CN109840543A - A kind of data monitoring and method for early warning based on neural network and sensitive information stream - Google Patents
A kind of data monitoring and method for early warning based on neural network and sensitive information stream Download PDFInfo
- Publication number
- CN109840543A CN109840543A CN201811536697.9A CN201811536697A CN109840543A CN 109840543 A CN109840543 A CN 109840543A CN 201811536697 A CN201811536697 A CN 201811536697A CN 109840543 A CN109840543 A CN 109840543A
- Authority
- CN
- China
- Prior art keywords
- sensitive information
- sensitive
- information
- neural network
- early warning
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a kind of data monitoring and method for early warning based on neural network and sensitive information stream, comprising: step 1, the susceptibility of information is divided, determines information sensing grade, sensitive information is obtained based on information sensing grade;Step 2, the monitoring on time domain and airspace is carried out to the flow direction of sensitive information, monitored results is carried out vectorization processing and obtain sensitive information to flow to vector;Step 3, disaggregated model is flowed to based on sensitive information vector is flowed to sensitive information and classified, identify abnormal flow direction.The present invention can be used for the monitoring to sensitive data, realize the detection and early warning flowed to extremely to sensitive data, establish solid foundation to construct safe and reliable data use environment.
Description
Technical field
The invention belongs to field of information security technology more particularly to a kind of data based on neural network and sensitive information stream
Monitoring and method for early warning.
Background technique
Monitoring to sensitive information outflow is to one of the important means of sensitive information protection.It is sensitive from macroscopically, holding
When information flows out, where is flowed to, and more intuitive, more easily to data operation maintenance personnel can bring decision-making foundation.
Sensitive information outflow detection early warning extremely is one of current data O&M, data protection urgent problem.It is right
The monitoring of sensitive information outflow is at present there are mainly two types of mode, one is the monitoring of the monitoring formula of implant data platform, one is
Bypass type monitoring except platform.The monitoring of monitoring formula can in real time, quickly provide data and spill out to but needing to be embedded in flat
Among platform, certain risk can be brought.Bypass type monitoring with will not impact to platform, but is needed except platform
Mirror image is done to access data, there is the overhead in performance and storage.
Summary of the invention
The object of the present invention is to provide a kind of data monitoring and method for early warning based on neural network and sensitive information stream, use
In data safety and data monitoring field, the omnibearing protection to sensitive data is realized.
The present invention provides a kind of data monitoring and method for early warning based on neural network and sensitive information stream, comprising:
Step 1, the susceptibility of information is divided, determines information sensing grade, obtained based on information sensing grade quick
Feel information;
Step 2, the monitoring on time domain and airspace is carried out to the flow direction of sensitive information, vectorization processing is carried out to monitored results
And it obtains sensitive information and flows to vector;
Step 3, disaggregated model is flowed to based on sensitive information vector is flowed to sensitive information and classified, identify exception stream
To.
Further, the step 1 includes:
The information sensing grade is divided into insensitive, general sensitive, sensitive and very sensitive four grades, and will be general
Sensitive, sensitive and very sensitive information is determined as sensitive information.
Further, carrying out time domain with the monitoring on airspace to the flow direction of sensitive information described in step 2 includes:
Monitor the sensitive information sensitive grade, sensitive information outflow time, sensitive information where flowed to.
Further, it is two layers neural network model that sensitive information described in step 3, which flows to disaggregated model, every layer
There are 64 concealed nodes, uses logistic regression as classification function.
Further, the step 3 includes:
Before flowing to disaggregated model using the sensitive information, disaggregated model is flowed to the sensitive information and is trained.
According to the above aspect of the present invention, can be used for by data monitoring and method for early warning based on neural network and sensitive information stream
The detection and early warning flowed to extremely to sensitive data is realized in monitoring to sensitive data, is used to construct safe and reliable data
Environment has established solid foundation.
The above description is only an overview of the technical scheme of the present invention, in order to better understand the technical means of the present invention,
And can be implemented in accordance with the contents of the specification, the following is a detailed description of the preferred embodiments of the present invention and the accompanying drawings.
Detailed description of the invention
Fig. 1 is a kind of overall flow of data monitoring and method for early warning based on neural network and sensitive information stream of the present invention
Figure;
Fig. 2 is a kind of vectorization stream of data monitoring and method for early warning based on neural network and sensitive information stream of the present invention
Cheng Tu;
Fig. 3 is a kind of sensitive information of data monitoring and method for early warning based on neural network and sensitive information stream of the present invention
Flow to disaggregated model structure chart.
Specific embodiment
With reference to the accompanying drawings and examples, specific embodiments of the present invention will be described in further detail.Implement below
Example is not intended to limit the scope of the invention for illustrating the present invention.
Present embodiments provide a kind of data monitoring and method for early warning based on neural network and sensitive information stream, comprising:
Step 1, the susceptibility of information is divided, determines information sensing grade, obtained based on information sensing grade quick
Feel information;
Step 2, the monitoring on time domain and airspace is carried out to the flow direction of sensitive information, vectorization processing is carried out to monitored results
And it obtains sensitive information and flows to vector;
Step 3, disaggregated model is flowed to based on sensitive information vector is flowed to sensitive information and classified, identify exception stream
To.
The data monitoring and method for early warning based on neural network and sensitive information stream, can be used for the prison to sensitive data
The detection and early warning flowed to extremely to sensitive data is realized in control, has been established for the safe and reliable data use environment of building solid
Basis.
Invention is further described in detail below.
As shown in Figure 1, Figure 2, Figure 3 shows, this method includes dividing to the susceptibility of information, determines information sensing grade,
Monitoring on time domain and airspace is carried out to the flow direction of sensitive information, vectorization processing is carried out to monitored results and obtains sensitive information
Vector is flowed to, disaggregated model is flowed to using sensitive information vector is flowed to sensitive information and classify, identify abnormal flow direction.
Information sensitivity divides, and is the measurement standard of information sensing, be divided into it is insensitive, general it is sensitive, sensitive with very
Sensitive four grades, are indicated with digital 0,1,2,3 respectively, wherein susceptibility is general sensitive, sensitive and very sensitive information
It is referred to as sensitive information.The criteria for classifying of susceptibility can voluntarily be divided according to service conditions.
Monitoring on time domain and airspace is carried out to the flow direction of sensitive information, is the outflow to sensitive information, carry out the time with
Spatially two-dimensional monitoring, sensitive grade, the time of sensitive information outflow including sensitive information, sensitive information flow to where.
In one embodiment, vectorization processing is carried out to monitored results to include the following steps:
(1) the 34 full zero row vector of dimension of note is initialization vector;
(2) it in initialization vector, the 0th to the 33rd, respectively represents: Beijing, Tianjin, Shanghai City, Chongqing City, mountain
Xi Sheng, Liaoning Province, Jilin Province, Heilongjiang Province, Jiangsu Province, Zhejiang Province, Anhui Province, Fujian Province, Jiangxi Province, Shandong Province, Henan Province,
Hubei Province, Hunan Province, Guangdong Province, Hainan Province, Sichuan Province, Guizhou Province, Yunnan Province, Shaanxi Province, Gansu Province, Qinghai Province, Taiwan Province;
(3) the requested place of sensitive information and number in one hour are counted, and divides object to save, it is quick that certain saves request
The number of sense information is denoted as Nt, by the vector value of the position of initialization vector position corresponding to the province, is revised as Nt x sensitivity grade.
Sensitive information flows to vector, is the vector result for obtain after vectorization processing to monitored results, being will be sensitive
Information monitoring problem is converted into the transfer of Solve problems mathematically, meanwhile, which can more intuitively show sensitive letter
The flowing situation of breath is prepared for subsequent anomalous identification.
Sensitive information flows to disaggregated model, is two layers neural network model, every layer has 64 concealed nodes, makes
It uses logistic regression as classification function, before flowing to disaggregated model using sensitive information, needs first to flow to sensitive information and classify
Model is trained.
In one embodiment, training process includes the following steps:
(1) collecting 10,000 manual examination and verification respectively is that abnormal and normal sensitive information flows to vector;
(2) 20,000 datas are divided into 625 pieces with 32 for a data block;
(3) 0.001 is set by learning rate, inputs a data block every time and is trained, and calculate current loss;
(4) step (3) are repeated, until losing convergence;
(5) it saves mould sensitive information and flows to disaggregated model parameter, training finishes.
The above is only a preferred embodiment of the present invention, it is not intended to restrict the invention, it is noted that for this skill
For the those of ordinary skill in art field, without departing from the technical principles of the invention, can also make it is several improvement and
Modification, these improvements and modifications also should be regarded as protection scope of the present invention.
Claims (5)
1. a kind of data monitoring and method for early warning based on neural network and sensitive information stream characterized by comprising
Step 1, the susceptibility of information is divided, determines information sensing grade, sensitive letter is obtained based on information sensing grade
Breath;
Step 2, the monitoring on time domain and airspace is carried out to the flow direction of sensitive information, vectorization processing is carried out to monitored results and obtained
Vector is flowed to sensitive information;
Step 3, disaggregated model is flowed to based on sensitive information vector is flowed to sensitive information and classified, identify abnormal flow direction.
2. the data monitoring and method for early warning according to claim 1 based on neural network and sensitive information stream, feature
It is, the step 1 includes:
The information sensing grade is divided into insensitive, general sensitive, sensitive and very sensitive four grades, and will be general quick
Sense, sensitive and very sensitive information are determined as sensitive information.
3. the data monitoring and method for early warning according to claim 2 based on neural network and sensitive information stream, feature
It is, carrying out time domain with the monitoring on airspace to the flow direction of sensitive information described in step 2 includes:
Monitor the sensitive information sensitive grade, sensitive information outflow time, sensitive information where flowed to.
4. the data monitoring and method for early warning according to claim 3 based on neural network and sensitive information stream, feature
It is, it is two layers neural network model that sensitive information described in step 3, which flows to disaggregated model, and every layer there are 64 to hide
Node uses logistic regression as classification function.
5. the data monitoring and method for early warning according to claim 4 based on neural network and sensitive information stream, feature
It is, the step 3 includes:
Before flowing to disaggregated model using the sensitive information, disaggregated model is flowed to the sensitive information and is trained.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811536697.9A CN109840543A (en) | 2018-12-15 | 2018-12-15 | A kind of data monitoring and method for early warning based on neural network and sensitive information stream |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811536697.9A CN109840543A (en) | 2018-12-15 | 2018-12-15 | A kind of data monitoring and method for early warning based on neural network and sensitive information stream |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109840543A true CN109840543A (en) | 2019-06-04 |
Family
ID=66883280
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811536697.9A Pending CN109840543A (en) | 2018-12-15 | 2018-12-15 | A kind of data monitoring and method for early warning based on neural network and sensitive information stream |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109840543A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110457349A (en) * | 2019-07-02 | 2019-11-15 | 北京人人云图信息技术有限公司 | The monitoring method and monitoring device of information outflow |
CN111726648A (en) * | 2020-06-28 | 2020-09-29 | 百度在线网络技术(北京)有限公司 | Method, device and equipment for detecting image data and computer readable storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20180174260A1 (en) * | 2016-12-08 | 2018-06-21 | Nuctech Company Limited | Method and apparatus for classifying person being inspected in security inspection |
CN108737138A (en) * | 2017-04-18 | 2018-11-02 | 腾讯科技(深圳)有限公司 | Service providing method and service platform |
CN109660512A (en) * | 2018-11-12 | 2019-04-19 | 全球能源互联网研究院有限公司 | A kind of sensitive information flows to vectorization method, abnormal flows to recognition methods and device |
-
2018
- 2018-12-15 CN CN201811536697.9A patent/CN109840543A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20180174260A1 (en) * | 2016-12-08 | 2018-06-21 | Nuctech Company Limited | Method and apparatus for classifying person being inspected in security inspection |
CN108737138A (en) * | 2017-04-18 | 2018-11-02 | 腾讯科技(深圳)有限公司 | Service providing method and service platform |
CN109660512A (en) * | 2018-11-12 | 2019-04-19 | 全球能源互联网研究院有限公司 | A kind of sensitive information flows to vectorization method, abnormal flows to recognition methods and device |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110457349A (en) * | 2019-07-02 | 2019-11-15 | 北京人人云图信息技术有限公司 | The monitoring method and monitoring device of information outflow |
CN110457349B (en) * | 2019-07-02 | 2022-04-05 | 北京人人云图信息技术有限公司 | Information outflow monitoring method and monitoring device |
CN111726648A (en) * | 2020-06-28 | 2020-09-29 | 百度在线网络技术(北京)有限公司 | Method, device and equipment for detecting image data and computer readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106502234B (en) | Industrial control system method for detecting abnormality based on double skeleton patterns | |
US20200250905A1 (en) | System and method for aircraft fault detection | |
CN103581186B (en) | A kind of network security situational awareness method and system | |
CN109379379A (en) | Based on the network inbreak detection method for improving convolutional neural networks | |
CN106411597A (en) | Network traffic abnormality detection method and system | |
CN109861845A (en) | A kind of data monitoring and method for early warning based on neural network and user access activity | |
CN109840543A (en) | A kind of data monitoring and method for early warning based on neural network and sensitive information stream | |
CN102663264B (en) | Semi-supervised synergistic evaluation method for static parameter of health monitoring of bridge structure | |
CN108490370A (en) | A kind of method and apparatus of fault diagnosis | |
CN103001328A (en) | Fault diagnosis and assessment method of intelligent substation | |
CN113780443B (en) | Threat detection-oriented network security situation assessment method | |
CN112087443B (en) | Sensing data anomaly detection method under physical attack of industrial sensing network information | |
CN106779215B (en) | Machine learning-based power grid global delay situation sensing method | |
Jiang et al. | A multisensor cycle-supervised convolutional neural network for anomaly detection on magnetic flux leakage signals | |
CN112134871A (en) | Abnormal flow detection device and method for energy internet information support network | |
CN109242746A (en) | One-dimensional instantaneous point source pollution sources source tracing method based on emergency monitoring data | |
CN111080005B (en) | Support vector machine-based public security risk early warning method and system | |
CN109242008B (en) | Compound fault identification method under incomplete sample class condition | |
CN109660512B (en) | Sensitive information flow direction vectorization method, abnormal flow direction identification method and device | |
Zhao et al. | Analysis of private industrial control protocol format based on LSTM-FCN model | |
CN109413036A (en) | Sensitive information outflow monitoring method, device and server extremely | |
Li et al. | Safety Boundary Extraction Using FCM and Prediction Using ELM for Aero-engine Performance Parameters | |
CN117692207B (en) | Instruction-level power system service protection method based on weighted similarity matching | |
Zhang | Diagnostic parameters selection of aero-engine based on BP network | |
Qiao et al. | Error prediction method of electronic transformers based on RBF neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190604 |