CN107193900A - A kind of identifying system and its application method of suspicious SMS - Google Patents
A kind of identifying system and its application method of suspicious SMS Download PDFInfo
- Publication number
- CN107193900A CN107193900A CN201710327688.8A CN201710327688A CN107193900A CN 107193900 A CN107193900 A CN 107193900A CN 201710327688 A CN201710327688 A CN 201710327688A CN 107193900 A CN107193900 A CN 107193900A
- Authority
- CN
- China
- Prior art keywords
- short message
- suspicious
- module
- feedback
- clouds
- 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
-
- 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
- H04W12/128—Anti-malware arrangements, e.g. protection against SMS fraud or mobile malware
-
- 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
- G06F16/335—Filtering based on additional data, e.g. user or group profiles
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/12—Messaging; Mailboxes; Announcements
- H04W4/14—Short messaging services, e.g. short message services [SMS] or unstructured supplementary service data [USSD]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Security & Cryptography (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The invention provides a kind of identifying system of suspicious SMS, it is related to information security field, including receiving module, pretreatment module, comparison module and feedback module, the short message received is transferred to the pretreatment module and pre-processed by the receiving module, the pretreatment module pre-processes to short message and is transferred to the comparison module, the number of the text feature of the short message received and transmission short message is compared and sends comparative result to the feedback module by the comparison module with deep learning model, the feedback module is fed back according to the comparative result of the comparison module to user.The present invention provides a kind of algorithm reliably, and recognition speed is fast, and Sample Storehouse updates the identifying system and its application method of timely suspicious SMS, in order to which cellphone subscriber recognizes fraud text message in time, reduces cheated risk.
Description
Technical field
The present invention relates to information security field, identifying system and its user more particularly, to a kind of suspicious SMS
Method.
Background technology
Mobile phone has been the indispensable instrument of people's daily life, is unequal in this case using the thing number of short message fraud
Number, and also have the impetus further expanded.
Currently for fraud text message, domestic well-known safe mobile phone manufacturer is substantially simple right using database progress
Than, or only with simple machine learning recognize fraud text message.But the preventing mobile phone swindle system means list used at present
One, it is impossible to be effectively protected user.Meanwhile, current preventing mobile phone swindle system can not be updated in time, it is impossible to obtain newest
Fraud information.
The content of the invention
Reliable the present invention solves the technical problem of a kind of algorithm is provided, recognition speed is fast, and Sample Storehouse updates timely
Suspicious SMS identifying system and its application method, in order to which cellphone subscriber recognizes fraud text message in time, reduction is taken advantage of
The risk deceived.
In order to solve the above technical problems, the invention provides a kind of identifying system of suspicious SMS, it is characterised in that
Including receiving module, pretreatment module, comparison module and feedback module, the short message received is transferred to described by the receiving module
Pretreatment module is pre-processed, and the pretreatment module pre-processes to short message and is transferred to the comparison module, the comparison
Module is by the text feature of the short message received and sends number and the deep learning model of short message and is compared and by comparative result
Send to the feedback module, the feedback module is fed back according to the comparative result of the comparison module to user.
Further, the short message includes short message word and/or URL (Uniform Resource Locator, unified money
Source finger URL).
Further, pretreatment of the pretreatment module to short message includes extracting the number for sending short message, extracts described
Short message word and/or URL simultaneously generate characteristic vector.
Further, the comparison module includes local part and high in the clouds part, in use, first by short message and local part
The suspicious sample of deep learning model be compared and judge, if it is determined that short message is suspicious short message, then further upload
High in the clouds part is further detected, if it is determined that short message is normal short message, then it is normal to send short message to the feedback module
As a result；The short message and the suspicious sample of high in the clouds deep learning model for uploading high in the clouds part are compared, if it is determined that short message is can
Short message is doubted, then the comparison module sends the suspicious result of short message to the feedback module, if it is determined that short message is normal short message,
Then the normal result of short message is sent to the feedback module；The high in the clouds part uses Spark algorithms.
Further, the feedback module receives the comparative result that the comparison module is sent, if normal short message,
Then keep silent, wait next short message；If suspicious short message, then remind user that short message is labeled as into fraud text message.
There is provided a kind of application method of the identifying system of suspicious SMS in the better embodiment of the present invention,
Including：Receive short message；Locally judge whether short message is suspicious short message in mobile phone, if locally judging short message as suspicious short message,
High in the clouds is uploaded to determine whether and feed back to user；If being locally determined as normal short message, feedback-less, silence waits next
Bar short message；The high in the clouds updates sample set according to the suspicious short message of upload, and the high in the clouds uses Spark algorithms；The short message bag
Include short message word and/or URL.
Further, when the short message be short message word when, mobile phone locally judge short message whether be suspicious short message step
Suddenly further comprise, extract the number and the short message word for sending short message, first compare the number and local depth for sending short message
The suspicious sample of the suspicious sample of learning model, then the short message word and local deep learning model；If compared
As a result it is yes, that is, judges short message as suspicious short message, and short message is uploaded into high in the clouds to determine whether；If it is judged that be it is no,
Then feedback-less content, not to user feedback.
Further, what the short message uploaded to high in the clouds determines whether that step includes, relatively more described short message word and cloud
The suspicious sample of deep learning model is held, if comparative result is yes, that is, short message is judged as suspicious short message, and sentenced to local feedback
Disconnected result is that short message is suspicious short message, and reminds user's short message to be suspicious short message；If it is judged that being no, then in feedback-less
Hold, not to user feedback.
Further, when the short message is URL, locally judge that the step of whether short message is suspicious short message enters one in mobile phone
Step includes, and extracts the number and the URL for sending short message, first compares in the number and local deep learning model that send short message
The suspicious sample of suspicious sample, the then URL and local deep learning model；If comparative result is yes, that is, judge
Short message is suspicious short message, and short message is uploaded into high in the clouds determined whether；If it is judged that being no, then feedback-less, and i.e. not
To user feedback content, keep silent and wait next short message to be received.
Further, what the short message uploaded to high in the clouds determines whether that step includes, relatively more described URL and high in the clouds depth
The suspicious sample of learning model, if comparative result is yes, that is, judges short message as suspicious short message, and feeds back judged result to local
It is suspicious short message for short message, and reminds user's short message to be suspicious short message；If it is judged that being no, then feedback-less content, no
To user feedback.
Compared with prior art, the beneficial effects of the invention are as follows：The present invention has high expansibility, using Spark cloud meters
Calculate, training pattern speed is greatly speeded up, can be by adding cheap calculate node improving performance, and expandability is strong.The module of algorithm
Change, the processing of each several part neural network moduleization is easy to locally and remote interaction；Algorithm reliability, using deep learning algorithm,
Automatic study sentence feature, webpage uses DBN, potential feature can be more excavated compared to shallow-layer algorithm.
The technique effect of the design of the present invention, concrete structure and generation is described further below with reference to accompanying drawing, with
It is fully understood from the purpose of the present invention, feature and effect.
Brief description of the drawings
Fig. 1 is a kind of identifying system structural representation of the suspicious SMS of preferred embodiment of the invention；
Fig. 2 is a kind of identifying system application method schematic flow sheet of the suspicious SMS of preferred embodiment of the invention；
Fig. 3 is the identifying system application method flow signal of the suspicious SMS of another preferred embodiment of the invention
Figure；
Fig. 4 is the identifying system application method flow signal of the suspicious SMS of another preferred embodiment of the invention
Figure.
Embodiment
A kind of preferred embodiment of the identifying system of suspicious SMS of the present invention is done below in conjunction with accompanying drawing
It is described in detail, but the present invention is not limited in the embodiment.Thoroughly understand in order that the public has to the present invention, in following hair
Concrete details is described in detail in bright preferred embodiment.
Embodiment 1：
As shown in figure 1, a kind of identifying system of suspicious SMS, including receiving module, pretreatment module, comparison module
And feedback module, the short message received is transferred to pretreatment module and pre-processed by receiving module, and pretreatment module is pre- to short message
Handle and be transferred to comparison module, comparison module by the text feature of the short message received and send short message number and deep learning
Model is compared and sends comparative result to feedback module, and feedback module enters according to the comparative result of comparison module to user
Row feedback, wherein, short message includes short message word and/or URL.
Pretreatment of the pretreatment module to short message includes extracting the number for sending short message, extracts short message word and/or URL simultaneously
Generate characteristic vector.
Comparison module includes local part and high in the clouds part, in use, first by short message and the deep learning mould of local part
The suspicious sample of type is compared and judged, if it is determined that short message is suspicious short message, then further uploads to high in the clouds part and carries out
Further detection, if it is determined that short message is normal short message, then sends the normal result of short message to feedback module；Upload high in the clouds part
Short message and the suspicious sample of high in the clouds deep learning model be compared, if it is determined that short message is suspicious short message, then comparison module
The suspicious result of short message is sent to feedback module, if it is determined that short message is normal short message, then it is normal to send short message to feedback module
Result；High in the clouds part uses Spark algorithms.
Feedback module receives the comparative result of comparison module transmission, if normal short message, then keeps silent, under wait
One short message；If suspicious short message, then remind user that short message is labeled as into fraud text message.
Embodiment 2：
As shown in Fig. 2 a kind of application method of the identifying system of suspicious SMS, including：
Step 100, short message is received, short message includes short message word and/or URL, into step S110；
Step S110, locally judges whether short message is suspicious short message in mobile phone, if locally judging short message as suspicious short message,
Then enter step S400；If being locally determined as normal short message, into step S300；
Step S400, uploads to high in the clouds by suspicious short message and determines whether and feed back to user, wherein, high in the clouds is according to upload
Suspicious short message update sample set, high in the clouds use Spark algorithms；
Step S300, feedback-less, silence waits next short message.
Embodiment 3：
As shown in figure 3, a kind of application method of the identifying system of suspicious SMS, including：
Step 100, short message is received, short message includes short message word, into step 200；
Step 200, the character features of short message are extracted and the number of short message is sent, into step 300；
Step 300, first compare the suspicious sample of the number and local deep learning model that send short message, then compare short message
The suspicious sample of word and local deep learning model, if comparative result is yes, that is, judges that short message as suspicious short message, then enters
Step 500；If it is judged that being no, then into step 400；
Step 400, feedback-less content, not to user feedback；
Step 500, suspicious short message is uploaded into high in the clouds, into step 600；
Step 600, the character features of short message and the suspicious sample of high in the clouds deep learning model are compared, into step 610；
Step 610, judge whether short message is suspicious according to comparative result, if yes then enter step 800, if otherwise entered
Step 700；
Step 700, it is suspicious short message to remind user's short message；
Step 800, feedback-less, silence waits next short message.
Embodiment 4：
As shown in figure 4, a kind of application method of the identifying system of suspicious SMS, including：
Step 100, short message is received, short message includes URL, into step 200；
Step 200, extract the URL of short message and send the number of short message, into step 300；
Step 300, first compare the suspicious sample of the number and local deep learning model that send short message, then compare URL
With the suspicious sample of local deep learning model, if comparative result is yes, that is, short message is judged as suspicious short message, then into step
500；If it is judged that being no, then into step 400；
Step 400, feedback-less content, not to user feedback；
Step 500, suspicious short message is uploaded into high in the clouds, into step 600；
Step S600, compares the URL of short message and the suspicious sample of high in the clouds deep learning model, into step 610；
Step 610, judge whether short message is suspicious according to comparative result, if yes then enter step 800, if otherwise entered
Step 700；
Step 700, it is suspicious short message to remind user's short message；
Step 800, feedback-less, silence waits next short message.
In summary, the present invention has high expansibility, and using Spark cloud computings, training pattern speed is greatly speeded up, can
By adding cheap calculate node improving performance, expandability is strong.The modularization of algorithm, at each several part neural network module
Reason, is easy to locally and remote interaction；Algorithm reliability, using deep learning algorithm, automatic study sentence feature, webpage is used
DBN, potential feature can be more excavated compared to shallow-layer algorithm.
Preferred embodiment of the invention described in detail above.It should be appreciated that one of ordinary skill in the art without
Need creative work just can make many modifications and variations according to the design of the present invention.Therefore, all technologies in the art
Personnel are available by logical analysis, reasoning, or a limited experiment on the basis of existing technology under this invention's idea
Technical scheme, all should be in the protection domain being defined in the patent claims.
Claims (10)
1. a kind of identifying system of suspicious SMS, it is characterised in that including receiving module, pretreatment module, comparison module
And feedback module, the short message received is transferred to the pretreatment module and pre-processed by the receiving module, the pretreatment
Module pre-processes to short message and is transferred to the comparison module, and the comparison module is by the text feature of the short message received and transmission
The number of short message is compared with deep learning model and sends comparative result to the feedback module, the feedback module root
Fed back according to the comparative result of the comparison module to user.
2. the identifying system of suspicious SMS as claimed in claim 1, it is characterised in that the short message includes short message word
And/or URL.
3. the identifying system of suspicious SMS as claimed in claim 2, it is characterised in that the pretreatment module is to short message
Pretreatment include extracting the number for sending short message, extract the short message word and/or URL and generate characteristic vector.
4. the identifying system of suspicious SMS as claimed in claim 3, it is characterised in that the comparison module includes local
Part and high in the clouds part, in use, first the suspicious sample of short message and the deep learning model of local part is compared and sentenced
It is disconnected, if it is determined that short message is suspicious short message, then further uploads to high in the clouds part and further detected, if it is determined that short message is
Normal short message, then send the normal result of short message to the feedback module；Upload the short message and high in the clouds deep learning of high in the clouds part
The suspicious sample of model is compared, if it is determined that short message is suspicious short message, then the comparison module is sent out to the feedback module
The result that short message is suspicious is sent, if it is determined that short message is normal short message, then the normal result of short message is sent to the feedback module；Institute
State high in the clouds part and use Spark algorithms.
5. the identifying system of suspicious SMS as claimed in claim 4, it is characterised in that the feedback module receives institute
The comparative result of comparison module transmission is stated, if normal short message, then keeps silent, waits next short message；If suspicious
Short message, then remind user that short message is labeled as into fraud text message.
6. a kind of application method of the identifying system of suspicious SMS as any one of claim 1-5, its feature
It is, including：
Receive short message；
Locally judge whether short message is suspicious short message in mobile phone；
If locally judging short message as suspicious short message, upload to high in the clouds and determine whether and feed back to user；
If being locally determined as normal short message, feedback-less, silence waits next short message；
The high in the clouds updates sample set according to the suspicious short message of upload, and the high in the clouds uses Spark algorithms；
The short message includes short message word and/or URL.
7. the application method of the identifying system of suspicious SMS as claimed in claim 6, it is characterised in that when the short message
During for short message word, locally judge that the step of whether short message is suspicious short message further comprises in mobile phone, extract and send short message
Number and the short message word, first compare the suspicious sample of the number and local deep learning model that send short message, then compare
The suspicious sample of the short message word and local deep learning model；If comparative result is yes, that is, judge short message to be suspicious short
Believe, and short message is uploaded into high in the clouds and determine whether；If it is judged that being no, then feedback-less content, not to user feedback.
8. the application method of the identifying system of suspicious SMS as claimed in claim 7, it is characterised in that in the short message
That passes to high in the clouds determines whether that step includes, the suspicious sample of relatively more described short message word and high in the clouds deep learning model, such as
Fruit comparative result is yes, that is, judges short message as suspicious short message, and is suspicious short message for short message to local feedback judged result, and is carried
User's short message wake up for suspicious short message；If it is judged that being no, then feedback-less content, not to user feedback.
9. the application method of the identifying system of suspicious SMS as claimed in claim 6, it is characterised in that when the short message
During for URL, locally judge that the step of whether short message is suspicious short message further comprises in mobile phone, extract send short message number and
The URL, first compare send short message number and local deep learning model in suspicious sample, then the URL with
The suspicious sample of local deep learning model；If comparative result is yes, that is, short message is judged as suspicious short message, and short message is uploaded
Determined whether to high in the clouds；If it is judged that being no, then feedback-less, and i.e. not to user feedback content, keep silent and wait
Next short message to be received.
10. the application method of the identifying system of suspicious SMS as claimed in claim 9, it is characterised in that the short message
Upload to high in the clouds determines whether that step includes, relatively more described URL and high in the clouds deep learning model suspicious sample, if than
Relatively result is yes, that is, judges short message as suspicious short message, and is suspicious short message for short message to local feedback judged result, and reminds use
The family short message is suspicious short message；If it is judged that being no, then feedback-less content, not to user feedback.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710327688.8A CN107193900A (en) | 2017-05-10 | 2017-05-10 | A kind of identifying system and its application method of suspicious SMS |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710327688.8A CN107193900A (en) | 2017-05-10 | 2017-05-10 | A kind of identifying system and its application method of suspicious SMS |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107193900A true CN107193900A (en) | 2017-09-22 |
Family
ID=59872989
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710327688.8A Pending CN107193900A (en) | 2017-05-10 | 2017-05-10 | A kind of identifying system and its application method of suspicious SMS |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107193900A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109982272A (en) * | 2019-02-13 | 2019-07-05 | 北京航空航天大学 | A kind of fraud text message recognition methods and device |
| CN110913353A (en) * | 2018-09-17 | 2020-03-24 | 阿里巴巴集团控股有限公司 | Short message classification method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104168548A (en) * | 2014-08-21 | 2014-11-26 | 北京奇虎科技有限公司 | Short message intercepting method and device and cloud server |
| CN104640092A (en) * | 2015-01-27 | 2015-05-20 | 北京奇虎科技有限公司 | Spam short message identifying method, client end, cloud server and system |
| CN105550374A (en) * | 2016-01-29 | 2016-05-04 | 湖南大学 | Random forest parallelization machine studying method for big data in Spark cloud service environment |
| CN105608512A (en) * | 2016-03-24 | 2016-05-25 | 东南大学 | Short-term load forecasting method |
| CN105740424A (en) * | 2016-01-29 | 2016-07-06 | 湖南大学 | Spark platform based high efficiency text classification method |
| CN106161209A (en) * | 2016-07-21 | 2016-11-23 | 康佳集团股份有限公司 | A kind of method for filtering spam short messages based on degree of depth self study and system |
-
2017
- 2017-05-10 CN CN201710327688.8A patent/CN107193900A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104168548A (en) * | 2014-08-21 | 2014-11-26 | 北京奇虎科技有限公司 | Short message intercepting method and device and cloud server |
| CN104640092A (en) * | 2015-01-27 | 2015-05-20 | 北京奇虎科技有限公司 | Spam short message identifying method, client end, cloud server and system |
| CN105550374A (en) * | 2016-01-29 | 2016-05-04 | 湖南大学 | Random forest parallelization machine studying method for big data in Spark cloud service environment |
| CN105740424A (en) * | 2016-01-29 | 2016-07-06 | 湖南大学 | Spark platform based high efficiency text classification method |
| CN105608512A (en) * | 2016-03-24 | 2016-05-25 | 东南大学 | Short-term load forecasting method |
| CN106161209A (en) * | 2016-07-21 | 2016-11-23 | 康佳集团股份有限公司 | A kind of method for filtering spam short messages based on degree of depth self study and system |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110913353A (en) * | 2018-09-17 | 2020-03-24 | 阿里巴巴集团控股有限公司 | Short message classification method and device |
| CN110913353B (en) * | 2018-09-17 | 2022-01-18 | 阿里巴巴集团控股有限公司 | Short message classification method and device |
| CN109982272A (en) * | 2019-02-13 | 2019-07-05 | 北京航空航天大学 | A kind of fraud text message recognition methods and device |
| CN109982272B (en) * | 2019-02-13 | 2020-08-28 | 北京航空航天大学 | Fraud short message identification method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103634473B (en) | Based on mobile phone method for filtering spam short messages and the system of Naive Bayes Classification | |
| CN109005145B (en) | Malicious URL detection system and method based on automatic feature extraction | |
| CN103037339B (en) | One kind is based on the short message filter method of " user's credit worthiness and short message spam degree " | |
| CN110309304A (en) | A kind of file classification method, device, equipment and storage medium | |
| CN110727766B (en) | Sensitive word detection method | |
| CN107395352A (en) | Personal identification method and device based on vocal print | |
| CN103024746A (en) | System and method for processing spam short messages for telecommunication operator | |
| CN108111478A (en) | A kind of phishing recognition methods and device based on semantic understanding | |
| CN101257671A (en) | Method for real time filtering large scale rubbish SMS based on content | |
| CN106446877B (en) | A kind of radio-frequency fingerprint feature extraction and recognition methods based on signal amplitude sequence | |
| CN101150756A (en) | A spam filtering method | |
| CN103795612A (en) | Method for detecting junk and illegal messages in instant messaging | |
| CN107609103A (en) | It is a kind of based on push away spy event detecting method | |
| CN106713579A (en) | Telephone number identification method and device | |
| CN102170447A (en) | Method for detecting phishing webpage based on nearest neighbour and similarity measurement | |
| CN104284306A (en) | Junk message filter method and system, mobile terminal and cloud server | |
| CN107193900A (en) | A kind of identifying system and its application method of suspicious SMS | |
| Sahmoud et al. | Spam detection using BERT | |
| Hosseinpour et al. | An ensemble learning approach for sms spam detection | |
| CN107533574A (en) | Email relationship finger system based on random index pattern match | |
| CN107239504A (en) | A kind of deep learning algorithm for being used to recognize fraud text message | |
| CN113965377A (en) | Attack behavior detection method and device | |
| CN102946400A (en) | Safety filtering method and system for mass short message content based on behavioural analysis | |
| CN106899947A (en) | Short message method for cleaning and device | |
| CN116596274A (en) | Police dispatch method, police dispatch equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170922 |
|
| RJ01 | Rejection of invention patent application after publication |