CN106919564A - A kind of influence power measure based on mobile subscriber's behavior - Google Patents
A kind of influence power measure based on mobile subscriber's behavior Download PDFInfo
- Publication number
- CN106919564A CN106919564A CN201510981171.1A CN201510981171A CN106919564A CN 106919564 A CN106919564 A CN 106919564A CN 201510981171 A CN201510981171 A CN 201510981171A CN 106919564 A CN106919564 A CN 106919564A
- Authority
- CN
- China
- Prior art keywords
- user
- represent
- influence power
- behavior
- mobile
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mobile Radio Communication Systems (AREA)
- Telephonic Communication Services (AREA)
Abstract
The present invention relates to a kind of influence power measure based on mobile subscriber's behavior, the method detailed process is：Interbehavior according to mobile subscriber builds community network, and the influence power of user itself is calculated using the topological structure of network and mobile subscriber's behavior；According to the similarity between the interbehavior and user preference between user, the influence power between user is calculated；Finally, the influence power of user itself is introduced, influence power between user is merged.According to the influence power measure that the present invention is implemented, the interbehavior and user preference similitude between status, user of the user in community network are considered, the influence power that calculating is waited until all has preferable application value to social network analysis, mobile commending system, customer requirement retrieval etc..
Description
Technical field
The present invention relates to social networks technical field, particularly a kind of influence power measurement side based on mobile subscriber's behavior
Method.
Background technology
The influence relation how excavated from mobile subscriber's behavioral data between user has turned into one in social network analysis
Individual study hotspot.All had very important significance in theoretical or social utility value.Used in mobile community network
The behavior at family can be influenceed by surrounding people, for example household, friend, interest identical other users etc..By analyzing mobile using
Influence relation between the historical behavior data at family, digging user, to social network analysis, personalized recommendation system, user's request
The fields such as acquisition all have preferable application value.
At present, the computational methods of user force are divided three classes：Sum based on network topology structure, based on user behavior
Based on customer interaction information.Computational methods based on network topology structure are main (such as to spend centrad, close centers from node
Degree, betweenness center degree etc.) and carry out measure user influence power at (while betweenness, common neighbours) two aspect.Du et al. considers every kind of
Centrad computational methods have its advantage and disadvantage, therefore calculate user force using various centrads.Hu et al. then draws side
Enter to user force in calculating.The calculating on side is either also based on based on node, they all do not account for the behavior of user
Information, simply the topological structure simply according to network is calculated, therefore the accuracy of the user force for obtaining is not high.
In order to improve the accuracy of the user force being calculated, researcher has introduced user behavior, such as user
Login behavior, the information content (such as comment, forward) that user produces.Hair et al. in microblogging in order to obtain preferably influence
Power, it is contemplated that factor of both social network structure and user behavior.In a mobile network, the base of user can not only be utilized
The number of times of this communication behavior, such as voice communication, duration, short message bar number carry out the influence power of measure user, can be combined with wechat,
Forwarding, comment behavior in the instant communication softwares such as QQ carry out the influence power of measure user.
What the influence power based on network topology structure and based on user behavior measured technique study is user from network
Middle had influence power, does not account for the influence power between user.Influence power measure based on customer interaction information according to
Mutual-action behavior between user, such as forwarding, the quantity of comment information carry out the influence power between measure user.What Anger et al. built
User exchange content and interactive information are considered in regional effection model.And Guo et al. is by analyzing the historical behavior day of user
Will, the influence power between user is calculated using the thought of maximal possibility estimation.In a mobile network can be by (the language that communicated between user
Speech, short message, wechat, QQ) number of times and duration and user preference the similarity influence power of coming between measure user.
But, although the influence power measure based on customer interaction information considers the interbehavior of user, but mostly
Number is applied in wechat or microblogging website, and correlative study in a mobile network is little.In sum, shadow in mobile network
The measure for ringing power is still present very big room for promotion in the degree of accuracy.
The content of the invention
It is an object of the invention to provide a kind of influence power measure based on mobile subscriber's behavior, the method is conducive to
The accuracy that influence power is calculated between user in the mobile community network of raising.
To achieve the above object, the technical scheme is that：A kind of influence power measurement side based on mobile subscriber's behavior
Method, comprises the following steps：
Step A：Mobile subscriber's historical behavior data are read, is constructed with mobile subscriber as node, customer relationship is the shifting on side
Dynamic community network figure；
Step B：Calculate user's itself affect power：Status of the analysis user in mobile community network, calculates user itself
Influence power；
Step C：Calculate influence power between the user based on user mutual behavior：According to interaction duration and interaction times to user
Various interbehaviors quantified, and assign different weighted values to the interbehavior under different contexts；
Step D：Calculate influence power between the user based on user preference similitude：Analysis mobile subscriber's historical behavior, uses
User preference under logarithmic function is constrained context quantifies, and mobile subscriber is calculated using improved Pearson correlation coefficients
The similitude of preference；
Step E：Influence power between fusion user.
Further, in above-mentioned steps A, the building process of mobile community network is as follows：Represented using non-directed graph G (V, E)
Mobile community network, V represents the node set in network, i.e. mobile subscriber's set；E represents the set on side in network.WithTable
Show the user mutual behavior after quantifying, whenDuring more than given threshold value, judge there is relation between user, otherwise judge user
Between it is irrelevant,Computing formula it is as follows：
Wherein, ui∈ U represent that mobile subscriber i, U represent that mobile subscriber gathers；wl∈WijRepresent uiAnd ujThe interaction side for using
Formula, WijRepresent uiAnd ujThe set of the interactive mode for using, Nw=| Wij| represent the quantity of the interactive mode that user uses；Represent uiAnd ujUse wlInteractive duration；Represent the average value of all user mutual durations in network；Represent
uiAnd ujUse wlInteractive number of times；Represent the average value of all user mutual number of times in network；a1And a2Interaction is represented respectively
The weighted value of duration and interaction times, and a1+a2=1.
Further, in above-mentioned steps B, the detailed process for calculating user's itself affect power is：First according to community network
The degree centrad of interior jointThe average interactive quantity of targeted customer and its neighbourAnd use of the user to mobile network service
AmountDetermine status of the user in community network, then calculate the influence power of user itself,Computing formula it is as follows：
Wherein, N=| U | represent the total quantity of nodes；UniRepresent node uiNeighborhood；Ni=| Uni| represent
Node uiNeighbours' quantity；
Computing formula it is as follows：
Computing formula it is as follows：
Wherein, sk∈ S represent that mobile network service k, S represent the set of mobile network service, NS=| S | represents mobile network
Network services total species；Represent uiThe species set of used mobile network service；Represent uiUsed shifting
The quantity of dynamic network service species；Represent uiUse skDuration；Represent uiUse skNumber of times；
The computing formula of user's itself affect power is as follows：
Further, in above-mentioned steps C, context refers to the environmental information residing for user, for example the time, place, activity,
Surrounding people etc., user can select different interactive modes according to surrounding environment, for example, use short message and other use in library
Family interacts, and selects voice and other users interaction, C ' to represent the set of context instance vector, C on playgroundr∈ C ' represent tool
The context instance vector of body, the quantitative formula of the lower user mutual behavior of context constraint is as follows：
Wherein,Represent uiAnd ujIn CrW is used under constraintlInteractive duration；Represent uiAnd ujIn CrAbout
W is used under beamlInteractive number of times；
Influence power computing formula is as follows between the user based on user mutual behavior：
Wherein,Represent CrIn user ujWeighted value in behavior.
Further, in the step D, the order of user preference generation is considered in improved Pearson correlation coefficients, is received
The preference of the targeted customer of other users influence will lag behind the preference of other users.Pearson correlation coefficients after improvement are such as
Under：
Wherein, SI, j, rRepresent uiIn CrEarlier than u under constraintjThe set of the mobile network service for using；Represent user
uiIn CrConstraint under to mobile network service skPreference；Represent user in CrMobile network service is put down under constraint
Equal preference, computing formula is as follows：
Influence power computing formula is as follows between the user based on user preference similitude：
Wherein, SI, rRepresent uiIn CrThe set of the mobile network service used under constraint.
Further, in the step E, by influence power between user's itself affect power, the user based on user mutual behavior and
Influence power is merged between the user based on user preference similitude, determines the weighted value of each influence factor.
Compared to prior art, the beneficial effects of the invention are as follows：Compared to existing influence power measure, for movement
The characteristics of community network, it is contemplated that influence power between user's itself affect power, user, make the influence power being calculated in accuracy
It is greatly enhanced.To sum up, algorithm of the invention can efficiently measure the influence power between user in mobile community network.
Brief description of the drawings
Fig. 1 realizes flow chart for the inventive method.
When the quantization threshold of user mutual behavior is set to different value, the weight value using duration and number of times is Fig. 2
When 0.5 (influence of the duration with number of times to user force is identical), the result that user's itself affect power is obtained, the party are only considered
Method is designated as Method 1.
Fig. 3 takes optimal value when the quantization threshold of user mutual behavior, and interval [0,1] is set to using the weight of duration, with
When step-length is 0.1 value, the result that user's itself affect power is obtained only is considered, the method is designated as Method 2.
Fig. 4 considers comparing result during different factors.Control methods includes：When taking optimal value using duration and number of times weight
Method 2；Influence power (does not consider context, i.e. context weight 1), to be designated as between the user based on user mutual behavior
Method 3；Influence power (consideration context factors), are designated as Method 4 between the user based on user mutual behavior；Based on user
Influence power between the user of preference similitude, is designated as Method 5；The method for merging Method 2 and Method 4, and weight phase
Together, respectively 0.5, it is designated as Method 6；Three kinds of influence powers of fusion, and weight is identical, respectively 1/3, it is designated as Method 7；Melt
Three kinds of influence powers, and weighted are closed, Method 8 is designated as.
Specific embodiment
By embodiment, the present invention is further detailed explanation below in conjunction with the accompanying drawings.
Fig. 1 is that a kind of influence power measure based on mobile subscriber's behavior of the invention realizes flow chart.Such as Fig. 1 institutes
Show, the described method comprises the following steps：
Step A：Mobile subscriber's historical behavior data are read, is constructed with mobile subscriber as node, customer relationship is the shifting on side
Dynamic community network figure.
In mobile community network, using each mobile subscriber an as node in community network, when between user
When interbehavior is more than given threshold value, judge that user has relation, i.e., have a line between users, the interbehavior of user is not
Only include the interbehavior using basic interactive mode (such as voice, short message), also including (QQ, wechat, micro- using interactive software
It is rich) interbehavior.
Specifically, in above-mentioned steps A, mobile community network is represented using non-directed graph G (V, E), V represents the section in network
Point set, i.e. mobile subscriber are gathered；E represents the set on side in network.WithThe user mutual behavior after quantifying is represented, whenDuring more than given threshold value, judge there is relation between user, it is otherwise irrelevant between judgement user,Computing formula
It is as follows：
Specifically, ui∈ u represent that mobile subscriber i, u represent that mobile subscriber gathers；wl∈WijRepresent uiAnd ujThe interaction for using
Mode, WijRepresent uiAnd ujThe set of the interactive mode for using, Nw=| Wij| represent the quantity of the interactive mode that user uses；Represent uiAnd ujUse wlInteractive duration；Represent the average value of all user mutual durations in network；Represent
uiAnd ujUse wlInteractive number of times；Represent the average value of all user mutual number of times in network；a1And a2Interaction is represented respectively
The weighted value of duration and interaction times, and a1+a2=1.
In the present embodiment, the True Data collection issued using rope made of hemp science and engineering builds mobile community network, VthresholdIt is use
The threshold value of interbehavior, V between familythresholdIt is set on [0,0.1] interval with the value of step-length 0.01；a1It is set as in [0,1]
On interval, with the value that step-length is 0.1, according to experimental result, optimal value is selected.
Step B：Calculate user's itself affect power：Status of the analysis user in mobile community network, calculates user itself
Influence power.
Specifically, in above-mentioned steps B, the detailed process for calculating user's itself affect power is：First according to mobile society
The degree centrad of nodesThe average interactive quantity of targeted customer and its neighbourAnd user is to mobile network service
Usage amountDetermine status of the user in community network, then calculate the influence power of user itself,Computing formula
It is as follows：
Specifically, N=| U | represent the total quantity of nodes；UniRepresent node uiNeighborhood；Ni=| Uni| table
Show node uiNeighbours' quantity；
Computing formula it is as follows：
Computing formula it is as follows：
Wherein, sk∈ S represent that mobile network service k, S represent the set of mobile network service, NS=| S | represents mobile network
Network services total species；Represent uiThe species set of used mobile network service；Represent uiUsed shifting
The quantity of dynamic network service species；Represent uiUse skDuration；Represent uiUse skNumber of times；
The computing formula of user's itself affect power is as follows：
Step C：Calculate influence power between the user based on user mutual behavior：According to interaction duration and interaction times to user
Various interbehaviors quantified, and assign different weighted values to the interbehavior under different contexts.
Specifically, in above-mentioned steps C, context refers to the environmental information residing for user, for example the time, place, activity,
Surrounding people etc., user can select different interactive modes according to surrounding environment, for example, use short message and other use in library
Family interacts, and selects voice and other users interaction, C ' to represent the set of context instance vector, C on playgroundr∈ C ' represent tool
The context instance vector of body, the quantitative formula of the lower user mutual behavior of context constraint is as follows：
Specifically,Represent uiAnd ujIn CrW is used under constraintlInteractive duration；Represent uiAnd ujIn Cr
W is used under constraintlInteractive number of times；
Influence power computing formula is as follows between the user based on user mutual behavior：
Specifically,Represent CrIn ujWeighted value in user behavior.
In this example,Value is on interval [0,1], with the value that step-length is 0.1, context to be determined using genetic algorithm
The value of weight.
Step D：Calculate influence power between the user based on user preference similitude：Analysis mobile subscriber's historical behavior, uses
User preference under logarithmic function is constrained context quantifies, and mobile subscriber is calculated using improved Pearson correlation coefficients
The similitude of preference.
Specifically, in above-mentioned steps D, the order of user preference generation is considered in improved Pearson correlation coefficients,
The preference of the targeted customer influenceed by other users will lag behind the preference of other users.Pearson correlation coefficients after improvement are such as
Under：
Specifically,Represent uiIn CrEarlier than u under constraintjThe set of the mobile network service for using；Represent user
uiIn CrConstraint under to mobile network service skPreference；Represent user in CrMobile network service is put down under constraint
Equal preference, computing formula is as follows：
Influence power computing formula is as follows between the user based on user preference similitude：
Specifically, SI, rRepresent uiIn CrThe set of the mobile network service used under constraint.
In this example, logarithm truth of a matter a when calculating user preference is set as on [1,2] interval, is 0.1 with step-length
Value, value of the selection user preference distribution closest to Pareto Law.
Step E：Influence power between fusion user.
Specifically, in above-mentioned steps E, by influence power between user's itself affect power, the user based on user mutual behavior
Influence power is merged and between the user based on user preference similitude, determines the weighted value of each influence factor.
Investigated in the present embodiment and True Data collection is used using the present invention, when the parameter being related to takes different value and consideration
When factor is different, the accuracy of influence power between the user being calculated.The present embodiment is evaluated using root-mean-square error RSME values
The degree of accuracy of influence power result of calculation, RSME is the influence power and true impact power close to journey for judging to be calculated according to the present invention
The judging basis of degree, rule is as follows：
Wherein, rU, iActual user's scoring is represented,User's scoring that prediction is obtained is represented, R is represented and need in test set pre-
User's scoring set of survey.
A kind of influence power measure based on mobile subscriber's behavior of the present invention, the influence power between user was calculated
Journey is divided into the mobile community network of structure, calculates influence power between user's itself affect power, calculating user, influence power four ranks of fusion
Section.First, mobile subscriber's behavioral data is read, is constructed with mobile subscriber as node, customer relationship is mobile society's figure on side.So
Afterwards, the topological structure according to mobile community network determine user degree centrad, calculated according to mobile subscriber's behavior user and its
, to the usage amount of mobile network service, summary three aspect factor determines user in movement for the average interactive quantity of neighbours, user
Status in community network, calculates the influence power of user itself.Secondly, it is contemplated that in mobile network, context is to user behavior
Influence it is more obvious, so quantify mobile subscriber's interbehavior when, it is contemplated that context factors, according to quantized result calculate
Influence power between the user based on user mutual behavior；Quantized contexts constrain usage amount of the lower user to mobile network service, root
Logarithmic function digging user preference is used according to quantized result, and using similar between Pearson correlation coefficients measure user preference
Property, influence power between the user for being based on user preference similitude is calculated on above-mentioned working foundation.Finally, by user's itself affect power
Influence power is merged and between user.
In order to prove the advantage of the inventive method, herein using True Data collection disclosed in rope made of hemp science and engineering (including 94 shifting
Employ behavior and contextual information, friend information of the family in 9 months) tested.Fig. 2 is the amount when user mutual behavior
Change threshold value when being set to different value, (duration and number of times are to user force when the use of the weight value of duration and number of times being 0.5
Influence it is identical), only consider the result that obtains of user's itself affect power.Figure it is seen that working as VthresholdWhen=0.02, meter
The influence power for obtaining can obtain optimal value.Fig. 4 is to work as Vthreshold=0.02, and weight using duration is set to interval
[0,1], when taking step-length as 0.1 value, only considers the result that user's itself affect power is obtained.From figure 3, it can be seen that working as a1=
When 0.03, the influence power being calculated can obtain optimal value, and compared to duration is used, access times are to customer impact for this explanation
Power has prior influence.Fig. 4 is the result obtained using distinct methods.Figure 4, it is seen that when a kind of only influence of consideration
During power, the result that influence power is obtained between the user based on user preference similitude is best, next to that based on user mutual behavior
Influence power between user, and it is worst according to the result that user's itself affect power is obtained；When three kinds of influence powers are merged, can obtain most
Good result, this explanation method proposed by the present invention can improve the accuracy of the influence power being calculated.
The foregoing is only presently preferred embodiments of the present invention, it is all carried out in invention jurisdictions mandate limited range change
Become, change, belong to protection scope of the present invention.
Claims (6)
1. a kind of influence power measure based on mobile subscriber's behavior, it is characterised in that the described method comprises the following steps：
Step A：Mobile subscriber's historical behavior data are read, is constructed with mobile subscriber as node, customer relationship is the mobile society on side
Can network；
Step B：Calculate user's itself affect power：Status of the analysis user in mobile community network, calculates user's itself affect
Power；
Step C：Calculate influence power between the user based on user mutual behavior：According to interaction duration and interaction times to each of user
Plant interbehavior to be quantified, and different weighted values are assigned to the interbehavior under different contexts；
Step D：Calculate influence power between the user based on user preference similitude：Analysis mobile subscriber's historical behavior, uses logarithm
User preference under the constraint of function pair context is quantified, and mobile subscriber's preference is calculated using improved Pearson correlation coefficients
Similitude；
Step E：Influence power between fusion user.
2. a kind of influence power measure based on mobile subscriber's behavior according to claim 1, it is characterised in that：
In above-mentioned steps A, mobile community network is represented using non-directed graph G (V, E), V represents the node set in network, that is, moves
Employ family set；E represents the set on side in network.WithThe user mutual behavior after quantifying is represented, whenMore than given
During threshold value, judge there is relation between user, it is otherwise irrelevant between judgement user,Computing formula it is as follows：
Wherein, ui∈ U represent that mobile subscriber i, U represent that mobile subscriber gathers；wl∈WijRepresent uiAnd ujThe interactive mode for using,
WijRepresent uiAnd ujThe set of the interactive mode for using, Nw=| Wij| represent the quantity of the interactive mode that user uses；Table
Show uiAnd ujUse wlInteractive duration；Represent the average value of all user mutual durations in network；Represent uiAnd ujMake
Use wlInteractive number of times；Represent the average value of all user mutual number of times in network；a1And a2Interaction duration and friendship are represented respectively
The weighted value of mutual number of times, and a1+a2=1.
3. a kind of influence power measure based on mobile subscriber's behavior according to claim 1, it is characterised in that：
In above-mentioned steps B, the detailed process for calculating user's itself affect power is：First according to mobile community network interior joint
Degree centradThe average interactive quantity of targeted customer and its neighbourAnd user is to the usage amount of mobile network service
Determine status of the user in community network, then calculate the influence power of user itself,Computing formula it is as follows：
Wherein, N=| U | represent the total quantity of nodes；UniRepresent node uiNeighborhood；Ni=| Uni| represent node
uiNeighbours' quantity；
Computing formula it is as follows：
Computing formula it is as follows：
Wherein, sk∈ S represent that mobile network service k, S represent the set of mobile network service, NS=| S | represents mobile network's clothes
The total species of business；Represent uiThe species set of used mobile network service；Represent uiUsed mobile network
The quantity of network type service；Represent uiUse skDuration；Represent uiUse skNumber of times；
The computing formula of user's itself affect power is as follows：
4. a kind of influence power measure based on mobile subscriber's behavior according to claim 1, it is characterised in that：
Refer to again up and down environmental information residing for user in above-mentioned steps C, such as time, place, activity, surrounding people etc.,
User can select different interactive modes according to surrounding environment, for example, interacted using short message and other users in library, and
Playground selects voice and other users interaction, and C ' represents the set of context instance vector, Cr∈ C ' represent specific context
Example vector, the quantitative formula of the lower user mutual behavior of context constraint is as follows：
Wherein,Represent uiAnd ujIn CrW is used under constraintlInteractive duration；Represent uiAnd ujIn CrUnder constraint
Use wlInteractive number of times；
Influence power computing formula is as follows between the user based on user mutual behavior：
Wherein,Represent CrIn user ujWeighted value in behavior.
5. a kind of influence power measure based on mobile subscriber's behavior according to claim 1, it is characterised in that：
In the step D, the order of user preference generation is considered in improved Pearson correlation coefficients, influenceed by other users
The preference of targeted customer to lag behind the preference of other users.Pearson correlation coefficients after improvement are as follows：
Wherein,Represent uiIn CrEarlier than u under constraintjThe set of the mobile network service for using；Represent user uiIn Cr
Constraint under to mobile network service skPreference；Represent user under Cr constraints to the average inclined of mobile network service
Good, computing formula is as follows：
Influence power computing formula is as follows between the user based on user preference similitude：
Wherein, SI, rRepresent uiIn CrThe set of the mobile network service used under constraint.
6. a kind of influence power measure based on mobile subscriber's behavior according to claim 1, it is characterised in that：
In the step E, by influence power between user's itself affect power, the user based on user mutual behavior and based on user preference
Influence power is merged between the user of similitude, determines the weighted value of each influence factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510981171.1A CN106919564A (en) | 2015-12-24 | 2015-12-24 | A kind of influence power measure based on mobile subscriber's behavior |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510981171.1A CN106919564A (en) | 2015-12-24 | 2015-12-24 | A kind of influence power measure based on mobile subscriber's behavior |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106919564A true CN106919564A (en) | 2017-07-04 |
Family
ID=59459892
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510981171.1A Pending CN106919564A (en) | 2015-12-24 | 2015-12-24 | A kind of influence power measure based on mobile subscriber's behavior |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106919564A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107577698A (en) * | 2017-07-21 | 2018-01-12 | 天津科技大学 | A kind of mobile subscriber's preference Forecasting Methodology based on influence power between user |
| CN109446171A (en) * | 2017-08-30 | 2019-03-08 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device |
| CN110020154A (en) * | 2017-12-04 | 2019-07-16 | 北京京东尚科信息技术有限公司 | For determining the method and device of user force |
| CN113127696A (en) * | 2021-03-21 | 2021-07-16 | 武汉大学深圳研究院 | Method for improving accuracy of influence measurement based on behaviors |
| CN114173159A (en) * | 2021-11-23 | 2022-03-11 | 武汉市烽视威科技有限公司 | Hot content prediction method, device, equipment and readable storage medium |
| CN114791982A (en) * | 2022-06-24 | 2022-07-26 | 百度在线网络技术（北京）有限公司 | Object recommendation method and device |
| WO2023286019A1 (en) * | 2021-07-16 | 2023-01-19 | Atphizyom Limited | Investment platform |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005015447A1 (en) * | 2003-08-11 | 2005-02-17 | Karl-Heinz Swoboda | Means for simulating the behavior of manipulators that can be freely positioned in a manual manner on influenceable particles having a different momentum |
| US20100121789A1 (en) * | 2008-11-11 | 2010-05-13 | Vladimir Bednyak | Interactive apparatus for assisting in encouraging or deterring of at least one predetermined human behavior |
| CN101883133A (en) * | 2010-04-26 | 2010-11-10 | 李爽 | Accurate influence marketing system based on signalling analysis and method thereof |
| CN103279512A (en) * | 2013-05-17 | 2013-09-04 | 湖州师范学院 | Method for using most influential node in social network to achieve efficient viral marketing |
| CN104537096A (en) * | 2015-01-09 | 2015-04-22 | 哈尔滨工程大学 | Microblog message influence measuring method based on microblog message propagation tree |
-
2015
- 2015-12-24 CN CN201510981171.1A patent/CN106919564A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005015447A1 (en) * | 2003-08-11 | 2005-02-17 | Karl-Heinz Swoboda | Means for simulating the behavior of manipulators that can be freely positioned in a manual manner on influenceable particles having a different momentum |
| US20100121789A1 (en) * | 2008-11-11 | 2010-05-13 | Vladimir Bednyak | Interactive apparatus for assisting in encouraging or deterring of at least one predetermined human behavior |
| CN101883133A (en) * | 2010-04-26 | 2010-11-10 | 李爽 | Accurate influence marketing system based on signalling analysis and method thereof |
| CN103279512A (en) * | 2013-05-17 | 2013-09-04 | 湖州师范学院 | Method for using most influential node in social network to achieve efficient viral marketing |
| CN104537096A (en) * | 2015-01-09 | 2015-04-22 | 哈尔滨工程大学 | Microblog message influence measuring method based on microblog message propagation tree |
Non-Patent Citations (2)
| Title |
|---|
| 史艳翠: "基于通信数据的上下文移动用户偏好动态获取方法研究", 《中国博士学位论文全文数据库 信息利技辑》 * |
| 史艳翠等: "一种上下文移动用户偏好自适应学习方法", 《软件学报》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107577698A (en) * | 2017-07-21 | 2018-01-12 | 天津科技大学 | A kind of mobile subscriber's preference Forecasting Methodology based on influence power between user |
| CN109446171A (en) * | 2017-08-30 | 2019-03-08 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device |
| CN110020154A (en) * | 2017-12-04 | 2019-07-16 | 北京京东尚科信息技术有限公司 | For determining the method and device of user force |
| CN113127696A (en) * | 2021-03-21 | 2021-07-16 | 武汉大学深圳研究院 | Method for improving accuracy of influence measurement based on behaviors |
| WO2023286019A1 (en) * | 2021-07-16 | 2023-01-19 | Atphizyom Limited | Investment platform |
| CN114173159A (en) * | 2021-11-23 | 2022-03-11 | 武汉市烽视威科技有限公司 | Hot content prediction method, device, equipment and readable storage medium |
| CN114791982A (en) * | 2022-06-24 | 2022-07-26 | 百度在线网络技术（北京）有限公司 | Object recommendation method and device |
| CN114791982B (en) * | 2022-06-24 | 2022-10-14 | 百度在线网络技术（北京）有限公司 | Object recommendation method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106919564A (en) | A kind of influence power measure based on mobile subscriber's behavior | |
| Zhan et al. | A learning-based incentive mechanism for federated learning | |
| Liu et al. | Energy-aware participant selection for smartphone-enabled mobile crowd sensing | |
| CN105207821B (en) | A kind of network synthesis performance estimating method of service-oriented | |
| CN103257921B (en) | Improved random forest algorithm based system and method for software fault prediction | |
| CN110417607B (en) | Flow prediction method, device and equipment | |
| CN101572623A (en) | Method for comprehensively evaluating network performance based on subjective and objective combination evaluation | |
| CN106875281B (en) | Social network node mining activation method based on greedy subgraph | |
| CN105809554A (en) | Prediction method of hot topics participated by users in social networks | |
| TWI505667B (en) | Method and system for carrying out predictive analysis relating to nodes of a communication network | |
| CN108280121B (en) | Method for obtaining social network opinion leader based on K-kernel decomposition | |
| CN103136331A (en) | Micro blog network opinion leader identification method | |
| CN103077247A (en) | Method for building friend relationship transitive tree in social network | |
| CN105678590A (en) | topN recommendation method for social network based on cloud model | |
| CN110297990A (en) | The associated detecting method and system of crowdsourcing marketing microblogging and waterborne troops | |
| CN104035987A (en) | Method for ranking microblog network user influence | |
| CN108229731A (en) | The user's behavior prediction system and method that more message mutually influence under a kind of much-talked-about topic | |
| Bo et al. | Establishment and application of fuzzy comprehensive evaluation of green building design based on data mining | |
| CN110009416A (en) | A kind of system based on big data cleaning and AI precision marketing | |
| CN105391490B (en) | A kind of satellite communication network selection algorithm based on cognition | |
| Postigo-Boix et al. | A social model based on customers’ profiles for analyzing the churning process in the mobile market of data plans | |
| CN107609717A (en) | A kind of model transfer amount Forecasting Methodology based on Kalman filter in social networks | |
| Zhang et al. | A dynamic web service composition algorithm based on TOPSIS | |
| Ye et al. | Bridging the micro and macro: Calibration of agent-based model using mean-field dynamics | |
| CN104579850A (en) | Quality of service (QoS) prediction method for Web service under mobile Internet environment |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170704 |