CN110312047A - The method and device of automatic shield harassing call - Google Patents
The method and device of automatic shield harassing call Download PDFInfo
- Publication number
- CN110312047A CN110312047A CN201910548703.0A CN201910548703A CN110312047A CN 110312047 A CN110312047 A CN 110312047A CN 201910548703 A CN201910548703 A CN 201910548703A CN 110312047 A CN110312047 A CN 110312047A
- Authority
- CN
- China
- Prior art keywords
- communicating data
- classification
- harassing
- data
- wrecking
- 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/436—Arrangements for screening incoming calls, i.e. evaluating the characteristics of a call before deciding whether to answer it
- H04M3/4365—Arrangements for screening incoming calls, i.e. evaluating the characteristics of a call before deciding whether to answer it based on information specified by the calling party, e.g. priority or subject
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
- H04W12/12—Detection or prevention of fraud
Abstract
The invention discloses a kind of method and device of automatic shield harassing call, method includes: the communicating data for obtaining client；Communicating data is acquired, cluster feature is obtained；According to cluster feature, classified using clustering algorithm to communicating data；Classification belonging to harassing and wrecking communicating data is filtered out, and includes that communicating data intercepts to the classification.This programme is in the air time for obtaining client's communicating data, after communication object, the communication frequency and the duration of call are as main cluster feature, classified using K-means clustering algorithm to communicating data, determine classification belonging to harassing call, and harassing call is effectively intercepted, to substantially reduce the harassing and wrecking frequency of phone, optimal user experience is provided for client.
Description
Technical field
The present invention relates to harassing calls to shield field, especially relate to a kind of automatic shield harassing call method and
Device.
Background technique
With the fast development of Communication in China business, communication quality is constantly promoted, and at the same time, some criminals utilize
The loophole of legal system, the defect of communication enterprise technology and management, dials harassing call without restraint on a communication network, these harassing calls
There are following features mostly:
1, high-volume time calling.Refer to and initiates calling or exhaling repeatedly to simple target number for batch mobile phone destination number
It cries, the subjective desire of mobile phone user has been violated in the big batch calling for sole user, and causes harassing and wrecking to user, is belonged to
Client's point-to-point communication dispute.
2, user's subjective desire is violated, this is one of main feature of harassing call, and harassing call number is to called subscriber
For be all strange number, or not existing virtual-number calls user by number pressure, these
User's subjective desire is all violated in calling behavior, is invalid calling for called subscriber.
3, user is caused to harass.This is the another important feature of harassing call, and harassing call is spy with short time connection
Sign, has just hung up before user's normal ON, has reversely been dialed to user, to reach its illegal purpose, this to
The normal communication at family causes harassing and wrecking.
For such harassing call, consumer very contradicts and dislikes, it is therefore desirable to using a kind of screen method to such
Harassing call carries out intelligent intercept.
Summary of the invention
In order to solve the defect of the above-mentioned prior art, the object of the present invention is to provide a kind of sides of automatic shield harassing call
Method and device.
In order to achieve the above objectives, the technical scheme is that
A kind of method of automatic shield harassing call, including,
Obtain the communicating data of client；
Communicating data is acquired, cluster feature is obtained；
According to cluster feature, classified using clustering algorithm to communicating data；
Classification belonging to harassing and wrecking communicating data is filtered out, and includes that communicating data intercepts to the classification.
Further, described that communicating data is acquired, cluster feature step is obtained, including,
The air time in communicating data is acquired, communication object communicates the frequency and the duration of call as cluster feature.
Further, described that classifying step is carried out to communicating data using clustering algorithm, including,
According to practical application scene, if defining K value size, as cluster centre；
Classified according to K value to all communicating datas, determines classification belonging to communicating data.
Further, described to be classified according to K value to all communicating datas, determine classifying step belonging to communicating data
Later, further include,
Determine new cluster centre point for each classification, according to new cluster centre point to communicating data again minute
Class；
Judge whether the cluster centre of classification changes；
If it is not, all communicating datas classified and respectively classified of output.
Further, described filter out harasses the affiliated classifying step of communicating data, including,
Obtain the harassing and wrecking communicating data for including in communicating data；
Obtain classification information belonging to harassing and wrecking communicating data；
According to classification information belonging to harassing and wrecking communicating data, classification belonging to harassing and wrecking communicating data is determined.
The invention also provides a kind of devices of automatic shield harassing call, including,
Data capture unit, for obtaining the communicating data of client；
Data acquisition unit obtains cluster feature for being acquired to communicating data；
Data sorting unit, for being classified to communicating data using clustering algorithm according to cluster feature；
Screen interception unit, for filter out harassing and wrecking communicating data belonging to classification, and to the classification include communicating data into
Row intercepts.
Further, the data acquisition unit further includes collection apparatus module, and collection apparatus module is for acquiring call
Air time in data, communication object communicate the frequency and the duration of call as cluster feature.
Further, the data sorting unit includes setup module and categorization module,
The setup module is used for according to practical application scene, if defining K value size, as cluster centre；
The categorization module determines and divides belonging to communicating data for being classified according to K value to all communicating datas
Class.
Further, the data sorting unit further includes having center to reset module, variation judgment module and classification output
Module,
The center resets module, for determining new cluster centre point for each classification, according to new cluster centre point
Communicating data is reclassified；
The variation judgment module, for judging whether the cluster centre of classification changes；
The classification output module exports all classification and respectively classifies logical if not changing for cluster centre
Talk about data.
Further, the screening interception unit includes the first acquisition module, and second obtains module and harassing and wrecking categorization module,
Described first obtains module, for obtaining the harassing and wrecking communicating data for including in communicating data；
Described second obtains module, for obtaining classification information belonging to harassing and wrecking communicating data；
The harassing and wrecking categorization module determines harassing and wrecking communicating data for the classification information according to belonging to harassing and wrecking communicating data
Affiliated classification.
The beneficial effects of the present invention are: in the air time for obtaining client's communicating data, communication object communicates the frequency and leads to
After duration is talked about as main cluster feature, is classified using K-means clustering algorithm to communicating data, determine harassing call
Affiliated classification, and harassing call is effectively intercepted, to substantially reduce the harassing and wrecking frequency of phone, best use is provided for client
Family experience.
Detailed description of the invention
Fig. 1 is a kind of flow chart of the method for automatic shield harassing call of the present invention；
Fig. 2 is the step flow chart that the present invention carries out classifying step using clustering algorithm to communicating data；
Fig. 3 is the step flow chart that the present invention filters out the harassing and wrecking affiliated classifying step of communicating data；
Fig. 4 is a kind of structural principle block diagram of the device of automatic shield harassing call of the present invention；
Fig. 5 is the structural block diagram of data sorting unit of the invention；
Fig. 6 is the structural block diagram of screening interception unit of the invention.
Specific embodiment
To illustrate thought and purpose of the invention, the present invention is done further below in conjunction with the drawings and specific embodiments
Explanation.
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiment is only a part of the embodiments of the present invention, instead of all the embodiments.Base
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts it is all its
His embodiment, shall fall within the protection scope of the present invention.
It is to be appreciated that the directional instruction (up, down, left, right, before and after etc.) of institute is only used in the embodiment of the present invention
It explains in relative positional relationship, the motion conditions etc. under a certain particular pose (as shown in the picture) between each component, if the spy
When determining posture and changing, then directionality instruction also correspondingly changes correspondingly, and the connection, which can be, to be directly connected to, can also
To be to be indirectly connected with.
In addition, the description for being such as related to " first ", " second " in the present invention is used for description purposes only, and should not be understood as
Its relative importance of indication or suggestion or the quantity for implicitly indicating indicated technical characteristic.Define as a result, " first ",
The feature of " second " can explicitly or implicitly include at least one of the features.In addition, the technical side between each embodiment
Case can be combined with each other, but must be based on can be realized by those of ordinary skill in the art, when the combination of technical solution
Conflicting or cannot achieve when occur will be understood that the combination of this technical solution is not present, also not the present invention claims guarantor
Within the scope of shield.
Unless otherwise instructed, "/" herein represents meaning as "or".
Referring to Fig.1-3, a proposition specific embodiment of the invention, a kind of method of automatic shield harassing call, including,
S10, the communicating data for obtaining client；
S20, communicating data is acquired, obtains cluster feature；
S30, according to cluster feature, classified using clustering algorithm to communicating data；
S40, classification belonging to harassing and wrecking communicating data is filtered out, and includes that communicating data intercepts to the classification.
For step S10, the communicating data of client is obtained, includes the call details of multiple and different clients in communicating data
Data, such as conversation object (incoming number), air time, duration of call, and the communication frequency etc. information, in addition, each
Client can also be associated with multiple message registrations simultaneously.
For step S20, the communicating data got is acquired, obtains the air time of communicating data, communication object, communication
The frequency and the duration of call are main cluster feature, and cluster feature can be used for subsequent progress communicating data classification, and then screen
Harassing call out.
Step S20 further include: step S21, acquire the air time in communicating data, communication object communicates the frequency and leads to
Duration is talked about as cluster feature.
For step S30, clustering algorithm is one of data mining algorithm most frequently, and be widely used range, main
A point heap sort is carried out according to cluster feature, make there are similar, same characteristic features users to assign to together, simultaneously in the different same clans
User's difference show maximization.Clustering algorithm is a kind of unsupervised learning, its number data divide into several classes, in same class
According to similitude answer as large as possible, the otherness of the data in inhomogeneity is answered as large as possible.By using clustering algorithm pair
The communicating data of client carries out classification processing, communicating data can be divided into different classification according to different characteristic, convenient for subsequent
Further screening belongs to the classification of harassing call.
It, usually can be by communicating data information to determine whether being harassing call in face of first phone.For example, high-end quotient
Business crowd, the common trait that should have: call is more frequent, and monthly average telephone expenses are relatively high, and the duration of call is generally long etc..This reality
It applies the air time in example, communication object, the communication frequency, the duration of call are sent a telegram here as main cluster feature with this to divide
Number whether be harassing call crowd transfer to phone.
Specifically, clustering algorithm is K-means clustering algorithm, by point for realizing K-means clustering algorithm harassing call
The process of class filtering is as follows:
It is assumed that the cluster feature number of the number behavior to be considered is n.
N number of sample { X1, X2 ... Xn } Xn ∈ Rn, i.e. each sample X is n-dimensional vector.
Step a1, k value is random value, randomly selects k point (k≤n), as the cluster centre of k different classifications, is used
CkTo indicate.
Step a2 traverses all data point Xn, by calculating distance, finds the nearest cluster centre point Ck of distance Xn,
At this time it may be said that n-th of communicating data belongs to kth class, and obtain k preliminary classification.
Judge that specific data point belongs to the formula of which cluster centre:
Ck=n:k=argmin | Xn-Xk|2}
Step a3 calculates separately the cluster centre point of each classification according to communicating data in different classifications, it can be understood as
It calculates the central point of each preliminary classification in step a2 or is particle, as such new cluster centre point, and according to
New cluster centre point, classifies to Xn.
Step a4, repeat the above steps a3, until the cluster centre of every one kind be no longer changed namely iteration until receive
It holds back, then communicating data classification is completed, using current class state as final classification.
Here the definition of iteration convergence is: cluster centre no longer changes；Each sample (communicating data) clusters to corresponding
There is no big variations for the sum of the distance at center.
Specifically, K-means cluster requires to be divided into N number of data point in k set (k≤n), so that most interior square
And minimum, cluster goal satisfaction following formula:
With reference to Fig. 2, step S30, comprising the following steps:
S31, according to practical application scene, k value size is set, as cluster centre；
S32, classified according to k value to all communicating datas, determine classification belonging to communicating data.
S33, new cluster centre point is determined for each classification, communicating data is carried out again according to new cluster centre point
Classification；
S34, judge whether the cluster centre of classification changes；
S35, if it is not, all classification of output and the communicating datas respectively classified.
For step S31-S32, as described above, by setting k value, and all data point Xn are further traversed, pass through
Data point is calculated at a distance from different cluster centre electricity, the nearest cluster centre point Ck of distance Xn is found, at this time it may be said that n-th
Communicating data belongs to kth class, obtains k preliminary classification, and all communicating datas are separately dispensed into pair according to above-mentioned condition
In the classification answered, the preliminary classification for communicating data is realized.
For step S33-S35, after communicating data is carried out preliminary classification, all communicating datas are each assigned to
It is accurate and reliable in order to verify the communicating data for including in all classification in suitable classification, it needs according to the institute in current class
There is communicating data from newly calculating a new cluster centre point (particle), as the cluster centre point of the classification, and it is again right
Communicating data is classified, if the cluster centre after reclassifying does not change, represents the classification accurate stable, can be with
It is used as final classification result, without modifying.If the cluster centre after reclassifying changes, then repeat
Step S33 is carried out, until cluster centre no longer changes, classification state at this time is final classification state, can be exported all
It include the classification of different communicating datas.
For step S40, classification belonging to harassing and wrecking communicating data is filtered out, and includes that communicating data blocks to the classification
It cuts.After all data classifications are good, the communicating data that part is determined as molestation is selected from original communicating data,
And navigate in the classification where being determined as the communicating data of molestation, determine whether classification is harassing and wrecking call according to accounting
Classification belonging to data, for example, 100 communicating datas for being determined as harassing call behavior are selected, if above-mentioned 100 logical
Words data have 75 to be located at kth1In classification, other 25 are scattered in other different classification, at this point, if presetting
The communicating data of harassing call behavior is more than 50 and represents this and be classified as classifying belonging to harassing call in good each classification, then
Kth1Classification had both been classification belonging to harassing call, it should be understood that above-mentioned 100,75 and 50 are only intended to illustrate
It is bright, practical judgement data are not represented.
Harassing and wrecking communicating data that is to say the communicating data of molestation, pass through the communicating data for the molestation that part determines
The call behavior properties of different classifications is reversely verified and positioned, the accuracy and effectively of molestation communicating data judgement is improved
Property, improving harassing call interception is efficiency, substantially reduces the harassing and wrecking frequency of phone, provides the user with optimal user experience.
With reference to Fig. 3, step S40 the following steps are included:
S41, the harassing and wrecking communicating data for including in communicating data is obtained；
S42, classification information belonging to harassing and wrecking communicating data is obtained；
S43, the classification information according to belonging to harassing and wrecking communicating data, determine classification belonging to harassing and wrecking communicating data.
This programme is in the air time for obtaining client's communicating data, and communication object, the communication frequency and the duration of call are as master
After the cluster feature wanted, classified using K-means clustering algorithm to communicating data, determines classification belonging to harassing call, and
Harassing call is effectively intercepted, to substantially reduce the harassing and wrecking frequency of phone, provides optimal user experience for client.
With reference to Fig. 4-6, the invention also provides a kind of devices of automatic shield harassing call, specifically include,
Data capture unit 10, for obtaining the communicating data of client；
Data acquisition unit 20 obtains cluster feature for being acquired to communicating data；
Data sorting unit 30, for being classified to communicating data using clustering algorithm according to cluster feature；
Interception unit 40 is screened, includes communicating data for filtering out classification belonging to harassing and wrecking communicating data, and to the classification
It is intercepted.
For data capture unit 10, the communicating data of client is obtained, includes multiple and different clients in communicating data
Call details data, such as conversation object (incoming number), the air time, duration of call, and the communication frequency etc. information, separately
Outside, each client can also be associated with multiple message registrations simultaneously.
For data acquisition unit 20, the communicating data got is acquired, obtains the air time of communicating data, communication pair
As, the communication frequency and the duration of call are main cluster feature, cluster feature can be used for subsequent progress communicating data classification, into
And filter out harassing call.
Data acquisition unit 20 further includes collection apparatus module 21, and collection apparatus module 21 is for acquiring in communicating data
Air time, communication object communicate the frequency and the duration of call as cluster feature.
For data sorting unit 30, clustering algorithm is one of data mining algorithm most frequently, is widely used
Range mainly carries out a point heap sort according to cluster feature, makes have similar, same characteristic features users to assign to together, simultaneously exists
User's difference of the different same clans shows maximization.Clustering algorithm is a kind of unsupervised learning, it is same data divide into several classes
The similitude of data in class is answered as large as possible, and the otherness of the data in inhomogeneity is answered as large as possible.By using poly-
Class algorithm carries out classification processing to the communicating data of client, communicating data can be divided into different classification according to different characteristic,
Belong to the classification of harassing call convenient for subsequent further screening.
It, usually can be by communicating data information to determine whether being harassing call in face of first phone.For example, high-end quotient
Business crowd, the common trait that should have: call is more frequent, and monthly average telephone expenses are relatively high, and the duration of call is generally long etc..This reality
It applies the air time in example, communication object, the communication frequency, the duration of call are sent a telegram here as main cluster feature with this to divide
Number whether be harassing call crowd transfer to phone.
Specifically, clustering algorithm is K-means clustering algorithm, by point for realizing K-means clustering algorithm harassing call
The process of class filtering is as follows:
It is assumed that the cluster feature number of the number behavior to be considered is n.
N number of sample { X1, X2 ... Xn } Xn ∈ Rn, i.e. each sample X is n-dimensional vector.
Step a1, k value is random value, randomly selects k point (k≤n), as the cluster centre of k different classifications, is used
CkTo indicate.
Step a2 traverses all data point Xn, by calculating distance, finds the nearest cluster centre point Ck of distance Xn,
At this time it may be said that n-th of communicating data belongs to kth class, and obtain k preliminary classification.
Judge that specific data point belongs to the formula of which cluster centre:
Ck=n:k=argmin | Xn-Xk|2}。
Step a3 calculates separately the cluster centre point of each classification according to communicating data in different classifications, it can be understood as
It calculates the central point of each preliminary classification in step a2 or is particle, as such new cluster centre point, and according to
New cluster centre point, classifies to Xn.
Step a4, repeat the above steps a3, until the cluster centre of every one kind be no longer changed namely iteration until receive
It holds back, then communicating data classification is completed, using current class state as final classification.
Here the definition of iteration convergence is: cluster centre no longer changes；Each sample (communicating data) clusters to corresponding
There is no big variations for the sum of the distance at center.
Specifically, K-means cluster requires to be divided into N number of data point in k set (k≤n), so that most interior square
And minimum, cluster goal satisfaction following formula:
With reference to Fig. 5, data sorting unit 30 includes setup module 31, categorization module 32, and center resets module 33, and variation is sentenced
Disconnected module 34 and classification output module 35.
Setup module 31 is used for according to practical application scene, if defining K value size, as cluster centre.
Categorization module 32 determines classification belonging to communicating data for classifying according to K value to all communicating datas.
Center resets module 33, for determining new cluster centre point for each classification, according to new cluster centre point pair
Communicating data is reclassified.
Change judgment module 34, for judging whether the cluster centre of classification changes.
Classification output module 35 exports all calls classified and respectively classified if not changing for cluster centre
Data.
For setup module 31 and categorization module 32, as described above, by setting k value, and further traverse all numbers
Strong point Xn finds the nearest cluster centre point Ck of distance Xn, at this time by calculating data point at a distance from different cluster centre electricity
It may be said that n-th of communicating data belongs to kth class, k preliminary classification is obtained, and by all communicating datas according to above-mentioned condition point
It is not assigned in corresponding classification, realizes the preliminary classification for communicating data.
Module 33, variation judgment module 34 and classification output module 35 are reset for center, carried out just by communicating data
After step classification, all communicating datas are each assigned in suitable classification, in order to verify the call for including in all classification
Data are accurate and reliable, need according to all communicating datas in current class from newly calculating a new cluster centre point (matter
Point), as the cluster centre point of the classification, and classify again to communicating data, if the cluster centre after reclassifying
It does not change, then represents the classification accurate stable, can be used as the use of final classification result, without modifying.If
Cluster centre after reclassifying changes, then repeats to determine new cluster centre point for each classification, according to new cluster
Central point reclassifies communicating data, and until cluster centre no longer changes, classification state at this time is final classification
State, can export it is all include different communicating datas classification.
For screening interception unit 40, classification belonging to harassing and wrecking communicating data is filtered out, and include communicating data to the classification
It is intercepted.After all data classifications are good, part is selected from original communicating data and is determined as the logical of molestation
Talk about data, and navigate in the classification where being determined as the communicating data of molestation, determined according to accounting classification whether be
Classification belonging to communicating data is harassed, for example, 100 communicating datas for being determined as harassing call behavior are selected, if on
Stating 100 communicating datas has 75 to be located at kth1In classification, other 25 are scattered in other different classification, at this point, such as
It is to represent this to be classified as harassing call institute that the communicating data of harassing call behavior, which is more than 50, in the pre-set each classification of fruit
Belong to classification, then kth1Classification had both been classification belonging to harassing call, it should be understood that above-mentioned 100,75 and 50 are
For for example, not representing practical judgement data.
The call behavior property of different classifications is reversely verified and positioned to the communicating data of the molestation determined by part,
The accuracy and validity for improving the judgement of molestation communicating data, improving harassing call interception is efficiency, is substantially reduced
The harassing and wrecking frequency of phone provides the user with optimal user experience.
With reference to Fig. 6, screening interception unit 40 includes the first acquisition module 41, and second obtains module 42 and harassing and wrecking categorization module
43。
First obtains module 41, for obtaining the harassing and wrecking communicating data for including in communicating data.
Second obtains module 42, for obtaining classification information belonging to harassing and wrecking communicating data.
Categorization module 43 is harassed, for the classification information according to belonging to harassing and wrecking communicating data, determines harassing and wrecking communicating data institute
Belong to classification.
This programme is in the air time for obtaining client's communicating data, and communication object, the communication frequency and the duration of call are as master
After the cluster feature wanted, classified using K-means clustering algorithm to communicating data, determines classification belonging to harassing call, and
Harassing call is effectively intercepted, to substantially reduce the harassing and wrecking frequency of phone, provides optimal user experience for client.
The above description is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all utilizations
Equivalent structure or equivalent flow shift made by description of the invention and accompanying drawing content is applied directly or indirectly in other correlations
Technical field, be included within the scope of the present invention.
Claims (10)
1. a kind of method of automatic shield harassing call, which is characterized in that including,
Obtain the communicating data of client；
Communicating data is acquired, cluster feature is obtained；
According to cluster feature, classified using clustering algorithm to communicating data；
Classification belonging to harassing and wrecking communicating data is filtered out, and includes that communicating data intercepts to the classification.
2. the method for automatic shield harassing call as described in claim 1, which is characterized in that described to be adopted to communicating data
Collection obtains cluster feature step, including,
The air time in communicating data is acquired, communication object communicates the frequency and the duration of call as cluster feature.
3. the method for automatic shield harassing call as described in claim 1, which is characterized in that described to use clustering algorithm to logical
It talks about data and carries out classifying step, including,
According to practical application scene, if defining K value size, as cluster centre；
Classified according to K value to all communicating datas, determines classification belonging to communicating data.
4. the method for automatic shield harassing call as claimed in claim 3, which is characterized in that described to be led to according to K value to all
Words data are classified, and after determining classifying step belonging to communicating data, further include,
New cluster centre point is determined for each classification, and communicating data is reclassified according to new cluster centre point；
Judge whether the cluster centre of classification changes；
If it is not, all communicating datas classified and respectively classified of output.
5. the method for automatic shield harassing call as described in claim 1, which is characterized in that described to filter out harassing and wrecking call number
According to affiliated classifying step, including,
Obtain the harassing and wrecking communicating data for including in communicating data；
Obtain classification information belonging to harassing and wrecking communicating data；
According to classification information belonging to harassing and wrecking communicating data, classification belonging to harassing and wrecking communicating data is determined.
6. a kind of device of automatic shield harassing call characterized by comprising
Data capture unit, for obtaining the communicating data of client；
Data acquisition unit obtains cluster feature for being acquired to communicating data；
Data sorting unit, for being classified to communicating data using clustering algorithm according to cluster feature；
Interception unit is screened, includes that communicating data blocks for filtering out classification belonging to harassing and wrecking communicating data, and to the classification
It cuts.
7. the device of automatic shield harassing call as claimed in claim 6, which is characterized in that the data acquisition unit also wraps
Collection apparatus module is included, collection apparatus module is used to acquire the air time in communicating data, and communication object communicates the frequency and leads to
Duration is talked about as cluster feature.
8. the device of automatic shield harassing call as claimed in claim 6, which is characterized in that the data sorting unit includes
Setup module and categorization module,
The setup module is used for according to practical application scene, if defining K value size, as cluster centre；
The categorization module determines classification belonging to communicating data for classifying according to K value to all communicating datas.
9. the device of automatic shield harassing call as claimed in claim 8, which is characterized in that the data sorting unit also wraps
The center of having included resets module, variation judgment module and classification output module,
The center resets module, for determining new cluster centre point for each classification, according to new cluster centre point to logical
Words data are reclassified；
The variation judgment module, for judging whether the cluster centre of classification changes；
The classification output module exports all call numbers classified and respectively classified if not changing for cluster centre
According to.
10. the device of automatic shield harassing call as claimed in claim 6, which is characterized in that the screening interception unit packet
The first acquisition module is included, second obtains module and harassing and wrecking categorization module,
Described first obtains module, for obtaining the harassing and wrecking communicating data for including in communicating data；
Described second obtains module, for obtaining classification information belonging to harassing and wrecking communicating data；
The harassing and wrecking categorization module determines belonging to harassing and wrecking communicating data for the classification information according to belonging to harassing and wrecking communicating data
Classification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910548703.0A CN110312047A (en) | 2019-06-24 | 2019-06-24 | The method and device of automatic shield harassing call |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910548703.0A CN110312047A (en) | 2019-06-24 | 2019-06-24 | The method and device of automatic shield harassing call |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110312047A true CN110312047A (en) | 2019-10-08 |
Family
ID=68076142
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910548703.0A Pending CN110312047A (en) | 2019-06-24 | 2019-06-24 | The method and device of automatic shield harassing call |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110312047A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110942783A (en) * | 2019-10-15 | 2020-03-31 | 国家计算机网络与信息安全管理中心 | Group call type crank call classification method based on audio multistage clustering |
| CN113163048A (en) * | 2021-04-02 | 2021-07-23 | 维沃移动通信有限公司 | Crank call identification method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20130113653A (en) * | 2012-04-06 | 2013-10-16 | 에스케이플래닛 주식회사 | System and method for analyzing cluster result of mass data |
| CN104469025A (en) * | 2014-11-26 | 2015-03-25 | 杭州东信北邮信息技术有限公司 | Clustering-algorithm-based method and system for intercepting fraud phone in real time |
| CN108462785A (en) * | 2017-02-21 | 2018-08-28 | 中国移动通信集团浙江有限公司 | A kind of processing method and processing device of malicious call phone |
| CN109451182A (en) * | 2018-10-19 | 2019-03-08 | 北京邮电大学 | A kind of detection method and device of fraudulent call |
| CN109587357A (en) * | 2018-11-14 | 2019-04-05 | 上海麦图信息科技有限公司 | A kind of recognition methods of harassing call |
| CN109600752A (en) * | 2018-11-28 | 2019-04-09 | 国家计算机网络与信息安全管理中心 | A kind of method and apparatus of depth cluster swindle detection |
-
2019
- 2019-06-24 CN CN201910548703.0A patent/CN110312047A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20130113653A (en) * | 2012-04-06 | 2013-10-16 | 에스케이플래닛 주식회사 | System and method for analyzing cluster result of mass data |
| CN104469025A (en) * | 2014-11-26 | 2015-03-25 | 杭州东信北邮信息技术有限公司 | Clustering-algorithm-based method and system for intercepting fraud phone in real time |
| CN108462785A (en) * | 2017-02-21 | 2018-08-28 | 中国移动通信集团浙江有限公司 | A kind of processing method and processing device of malicious call phone |
| CN109451182A (en) * | 2018-10-19 | 2019-03-08 | 北京邮电大学 | A kind of detection method and device of fraudulent call |
| CN109587357A (en) * | 2018-11-14 | 2019-04-05 | 上海麦图信息科技有限公司 | A kind of recognition methods of harassing call |
| CN109600752A (en) * | 2018-11-28 | 2019-04-09 | 国家计算机网络与信息安全管理中心 | A kind of method and apparatus of depth cluster swindle detection |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110942783A (en) * | 2019-10-15 | 2020-03-31 | 国家计算机网络与信息安全管理中心 | Group call type crank call classification method based on audio multistage clustering |
| CN110942783B (en) * | 2019-10-15 | 2022-06-17 | 国家计算机网络与信息安全管理中心 | Group call type crank call classification method based on audio multistage clustering |
| CN113163048A (en) * | 2021-04-02 | 2021-07-23 | 维沃移动通信有限公司 | Crank call identification method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106686261A (en) | Information processing method and system | |
| CN109474923B (en) | Object recognition method and device, and storage medium | |
| CN106255113A (en) | The recognition methods of harassing call and device | |
| CN108462785B (en) | Method and device for processing malicious call | |
| CN110312047A (en) | The method and device of automatic shield harassing call | |
| CN106331390A (en) | Method and system for identifying fraud number based on call data | |
| US8953877B2 (en) | Noise estimation for images | |
| CN104320354B (en) | A kind of processing method and processing device of data | |
| Bertini et al. | Profile resolution across multilayer networks through smartphone camera fingerprint | |
| CN106454195A (en) | Anti-peeping method and system for video chats based on VR | |
| CN103581976A (en) | Cell identifying method and device | |
| CN107194815B (en) | Client segmentation method and system | |
| CN106227409A (en) | Method for information display and device | |
| CN110167030B (en) | Method, device, electronic equipment and storage medium for identifying crank calls | |
| US20120108228A1 (en) | Method for identifying a mobile telephone | |
| CN112866192B (en) | Method and device for identifying abnormal aggregation behaviors | |
| CN108696873B (en) | False user identification method and device | |
| CN105847506B (en) | Communication means, device and the mobile terminal of mobile terminal | |
| CN110536302A (en) | Telecommunication fraud based reminding method and device | |
| CN108270753A (en) | The method and device of logging off users account | |
| CN102256255A (en) | Detection method for parallel-used-card proof based on time and geographic location collisions | |
| CN110213449B (en) | Method for identifying roaming fraud number | |
| CN109587357A (en) | A kind of recognition methods of harassing call | |
| CN109600744B (en) | Voice processing method and system | |
| CN109951609A (en) | A kind of malicious call number processing method and device |
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: 20191008 |