j row of i in the matrix.
9. according to any one described device among the claim 6-8, wherein said device also comprises the description of microblogging user group behavioural characteristic and update module, is used for making up microblogging user group behavioural characteristic and describes.
10. a microblogging fashion trend prognoses system comprises as the described microblogging fashion trend of claim 6-9 prediction unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310190225.3A CN103258248B (en) | 2013-05-21 | 2013-05-21 | A kind of microblog fashion trend Forecasting Methodology, Apparatus and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310190225.3A CN103258248B (en) | 2013-05-21 | 2013-05-21 | A kind of microblog fashion trend Forecasting Methodology, Apparatus and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103258248A true CN103258248A (en) | 2013-08-21 |
| CN103258248B CN103258248B (en) | 2016-12-07 |
Family
ID=48962149
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310190225.3A Active CN103258248B (en) | 2013-05-21 | 2013-05-21 | A kind of microblog fashion trend Forecasting Methodology, Apparatus and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103258248B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103678613A (en) * | 2013-12-17 | 2014-03-26 | 北京启明星辰信息安全技术有限公司 | Method and device for calculating influence data |
| CN103984701A (en) * | 2014-04-16 | 2014-08-13 | 北京邮电大学 | Micro-blog forwarding quantity prediction model generation method and micro-blog forwarding quantity prediction method |
| CN104281882A (en) * | 2014-09-16 | 2015-01-14 | 中国科学院信息工程研究所 | Method and system for predicting social network information popularity on basis of user characteristics |
| CN104516962A (en) * | 2014-12-18 | 2015-04-15 | 北京牡丹电子集团有限责任公司数字电视技术中心 | Monitoring method and system for microblogging public opinion |
| 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 |
| WO2016041376A1 (en) * | 2014-09-18 | 2016-03-24 | 华为技术有限公司 | Method and device for predicting information propagation in social network |
| CN106776667A (en) * | 2015-11-18 | 2017-05-31 | 财团法人资讯工业策进会 | Method and device for predicting influence of community articles |
| CN108717587A (en) * | 2018-05-25 | 2018-10-30 | 杭州知智能科技有限公司 | A method of text prediction forwarding task is pushed away based on the solution of multi-panel sorting network |
| CN110458360A (en) * | 2019-08-13 | 2019-11-15 | 腾讯科技（深圳）有限公司 | Prediction technique, device, equipment and the storage medium of hot resource |
| CN110795554A (en) * | 2019-10-29 | 2020-02-14 | 北京字节跳动网络技术有限公司 | Target information analysis method, device, equipment and storage medium |
| US10860941B2 (en) | 2014-09-18 | 2020-12-08 | Huawei Technologies Co., Ltd. | Method and device for predicting information propagation in social network |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102012929A (en) * | 2010-11-26 | 2011-04-13 | 北京交通大学 | Network consensus prediction method and system |
| US20110302103A1 (en) * | 2010-06-08 | 2011-12-08 | International Business Machines Corporation | Popularity prediction of user-generated content |
| CN102394798A (en) * | 2011-11-16 | 2012-03-28 | 北京交通大学 | Multi-feature based prediction method of propagation behavior of microblog information and system thereof |
| CN102999539A (en) * | 2011-09-13 | 2013-03-27 | 富士通株式会社 | Method and device for forecasting future development trend of given topic |
| CN103092921A (en) * | 2012-12-26 | 2013-05-08 | 中国科学院深圳先进技术研究院 | Dynamic prediction method facing microblog hot topic community and system |
-
2013
- 2013-05-21 CN CN201310190225.3A patent/CN103258248B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110302103A1 (en) * | 2010-06-08 | 2011-12-08 | International Business Machines Corporation | Popularity prediction of user-generated content |
| CN102012929A (en) * | 2010-11-26 | 2011-04-13 | 北京交通大学 | Network consensus prediction method and system |
| CN102999539A (en) * | 2011-09-13 | 2013-03-27 | 富士通株式会社 | Method and device for forecasting future development trend of given topic |
| CN102394798A (en) * | 2011-11-16 | 2012-03-28 | 北京交通大学 | Multi-feature based prediction method of propagation behavior of microblog information and system thereof |
| CN103092921A (en) * | 2012-12-26 | 2013-05-08 | 中国科学院深圳先进技术研究院 | Dynamic prediction method facing microblog hot topic community and system |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103678613B (en) * | 2013-12-17 | 2017-01-25 | 北京启明星辰信息安全技术有限公司 | Method and device for calculating influence data |
| CN103678613A (en) * | 2013-12-17 | 2014-03-26 | 北京启明星辰信息安全技术有限公司 | Method and device for calculating influence data |
| CN103984701A (en) * | 2014-04-16 | 2014-08-13 | 北京邮电大学 | Micro-blog forwarding quantity prediction model generation method and micro-blog forwarding quantity prediction method |
| CN104281882A (en) * | 2014-09-16 | 2015-01-14 | 中国科学院信息工程研究所 | Method and system for predicting social network information popularity on basis of user characteristics |
| CN104281882B (en) * | 2014-09-16 | 2017-09-15 | 中国科学院信息工程研究所 | The method and system of prediction social network information stream row degree based on user characteristics |
| US10860941B2 (en) | 2014-09-18 | 2020-12-08 | Huawei Technologies Co., Ltd. | Method and device for predicting information propagation in social network |
| WO2016041376A1 (en) * | 2014-09-18 | 2016-03-24 | 华为技术有限公司 | Method and device for predicting information propagation in social network |
| CN104516962A (en) * | 2014-12-18 | 2015-04-15 | 北京牡丹电子集团有限责任公司数字电视技术中心 | Monitoring method and system for microblogging public opinion |
| 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 |
| CN106776667A (en) * | 2015-11-18 | 2017-05-31 | 财团法人资讯工业策进会 | Method and device for predicting influence of community articles |
| CN108717587A (en) * | 2018-05-25 | 2018-10-30 | 杭州知智能科技有限公司 | A method of text prediction forwarding task is pushed away based on the solution of multi-panel sorting network |
| CN108717587B (en) * | 2018-05-25 | 2022-03-15 | 杭州一知智能科技有限公司 | Method for solving tweet prediction forwarding task based on multi-face sequencing network |
| CN110458360A (en) * | 2019-08-13 | 2019-11-15 | 腾讯科技（深圳）有限公司 | Prediction technique, device, equipment and the storage medium of hot resource |
| CN110458360B (en) * | 2019-08-13 | 2023-07-18 | 腾讯科技（深圳）有限公司 | Method, device, equipment and storage medium for predicting hot resources |
| CN110795554A (en) * | 2019-10-29 | 2020-02-14 | 北京字节跳动网络技术有限公司 | Target information analysis method, device, equipment and storage medium |
| CN110795554B (en) * | 2019-10-29 | 2022-07-26 | 北京字节跳动网络技术有限公司 | Target information analysis method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103258248B (en) | 2016-12-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103258248A (en) | Method, device and system for predicting microblog fashion trend | |
| Agliamzanov et al. | Hydrology@ Home: a distributed volunteer computing framework for hydrological research and applications | |
| CN105488216A (en) | Recommendation system and method based on implicit feedback collaborative filtering algorithm | |
| CN102866992B (en) | A kind of method and device showing product information in webpage | |
| Zhen et al. | Distributed knowledge sharing for collaborative product development | |
| CN102238045A (en) | System and method for predicting user behavior in wireless Internet | |
| CN101505311B (en) | Information transmission method and system based on socialized network | |
| Dong et al. | Balancing and sequencing of stochastic mixed-model assembly U-lines to minimise the expectation of work overload time | |
| US20120102053A1 (en) | Digital analytics system | |
| CN102609533A (en) | Kernel method-based collaborative filtering recommendation system and method | |
| CN103971170A (en) | Method and device for forecasting changes of feature information | |
| CN107292326A (en) | The training method and device of a kind of model | |
| CN104182457A (en) | Poisson-process-model-based method for predicting event popularity in social network | |
| CN109409738A (en) | Method, the electronic device of deep learning are carried out based on block platform chain | |
| Alexopoulos et al. | DESYMA: assessing flexibility for the lifecycle of manufacturing systems | |
| JP2021517673A (en) | Distributed resource allocation | |
| Yan et al. | Cash transportation vehicle routing and scheduling under stochastic travel times | |
| CN103064778A (en) | Method, device and system of server performances testing | |
| CN103064670A (en) | Method and system for innovation platform data management based on place net | |
| CN104348892A (en) | Information display method and device | |
| CN108133390A (en) | For predicting the method and apparatus of user behavior and computing device | |
| Meng et al. | Batch size modeling in a multi-item, discrete manufacturing system via an open queuing network | |
| Song et al. | A dynamic perspective: local interactions driving the spread of social networks | |
| Fosgerau et al. | Bikeability and the induced demand for cycling | |
| Pınarbaşı et al. | Balancing stochastic type-II assembly lines: chance-constrained mixed integer and constraint programming models |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20130821 Assignee: Branch DNT data Polytron Technologies Inc Assignor: Institute of Computing Technology, Chinese Academy of Sciences Contract record no.: 2018110000033 Denomination of invention: Method, device and system for predicting microblog fashion trend Granted publication date: 20161207 License type: Common License Record date: 20180807 |
|
| EE01 | Entry into force of recordation of patent licensing contract |