CN107135314A - Harass detection method, system, mobile terminal and the server of short message - Google Patents
Harass detection method, system, mobile terminal and the server of short message Download PDFInfo
- Publication number
- CN107135314A CN107135314A CN201710476918.7A CN201710476918A CN107135314A CN 107135314 A CN107135314 A CN 107135314A CN 201710476918 A CN201710476918 A CN 201710476918A CN 107135314 A CN107135314 A CN 107135314A
- Authority
- CN
- China
- Prior art keywords
- file
- short message
- server
- mobile terminal
- merging
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 24
- 238000000034 method Methods 0.000 claims abstract description 52
- 230000005540 biological transmission Effects 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 7
- 230000006835 compression Effects 0.000 claims description 3
- 238000007906 compression Methods 0.000 claims description 3
- 239000000725 suspension Substances 0.000 abstract description 6
- 238000012549 training Methods 0.000 description 18
- 239000013598 vector Substances 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 239000000284 extract Substances 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 241001269238 Data Species 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 241000251730 Chondrichthyes Species 0.000 description 1
- 208000027534 Emotional disease Diseases 0.000 description 1
- 230000000172 allergic effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001010 compromised effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 229910052709 silver Inorganic materials 0.000 description 1
- 239000004332 silver Substances 0.000 description 1
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
-
- 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]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/72—Mobile telephones; Cordless telephones, i.e. devices for establishing wireless links to base stations without route selection
- H04M1/724—User interfaces specially adapted for cordless or mobile telephones
- H04M1/72403—User interfaces specially adapted for cordless or mobile telephones with means for local support of applications that increase the functionality
- H04M1/7243—User interfaces specially adapted for cordless or mobile telephones with means for local support of applications that increase the functionality with interactive means for internal management of messages
- H04M1/72436—User interfaces specially adapted for cordless or mobile telephones with means for local support of applications that increase the functionality with interactive means for internal management of messages for text messaging, e.g. short messaging services [SMS] or e-mails
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Security & Cryptography (AREA)
- Business, Economics & Management (AREA)
- General Business, Economics & Management (AREA)
- Human Computer Interaction (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The embodiment of the invention discloses a kind of detection method, system, mobile terminal and server for harassing short message, wherein, applied to the method for mobile terminal, including:Cutting is carried out based on the merging file in memory to the short message of reception, at least one information characteristics is obtained;Merge file from server downloads to memory;Described information feature is inputted into SMS classified model, with reference to file is merged, the corresponding safe class of the short message is exported.Present invention method combines the merging file downloaded from server, realizes without uploading, only just carries out cutting to short message in the terminal;Information characteristics are inputted into SMS classified model, with reference to file is merged, the corresponding safe class of short message are exported;The safe class of short message is differentiated by merging file and SMS classified model realization, foundation is provided for how user's next step handles short message;In suspension environment, in storage resource only hundreds of KB situation, fraud text message, pseudo-base station note and refuse messages are effectively recognized.
Description
Technical field
The present invention relates to internet security technology, especially a kind of detection method for harassing short message, system, mobile terminal and
Server.
Background technology
With the popularization of the portable mobile termianls such as mobile phone and developing rapidly for short message service, people continually use short message
Communicated, however, while people enjoy fast and easily short message communication means, it is incident to spread unchecked increasingly
Fraud text message, refuse messages and pseudo-base station note.Fraud text message be using SMS mode shark up money or finance it is short
Letter form, the form of common fraud text message includes:Bank card integration related content, deduction annual fee related content, tickets refunding and change phase
Inside the Pass appearance, participation activity related content, open wholesale credit card related content etc., deceived by the false content of these great temptation
The important informations such as payment cipher, the payment cipher of Net silver of family such as bank card are taken, the vital interests of user are seriously compromised.
And refuse messages take the device space, puzzlement is caused to user's identification useful information, pseudo-base station note has a strong impact on making for user
With experience.
Fraud text message is that maximum short message is endangered user in all refuse messages.By being sampled to swindle class refuse messages
Analysis, find wherein, pretend to be bank short message at most, accounting is up to 44.7%;Taken advantage of next to that pretending to be electric business businessman to cheat fascination
Deceive consumer's (24.3%), the 3rd be pretend to be telecom operators' (13.1%).Fraud text message is also that change is the most frequent simultaneously
Short message.Cell-phone number, fixed line or URL its life cycle either left in fraud text message is short then one day, long also with regard to one week
Left and right.
The content of the invention
A technical problem to be solved of the embodiment of the present invention is:The harassing and wrecking short message of identification short message safe class is provided
Detection technique.
A kind of detection method for harassing short message provided in an embodiment of the present invention, applied to mobile terminal, including:
Cutting is carried out based on the merging file in memory to the short message of reception, at least one information characteristics is obtained;It is described
Merge file from server downloads to memory;
Described information feature is inputted into SMS classified model, with reference to file is merged, corresponding safety of the short message etc. is exported
Level.
In another embodiment based on the above method, in addition to:Safe class based on output, will meet preset security etc.
The short message of level and the relevant information of the short message are sent to server.
In another embodiment based on the above method, in addition to:Downloaded from server and merge file, to the merging file
Decompressed and decomposed, obtain dictionary library file and disaggregated model file.
In another embodiment based on the above method, the short message of described pair of reception is carried out based on the merging file in memory
Cutting, obtains at least one information characteristics, including:
Short message is received, the corresponding content of text of short message is obtained, the dictionary library file merged in memory in file is called, will
The content of text of the short message is matched with the Feature Words in the dictionary library file, determines including accordingly in the short message
Multiple information characteristics of Feature Words;
Wherein, the length of the Feature Words in the dictionary includes one or more characters, and the Feature Words are with aobvious
Write the word of distinction.
In another embodiment based on the above method, described information feature is inputted into SMS classified model, it is literary with reference to merging
Part, exports the corresponding safe class of the short message, including:
By based on merging in the SMS classified model of disaggregated model file input that file is obtained, SMS classified model is carried out
Training, the SMS classified model trained;
It is described based on the SMS classified model output in the SMS classified model that will be trained described in information characteristics input
The corresponding safe class of short message.
It is described corresponding based on the SMS classified model output short message in another embodiment based on the above method
Safe class, including:
The content of text of the short message is estimated according to described information feature, to obtain the content of text of the short message
Assessed value, corresponding safe class is worth to based on assessment.
In another embodiment based on the above method, the content of text of the short message is commented according to described information feature
Estimate, including:
It is determined that the corresponding weight of each described information feature;
The content of text of the short message is estimated according to the weight of the determination and described information feature.
Other side according to embodiments of the present invention there is provided a kind of harassing and wrecking short message detection method, applied to service
Device, including:
Receive the file download request that mobile terminal is sent;
Dictionary library file and disaggregated model file are combined, and the dictionary library file after combination and disaggregated model is literary
Part, which is compressed, to be obtained merging file;
Merging file is issued to the mobile terminal for sending file download request.
Other side according to embodiments of the present invention there is provided a kind of mobile terminal, be applied to, including:
Feature extraction unit, carries out cutting based on the merging file in memory for the short message to reception, obtains at least
One information characteristics;The merging file is from server downloads to memory;
Ranking unit, for described information feature to be inputted into SMS classified model, with reference to file is merged, output is described
The corresponding safe class of short message.
Other side according to embodiments of the present invention there is provided a kind of server, including:
Request reception unit, the file download request for receiving mobile terminal transmission;
Piece file mergence unit, for dictionary library file and disaggregated model file to be combined, and by the dictionary after combination
Library file and disaggregated model file, which are compressed, to be obtained merging file;
File issuance unit, for will merge file be issued to send file download request the mobile terminal.
Other side according to embodiments of the present invention there is provided a kind of harassing and wrecking short message detecting system, including:One such as
Upper described server and at least one mobile terminal as described above.
Detection method, system, mobile terminal and the service of a kind of harassing and wrecking short message provided based on the above embodiment of the present invention
Device, carries out cutting based on the merging file in memory to the short message of reception, obtains at least one information characteristics;Merge file from
Server is downloaded in memory;With reference to the merging file downloaded from server, realize without uploading, only in the terminal
Cutting is carried out to short message;Information characteristics are inputted into SMS classified model, with reference to file is merged, corresponding safety of short message etc. are exported
Level;The safe class of short message is differentiated by merging file and SMS classified model realization, is user's next step such as where
The letter that is in the wrong provides foundation;In suspension environment, in storage resource used only hundreds of KB situation, accomplish effectively to recognize fraud text message,
Pseudo-base station note and refuse messages.
Below by drawings and examples, technical scheme is described in further detail.
Brief description of the drawings
The accompanying drawing for constituting a part for specification describes embodiments of the invention, and together with description for explaining
The principle of the present invention.
Referring to the drawings, according to following detailed description, the present invention can be more clearly understood from, wherein:
Fig. 1 harasses the flow chart of detection method one embodiment of short message for the present invention.
Fig. 2 harasses the flow chart of another embodiment of the detection method of short message for the present invention.
Fig. 3 is the structural representation of mobile terminal one embodiment of the present invention.
Fig. 4 is the structural representation of another embodiment of mobile terminal of the present invention.
Fig. 5 harasses the flow chart of detection method one embodiment of short message for the present invention.
Fig. 6 is the structural representation of server one embodiment of the present invention.
Embodiment
The various exemplary embodiments of the present invention are described in detail now with reference to accompanying drawing.It should be noted that:Unless had in addition
Body illustrates that the part and the positioned opposite of step, numerical expression and numerical value otherwise illustrated in these embodiments does not limit this
The scope of invention.
Simultaneously, it should be appreciated that for the ease of description, the size of the various pieces shown in accompanying drawing is not according to reality
Proportionate relationship draw.
The description only actually at least one exemplary embodiment is illustrative below, never as to the present invention
And its any limitation applied or used.
It may be not discussed in detail for technology, method and apparatus known to person of ordinary skill in the relevant, but suitable
In the case of, the technology, method and apparatus should be considered as a part for specification.
It should be noted that:Similar label and letter represents similar terms in following accompanying drawing, therefore, once a certain Xiang Yi
It is defined, then it need not be further discussed in subsequent accompanying drawing in individual accompanying drawing.
The embodiment of the present invention can apply to computer system/server, and it can be with numerous other universal or special calculating
System environments or configuration are operated together.Suitable for be used together with computer system/server well-known computing system, ring
The example of border and/or configuration includes but is not limited to:Personal computer system, server computer system, thin client, thick client
Machine, hand-held or laptop devices, the system based on microprocessor, set top box, programmable consumer electronics, NetPC Network PC,
Little types Ji calculates machine Xi Tong ﹑ large computer systems and the distributed cloud computing technology environment including any of the above described system, etc..
Computer system/server can be in computer system executable instruction (such as journey performed by computer system
Sequence module) general linguistic context under describe.Generally, program module can include routine, program, target program, component, logic, number
According to structure etc., they perform specific task or realize specific abstract data type.Computer system/server can be with
Implement in distributed cloud computing environment, in distributed cloud computing environment, task is by by the long-range of communication network links
Manage what equipment was performed.In distributed cloud computing environment, program module, which can be located at, includes the Local or Remote meter of storage device
On calculation system storage medium.
Fig. 1 harasses the flow chart of detection method one embodiment of short message for the present invention.As shown in figure 1, this method application
In mobile terminal, the embodiment method includes:
Step 101, cutting is carried out based on the merging file in memory to the short message of reception, obtains at least one information special
Levy;Merge file from server downloads to memory.
It can be that transmission file download is asked after receiving short message to download merging file from server, and the reception server
Merging in file deposit memory for feedback carries out cutting to the short message of reception;It can also be before short message is received, send out in advance
File download request is sent into server, the merging file deposit memory of the reception server feedback, in case subsequently being connect to terminal
The short message received carries out cutting;Merge file if receiving in advance, then, when merging file is regularly updated in server,
It is synchronous that merging file in mobile terminal is updated.
Step 102, information characteristics are inputted into SMS classified model, with reference to file is merged, exports corresponding safety of short message etc.
Level.
The safe class of output can include but is not limited to dangerous, harassing and wrecking and level of security, be mainly used in for user it is next
Step performs operation to the short message and provides foundation, and the short message for harmful grade is generally isolated and/or delete processing;For disturbing
The short message for disturbing rank can be using isolation and/or deletion and/or prompting user's processing;Short message for level of security is typically not
Process, be directly displayed to user.Certainly the first safety can also be such as set according to specifically needing to be configured safe class
Rank, second level of security etc., can be set at least two safe class ranks, and no ceiling is suitable for various situations.
A kind of detection method of the harassing and wrecking short message provided based on the above embodiment of the present invention, the short message to reception is based on storage
Merging file in device carries out cutting, obtains at least one information characteristics;Merge file from server downloads to memory;Knot
The merging file downloaded from server is closed, realizes without uploading, cutting just only is carried out to short message in the terminal;Information is special
The SMS classified model of input is levied, with reference to file is merged, the corresponding safe class of short message is exported;By merging file and SMS classified
Model realization is differentiated to the safe class of short message, and foundation is provided for how user's next step handles short message;In suspension environment,
In storage resource used only hundreds of KB situation, accomplish effectively to recognize fraud text message, pseudo-base station note and refuse messages.
In a specific example of the inventive method above-described embodiment, in addition to:Safe class based on output, will be accorded with
The relevant information of the short message and short message of closing preset security grade is sent to server.
Clothes can be reported to by advertisement SMS, fraud text message or illegal short message by monitoring user by the safe class of mobile terminal
Business device end, this reporting schemes can ask whether the refuse messages that can be reported with the dynamic transmission of householder or application program.Such short message
Predominantly mobile terminal is installed before safety monitoring application program, the short message of mobile terminal storage.
After the short message of received server-side reporting of user, the number information in short message text is extracted, for building black number
Storehouse.Wherein, user can also report the fraud of black number and its correlation, and the black number is related to certified malfeasance, have
There is true victim (such as:Individual, company, group etc.), black number and related fraud information are documented to black number by server end
In storehouse.Exemplified by reporting fraud text message, Bank Account Number in fraud text message is the account of false prize information, and the account is to lead
The name of prize guarantee fund gains some members by cheating, and provides relevant evidence.
Wherein, user can also report doubtful refuse messages.
The doubtful short message that server end judges:When server end is matched to the number in short message, if the number is doubted
Like being black number, then also the number is taken in black number storehouse.
The refuse messages collected using Internet resources:Server end extracts the number information in refuse messages, and records extremely
Black number storehouse.
In other embodiments, the mode of junk information is collected specifically, can be in part-time, taxi, goods using Internet resources
The information promulgating platforms such as thing transaction, gaming platform, communication group are collected, and the keyword for being related to wealth in message is carried out first
Crawl, then by default rule (such as:Transaction value exception etc.) judged, to determine whether the message is rubbish letter
Breath, if so, the number information in the message is recorded to black number storehouse.
For black number storehouse, black number storehouse can be classified, such as blacklist, high-risk list and doubtful list, in different lists
The degree of danger of black number is different.
Black number in blacklist is mainly the number in the certified number information for producing true victim, this list
With serious danger.
Black number in high-risk list is mainly to be accused of malfeasant number information, and the number information in this list does not have also
There is the illegal act for occurring essence or do not reported by victim also, with high risk.As number information is related to down tick
Deng.
Black number in doubtful list be mainly it is doubtful be accused of malfeasant number information, the number information in this list
The number information or user that are predominantly related to sensitive information ask the number information of safety detection, in its dangerous more high-risk list
Black number it is low.As submitted safety detection request, certain number information to be largely related to by a large number of users in certain number information short time
It is part-time, the sensitive information for being related to wealth such as supplement with money.
Except black number storehouse, also can be set number information predominantly safe in corresponding white list, white list, generally performance
For the message of the push such as movement, telecommunications, UNICOM, bank.
In a specific example of the inventive method above-described embodiment, in addition to:Downloaded from server and merge file, it is right
Merge file to be decompressed and decomposed, obtain dictionary library file and disaggregated model file.
In the present embodiment, merging file includes dictionary library file and disaggregated model file, so as to accomplish in limited space
Under, by participle dependent file, sorting algorithm dependent file is issued simultaneously.Participle can be accurately so carried out, can be with profit
With the sorting algorithm of machine learning, natural language processing is combined with machine learning.
In an embodiment of the present invention, the label information in the dictionary library file that server is provided can be server from mutual
Join online collection or that reception is submitted from mobile terminal, the invention is not limited in this regard.For example, server from
In the web page contents of internet extract number label information, the label information can be the classification of number, such as express delivery, food delivery,
Education etc. or directly mark the number be malice or non-malicious, etc..And for example, number is submitted in mobile terminal
Label information before, hand labeled can be carried out by user, such as user, can at the interface of mobile terminal after phone has been answered
On show multiple options, such as food delivery phone, express delivery phone, sale call, the house property medium, user is according to receiving calls
Content selection goes out the type of the telephone number, and the information such as time, area, classification of these labeled numbers can be number
Label information.
In a specific example of the inventive method above-described embodiment, step 101 can include:
Short message is received, the corresponding content of text of short message is obtained, the dictionary library file merged in memory in file is called, will
The content of text of short message is matched with the Feature Words in dictionary library file, is determined multiple comprising individual features word in short message
Information characteristics;
Wherein, the length of the Feature Words in dictionary includes one or more characters, and Feature Words are with notable distinction
Word.
In the present embodiment, dictionary library file includes a large amount of based on known refuse messages, fraud text message and pseudo-base station
The Feature Words with identification function that short message is obtained, now by the content of text in the short message received and the progress of all Feature Words
Match somebody with somebody, the contents extraction matched in content of text with Feature Words is come out, configuration information feature is subsequently needed according to these information
Feature is judged the rank of the short message.
Fig. 2 harasses the flow chart of another embodiment of the detection method of short message for the present invention.As shown in Fig. 2 the present embodiment
Method includes:
Step 101, cutting is carried out based on the merging file in memory to the short message of reception, obtains at least one information special
Levy;Merge file from server downloads to memory.
It can be that transmission file download is asked after receiving short message to download merging file from server, and the reception server
Merging in file deposit memory for feedback carries out cutting to the short message of reception;It can also be before short message is received, send out in advance
File download request is sent into server, the merging file deposit memory of the reception server feedback, in case subsequently being connect to terminal
The short message received carries out cutting;Merge file if receiving in advance, then, when merging file is regularly updated in server,
It is synchronous that merging file in mobile terminal is updated.
Step 202, it will be inputted based on the disaggregated model file for merging file acquisition in SMS classified model, to SMS classified
Model is trained, the SMS classified model trained.
Disaggregated model file is to be obtained based on lexicon file by processing, and disaggregated model file is included based on dictionary text
The attributive character such as term vector, the theme vector of corresponding short message that the Feature Words in part are obtained, short message point is input to by attributive character
In class model, SMS classified model can be trained, trained SMS classified model just can be defeated to the short message of input
Go out correct classification type.
Step 203, information characteristics are inputted in the SMS classified model trained, it is described based on the output of SMS classified model
The corresponding safe class of short message.
Polytypic SMS classified model can be constructed using Libsvm (support vector cassification storehouse), it would however also be possible to employ
Liblinear (linear classification storehouse) constructs polytypic SMS classified model.Polytypic short message is being constructed using Liblinear
During disaggregated model, the dynamic cataloging storehouse in Liblinear can be called, it is polytypic short using the dynamic cataloging storehouse construction called
Believe disaggregated model.
In the present embodiment, the content of text cutting of the short message received is obtained into information characteristics, these features are input to
In the SMS classified model trained by the Feature Words of known classification type, due to the SMS classified skilled material of model
All it is the corresponding disaggregated model file of short message of rank known to difference, therefore, the SMS classified model that training is completed is for input
Information characteristics can accurately provide correspondence short message level of security.
Target message can be obtained by LDA (Latent Dirichlet Allocation, semantic topic generation model)
Affiliated theme vector is theme.Specifically, all participles at least one participle of target message are inputted into LDA models,
LDA models can correspond to each participle and obtain the vector for representing its affiliated theme topic probability, then by all participles of input
The addition of vectors of its affiliated theme topic probability is represented, is obtained in the theme matrix multiple that will add up obtained vector and LDA models
Obtain theme vector and the output of target message.
The training sample of above-mentioned generation is subjected to handmarking, such as the spending amount, the remaining sum that judge bank's transmission
Etc. information, then positive sample will be labeled as based on the characteristic vector generated on the basis of the numeral for characterizing the amount of money, remaining sum etc. in short message;It is short
Other digital informations in letter, such as beat a few folding numerals, and the characteristic vector generated on the basis of those numerals is labeled as negative sample.
Also, the object information of positive sample is set to correctly, output result value is 1, and the object information of negative sample is set into mistake, exported
End value is 0.
In a specific example of the inventive method above-described embodiment, step 203 can specifically include:It is special according to information
Levy and the content of text of short message is estimated, with the assessed value for the content of text for obtaining short message, be worth to based on assessment corresponding
Safe class.
After short message is judged based on safe class for harassing and wrecking short message, then the short message is intercepted and be stored in blacklist
In.In other embodiments, after judging short message for harassing and wrecking short message, the short message is deleted from the message list of user, to avoid
Harassing and wrecking short message is the inconvenience that user brings, and improves the Experience Degree that user is communicated using short message.
In a specific example of the inventive method above-described embodiment, the content of text of short message is entered according to information characteristics
Row is assessed, and can be included:
It is determined that the corresponding weight of each information characteristics;
The content of text of short message is estimated according to the weight and information characteristics of determination.
In the present embodiment, the content of text of short message is estimated according to information characteristics, can be according to information characteristics
Weight is estimated, and determines that the weight of each information characteristics can be realized in several ways, such as according to expertise or
Person carries out weight according to actual conditions and determines to operate;And the acquisition of assessed value can include:Obtain what is pre-established in server
The Feature Words of dictionary library file and the corresponding relation of weight, according to the corresponding relation, search the weight of each information characteristics, with
Afterwards, the content of text of short message can be estimated according to weight and each information characteristics.
One of ordinary skill in the art will appreciate that:Realizing all or part of step of above method embodiment can pass through
Programmed instruction related hardware is completed, and foregoing program can be stored in a computer read/write memory medium, the program
Upon execution, the step of including above method embodiment is performed;And foregoing storage medium includes:ROM, RAM, magnetic disc or light
Disk etc. is various can be with the medium of store program codes.
Fig. 3 is the structural representation of mobile terminal one embodiment of the present invention.The mobile terminal of the embodiment can be used for real
Existing above-mentioned each method embodiment of the invention.As shown in figure 3, the mobile terminal of the embodiment includes:
Feature extraction unit 31, for the short message to reception based in memory merging file carry out cutting, obtain to
Few information characteristics.
Wherein, file is merged from server downloads to memory.
Ranking unit 32, for information characteristics to be inputted into SMS classified model, with reference to file is merged, exports short message pair
The safe class answered.
A kind of mobile terminal provided based on the above embodiment of the present invention, to the short message of reception based on the merging in memory
File carries out cutting, obtains at least one information characteristics;Merge file from server downloads to memory;With reference to from server
The merging file of download, realizes without uploading, only just carries out cutting to short message in the terminal;Information characteristics are inputted into short message
Disaggregated model, with reference to file is merged, exports the corresponding safe class of short message;By merging file and SMS classified model realization pair
The safe class of short message is differentiated, and foundation is provided for how user's next step handles short message;In suspension environment, in storage used
Resource only hundreds of KB situation, accomplishes effectively to recognize fraud text message, pseudo-base station note and refuse messages.
In a specific example of mobile terminal above-described embodiment of the present invention, in addition to short message sending unit, for base
In the safe class of output, the relevant information of the short message and short message that meet preset security grade is sent to server.
In a specific example of mobile terminal above-described embodiment of the present invention, in addition to:Download unit, for from service
Device, which is downloaded, merges file, and the merging file is decompressed and decomposed, dictionary library file and disaggregated model file is obtained.
In a specific example of mobile terminal above-described embodiment of the present invention, feature extraction unit 31, specifically for connecing
Short message is received, the corresponding content of text of short message is obtained, the dictionary library file merged in memory in file is called, by the text of short message
Content is matched with the Feature Words in dictionary library file, determines the multiple information characteristics for including individual features word in short message.
Wherein, the length of the Feature Words in dictionary includes one or more characters, and Feature Words are with notable distinction
Word.
Fig. 4 is the structural representation of another embodiment of mobile terminal of the present invention.As shown in figure 4, the present embodiment movement is eventually
Middle grade evaluation unit 32 is held, including:
Training module 321, it is right for will be inputted based on the disaggregated model file for merging file acquisition in SMS classified model
SMS classified model is trained, the SMS classified model trained.
Prediction module 322, for information characteristics to be inputted in the SMS classified model trained, based on SMS classified model
Export the corresponding safe class of short message.
In the present embodiment, disaggregated model file is to be obtained based on lexicon file by processing, and disaggregated model file is
Know the file that arrives of the short message of its class categories after sorting out and dividing, disaggregated model file is input to SMS classified model
In, SMS classified model can be trained, trained SMS classified model just can export correct to the short message of input
Classification type.The content of text cutting of the short message received is obtained into information characteristics, these features are input to by known
In the SMS classified model of the Feature Words training of classification type, because the SMS classified skilled material of model is all different
The corresponding disaggregated model file of short message of known rank, therefore, the SMS classified model that training is completed are special for the information of input
The level of security of correspondence short message can accurately be provided by levying.
In a specific example of mobile terminal above-described embodiment of the present invention, prediction module 322, specifically for according to letter
Breath feature is estimated to the content of text of the short message, with the assessed value for the content of text for obtaining short message, is obtained based on assessed value
To corresponding safe class.
In a specific example of mobile terminal above-described embodiment of the present invention, prediction module 322 is additionally operable to determine each
The corresponding weight of information characteristics;The content of text of short message is estimated according to the weight and information characteristics of determination.
Involved short message, which refers not only to the mobile terminals such as short message, in addition to wechat short message, QQ short messages, in the present invention to receive
That arrives includes all short messages of content of text.
Fig. 5 harasses the flow chart of detection method one embodiment of short message for the present invention.As shown in figure 5, this method application
In server, the embodiment method includes:
Step 501, the file download request that mobile terminal is sent is received.
Mobile terminal limited storage space, therefore, when that need not differentiate to short message, it is possible to by what is stored before
Merge file to delete, therefore, received in mobile terminal when new short message needs discriminating, it is necessary to send file download to server
Request, server is issued to mobile terminal according to request by file is merged;And work as in mobile terminal and there is existing merging file
When, now file download is updated to mobile terminal to existing merging file.
Step 502, dictionary library file and disaggregated model file are combined, and by the dictionary library file after combination and divided
Class model file, which is compressed, to be obtained merging file.
Step 503, merging file is issued to the mobile terminal for sending file download request.
The detection method of a kind of harassing and wrecking short message provided based on the above embodiment of the present invention, in the server by dictionary library text
Part and disaggregated model file are combined and compressed, make merging file be sized to normally run in the terminal, it is mobile eventually
In the case that end only provides hundreds of KB memory space, in suspension state, mobile terminal only relies on the merging downloaded from server
File can be achieved with the identification to refuse messages, fraud text message and pseudo-base station note.
In an embodiment of the present invention, the label information in the signature library of server can be that server is received from internet
It is collection or that reception is submitted from mobile terminal, the invention is not limited in this regard.For example, server is from internet
The label information of number is extracted in web page contents, the label information can be the classification of number, such as express delivery, food delivery, education,
Can be directly mark the number be malice or non-malicious, etc..And for example, the label information of number is submitted in mobile terminal
Before, hand labeled can be carried out by user, such as user can show many after phone has been answered on the interface of mobile terminal
Individual option, such as food delivery phone, express delivery phone, sale call, house property medium, user goes out according to the content selection received calls
The type of the telephone number, the information such as time, area, the classification of these labeled numbers can be believed for the mark of number
Breath.
In a specific example of the inventive method above-described embodiment, disaggregated model file obtains wherein in step 502
Process is taken to include:
Participle is carried out to the training data of collection based on the Feature Words in dictionary library file and feature selecting obtains participle spy
Levy data;
Participle characteristic is classified based on sorting algorithm, disaggregated model file is obtained.
In the present embodiment, the Feature Words in dictionary library file are all to extract to obtain by the short message to known rank,
Participle and feature selecting are carried out by obtained participle characteristic to a large amount of training datas of collection based on these Feature Words, wherein
A large amount of training datas are the short messages with clear and definite classification type gathered for classification based training, by what is obtained based on training data
After participle characteristic input sorting algorithm, so that it may obtain the corresponding class categories information of each training data, and based on classification
Classification information is classified to all participle characteristics, and then obtains disaggregated model file.
In a specific example of the inventive method above-described embodiment, in addition to:Receive the short message that mobile terminal is sent
With the relevant information of short message, short message and relevant information are stored in associated databases;Data include black list database, and/or
White list database.
After the short message of received server-side reporting of user, the number information in short message text is extracted, for building black number
Storehouse.Wherein, user can also report the fraud of black number and its correlation, and the black number is related to certified malfeasance, have
There is true victim (such as:Individual, company, group etc.), black number and related fraud information are documented to black number by server end
In storehouse.Exemplified by reporting fraud text message, Bank Account Number in fraud text message is the account of false prize information, and the account is to lead
The name of prize guarantee fund gains some members by cheating, and provides relevant evidence.
Wherein, user can also report doubtful refuse messages.
The doubtful short message that server end judges:When server end is matched to the number in short message, if the number is doubted
Like being black number, then also the number is taken in black number storehouse.
The refuse messages collected using Internet resources:Server end extracts the number information in refuse messages, and records extremely
Black number storehouse.
In other embodiments, the mode of junk information is collected specifically, can be in part-time, taxi, goods using Internet resources
The information promulgating platforms such as thing transaction, gaming platform, communication group are collected, and the keyword for being related to wealth in message is carried out first
Crawl, then by default rule (such as:Transaction value exception etc.) judged, to determine whether the message is rubbish letter
Breath, if so, the number information in the message is recorded to black number storehouse.
For black number storehouse, black number storehouse can be classified, such as blacklist, high-risk list and doubtful list, in different lists
The degree of danger of black number is different.
Black number in blacklist is mainly the number in the certified number information for producing true victim, this list
With serious danger.
Black number in high-risk list is mainly to be accused of malfeasant number information, and the number information in this list does not have also
There is the illegal act for occurring essence or do not reported by victim also, with high risk.As number information is related to down tick
Deng.
Black number in doubtful list be mainly it is doubtful be accused of malfeasant number information, the number information in this list
The number information or user that are predominantly related to sensitive information ask the number information of safety detection, in its dangerous more high-risk list
Black number it is low.As submitted safety detection request, certain number information to be largely related to by a large number of users in certain number information short time
It is part-time, the sensitive information for being related to wealth such as supplement with money.
Except black number storehouse, also can be set number information predominantly safe in corresponding white list, white list, generally performance
For the message of the push such as movement, telecommunications, UNICOM, bank.
One of ordinary skill in the art will appreciate that:Realizing all or part of step of above method embodiment can pass through
Programmed instruction related hardware is completed, and foregoing program can be stored in a computer read/write memory medium, the program
Upon execution, the step of including above method embodiment is performed;And foregoing storage medium includes:ROM, RAM, magnetic disc or light
Disk etc. is various can be with the medium of store program codes.
Fig. 6 is the structural representation of server one embodiment of the present invention.The server of the embodiment can be used for realizing this
Invent above-mentioned each method embodiment.As shown in fig. 6, the server of the embodiment includes:
Request reception unit 61, the file download request for receiving mobile terminal transmission.
Mobile terminal limited storage space, therefore, when that need not differentiate to short message, it is possible to by what is stored before
Merge file to delete, therefore, received in mobile terminal when new short message needs discriminating, it is necessary to send file download to server
Request, server is issued to mobile terminal according to request by file is merged;And work as in mobile terminal and there is existing merging file
When, now file download is updated to mobile terminal to existing merging file.
Piece file mergence unit 62, for dictionary library file and disaggregated model file to be combined, and by the word after combination
Allusion quotation library file and disaggregated model file, which are compressed, to be obtained merging file.
File issuance unit 63, for will merge file be issued to send file download request mobile terminal.
A kind of server provided based on the above embodiment of the present invention, in the server by dictionary library file and disaggregated model
File is combined and compressed, make merging file be sized to normally run in the terminal, only provide hundreds of in mobile terminal
In the case of KB memory space, in suspension state, mobile terminal only relies on the merging file downloaded from server and can be achieved with
Identification to refuse messages, fraud text message and pseudo-base station note.
In a specific example of server above-described embodiment of the present invention, in addition to:Classify acquiring unit, for based on
Feature Words in dictionary library file carry out participle to the training data of collection and feature selecting obtains participle characteristic;Based on point
Class algorithm is classified to participle characteristic, obtains disaggregated model file.
In a specific example of server above-described embodiment of the present invention, in addition to memory cell, for receiving movement
Short message and the relevant information of short message that terminal is sent, short message and relevant information are stored in associated databases;Data include black name
Single database, and/or white list database.
The various embodiments described above of the present invention, a lexicon file, dictionary text are obtained in the study of server by utilizing mass data
Part is that the algorithm based on a global analysis and utilization information gain, cross entropy calculates the word for obtaining most notable distinction.
Using these words, participle, feature selecting are carried out with reference to the large-scale training corpus of server, is classified in conjunction with sorting algorithm
Model file.Then lexicon file and model file are effectively combined again, while carrying out utilizing efficient compression algorithm
Generate the merging file finally issued.Mobile terminal is cut when a short message is received using merging file and first carrying out short message
Point, then SMS classified model receives the feature after cutting, with reference to the safe class for merging file final output short message.
Knowledge method for distinguishing for fraud text message is not necessarily to carry out, be taken the lead in for fraud text message to fast mode soon
Proposition non-white be black strategy.For integration fraud text message, aviation swindle, bank card credit line extracts short message, and QQ is part-time short
Letter etc. common type swindle we all can be black strategy using non-white.User input short message first using URL strategy and number
Code strategy is handled, and the powerful white library collected with reference to server, and machine learning algorithm provide safe class jointly.
Non-white is that black strategy is not simply to contrast white storehouse for URL, can not be met for fraud text message needs high accuracy
The requirement quoted.So using machine learning algorithm, generating the correlation model of fraud text message, further being lifted by algorithm model
Accuracy rate, is that black accuracy rate is lifted to 99% by non-white.So as to the interception fraud text message of precise and high efficiency, realize with not allergic effect ten thousand
Become.
Other side according to embodiments of the present invention there is provided a kind of harassing and wrecking short message detecting system, including:The present invention
Above-mentioned any one embodiment server and at least one above-mentioned any one embodiment mobile terminal of the invention.
1st, a kind of detection method for harassing short message, applied to mobile terminal, including:
Cutting is carried out based on the merging file in memory to the short message of reception, at least one information characteristics is obtained;It is described
Merge file from server downloads to memory;
Described information feature is inputted into SMS classified model, with reference to file is merged, corresponding safety of the short message etc. is exported
Level.
2nd, the method according to 1, in addition to:Safe class based on output, will meet the short message of preset security grade
Server is sent to the relevant information of the short message.
3rd, the method according to 1 or 2, in addition to:Downloaded from server and merge file, the merging file is solved
Compression and decomposition, obtain dictionary library file and disaggregated model file.
4th, the method according to 3, the short message of described pair of reception carries out cutting based on the merging file in memory, obtains
At least one information characteristics, including:
Short message is received, the corresponding content of text of short message is obtained, the dictionary library file merged in memory in file is called, will
The content of text of the short message is matched with the Feature Words in the dictionary library file, determines including accordingly in the short message
Multiple information characteristics of Feature Words;
Wherein, the length of the Feature Words in the dictionary includes one or more characters, and the Feature Words are with aobvious
Write the word of distinction.
5th, the method according to 3 or 4, SMS classified model is inputted by described information feature, with reference to file is merged, is exported
The corresponding safe class of the short message, including:
By based on merging in the SMS classified model of disaggregated model file input that file is obtained, SMS classified model is carried out
Training, the SMS classified model trained;
It is described based on the SMS classified model output in the SMS classified model that will be trained described in information characteristics input
The corresponding safe class of short message.
6th, the method according to 5, it is described that the corresponding safe class of the short message is exported based on the SMS classified model,
Including:
The content of text of the short message is estimated according to described information feature, to obtain the content of text of the short message
Assessed value, corresponding safe class is worth to based on assessment.
7th, the method according to 6, is estimated according to described information feature to the content of text of the short message, including:
It is determined that the corresponding weight of each described information feature;
The content of text of the short message is estimated according to the weight of the determination and described information feature.
8th, a kind of detection method for harassing short message, applied to server, including:
Receive the file download request that mobile terminal is sent;
Dictionary library file and disaggregated model file are combined, and the dictionary library file after combination and disaggregated model is literary
Part, which is compressed, to be obtained merging file;
Merging file is issued to the mobile terminal for sending file download request.
9th, the method according to 8, the acquisition of the disaggregated model file, including:
Participle is carried out to the training data of collection based on the Feature Words in dictionary library file and feature selecting obtains participle spy
Levy data;
Participle characteristic is classified based on sorting algorithm, disaggregated model file is obtained.
10th, the method according to 8 or 9, in addition to:Receive the short message letter related to the short message that mobile terminal is sent
Breath, the short message and relevant information are stored in associated databases;The data include black list database, and/or white list
Database.
11st, a kind of mobile terminal, including:
Feature extraction unit, carries out cutting based on the merging file in memory for the short message to reception, obtains at least
One information characteristics;The merging file is from server downloads to memory;
Ranking unit, for described information feature to be inputted into SMS classified model, with reference to file is merged, output is described
The corresponding safe class of short message.
12nd, the mobile terminal according to 11, in addition to short message sending unit, will for the safe class based on output
The relevant information of the short message and the short message that meet preset security grade is sent to server.
13rd, the mobile terminal according to 11 or 12, in addition to:Download unit, merges file for being downloaded from server,
The merging file is decompressed and decomposed, dictionary library file and disaggregated model file is obtained.
14th, the mobile terminal according to 13, the feature extraction unit, specifically for receiving short message, obtains short message pair
The content of text answered, calls the dictionary library file merged in memory in file, by the content of text of the short message and institute's predicate
Feature Words in allusion quotation library file are matched, and determine the multiple information characteristics for including individual features word in the short message;
Wherein, the length of the Feature Words in the dictionary includes one or more characters, and the Feature Words are with aobvious
Write the word of distinction.
15th, the mobile terminal according to 13 or 14, the ranking unit, including:
Training module, for will be inputted based on the disaggregated model file for merging file acquisition in SMS classified model, to short
Letter disaggregated model is trained, the SMS classified model trained;
Prediction module, for will be divided based on the short message in the information characteristics input SMS classified model trained
Class model exports the corresponding safe class of the short message.
16th, the mobile terminal according to 15, the prediction module, specifically for according to described information feature to described short
The content of text of letter is estimated, and with the assessed value for the content of text for obtaining the short message, corresponding peace is worth to based on assessment
Congruent level.
17th, the mobile terminal according to 16, the prediction module is additionally operable to determine that each described information feature is corresponding
Weight;The content of text of the short message is estimated according to the weight of the determination and described information feature.
18th, a kind of server, including:
Request reception unit, the file download request for receiving mobile terminal transmission;
Piece file mergence unit, for dictionary library file and disaggregated model file to be combined, and by the dictionary after combination
Library file and disaggregated model file, which are compressed, to be obtained merging file;
File issuance unit, for will merge file be issued to send file download request the mobile terminal.
19th, the server according to 18, in addition to:Classification acquiring unit, for based on the feature in dictionary library file
Word carries out participle to the training data of collection and feature selecting obtains participle characteristic;Based on sorting algorithm to participle characteristic
According to being classified, disaggregated model file is obtained.
20th, the server according to 18 or 19, in addition to memory cell, for receive mobile terminal transmission short message and
The relevant information of the short message, the short message and relevant information are stored in associated databases;The data include blacklist number
According to storehouse, and/or white list database.
21st, a kind of detecting system for harassing short message, including:One server as described in 18 to 20 any one and at least
One mobile terminal as described in 11 to 17 any one.
The embodiment of each in this specification is described by the way of progressive, and what each embodiment was stressed is and it
Same or analogous part cross-reference between the difference of its embodiment, each embodiment.For system embodiment
For, because it is substantially corresponding with embodiment of the method, so description is fairly simple, referring to the portion of embodiment of the method in place of correlation
Defend oneself bright.
Methods and apparatus of the present invention may be achieved in many ways.For example, can by software, hardware, firmware or
Software, hardware, any combinations of firmware realize methods and apparatus of the present invention.The said sequence of the step of for methods described
Order described in detail above is not limited to merely to illustrate, the step of method of the invention, it is special unless otherwise
Do not mentionlet alone bright.In addition, in certain embodiments, the present invention can be also embodied as recording to program in the recording medium, these programs
Including the machine readable instructions for realizing the method according to the invention.Thus, the present invention also covering storage is used to perform basis
The recording medium of the program of the method for the present invention.
Description of the invention is provided for the sake of example and description, and is not exhaustively or by the present invention
It is limited to disclosed form.Many modifications and variations are obvious for the ordinary skill in the art.Select and retouch
State embodiment and be more preferably to illustrate the principle and practical application of the present invention, and one of ordinary skill in the art is managed
The solution present invention is so as to design the various embodiments with various modifications suitable for special-purpose.
Claims (10)
1. a kind of detection method for harassing short message, it is characterised in that applied to mobile terminal, including:
Cutting is carried out based on the merging file in memory to the short message of reception, at least one information characteristics is obtained;It is described to merge
File is from server downloads to memory;
Described information feature is inputted into SMS classified model, with reference to file is merged, the corresponding safe class of the short message is exported.
2. according to the method described in claim 1, it is characterised in that also include:Safe class based on output, it is default by meeting
The relevant information of the short message of safe class and the short message is sent to server.
3. method according to claim 1 or 2, it is characterised in that also include:Downloaded from server and merge file, to institute
State merging file to be decompressed and decomposed, obtain dictionary library file and disaggregated model file.
4. method according to claim 3, it is characterised in that the short message of described pair of reception is based on the merging text in memory
Part carries out cutting, obtains at least one information characteristics, including:
Short message is received, the corresponding content of text of short message is obtained, the dictionary library file merged in memory in file is called, will be described
The content of text of short message is matched with the Feature Words in the dictionary library file, determines to include individual features in the short message
Multiple information characteristics of word;
Wherein, the length of the Feature Words in the dictionary includes one or more characters, and the Feature Words are with notable area
Divide the word of property.
5. the method according to claim 3 or 4, it is characterised in that described information feature is inputted into SMS classified model, knot
Merge file, export the corresponding safe class of the short message, including:
By based on merging in the SMS classified model of disaggregated model file input that file is obtained, SMS classified model is instructed
Practice, the SMS classified model trained;
In the SMS classified model that will be trained described in information characteristics input, the short message is exported based on the SMS classified model
Corresponding safe class.
6. method according to claim 5, it is characterised in that described that the short message is exported based on the SMS classified model
Corresponding safe class, including:
The content of text of the short message is estimated according to described information feature, commented with the content of text for obtaining the short message
Valuation, corresponding safe class is worth to based on assessment.
7. a kind of detection method for harassing short message, it is characterised in that applied to server, including:
Receive the file download request that mobile terminal is sent;
Dictionary library file and disaggregated model file are combined, and the dictionary library file and disaggregated model file after combination are entered
Row compression obtains merging file;
Merging file is issued to the mobile terminal for sending file download request.
8. a kind of mobile terminal, it is characterised in that including:
Feature extraction unit, carries out cutting based on the merging file in memory for the short message to reception, obtains at least one
Information characteristics;The merging file is from server downloads to memory;
Ranking unit, for described information feature to be inputted into SMS classified model, with reference to file is merged, exports the short message
Corresponding safe class.
9. a kind of server, it is characterised in that including:
Request reception unit, the file download request for receiving mobile terminal transmission;
Piece file mergence unit, for dictionary library file and disaggregated model file to be combined, and by the dictionary library text after combination
Part and disaggregated model file, which are compressed, to be obtained merging file;
File issuance unit, for will merge file be issued to send file download request the mobile terminal.
10. a kind of detecting system for harassing short message, it is characterised in that including:One server as claimed in claim 9 and extremely
A few mobile terminal as claimed in claim 8.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710476918.7A CN107135314A (en) | 2017-06-21 | 2017-06-21 | Harass detection method, system, mobile terminal and the server of short message |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710476918.7A CN107135314A (en) | 2017-06-21 | 2017-06-21 | Harass detection method, system, mobile terminal and the server of short message |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107135314A true CN107135314A (en) | 2017-09-05 |
Family
ID=59735419
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710476918.7A Pending CN107135314A (en) | 2017-06-21 | 2017-06-21 | Harass detection method, system, mobile terminal and the server of short message |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107135314A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108898167A (en) * | 2018-06-14 | 2018-11-27 | 北京奇虎科技有限公司 | It breaks one's promise the display methods and device of number |
CN109462582A (en) * | 2018-10-30 | 2019-03-12 | 腾讯科技(深圳)有限公司 | Text recognition method, device, server and storage medium |
CN109544904A (en) * | 2018-12-29 | 2019-03-29 | 上海易点时空网络有限公司 | For information processing method and device, server violating the regulations |
CN111563495A (en) * | 2020-05-09 | 2020-08-21 | 北京奇艺世纪科技有限公司 | Method and device for recognizing characters in image and electronic equipment |
CN111597805A (en) * | 2020-05-21 | 2020-08-28 | 上海创蓝文化传播有限公司 | Method and device for auditing short message text links based on deep learning |
CN112492606A (en) * | 2020-11-10 | 2021-03-12 | 恒安嘉新(北京)科技股份公司 | Classification and identification method and device for spam messages, computer equipment and storage medium |
CN113037715A (en) * | 2017-12-08 | 2021-06-25 | 谷歌有限责任公司 | Transmission system and method for limiting manipulation of content in a network environment |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103813279A (en) * | 2012-11-14 | 2014-05-21 | ***通信集团设计院有限公司 | Junk short message detecting method and device |
CN105138611A (en) * | 2015-08-07 | 2015-12-09 | 北京奇虎科技有限公司 | Short message type identification method and device |
CN105516499A (en) * | 2015-12-14 | 2016-04-20 | 北京奇虎科技有限公司 | Method and device for classifying short messages, communication terminal and server |
CN106878347A (en) * | 2017-04-28 | 2017-06-20 | 北京奇虎科技有限公司 | Information processing method, system, mobile terminal and server |
-
2017
- 2017-06-21 CN CN201710476918.7A patent/CN107135314A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103813279A (en) * | 2012-11-14 | 2014-05-21 | ***通信集团设计院有限公司 | Junk short message detecting method and device |
CN105138611A (en) * | 2015-08-07 | 2015-12-09 | 北京奇虎科技有限公司 | Short message type identification method and device |
CN105516499A (en) * | 2015-12-14 | 2016-04-20 | 北京奇虎科技有限公司 | Method and device for classifying short messages, communication terminal and server |
CN106878347A (en) * | 2017-04-28 | 2017-06-20 | 北京奇虎科技有限公司 | Information processing method, system, mobile terminal and server |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113037715A (en) * | 2017-12-08 | 2021-06-25 | 谷歌有限责任公司 | Transmission system and method for limiting manipulation of content in a network environment |
CN108898167A (en) * | 2018-06-14 | 2018-11-27 | 北京奇虎科技有限公司 | It breaks one's promise the display methods and device of number |
CN109462582A (en) * | 2018-10-30 | 2019-03-12 | 腾讯科技(深圳)有限公司 | Text recognition method, device, server and storage medium |
CN109462582B (en) * | 2018-10-30 | 2020-11-20 | 腾讯科技(深圳)有限公司 | Text recognition method, text recognition device, server and storage medium |
CN109544904A (en) * | 2018-12-29 | 2019-03-29 | 上海易点时空网络有限公司 | For information processing method and device, server violating the regulations |
CN111563495A (en) * | 2020-05-09 | 2020-08-21 | 北京奇艺世纪科技有限公司 | Method and device for recognizing characters in image and electronic equipment |
CN111563495B (en) * | 2020-05-09 | 2023-10-27 | 北京奇艺世纪科技有限公司 | Method and device for recognizing characters in image and electronic equipment |
CN111597805A (en) * | 2020-05-21 | 2020-08-28 | 上海创蓝文化传播有限公司 | Method and device for auditing short message text links based on deep learning |
CN111597805B (en) * | 2020-05-21 | 2021-01-05 | 上海创蓝文化传播有限公司 | Method and device for auditing short message text links based on deep learning |
CN112492606A (en) * | 2020-11-10 | 2021-03-12 | 恒安嘉新(北京)科技股份公司 | Classification and identification method and device for spam messages, computer equipment and storage medium |
CN112492606B (en) * | 2020-11-10 | 2024-05-17 | 恒安嘉新(北京)科技股份公司 | Classification recognition method and device for spam messages, computer equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107135314A (en) | Harass detection method, system, mobile terminal and the server of short message | |
CN103279868B (en) | A kind of method and apparatus of automatic identification swindle order | |
CN104408093B (en) | A kind of media event key element abstracting method and device | |
CN108595519A (en) | Focus incident sorting technique, device and storage medium | |
CN106453061B (en) | A kind of method and system identifying network fraudulent act | |
CN105512938A (en) | Online credit risk assessment method based on long-term using behavior of user | |
CN107547555A (en) | A kind of web portal security monitoring method and device | |
CN107566358A (en) | A kind of Risk-warning reminding method, device, medium and equipment | |
CN106713579B (en) | Telephone number identification method and device | |
CN108134784A (en) | web page classification method and device, storage medium and electronic equipment | |
CN106296195A (en) | A kind of Risk Identification Method and device | |
CN109214914A (en) | A kind of loan information checking method and device based on communication open platform | |
CN110598070B (en) | Application type identification method and device, server and storage medium | |
CN104156490A (en) | Method and device for detecting suspicious fishing webpage based on character recognition | |
CN106529773A (en) | Online credit and fraud risk evaluation method based on identifying code type question answering | |
CN111078880B (en) | Sub-application risk identification method and device | |
CN106878347A (en) | Information processing method, system, mobile terminal and server | |
CN102567534B (en) | Interactive product user generated content intercepting system and intercepting method for the same | |
CN107341399A (en) | Assess the method and device of code file security | |
CN106096439A (en) | A kind of intimacy protection system obscured based on mobile user profile and method | |
CN109978020A (en) | A kind of social networks account vest identity identification method based on multidimensional characteristic | |
CN113011889A (en) | Account abnormity identification method, system, device, equipment and medium | |
CN111782793A (en) | Intelligent customer service processing method, system and equipment | |
CN109947944A (en) | Short message display method, device and storage medium | |
CN115577172A (en) | Article recommendation method, device, equipment and 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: 20170905 |
|
RJ01 | Rejection of invention patent application after publication |