CN107230158A - Social network user relative influence measure - Google Patents
Social network user relative influence measure Download PDFInfo
- Publication number
- CN107230158A CN107230158A CN201710439921.1A CN201710439921A CN107230158A CN 107230158 A CN107230158 A CN 107230158A CN 201710439921 A CN201710439921 A CN 201710439921A CN 107230158 A CN107230158 A CN 107230158A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- user
- field direction
- represent
- 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
- 230000008901 benefit Effects 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000005259 measurement Methods 0.000 claims description 8
- 238000000034 method Methods 0.000 claims description 7
- 239000012141 concentrate Substances 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 12
- 230000009467 reduction Effects 0.000 abstract description 3
- 238000004364 calculation method Methods 0.000 description 11
- 230000009471 action Effects 0.000 description 9
- 238000004891 communication Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 235000010627 Phaseolus vulgaris Nutrition 0.000 description 1
- 244000046052 Phaseolus vulgaris Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003012 network analysis Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000007480 spreading Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of measure of social network user relative influence, including：1 builds network user's relative effect Measure Indexes；2 calculate each index weights；3 build user version characteristic vector and each field direction Text eigenvector, calculate user and the degree of correlation in each field direction；4, from itself relative and global relative two angle, build network user's relative influence measure model, calculate the relative influence of user.The present invention can define the territory of network user's influence power effect, influence power size of the measure user based on different field direction, so that the advantage function field direction of user force is recognized, and then the reduction network user selects difficulty, lifts Decision Quality.
Description
Technical field
The present invention relates to a kind of network influence measure, specifically a kind of measurement social network user is with respect to shadow
Ring the method for power.
Technical background
Web2.0 epoch, UGC turns into the major channel that Web content is produced, and increasing user passes through internet active
Ground retrieval, issue and propagation information, accumulate own net speech effect consciously.Substantial amounts of achievement in research shows, in society
Media platform can be changed, the influential network user is most important in terms of Information Communication or information guidance.Mutually
The opening and complex of networking mean the otherness of the network user and the diversity of Web content.Heterogeneous networks user Yin Wen
Change the difference in terms of background, Knowledge Construction, experience experience, social relationships and social activities ability, its field paid close attention to and words
Topic is also different, causes the sphere of action and action intensity of its network influence also to vary.
The research measured and recognized on network user's influence power at present, is focused primarily upon：(1) it is deep from physics angle
Enter to study network structure, link level and the social networks of user, using the method for social network analysis, describe its physical topology
Relation between structure and node, to calculate the network influence for obtaining user；(2) from Information Communication angle, deeply grind
Study carefully the information housing choice behavior and concern relation to each other of the network user, the net for obtaining user is calculated with PageRank algorithms
Network influence power；(3) from dissemination angle, the propagation characteristic and coverage of network user's influence power is analyzed, web influence is set up
Power probability of spreading model, to obtain the network influence of user；(4) from Information Management angle, according to the information between user
Issue and housing choice behavior, portray its people and social attributive character, build user force multidimensional information entropy measure model, with
Calculate the network influence for obtaining user.
It is measure user mostly although existing study the measure that have studied network user's influence power with different view
The overall size of influence power, relative sex expression of the fresh user force of analysis less in sphere of action and action intensity, it is impossible to complete
The value utilitarian of user force is assessed to face, the decision-making difficulty for selecting the network user is added.
The content of the invention
There is provided a kind of social network user relative influence measurement side to overcome the shortcomings of prior art presence by the present invention
Method, to define the territory of network user's influence power effect, influence power of the measure user based on different field direction is big
It is small, so that the advantage function field direction of user force is recognized, and then the reduction network user selects difficulty, lifts decision-making matter
Amount.
In order to achieve the above object, the present invention is adopted the following technical scheme that：
A kind of the characteristics of measure of social network user relative influence of the present invention is to carry out as follows：
Step 1, the degree of correlation based on personal attribute's feature, social attributive character and with each field direction, build society
The network user is handed over to concentrate any user UiRelative influence measurement index collection be { Pi,Si,Ri, wherein, PiRepresent user Ui's
Personal attribute, and have Pi={ pi1,pi2,…,pix,…,pih, pixRepresent user UiXth personal attribute's index, h is individual
The quantity of ATTRIBUTE INDEX, and x=1,2 ..., h；SiRepresent user UiSocial attribute, and have Si={ si1,si2,…,
siy,…,sig, siyRepresent user UiY social ATTRIBUTE INDEXs, g is the quantity of social ATTRIBUTE INDEX, and y=
1,2,…,g；RiRepresent user UiWith each field directional correlations set, and there is Ri={ ri1,ri2,…,rik,…,rim, rikTable
Show user UiWith k-th of field direction f in the F of fieldkThe degree of correlation, m is the quantity in field direction, and k=1,2 ..., m；
Step 2, with subjective and objective combination weights method, determine user U in the social networksiPersonal attribute PiAnd society
Meeting property attribute SiWeight and its corresponding index weight：
Step 2.1, the significance level progress experience using expert graded respectively to personal attribute and social attribute are beaten
Point, obtain the average of all expert estimations and be normalized, obtain the weight α of personal attribute and the power of social attribute
Weight β；
Step 2.2, the personal attribute's index and social ATTRIBUTE INDEX collected based on the user, set up initial matrix D, root
According to the positive and negative of indices benefit, initial matrix D is carried out nondimensionalization processing obtain dimensionless matrix D ', utilize entropy assessment pair
The dimensionless matrix D ' calculated, obtain personal attribute's index item weight setAnd society
Property ATTRIBUTE INDEX set
Step 3, structure user UiText eigenvector VUiWith each field direction Text eigenvector Vfk, calculate user
UiWith the degree of correlation in each field direction：
Step 3.1, acquisition k-th of field direction fkUnder all users issue text, utilize participle toolmark and system
Keyword is counted, k-th of field direction f is obtainedkFeature lexical item set Tk={ tk1,tk2,…,tka,…,tkA, wherein, tkaTable
Show k-th of field direction fkA-th of feature lexical item, A represents k-th of field direction fkFeature lexical item quantity；So as to obtain m
Feature lexical item set { the T in individual field direction1,T2,…,Tk,…,Tm}；
Step 3.2, the feature lexical item set { T to m field direction1,T2,…,Tk,…,TmIn feature lexical item duplicate removal,
Construct the feature lexical item set T={ t of the field F1,t2,…,tp,…tn, wherein tpP-th of feature lexical item is represented, then is tieed up with n
K-th of field of vector representation direction fkWith user UiText eigenvector, is designated asWithWherein,Represent p-th of feature lexical item tpK-th of field side
To fkIn weight,Represent p-th of feature lexical item tpIn user UiWeight in text；
Step 3.3, formula (1) and formula (2) is utilized to calculate p-th of feature lexical item tpIn k-th of field direction fkIn weightWith p-th of feature lexical item tpIn user UiWeight in text
In formula (1), NfkRepresent k-th of field direction fkUnder user version quantity, djRepresent k-th of field direction fkUnder
J-th of user version, TF-IDF (tp,dj) represent to calculate p-th obtained of feature lexical item t using TF-IDF formulapIn jth
Individual user version djIn weight,Represent p-th of feature lexical item tpOverall significance level in the field F, wherein,Represent p-th of feature lexical item tpAppear in the feature lexical item set { T in m field direction1,T2,…,Tk,…,TmIn time
Number,
In formula (2), NUiRepresent user UiThe textual data of issue, dlFor user UiL-th of text of issue, TF-IDF (tp,
dl) it is to calculate to obtain p-th of feature lexical item t using TF-IDF formulapIn l-th of text dlIn weight；
Step 3.4, formula (3) is utilized to calculate user UiWith k-th of field direction fkDegree of correlation rik, so as to obtain user Ui
With the degree of correlation set R in m field directioni：
User U in step 4, the structure social networksiRelative influence measure model, calculate user UiRelative shadow
Ring power：
Step 4.1, formula (4) and formula (5) is utilized to calculate user UiPersonal attribute value IP (Ui) and social property value IS
(Ui)：
Step 4.2, formula (6) is utilized to build user UiIndividual influence total amount measure model：
C(Ui)=IP (Ui)+IS(Ui) (6)
In formula (6), C (Ui) represent user UiIndividual influence total amount size；
Step 4.3, formula (7) is utilized to build user UiBased on k-th of field direction fkInfluence power measure model：
In formula (7),Represent user UiBased on k-th of field direction fkInfluence power size；
Step 4.4, using formula (8) and formula (9) user U is calculated respectivelyiRelative influence：
In formula (8), RCU(i)Represent user UiThe relative size of influence power total amount,Represent user's collection influence power total amount
Average；
In formula (9), RCU(ik)Represent user UiIn k-th of field direction fkUnder relative influence,Represent user's collection
Based on k-th of field direction fkInfluence power average.
Compared to prior art, beneficial effects of the present invention are embodied in：
1st, the degree of correlation of the invention by analyzing network user's content of text and each field direction, defines user's concern or relates to
And field direction, and combine social network user in terms of personal attribute and social attribute feature performance, from user from
Body and global two angles of network, devise the measurement model of network user's relative influence, have both considered user's entire effect
Power size, also apparent influence power size of the user under different field direction, improves the applicability of network user's influence power,
Reduce influence power user and select difficulty.
2nd, the feature lexical item set of the invention based on each field direction, builds the feature lexical item vector for characterizing overall field,
The Text eigenvector in user version characteristic vector and each field direction is built by the vector, it is ensured that while the degree of accuracy,
Also reduce operation complexity.
3rd, the present invention considers significance level of the feature lexical item in a certain field direction and overall field, obtains feature
Weight of the lexical item in the Text eigenvector of each field direction, has highlighted the relevance between each field direction, has improved weight meter
Calculate exactness accurately.
4th, network user's relative influence measurement model that the present invention is built, can be relative from itself and global relative two
Angle, assesses the overall relative influence of user and its relative influence under different field direction, overcomes the single degree of tradition
The limitation of influence power size is measured, the applicability and practicality of network user's influence power is greatly improved, reduces and select net
The decision-making difficulty of network user.
Brief description of the drawings
Fig. 1 is the flow chart of measure of the present invention；
Fig. 2 is user force total amount size distribution effect curve；
Fig. 3 is influence power distributed effect figure of the user based on different field direction.
Embodiment
In the present embodiment, as shown in figure 1, a kind of measure of social network user relative influence, can effectively be defined
The territory of network user's influence power effect, and influence power size of the measure user in different range, overcame in the past only
Consider the limitation of user force total amount, help to recognize the advantage areas direction of the network user, the reduction network user's selects
Difficulty, available for fields such as social network advertisement putting, community management, public sentiment monitorings.Specifically, this method is by as follows
Step is carried out：
Step 1, the degree of correlation based on personal attribute's feature, social attributive character and with each field direction, build society
The network user is handed over to concentrate any user UiRelative influence measurement index collection be { Pi,Si,Ri, wherein, PiRepresent user Ui's
Personal attribute, and have Pi={ pi1,pi2,…,pix,…,pih, pixRepresent user UiXth personal attribute's index, h is individual
The quantity of ATTRIBUTE INDEX, and x=1,2 ..., h；SiRepresent user UiSocial attribute, and have Si={ si1,si2,…,
siy,…,sig, siyRepresent user UiY social ATTRIBUTE INDEXs, g is the quantity of social ATTRIBUTE INDEX, and y=
1,2,…,g；RiRepresent user UiWith each field directional correlations set, and there is Ri={ ri1,ri2,…,rik,…,rim, rikTable
Show user UiWith k-th of field direction f in the F of fieldkThe degree of correlation, m is the quantity in field direction, and k=1,2 ..., m；
" family of automobile " forum is using brand community as organizational form, and boundary line between various brands community is than more visible, this hair
The family Jeep brand communities data of bright use automobile, Jeep brand communities are considered as in the field F of user's topic focusing, community
Product forum is considered as each field direction fk.By web crawlers, the institute that user issues in statistical time range Nei Ge products forum is obtained
There are text, individual subscriber ATTRIBUTE INDEX data, the social ATTRIBUTE INDEX data of user and the text of user's issue.
In the present embodiment, user U is builtiRelative influence measurement index collection be { Pi,Si,Ri, wherein, PiRepresent user
UiPersonal attribute, Pi={ pi1,pi2,…,pix,…,pi5, pi1Represent network hierarchy of the user in Jeep brand communities；pi2
Authenticating user identification state is represented, by being designated as 1, not by being designated as 0；pi3Represent the bean vermicelli number of user；pi4Represent dispatch number
Amount, refers to the amount of text that user issues in statistical time range；pi5High-quality text quantity is represented, refers to user in statistical time range and issues
High-quality text quantity, this example refer to user issue essence note quantity；Si={ si1,si2,si3, SiRepresent user UiSociety
Property attribute, si1The average click volume of text is represented, refers to total number of clicks/dispatch quantity that user in statistical time range issues text；si2Table
Show the average reply volume of text, refer to reply sum/dispatch quantity that user in statistical time range issues text；si3Represent high ranked user
Number is replied, the high ranked user for referring to user's issue text in statistical time range replys sum；Ri={ ri1,ri2,ri3,ri4, ri1,
ri2,ri3,ri4User and herdsman's product, guide person product, big Cherokee product, freedom under Jeep brand communities are represented respectively
The degree of correlation of light product.
Step 2, with subjective and objective combination weights method, determine user U in the social networksiPersonal attribute PiAnd society
Meeting property attribute SiWeight and its corresponding index weight：
Step 2.1, the significance level progress experience using expert graded respectively to personal attribute and social attribute are beaten
Point, obtain the average of all expert estimations and be normalized, obtain the weight α of personal attribute and the power of social attribute
Weight β；
Step 2.2, the personal attribute's index and social ATTRIBUTE INDEX collected based on the user, set up initial matrix D, root
According to the positive and negative of indices benefit, initial matrix D is carried out nondimensionalization processing obtain dimensionless matrix D ', utilize entropy assessment pair
The dimensionless matrix D ' calculated, obtain personal attribute's index item weight setAnd society
Property ATTRIBUTE INDEX set
In the present embodiment, each index weights result of calculation is shown in Table 1.
Each index weights result of calculation of table 1
Step 3, structure user UiText eigenvector VUiWith each field direction Text eigenvector Vfk, calculate user
UiWith the degree of correlation in each field direction：
Step 3.1, acquisition k-th of field direction fkUnder all users issue text, utilize participle toolmark and system
Keyword is counted, k-th of field direction f is obtainedkFeature lexical item set Tk={ tk1,tk2,…,tka,…,tkA, wherein, tkaTable
Show k-th of field direction fkA-th of feature lexical item, A represents k-th of field direction fkFeature lexical item quantity；So as to obtain m
Feature lexical item set { the T in individual field direction1,T2,…,Tk,…,Tm}；
Step 3.2, the feature lexical item set { T to m field direction1,T2,…,Tk,…,TmIn feature lexical item go
Weight, constructs institute field F feature lexical item set T={ t1,t2,…,tp,…tn, wherein tpP-th of feature lexical item is represented, then uses n
Dimensional vector represents k-th of field direction fkWith user UiText eigenvector, is designated as
WithWherein,Represent p-th of feature lexical item tpIn k-th of field
Direction fkIn weight,Represent p-th of feature lexical item tpIn user UiWeight in text；
Step 3.3, formula (1) and formula (2) is utilized to calculate p-th of feature lexical item tpIn k-th of field direction fkIn weightWith p-th of feature lexical item tpIn user UiWeight in text
In formula (1), NfkRepresent k-th of field direction fkUnder user version quantity, djRepresent k-th of field direction fkUnder
J-th of user version, TF-IDF (tp,dj) represent to calculate p-th obtained of feature lexical item t using TF-IDF formulapIn jth
Individual user version djIn weight,Represent p-th of feature lexical item tpOverall significance level in the field F, wherein,Represent p-th of feature lexical item tpAppear in the feature lexical item set { T in m field direction1,T2,…,Tk,…,TmIn time
Number,
In formula (2), NUiRepresent user UiThe textual data of issue, dlFor user UiL-th of text of issue, TF-IDF (tp,
dl) it is to calculate to obtain p-th of feature lexical item t using TF-IDF formulapIn l-th of text dlIn weight；
Step 3.4, formula (3) is utilized to calculate user UiWith k-th of field direction fkDegree of correlation rik, so as to obtain user Ui
With the degree of correlation set R in m field directioni：
In the present embodiment, user Ui2 are shown in Table with each field directional correlations result of calculation.
The user U of table 2iWith each field directional correlations result of calculation
User U in step 4, the structure social networksiRelative influence measure model, calculate user UiRelative shadow
Ring power：
Step 4.1, formula (4) and formula (5) is utilized to calculate user UiPersonal attribute value IP (Ui) and social property value IS
(Ui)：
Step 4.2, formula (6) is utilized to build user UiIndividual influence total amount measure model：
C(Ui)=IP (Ui)+IS(Ui) (6)
In formula (6), C (Ui) represent user UiIndividual influence total amount size；
In the present embodiment, the influence power total amount result of calculation of each user is shown in Table 3.
The user U of table 3iInfluence power total amount result of calculation
Each user force total amount size distribution curve as shown in Fig. 2 different user influence power total amount has relativity, this
Outside, even if the influence power total amount of two users is in the same size, its personal attribute's value and social property value also likely to be present difference,
Such as user U7And U9。
Step 4.3, formula (7) is utilized to build user UiBased on k-th of field direction fkInfluence power measure model：
In formula (7),Represent user UiBased on k-th of field direction fkInfluence power size；
In the present embodiment, influence power result of calculation of the user based on each field direction is shown in Table 4.
The user U of table 4iInfluence power result of calculation based on each field direction
Influence power distributed effect of the different user based on each field direction is as shown in figure 3, result shows its shadow of different user
Ringing the sphere of action and action intensity of power has larger difference, user U10Four fields direction is all related to, and in big Cherokee
Middle influence power is maximum；And user U2、U7Field direction is only related to, and influence power effect field direction is inconsistent；U7And U9Shadow
Sound power total amount size is basically identical, but the sphere of action and action intensity of both influence powers are also far from each other.Meanwhile, influence power is total
The high user of amount, specific on a certain field direction, influence power action effect may not be optimal, by taking herdsman as an example, customer impact
Power total amount, descending is U successively2>U1>U9>U10, and under herdsman direction, the influence power of user is descending to be followed successively by：
U1>U2>U10>U9。
Step 4.4, using formula (8) and formula (9) user U is calculated respectivelyiRelative influence：
In formula (8), RCU(i)Represent user UiThe relative size of influence power total amount,Represent user's collection influence power total amount
Average；
In formula (9), RCU(ik)Represent user UiIn k-th of field direction fkUnder relative influence,Represent user's collection
Based on k-th of field direction fkInfluence power average.
In the present embodiment, user UiOverall relative influence and relative influence result of calculation based on each field direction are shown in
Table 5 and table 6.
The user U of table 5iOverall relative influence result of calculation
The user U of table 6iRelative influence result of calculation based on each field direction
Above is presently preferred embodiments of the present invention, all changes made according to technical solution of the present invention, produced function is made
During with scope without departing from technical solution of the present invention, protection scope of the present invention is belonged to.
Claims (1)
1. a kind of measure of social network user relative influence, it is characterized in that carrying out as follows：
Step 1, the degree of correlation based on personal attribute's feature, social attributive character and with each field direction, build social network
Network user concentrates any user UiRelative influence measurement index collection be { Pi,Si,Ri, wherein, PiRepresent user UiIndividual
Attribute, and have Pi={ pi1,pi2,…,pix,…,pih, pixRepresent user UiXth personal attribute's index, h is personal attribute
The quantity of index item, and x=1,2 ..., h；SiRepresent user UiSocial attribute, and have Si={ si1,si2,…,siy,…,
sig, siyRepresent user UiY social ATTRIBUTE INDEXs, g is the quantity of social ATTRIBUTE INDEX, and y=1,2 ...,
g；RiRepresent user UiWith each field directional correlations set, and there is Ri={ ri1,ri2,…,rik,…,rim, rikRepresent user Ui
With k-th of field direction f in the F of fieldkThe degree of correlation, m is the quantity in field direction, and k=1,2 ..., m；
Step 2, with subjective and objective combination weights method, determine user U in the social networksiPersonal attribute PiWith it is social
Attribute SiWeight and its corresponding index weight：
Step 2.1, the significance level using expert graded respectively to personal attribute and social attribute carry out experience marking, obtain
To all expert estimations average and be normalized, obtain personal attribute weight α and social attribute weight beta；
Step 2.2, the personal attribute's index and social ATTRIBUTE INDEX collected based on the user, set up initial matrix D, according to each
Index benefit it is positive and negative, initial matrix D is carried out nondimensionalization processing obtain dimensionless matrix D ', using entropy assessment to described
Dimensionless matrix D ' calculated, obtain personal attribute's index item weight setWith social category
Property index item set
Step 3, structure user UiText eigenvector VUiWith each field direction Text eigenvector Vfk, calculate user UiWith it is each
The degree of correlation in field direction：
Step 3.1, acquisition k-th of field direction fkUnder all users issue text, utilize participle toolmark and statistics crucial
Word, obtains k-th of field direction fkFeature lexical item set Tk={ tk1,tk2,…,tka,…,tkA, wherein, tkaRepresent k-th
Field direction fkA-th of feature lexical item, A represents k-th of field direction fkFeature lexical item quantity；So as to obtain m field side
To feature lexical item set { T1,T2,…,Tk,…,Tm}；
Step 3.2, the feature lexical item set { T to m field direction1,T2,…,Tk,…,TmIn feature lexical item duplicate removal, construction
The feature lexical item set T={ t of the field F1,t2,…,tp,…tn, wherein tpRepresent p-th of feature lexical item, then tieed up with n to
Amount represents k-th of field direction fkWith user UiText eigenvector, is designated as
WithWherein,Represent p-th of feature lexical item tpIn k-th of field
Direction fkIn weight,Represent p-th of feature lexical item tpIn user UiWeight in text；
Step 3.3, formula (1) and formula (2) is utilized to calculate p-th of feature lexical item tpIn k-th of field direction fkIn weightWith
P-th of feature lexical item tpIn user UiWeight in text
<mrow>
<msubsup>
<mi>wf</mi>
<mi>k</mi>
<msub>
<mi>t</mi>
<mi>p</mi>
</msub>
</msubsup>
<mo>=</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<msub>
<mi>Nf</mi>
<mi>k</mi>
</msub>
</mrow>
</munderover>
<mrow>
<mo>(</mo>
<mi>T</mi>
<mi>F</mi>
<mo>-</mo>
<mi>I</mi>
<mi>D</mi>
<mi>F</mi>
<mo>(</mo>
<mrow>
<msub>
<mi>t</mi>
<mi>p</mi>
</msub>
<mo>,</mo>
<msub>
<mi>d</mi>
<mi>j</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>Nf</mi>
<mi>k</mi>
</msub>
</mrow>
</mfrac>
<mo>&times;</mo>
<mrow>
<mo>(</mo>
<mfrac>
<msub>
<mi>M</mi>
<msub>
<mi>t</mi>
<mi>p</mi>
</msub>
</msub>
<mi>m</mi>
</mfrac>
<mo>+</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msubsup>
<mi>wU</mi>
<mi>i</mi>
<msub>
<mi>t</mi>
<mi>p</mi>
</msub>
</msubsup>
<mo>=</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<msub>
<mi>NU</mi>
<mi>i</mi>
</msub>
</mrow>
</munderover>
<mrow>
<mo>(</mo>
<mi>T</mi>
<mi>F</mi>
<mo>-</mo>
<mi>I</mi>
<mi>D</mi>
<mi>F</mi>
<mo>(</mo>
<mrow>
<msub>
<mi>t</mi>
<mi>p</mi>
</msub>
<mo>,</mo>
<msub>
<mi>d</mi>
<mi>l</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>NU</mi>
<mi>i</mi>
</msub>
</mrow>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (1), NfkRepresent k-th of field direction fkUnder user version quantity, djRepresent k-th of field direction fkUnder jth
Individual user version, TF-IDF (tp,dj) represent to calculate p-th obtained of feature lexical item t using TF-IDF formulapIn j-th of user
Text djIn weight,Represent p-th of feature lexical item tpOverall significance level in the field F, wherein,Represent
P-th of feature lexical item tpAppear in the feature lexical item set { T in m field direction1,T2,…,Tk,…,TmIn number of times,
In formula (2), NUiRepresent user UiThe textual data of issue, dlFor user UiL-th of text of issue, TF-IDF (tp,dl) be
Calculated using TF-IDF formula and obtain p-th of feature lexical item tpIn l-th of text dlIn weight；
Step 3.4, formula (3) is utilized to calculate user UiWith k-th of field direction fkDegree of correlation rik, so as to obtain user UiWith m
The degree of correlation set R in field directioni：
<mrow>
<msub>
<mi>r</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>Vf</mi>
<mi>k</mi>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>VU</mi>
<mi>i</mi>
</msub>
</mrow>
<mrow>
<mo>|</mo>
<mo>|</mo>
<mi>V</mi>
<mi>f</mi>
<mo>|</mo>
<mo>|</mo>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>VU</mi>
<mi>i</mi>
</msub>
<mo>|</mo>
<mo>|</mo>
</mrow>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>3</mn>
<mo>)</mo>
</mrow>
</mrow>
User U in step 4, the structure social networksiRelative influence measure model, calculate user UiRelative influence：
Step 4.1, formula (4) and formula (5) is utilized to calculate user UiPersonal attribute value IP (Ui) and social property value IS (Ui)：
<mrow>
<mi>I</mi>
<mi>P</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>U</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>&alpha;</mi>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>x</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>h</mi>
</munderover>
<msub>
<mi>&omega;</mi>
<mi>x</mi>
</msub>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mi>x</mi>
</mrow>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>I</mi>
<mi>S</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>U</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>&beta;</mi>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>y</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</munderover>
<msub>
<mi>&omega;</mi>
<mi>y</mi>
</msub>
<msub>
<mi>s</mi>
<mrow>
<mi>i</mi>
<mi>y</mi>
</mrow>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 4.2, formula (6) is utilized to build user UiIndividual influence total amount measure model：
C(Ui)=IP (Ui)+IS(Ui) (6)
In formula (6), C (Ui) represent user UiIndividual influence total amount size；
Step 4.3, formula (7) is utilized to build user UiBased on k-th of field direction fkInfluence power measure model：
<mrow>
<msub>
<mi>F</mi>
<mi>k</mi>
</msub>
<msub>
<mi>C</mi>
<msub>
<mi>U</mi>
<mi>i</mi>
</msub>
</msub>
<mo>=</mo>
<mi>C</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>U</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>&times;</mo>
<msub>
<mi>r</mi>
<mi>ik</mi>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (7),Represent user UiBased on k-th of field direction fkInfluence power size；
Step 4.4, using formula (8) and formula (9) user U is calculated respectivelyiRelative influence：
<mrow>
<msub>
<mi>RC</mi>
<mrow>
<mi>U</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mi>C</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>U</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mover>
<mrow>
<mi>C</mi>
<mrow>
<mo>(</mo>
<mi>U</mi>
<mo>)</mo>
</mrow>
</mrow>
<mo>&OverBar;</mo>
</mover>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>8</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>RC</mi>
<mrow>
<mi>U</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mi>k</mi>
<mo>)</mo>
</mrow>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>F</mi>
<mi>k</mi>
</msub>
<msub>
<mi>C</mi>
<msub>
<mi>U</mi>
<mi>i</mi>
</msub>
</msub>
</mrow>
<mover>
<mrow>
<msub>
<mi>F</mi>
<mi>k</mi>
</msub>
<msub>
<mi>C</mi>
<mi>U</mi>
</msub>
</mrow>
<mo>&OverBar;</mo>
</mover>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>9</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (8), RCU(i)Represent user UiThe relative size of influence power total amount,Represent the equal of user's collection influence power total amount
Value；
In formula (9), RCU(ik)Represent user UiIn k-th of field direction fkUnder relative influence,Represent that user's collection is based on
K-th of field direction fkInfluence power average.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710439921.1A CN107230158A (en) | 2017-06-12 | 2017-06-12 | Social network user relative influence measure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710439921.1A CN107230158A (en) | 2017-06-12 | 2017-06-12 | Social network user relative influence measure |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107230158A true CN107230158A (en) | 2017-10-03 |
Family
ID=59935407
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710439921.1A Pending CN107230158A (en) | 2017-06-12 | 2017-06-12 | Social network user relative influence measure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107230158A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109451332A (en) * | 2018-11-16 | 2019-03-08 | 武汉斗鱼网络科技有限公司 | A kind of user property labeling method, device, computer equipment and medium |
| CN110941598A (en) * | 2019-12-02 | 2020-03-31 | 北京锐安科技有限公司 | Data deduplication method, device, terminal and storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102314489A (en) * | 2011-08-15 | 2012-01-11 | 哈尔滨工业大学 | Method for analyzing opinion leader in network forum |
| CN103309962A (en) * | 2013-05-31 | 2013-09-18 | 华东师范大学 | Microblog service expert positioning method based on content relevance and social contact influence |
| CN103309990A (en) * | 2013-06-18 | 2013-09-18 | 上海晶樵网络信息技术有限公司 | User multidimensional analysis and monitoring method based on public information of Internet user |
| CN104035968A (en) * | 2014-05-20 | 2014-09-10 | 微梦创科网络科技（中国）有限公司 | Method and device for constructing training corpus set based on social network |
| CN104035967A (en) * | 2014-05-20 | 2014-09-10 | 微梦创科网络科技（中国）有限公司 | Method and system for finding domain expert in social network |
| US20150262199A1 (en) * | 2014-03-11 | 2015-09-17 | Marketo, Inc. | Cross-domain identity service |
| CN105913159A (en) * | 2016-04-29 | 2016-08-31 | 北京邮电大学 | Social network event based user's influence prediction method |
| CN106372117A (en) * | 2016-08-23 | 2017-02-01 | 电子科技大学 | Word co-occurrence-based text classification method and apparatus |
-
2017
- 2017-06-12 CN CN201710439921.1A patent/CN107230158A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102314489A (en) * | 2011-08-15 | 2012-01-11 | 哈尔滨工业大学 | Method for analyzing opinion leader in network forum |
| CN103309962A (en) * | 2013-05-31 | 2013-09-18 | 华东师范大学 | Microblog service expert positioning method based on content relevance and social contact influence |
| CN103309990A (en) * | 2013-06-18 | 2013-09-18 | 上海晶樵网络信息技术有限公司 | User multidimensional analysis and monitoring method based on public information of Internet user |
| US20150262199A1 (en) * | 2014-03-11 | 2015-09-17 | Marketo, Inc. | Cross-domain identity service |
| CN104035968A (en) * | 2014-05-20 | 2014-09-10 | 微梦创科网络科技（中国）有限公司 | Method and device for constructing training corpus set based on social network |
| CN104035967A (en) * | 2014-05-20 | 2014-09-10 | 微梦创科网络科技（中国）有限公司 | Method and system for finding domain expert in social network |
| CN105913159A (en) * | 2016-04-29 | 2016-08-31 | 北京邮电大学 | Social network event based user's influence prediction method |
| CN106372117A (en) * | 2016-08-23 | 2017-02-01 | 电子科技大学 | Word co-occurrence-based text classification method and apparatus |
Non-Patent Citations (3)
| Title |
|---|
| WUHUI CHEN等: "Awareness of social influence on linked social service", 《 2013 IEEE INTERNATIONAL CONFERENCE ON CYBERNETICS (CYBCO)》 * |
| 仇呈燕: "在线社交网络用户影响力量化研究", 《中国优秀硕士学位论文全文数据库》 * |
| 邹雪晴: "特定话题域的微博用户影响力研究与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109451332A (en) * | 2018-11-16 | 2019-03-08 | 武汉斗鱼网络科技有限公司 | A kind of user property labeling method, device, computer equipment and medium |
| CN110941598A (en) * | 2019-12-02 | 2020-03-31 | 北京锐安科技有限公司 | Data deduplication method, device, terminal and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102902691B (en) | Recommend method and system | |
| CN102314489B (en) | Method for analyzing opinion leader in network forum | |
| CN104599084B (en) | Method of quality control and device that one kind of groups is calculated | |
| Amancio et al. | Comparing intermittency and network measurements of words and their dependence on authorship | |
| CN103823888B (en) | Node-closeness-based social network site friend recommendation method | |
| CN106682172A (en) | Keyword-based document research hotspot recommending method | |
| CN106610955A (en) | Dictionary-based multi-dimensional emotion analysis method | |
| CN108595461A (en) | Interest heuristic approach, storage medium, electronic equipment and system | |
| CN110347897A (en) | Micro blog network emotion community detection method based on event detection | |
| CN107885778A (en) | A kind of personalized recommendation method based on dynamic point of proximity spectral clustering | |
| CN108090197A (en) | A kind of community discovery method of multidimensional social networks | |
| CN104598648B (en) | A kind of microblog users interactive mode gender identification method and device | |
| CN111078873A (en) | Domain expert selection method based on citation network and scientific research cooperation network | |
| CN106886579A (en) | Real-time streaming textual hierarchy monitoring method and device | |
| CN107679239A (en) | Recommend method in a kind of personalized community based on user behavior | |
| Alzaabi et al. | CISRI: A crime investigation system using the relative importance of information spreaders in networks depicting criminals communications | |
| CN104794222B (en) | Network form semanteme restoration methods | |
| CN105045822A (en) | Method for monitoring similar users of specific user in microblog | |
| Lahoti et al. | Finding topical experts in Twitter via query-dependent personalized PageRank | |
| CN105978729A (en) | System and method for pushing mobile phone information based on user surfing log and position | |
| CN107230158A (en) | Social network user relative influence measure | |
| CN110533488A (en) | A kind of business big data analysis system | |
| CN110008411A (en) | It is a kind of to be registered the deep learning point of interest recommended method of sparse matrix based on user | |
| CN108491477A (en) | Neural network recommendation method based on multidimensional cloud and user's dynamic interest | |
| CN112560105A (en) | Joint modeling method and device for protecting multi-party data privacy |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171003 |