CN103345530B - A kind of social networks blacklist automatic fitration model based on semantic net - Google Patents
A kind of social networks blacklist automatic fitration model based on semantic net Download PDFInfo
- Publication number
- CN103345530B CN103345530B CN201310318042.5A CN201310318042A CN103345530B CN 103345530 B CN103345530 B CN 103345530B CN 201310318042 A CN201310318042 A CN 201310318042A CN 103345530 B CN103345530 B CN 103345530B
- Authority
- CN
- China
- Prior art keywords
- information
- malice
- value
- fallacious message
- user
- 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.)
- Expired - Fee Related
Links
- 230000005540 biological transmission Effects 0.000 claims abstract description 3
- 230000003993 interaction Effects 0.000 claims abstract description 3
- 238000000034 method Methods 0.000 claims description 10
- 238000000605 extraction Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 6
- 230000000873 masking effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 2
- 239000000203 mixture Substances 0.000 claims description 2
- 238000003012 network analysis Methods 0.000 claims description 2
- 230000002452 interceptive effect Effects 0.000 description 4
- 239000000284 extract Substances 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 238000012098 association analyses Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
Landscapes
- Information Transfer Between Computers (AREA)
Abstract
The invention discloses a kind of social networks blacklist automatic fitration model based on semantic net, the malice junk information of the model can be labeled as junk information by automatic shield, ability visible this type of information when only user actively checks；Fallacious message publisher can be also automatically flagged as in the blacklist of information receiver, shield the fallacious message of its transmission；The model is also provided with User Defined mask information, and this type of information is marked as fallacious message weights highest, that is the least patient information of user, once it is judged to this type of information, then the probability of sender and recipient's friend relation back to normal is with regard to very little, the characteristics of this just inherits traditional blacklist, on this basis, the model also has the relatively low info class of some malice weights, when being judged as this type of information, if recipient has actively carried out data interaction with sender, the user for being then judged as malice publisher is then possible to recover blacklist by system, the two is set to turn into normal relation.
Description
Technical field
The present invention relates to information security and Computer Applied Technology field, more particularly to a kind of social network based on semantic net
Network blacklist automatic fitration model.
Technical background
Social networks（SNS）By promoting on the line between people exchange and information sharing, have become a kind of main
Applied on line, obtain very huge customer group quantity.At the same time, in order to provide good Consumer's Experience to so a large amount of
User, the function of social networks itself also becomes increasingly to tend to complicated, interactive information text message single since most,
Have turned to more digital contents, such as video.This to social networks security and privacy protection propose new challenge.
According to investigations, some criminals utilize the opening and the information of exposure during user's use of social networks, such as
Interactive information of photo and friend etc., by targetedly propagating fraud information, the safety to social network user is caused
Serious to threaten, social networks has become a kind of main path that wealth is invaded in criminal's fraud.Social networks mistake is used in user
Cheng Zhong, can more or less receive the fallacious message of some criminals, and to some obvious fraud informations, user has the ability certainly
Row judges, but to some through camouflage it is hidden, show unconspicuous information, then its directly endanger people the person,
Property safety.Although some social networks, such as Renren Network have provided the user the function of some controls of authority, these controls
Function processed is all fairly simple, underaction.
In recent years, with the development of semantic net, increasing body is used to set up the representation of knowledge in specific area
Model, but it is huge due to knowledge knowledge network, also there are many white spaces and be not related to, particularly with current emerging application
Security fields, semantic network technology is also seldom employed.
By analysis, dynamic, personalized using semantic net itself the, characteristic such as take the initiative in offering a hand, is especially suitable for solving to work as
It is preceding it is proposed that the problem of.And the present invention the problem of can solve above well.
The content of the invention
Present invention aims at the problem of above-mentioned prior art is present is solved, there is provided a kind of social activity based on semantic net
Network blacklist automatic fitration model.The model is extracted and the scheme of integrated management and its related based on potential fallacious message
What supporting mechanism was realized.
The technical solution adopted for the present invention to solve the technical problems is：The present invention is to existing social network ontologies model
On the basis of improved, devise a kind of interactive information to user in social networks and carry out analysis judgement, extract possible crime
The automatic fitration model of membership credentials.The model is divided into information analysis extraction module, information malice index scoring module, score knot
Fruit blacklist relating module, it specifically includes as follows：
First, information analysis extraction module
The present invention proposes a kind of information analysis extraction module, it can be seen that in information stream part, will believe in the module
Breath is divided for two parts, and a portion is normal, harmless information, and another part is fallacious message
(Viciousness), according to the analysis to fallacious message feature, now it is divided into 4 classes by fallacious message is abstract, is respectively：
● meaningless information (meaningless)：This category information is mainly the alphabetical additional character of shape such as random alignment, mistake
Filter those and be likely to be the harassing and wrecking content that machine is issued at random, if malice publisher is taken a walk letter in large quantities using issue machine
Breath, not only brings flow pressure to social networks itself, invisible puzzlement is also brought to user.
● fraud information (fraud)：Fraud information is mainly what is extracted in the typical case counted by public security department
Some are accused of the feature sentence of fraud, such as inquiry passport NO., home address etc..
● violence information (violence)：Violence information refer to the sentence that social safety is endangered comprising vocabulary is abused and
It is accused of too drastic speech of spirit harassing and wrecking etc..Filtering this category information can avoid causing dyadic conflict, be that social networks builds one
Good " ecological environment ".
● customized information (user defined)：Customized information can allow user individual to set
The content for the content received, such as maskable advertisement and non-malicious push of oneself being unwilling so that middleware mentioned above
Function is more perfect.
In summary, the present invention proposes fallacious message ontology model, and the model is the extension to social networks model, is
Clearer description model, introduces the concept of sensitive semantic base, sensitive semantic base is exactly that 4 described above are big here
The set of category information composition, when information, which is detected, meets the rule in the semantic base, is then classified as in fallacious message
A certain class, namely fallacious message.Even if but be now classified as fallacious message, this would not be received by also not representing user
Information, the judge simply rough division to the information of this step, by the analyze data stream both sides always information record in later stage,
It can just further determine whether to shield this information completely.
2nd, information malice index scoring module
The method for the identification fallacious message that the present invention is provided, has considered not only the malice degree of information itself, has also combined
Send the malice record of the user of fallacious message.For clearer this process of description, the present invention proposes following general
Read：
The malice factor：Represented with fa, represent the malice degree for the infobit that user sends.The value of the malice factor such as table
Shown in 1：
Table 1
The above-mentioned value of table 1 is a reference, and such value, which is mainly, makes whole system more flexible, and a user issued
Fallacious message, but this user is not represented with that can not possibly constitute friend relation after information receiver, routine information value is
0.99, so it is multiplied every time with malice record index, malice record index is less than 1, if both sides repeatedly carry out conventional interaction,
Then malice record index is consistently less than 1.If system detectio is to fallacious message, according to upper table rule, fa values are according to fallacious message
Malice degree, record exponent arithmetic with existing malice, it is in increase tendency to make malice record index, such as current malice record refers to
Number is 0.5, as long as then sender retransmits a fallacious message, such as violence information, so that it may so that malice index is more than 1, then had
Safe range may be crossed, as malice sender.
Malice record index：R, S, V represent recipient, sender, malice record index respectively, then triple can be used
（R,S,V）Identify a data and sender, the relation of recipient and determine that can this data stream connect from sender's flow direction
Receipts person.The initial value of malice record index is 1, represents that both sides did not had data exchange also.Malice record index is smaller also indirect
The level of intimate of both sides is reflected, then the probability for constituting malice relation is also just smaller.
Malice records the safe range of index：This scope can determine according to system to the acceptance level of fallacious message, if
Put a value more than 1.View of the above, it will be seen that routine information makes malice record index tend to 0, fallacious message makes evil
Meaning record index tends to be infinitely great, makes it away from 1, if the information flow of both sides includes both the above, malice record index
All the time swung around 1, system will not be such that sender is in forever in blacklist, and the value of routine information directly influences transmission
Person departs from the difficulty of blacklist.
In summary, if regarding whole process as a function, the usable sentence described function formalized as follows
Input and output valve.It is user's set in social networks to define Usns first, and T represents the safe threshold that system is set, S, R, V
It is defined as above, then what this process can be vivider is expressed as S->R=V, namely sender, the current malice record product of recipient
Tired value is V, it is possible to use P (S, R)=V represents that the domain of definition of input value is represented by here：I={r:R,s:S|r,s∈Usns,
R ≠ s }, output valve can then be divided into virtual value and masking value scope, and valid value range is represented by：Z={v:V|0<v<T }, shielding
The scope of value is thenZ.
Following predicate logic can be used to represent for the condition of system mask：If
F(r,s):R and s is friend relation
C(x):The Msg that r is sent includes fallacious message
V(r,s,x):The information this time sent and the calculated value of malice record index are more than system safe threshold.
B(r,s):System mask s->R information.
U(r,s):System update V (r, s, x) value.
Then have：
If the fa values of routine information and fallacious message are respectively far, fav, active user a, b malice record exponential representation
For (a, b, base), then sender will make that oneself information is not needed then by system mask and recipient carries out n=-logfarBase times
Routine information is interacted.Here is the reasonability that one group of data represents data above, if current base values are that 20, far values are 0.99,
Then sender needs to interact with recipient's 300 effective routine informations of progress, and also even user has only issued 4 times and connect at the beginning
The information of receipts person's filtering, just makes fallacious message index quite big.
3rd, score result blacklist relating module
Need exist for introducing the concept of a safe threshold, this threshold values needs to rely on number present in existing social networks
According to by statistics, calculating obtains a rational value, this value not immutable, fixed value, but is answered according to different
Demand, is rational as long as maintaining within the scope of one.
By the judged result of network analysis previous step, if the safe threshold that information is set beyond system, by information
Be filtered into junk information, and the user of this information will be sent and add in the blacklist of recipient, the next sender again to
When secondary recipient sends information, information can be junk information by automatic fitration, and recipient can only look into junk information Xiang Zhongcai
See this type of information.
The information that only shielding person user is actively checked in rubbish, with being interacted certain normal information amount by shielding person
Afterwards, shielded this to be possible to be returned to outside blacklist by system, system will be considered that the two has recovered friend relation.
Beneficial effect：
1st, the present invention can carry out judge score to the malice degree of information flow, and automatic fitration exceeds the information of safe range
Stream, and related information sender and recipient, its is constituted potential isolated relation.
2nd, the present invention can make system automatically terminate two in the user in blacklist by improving the relation with recipient
Person's shielding relation.
3rd, The present invention gives a set of method more reasonably scored to fallacious message.
Brief description of the drawings
Fig. 1 is social network ontologies illustraton of model.
Fig. 2 is fallacious message ontology model figure.
Fig. 3 is system framework figure.
Embodiment
The invention is described in further detail below in conjunction with Figure of description.
Whole system framework is based on an existing social networks open platform, namely the SNS in Fig. 3
Infrastructure, the optional middleware that main functionality of the present invention belongs in this network platform, in this
Between part by the sensitive semantic base of maintenance management, information malice degree scoring rule, be responsible for customer relationship and interacted with information flow
Between association analysis.The embodiment of each part is just described from the following aspects here.
1st, the interception classification of information
The exchange of substantial amounts of data is there is in social networks, the data interception stream between certain two data terminal can join
The file classification method using Naive Bayes Classifier is examined, the rule defined with reference to sensitive semantic base is carried out to data content
Classification.When being detected as normal flow, then it is cleared and passes through, and then by information persistence and recipient can be pointed out this information,
If not passing through, malice exponential integral module is sent to, score computing is carried out to this information, with reference to（Recipient, sends
Person, accumulates malice exponential quantity）Triple, the malice fraction with current information judges that can fallacious message be intercepted by system, if
Malice degree is not obvious, does not reach the safe threshold of system setting, then this information of letting pass.Filter criteria, refers to fire wall
Filter type, be divided into two kinds, Yi Zhongwei：Every information being located in sensitive semantic base is shielded without exception, it is another for it is every not
The information defined by information bank is all shielded, and the system uses the first, then needs the support of a sensitive semantic base.
2nd, sensitive semantic base
User needs to go out a whole set of letter according to the frequency of occurrences analysis and summary of the malicious data of daily flowing in website
Storehouse is ceased, with reference to classifying rules proposed by the present invention, these information is loaded into sensitive semantic base, used for system.Sensitive semantic base
Structure directly affect the significant degree of whole system, if the information content that sensitive semantic base is defined is few, or not comprehensively, then
The use practical significance of system is not then obvious, so needing the special expert for field of social network to put forward information
Take, be rule of thumb put in storage classification information.Similar with the domain expert of semantic net, the assistance that domain expert is needed also exist for here is determined
Justice.
3rd, information malice index scoring module
Present system is not merely simply to be compared the information extracted with the key vocabularies in sensitive semantic base
Compared with can also utilize scoring module, carry out comprehensive analysis to the information that extracts, provide an analysis result, determine to believe for system
The relation of the user at breath stream two ends.System is responsible for safeguarding one（Information transmitter, information receiver, fallacious message accumulation index）
Triple data, a finger of flag information sender and information receiver always fallacious message present in interactive information
The real relation of number, indirectly the reflection end subscriber of information flow two.Information malice index scoring module will inquire this in score
Triple data, and the different malice weights set according to various fallacious message systems, by the score of obtained new data
Value, with the value computing, updates triple data, and returns to score result, judges for system.
4th, blacklist management module
Blacklist management module is the result set of whole system analysis, and the module can be responsible for the operation pipe of lasting layer data
Reason.Oracle database can be used, facilitates the use of system.System can provide one for each user in social networks
There are three kinds of relations between any two user in customer relationship in black list user's table, mark social networks, network：
Friend relation（isFriendOf）, strange relation（notFriendOf）, and shielding relation（isInBlacklistOf）, the system
Concern is primarily with the 3rd shielding relation, namely blacklist.
Claims (3)
1. a kind of social networks blacklist automatic filtering unit based on semantic net, it is characterised in that：Described device is divided into information
Analyze extraction module, information malice index scoring module, score result blacklist relating module；Information analysis extraction module will be believed
It is two parts to cease flow point, and a part is normal, harmless information；Another part is fallacious message；
Described information analyzes extraction module according to the analysis to fallacious message feature, is divided into 4 classes by fallacious message is abstract, respectively
For：
1) meaningless information (meaningless)：This category information refers to the alphabetical additional character of shape such as random alignment, filters those
It is likely to be the harassing and wrecking content that machine is issued at random；
2) fraud information (fraud)：Fraud information is mainly some extracted in the typical case counted by public security department
It is accused of the feature sentence of fraud, including inquiry passport NO., home address；
3) violence information (violence)：Violence information refers to, comprising vocabulary is abused, endanger the sentence of social safety and be accused of
The too drastic speech of spirit harassing and wrecking, filters this category information and avoids causing dyadic conflict, is that social networks builds a good " ecology
Environment "；
4) customized information (user defined)：Customized information sets some oneself to be reluctant by user individual
The content that the content received, i.e. shielding advertisement and the non-malicious of anticipating are pushed；
Fallacious message ontology model is the extension to social networks model, for clearer description model, is introduced here
The concept of sensitive semantic base, sensitive semantic base is exactly the set of the big category information composition in 4 described above, when information is detected
When meeting the rule in the semantic base, then a certain class in fallacious message, namely fallacious message are classified as, even if but now
Fallacious message is classified as, this information would not be received by also not representing user, and the judge of this step is simply to the information
It is rough to divide, by the analyze data stream both sides always information record in later stage, it can just further determine whether to shield this completely
Information；
Described information malice index scoring module includes the method for the identification fallacious message provided, has considered not only information itself
Malice degree, also combine send fallacious message user malice record, in order to it is clearer description this process, it is proposed that
Following concept, including：
The malice factor：Represented with fa, represent the malice degree for the infobit that user sends；The value of the malice factor such as institute of table 1
Show：
Table 1
One user issued fallacious message, but did not represented this user with that can not possibly constitute good friend pass after information receiver
System, routine information value is 0.99, is so multiplied every time with malice record index, malice record index is less than 1, if both sides
Conventional interaction is repeatedly carried out, then malice record index is consistently less than 1；If system detectio is to fallacious message, according to upper table rule,
Fa values record exponent arithmetic according to the malice degree of fallacious message with existing malice, and it is in increase tendency to make malice record index,
It is 0.5 as current malice records index, as long as then sender retransmits a fallacious message, allows for malice index more than 1, then
Safe range is crossed, as malice sender；
Malice record index：R, S, V represent recipient, sender, malice record index respectively, then using triple (R, S, V)
Identify a data and sender, the relation of recipient and determine that can this data stream flow to recipient from sender, dislike
The initial value of meaning record index is 1, represents that both sides did not had data exchange also；Malice record index is smaller also reflect indirectly it is double
The level of intimate of side, the then probability for constituting malice relation is also just smaller；
Malice records the safe range of index：This scope is determined according to system to the acceptance level of fallacious message, sets one
Value more than 1；Routine information makes malice record index tend to 0, and fallacious message makes malice record index tend to be infinitely great, makes it
Away from 1, if the information flow of both sides includes both the above, malice record index is swung around 1 all the time, and system will not make transmission
Person is in blacklist forever, and the value of routine information directly influences the difficulty that sender departs from blacklist；
In summary, regard whole process as a function, use the input and output of the sentence described function formalized as follows
Value；It is user's set in social networks to define Usns first, and T represents the safe threshold that system is set, and S, R, V definition is same
On, then what this process was vivider is expressed as S->R=V, namely sender, the current malice record accumulating value of recipient is V,
Also represent that the domain of definition of input value is expressed as here using P (S, R)=V：I={ r:R,s:S | r, s ∈ Usns, r ≠ s }, output
Value is then divided into virtual value and masking value scope, and valid value range is expressed as：Z={ v:V|0<v<T }, the scope of masking value then for-
Z；
The condition of system mask is represented using following predicate logic：If：
F(r,s):R and s is friend relation；
C(x):The Msg that r is sent includes fallacious message；
V(r,s,x):The information this time sent and the calculated value of malice record index are more than system safe threshold；
B(r,s):System mask s->R information；
U(r,s):System update V (r, s, x) value；
Then have：
If the fa values of routine information and fallacious message are respectively far, fav, active user a, b malice record exponential representation are
(a, b, base), then sender to make oneself information do not needed then by system mask and recipient carry out n=-logfarBase times
Routine information is interacted；
The score result blacklist relating module introduces the concept of a safe threshold, and this threshold values needs to rely on existing social activity
Data present in network, by statistics, calculating obtains a rational value, this value not immutable, fixed value,
But be rational as long as maintaining within the scope of one according to the demand of different application；
By the judged result of network analysis previous step, if the safe threshold that information is set beyond system, by information filtering
For junk information, and the user of this information will be sent add in the blacklist of recipient, the next sender connects to this again
When receipts person sends information, information can be junk information by automatic fitration, and recipient only views this in junk information Xiang Zhongcai
Category information；
The information that only shielding person user is actively checked in rubbish, is interacted after certain normal information amount, quilt with by shielding person
Shielding person could be returned to outside blacklist by system, and system will be considered that the two has recovered friend relation.
2. a kind of social networks blacklist automatic filtering unit based on semantic net according to claim 1, its feature exists
In：The every same shielding being located in sensitive semantic base of described information stream.
3. a kind of social networks blacklist automatic filtering unit based on semantic net according to claim 1, its feature exists
In：Described information malice index scoring module will inquire this triple data in score, and according to various fallacious messages
The different malice weights that system is set, by the scoring value of obtained new data, with the value computing, update triple data,
And return to score result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310318042.5A CN103345530B (en) | 2013-07-25 | 2013-07-25 | A kind of social networks blacklist automatic fitration model based on semantic net |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310318042.5A CN103345530B (en) | 2013-07-25 | 2013-07-25 | A kind of social networks blacklist automatic fitration model based on semantic net |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103345530A CN103345530A (en) | 2013-10-09 |
| CN103345530B true CN103345530B (en) | 2017-07-14 |
Family
ID=49280325
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310318042.5A Expired - Fee Related CN103345530B (en) | 2013-07-25 | 2013-07-25 | A kind of social networks blacklist automatic fitration model based on semantic net |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103345530B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103813020A (en) * | 2014-02-28 | 2014-05-21 | 联想(北京)有限公司 | Information processing method and electronic device |
| CN105491542B (en) * | 2015-12-01 | 2018-12-18 | 小米科技有限责任公司 | number identification method and device |
| CN105450510B (en) * | 2015-12-18 | 2019-03-01 | 小米科技有限责任公司 | Friend management method, device and server for social network-i i-platform |
| CN106095755B (en) * | 2016-06-12 | 2019-02-26 | 北京师范大学 | A kind of fake monitoring and method for early warning based on semantic temporal figure |
| CN106649672B (en) * | 2016-12-14 | 2020-08-11 | 华南师范大学 | Secure data semantic analysis method and system based on semantic network |
| US11575715B2 (en) | 2019-10-28 | 2023-02-07 | International Business Machines Corporation | Dynamically customized cognitive security filter |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6539430B1 (en) * | 1997-03-25 | 2003-03-25 | Symantec Corporation | System and method for filtering data received by a computer system |
| CN101114907B (en) * | 2006-07-28 | 2010-08-11 | 腾讯科技（深圳）有限公司 | Method and system for managing and filtering black list |
| CN101996203A (en) * | 2009-08-13 | 2011-03-30 | 阿里巴巴集团控股有限公司 | Web information filtering method and system |
| CN102208992B (en) * | 2010-06-13 | 2015-09-02 | 天津海量信息技术有限公司 | The malicious information filtering system of Internet and method thereof |
| CN102567340B (en) * | 2010-12-09 | 2016-02-24 | 腾讯科技（深圳）有限公司 | A kind of method and device filtering micro-blog information |
| CN102098332B (en) * | 2010-12-30 | 2014-04-16 | 北京新媒传信科技有限公司 | Method and device for examining and verifying contents |
| CN102880636A (en) * | 2012-08-03 | 2013-01-16 | 深圳证券信息有限公司 | Bad information detection method and server |
-
2013
- 2013-07-25 CN CN201310318042.5A patent/CN103345530B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN103345530A (en) | 2013-10-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103345530B (en) | A kind of social networks blacklist automatic fitration model based on semantic net | |
| Resende et al. | Analyzing textual (mis) information shared in WhatsApp groups | |
| CN106570144B (en) | The method and apparatus of recommendation information | |
| Heidari et al. | Bert model for fake news detection based on social bot activities in the covid-19 pandemic | |
| Verma et al. | Semantic feature selection for text with application to phishing email detection | |
| CN101784022A (en) | Method and system for filtering and classifying short messages | |
| CN103853841A (en) | Method for analyzing abnormal behavior of user in social networking site | |
| CN103037339A (en) | Short message filtering method based on user creditworthiness and short message spam degree | |
| CN103793503A (en) | Opinion mining and classification method based on web texts | |
| CN108765179A (en) | A kind of credible social networks analysis method calculated based on figure | |
| Xu et al. | Characterizing the roles of bots on Twitter during the COVID-19 infodemic | |
| Apte et al. | Frauds in online social networks: A review | |
| Bera et al. | Towards a thematic dimensional framework of online fraud: An exploration of fraudulent email attack tactics and intentions | |
| Dutta et al. | A system for intergroup prejudice detection: The case of microblogging under terrorist attacks | |
| Karimi et al. | Automated detection of doxing on twitter | |
| Takuro et al. | Codewords detection in microblogs focusing on differences in word use between two corpora | |
| Desai | Social media, misinformation and Covid-19 | |
| Kikerpill | Crime-as-communication: detecting diagnostically useful information from the content and context of social engineering attacks | |
| Cristani et al. | The Spider-man Behavior Protocol: Exploring Both Public and Dark Social Networks for Fake Identity Detection in Terrorism Informatics. | |
| Kim et al. | Scam detection assistant: Automated protection from scammers | |
| Rozo et al. | Cyber democracy in the digital age: Characterizing hate networks in the 2022 US midterm elections | |
| Kågström | Improving naive bayesian spam filtering | |
| Wright | Towards a response to epistemic nihilism | |
| Zhu et al. | Building of safer urban hubs: Insights from a comparative study on cyber telecom scams and early warning design | |
| Cao et al. | Ontology-based spam detection filtering system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170714 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |