CN111291962A - Method for preventing and attacking AI crime and AI data infringement - Google Patents
Method for preventing and attacking AI crime and AI data infringement Download PDFInfo
- Publication number
- CN111291962A CN111291962A CN201911318745.1A CN201911318745A CN111291962A CN 111291962 A CN111291962 A CN 111291962A CN 201911318745 A CN201911318745 A CN 201911318745A CN 111291962 A CN111291962 A CN 111291962A
- Authority
- CN
- China
- Prior art keywords
- artificial intelligence
- data
- crime
- infringement
- protection
- 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
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 121
- 238000001514 detection method Methods 0.000 claims abstract description 29
- 230000002265 prevention Effects 0.000 claims abstract description 15
- 238000005516 engineering process Methods 0.000 claims description 34
- 230000006399 behavior Effects 0.000 claims description 15
- 230000000694 effects Effects 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 14
- 230000002787 reinforcement Effects 0.000 claims description 14
- 238000013135 deep learning Methods 0.000 claims description 13
- 238000007405 data analysis Methods 0.000 claims description 12
- 238000004458 analytical method Methods 0.000 claims description 11
- 238000012544 monitoring process Methods 0.000 claims description 11
- 238000013475 authorization Methods 0.000 claims description 9
- 230000008901 benefit Effects 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 9
- 239000008358 core component Substances 0.000 claims description 9
- 241000700605 Viruses Species 0.000 claims description 8
- 206010033799 Paralysis Diseases 0.000 claims description 7
- 238000004891 communication Methods 0.000 claims description 7
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000007418 data mining Methods 0.000 claims description 5
- 238000011835 investigation Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 3
- 208000035475 disorder Diseases 0.000 claims description 3
- 238000010219 correlation analysis Methods 0.000 claims description 2
- 230000009545 invasion Effects 0.000 claims 1
- 239000000243 solution Substances 0.000 description 39
- 238000004422 calculation algorithm Methods 0.000 description 19
- 238000007689 inspection Methods 0.000 description 13
- 230000001364 causal effect Effects 0.000 description 8
- 238000011156 evaluation Methods 0.000 description 7
- 230000000875 corresponding effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 231100000518 lethal Toxicity 0.000 description 4
- 230000001665 lethal effect Effects 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 210000000056 organ Anatomy 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 230000008451 emotion Effects 0.000 description 3
- 238000002360 preparation method Methods 0.000 description 3
- 108090000623 proteins and genes Proteins 0.000 description 3
- 241001622623 Coeliadinae Species 0.000 description 2
- 230000009471 action Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 230000002596 correlated effect Effects 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000009527 percussion Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 235000009421 Myristica fragrans Nutrition 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 230000009193 crawling Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000002427 irreversible effect Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 239000001115 mace Substances 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 244000062645 predators Species 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
- 238000013526 transfer learning Methods 0.000 description 1
- 210000003462 vein Anatomy 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the field of AI prediction, detection, protection and attack systems, including the field of military AI, and relates to a method for preventing and attacking AI crimes and AI data infringement. The method comprises the following steps: when an artificial intelligence data infringement or crime occurs, the artificial intelligence is detected, monitored and detected, and early warning judgment is carried out when a monitored object reaches an AI crime and data infringement threshold by obtaining a prediction prevention index of the AI crime or the data infringement; secondly, analyzing an artificial intelligent crime scene, a crime mode, data infringement and infringement range, judging the danger level, proposing a protection solution, and self-adaptively improving the protection solution; thirdly, striking is carried out according to the early warning grade and the danger grade, a striking scheme and a technical solution are proposed, and the striking scheme and the technical solution are distributed to a judicial system or a military agency or are directly disposed according to a working procedure. The method collects AI crime and AI data infringement evidences, and provides a protection or attack scheme to realize the early discrimination, detection, prevention and attack of AI crime and AI data infringement.
Description
Technical Field
The invention belongs to the field of artificial intelligence prediction, detection, protection and attack systems, and particularly relates to a method for preventing and attacking AI crimes and AI data infringement, which belongs to the field of artificial intelligence prediction, detection, protection and attack systems, including artificial intelligence in the military field.
Background
Artificial intelligence (AI in english) is a technological science for researching and developing simulators, and is mainly reflected in intelligent robots and software technologies. Nowadays, AI technology is more and more deep into our lives, especially in 5G today, AI has already passed through an academic research stage and is in a real application stage, and is widely applied to the fields of medical treatment, transportation, commerce, military and the like, such as automatic driving, face recognition, robot police, Lethal Autonomous Weapon System (LAWS), intelligent recommendation, intelligent weapon, robot fighter, unmanned aerial vehicle, unmanned submarine and the like. In the present day of the rapid development of information technology, data is ubiquitous, and artificial intelligence relies on big data for data collection, mining and analysis, but this will be widely related to personal privacy, business secrets, financial secrets, military secrets, national secrets, etc. Particularly in the field of public data security, some enterprises or individuals infringe national interests such as personal privacy rights, enterprise business secrets, national property, intellectual property rights, military secrets, national defense interests and the like, and then the intelligent machine clears the action traces and destroys evidences, so that evidence collection and evidence positioning are difficult. In addition, while the artificial intelligence crime will become an issue that must be faced as technology evolves, as information technology is iterated at the geometric level, AI will be able to iterate automatically, authorize autonomously, wake-up consciously, etc. The AI will learn autonomously, think autonomously, judge autonomously, update autonomously, which will happen that "thinking AI" violates existing laws causing crimes, data infringement, traitors, etc., will compromise national, public and personal interests.
Particularly, when the AI technology is more and more closely combined with judicial authorities, the assistant case handling systems such as "AI justice officer", "AI inspector", "AI attorney" and the like, which are innovated by various units, have been developed in various aspects of judicial work, and "intelligent judicial" has become an indispensable part of case handling gradually. The AI technology is also combined with the existing weaponry more and more closely, and the intelligent systems such as artificial intelligence weapons and the like which are continuously innovated by the army of various countries, such as unmanned aerial vehicles, unmanned submarines, unmanned vehicles, robot soldiers and the like, the development speed of the 'fatal autonomous weapon system' is beyond imagination, and the artificial intelligence weapons become an important part which is indispensable for capturing the high points of future war and become killer mace weapons gradually.
While the AI technology is developed at a high speed, the application of each field is in a blowout state, and each company publicizes how intelligent and wide the application of the AI product, but does not prompt the AI safety problem. Especially for the military field and the judicial field, the problems of future AI crime and data infringement are very serious.
In the prior art, Chinese patent application 201510442984.3 discloses a method for calculating the suspected crime degree of network speech data, which belongs to the technical field of intelligent security, the invention provides a concept of the suspected crime degree of network speech, defines the suspected crime degree as crime possibility expressed by certain ID through the speech on a social network, summarizes the speech characteristics expressed by the criminal psychology by taking the criminal psychology as a theoretical basis, and provides a demand factor, an emotion factor and a preparation factor influence model of the crime degree of network speech; judging demand factors by using a text analysis technical means and a naive Bayes classifier, judging emotion factors by using an emotion dictionary, constructing a crime-sensitive word dictionary, judging preparation factors by combining a machine learning method, and establishing a network theory suspected crime degree theoretical framework and a mathematical model; the invention can advance the early warning to the stage of criminal psychology formation and criminal preparation, can automatically analyze and predict a large amount of data in the whole course when being applied to a practical network, does not need human intervention, and can intelligently promote a security system to a higher level.
Chinese patent application 201810998920.5 relates to a method for evaluating a C2C travel trust index and crime rate, belonging to the technical field of big data and social security. The invention abandons the mode of determining the trust index of a C2C traveling driver side only according to the evaluation factors of net friends, innovatively and fully utilizes the advantages of data, considers the income level, the education level, the driving age, the credit of payment treasures and a bank system and the relevant data of criminal presidential departments, compares an evaluation number set (evaluation information of the net friends) by a text similarity algorithm, checks the information of proportional contents such as the ambiguous evaluation passive words and the like to obtain the relative trust index and the relative crime rate of the evaluation numbers, synthesizes the information of danger coefficients and the like judged by passenger routes according to the regional attributes, and estimates the final trust value and the crime rate of the evaluation numbers. The invention solves the problems of monitoring the driver side of the C2C trip, enhancing the understanding of the customers to the drivers, restricting the behaviors of the C two sides, and the like, increases the management of the C2C company, and screens the qualification of the trip drivers and guarantees the social security.
The two related technologies have the defects that the problems of potential safety hazard, infringement, crime, traitor in military field AI, traitor escape, predator and the like of the artificial intelligence can not be solved only by utilizing an artificial intelligence screening method to filter social insecure factors.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a method for preventing and fighting AI crime and AI data infringement, which can detect and identify artificial intelligence potential safety hazards.
The invention is realized in this way, a method for preventing and fighting AI crime and AI data infringement, which is characterized in that: comprises the following steps:
when an artificial intelligence data infringement or an artificial intelligence crime occurs, carrying out detection, monitoring and detection on the artificial intelligence, designing a prediction prevention index, and carrying out early warning judgment when a monitored object reaches an index threshold;
analyzing the artificial intelligent crime scene, the crime mode, the data infringement type and the infringement range, judging the danger level, proposing a protection solution, and perfecting the protection solution through reinforcement learning;
and thirdly, striking according to the early warning grade and the danger grade, proposing a striking scheme and a technical solution, and distributing to a judicial system or a military institution or lifting and disposing by a detection institution and a military detection institution according to a working procedure.
The AI crime and AI data infringement prevention and attack system is to be operated in the whole field, the whole time domain and the whole process of AI crime and AI data infringement prevention and attack, and the AI crime and AI data infringement are prevented and attack by constructing a prediction subsystem, a prevention subsystem and an attack subsystem.
One is to predict AI data infringement using a prior probability model or causal inference model. And (3) establishing a data model by utilizing an algorithm for collection and analysis, researching a prevention method for dealing with artificial intelligence data infringement, and designing the algorithm to prevent, detect and attack strong artificial intelligence autonomous crimes. AI data infringement, whether people control artificial intelligence or artificial intelligence autonomous crime, all needs a function of early warning in advance, especially can predict and prejudge AI data infringement. The current predicted value is a priority judgment for future AI data infringement crime through previous experience accumulation. The key technology of the algorithm is as follows: and judging the availability of the AI data infringement crime index, calculating the AI data infringement probability, and judging whether the AI carries out data infringement or not. In the AI autonomous infringement crime field, the model can set crime prediction indexes, when a monitored object reaches the index model, the model carries out early warning judgment, carries out real-time monitoring on AI, and carries out prejudgment, prediction, detection and investigation on AI data infringement and AI crime.
The second is to analyze and judge the protection solution to be used. The key technology of the algorithm is as follows: according to problems discovered by historical data and a prediction subsystem and early warning levels, a block chain technology, quantum computation, quantum communication, a deep learning technology, reinforcement learning, a cause and effect inference model and big data analysis AI crime scene, crime modes, data infringement types, infringement ranges and influences are utilized, in the field of AI autonomous crime and autonomous data infringement, the model can judge the danger degree of AI through correlation analysis, provide different protection and solutions or submit the solutions to a attack solution subsystem, and can improve the protection scheme through self-learning.
Third, crime fighting AI and data infringement AI. The key technology of the algorithm is as follows: methods such as block chain technology, quantum computing, generation of a countermeasure network (GAN), AutoML, non-learning algorithm tools (non-learning algorithm tools) and the like include, but are not limited to, simulation and optimization of construction of an enhanced version attack AI, or injection of viruses by technical means, or destruction of an energy system, a perception function, core components and the like of a problem AI, so that software and hardware of the problem AI are destroyed, trapped in confusion or paralysis, and artificial intelligence for terminating, destroying, deleting, modifying or replacing the artificial intelligence and data infringement of the crime.
The preferable scheme is as follows:
the system also comprises an artificial intelligence system for destroying the crime and an artificial intelligence system for data infringement by the authorization of judicial authorities, military authorities or law enforcement authorities, and the artificial intelligence system comprises artificial intelligence and data infringement artificial intelligence for simulating and optimizing and constructing an enhanced version attack AI or injecting virus genes to destroy, delete, modify or replace the crime.
The treatment includes public welfare litigation, public complaints and/or inspection advice.
The method comprises the following steps that firstly, according to a prior probability model and a causal inference model, usability judgment of artificial intelligence data infringement crime indexes is conducted, probability calculation of artificial intelligence data infringement and artificial intelligence crime is conducted, whether artificial intelligence conducts data infringement and crime or not is judged, and specifically:
(1) firstly, automatically adding tags to related data according to initial tag setting and keywords;
(2) according to historical data, Python data analysis is utilized, an artificial intelligence data infringement and artificial intelligence crime feature map, behavior features, trends and the like are drawn through technical means such as a cause and effect inference model and the like, the behavior features are tracked by utilizing a block chain technology, network monitoring and communication investigation are implemented, data of all aspects of artificial intelligence of the problem are correlated and compared, and indexes and probabilities of various types of artificial intelligence crimes and artificial intelligence data infringement are designed by utilizing the cause and effect inference model;
(3) carrying out usability judgment on artificial intelligence crimes and artificial intelligence data infringement indexes;
(4) when the current artificial intelligence crime and artificial intelligence data infringement indexes are available, selecting N types of indexes with the maximum probability;
(5) index detection is carried out, and the detected problem data are labeled, so that the detection accuracy is improved;
(6) index identification, namely judging whether the report is missed, false or wrong, and eliminating problem early warning;
(7) calculating the probability of artificial intelligence data infringement and artificial intelligence crime, and judging whether the artificial intelligence carries out the data infringement and the crime or not;
(8) and if data infringement or crime occurs, early warning, precaution and monitoring are carried out.
And step two, analyzing the artificial intelligent crime scene, crime mode and data infringement type, infringement range and influence by using technical means such as a block chain technology, a quantum calculation, a deep learning technology, a reinforcement learning, a cause and effect inference model, big data analysis and the like according to the discovered problems and the early warning level, carrying out relevance analysis, judging whether the artificial intelligent independent thinking exists or not, judging the degree of danger, proposing different protection and solution schemes for the reinforcement learning or submitting the protection and solution schemes to a strike solution subsystem, perfecting the protection scheme through self-learning, and detecting the integrity of the protection and solution.
The second step is specifically as follows:
(3) selecting a protection method according to the early warning grade result;
(4) according to the judgment result, deep learning and data mining are carried out on the problem data through a neural network, an image and a video are searched by using a self-encoder, non-mechanization data are processed by using a simple understanding network, structured data are processed by using a deep confidence network, and unmarked data are automatically identified and added with labels;
(6) discriminating various conditions of artificial intelligent crimes and artificial intelligent data infringement by utilizing RNN deep learning and Python data analysis, constructing and reproducing related crime scene learning, carrying out correlation comparison analysis on data, and determining evidences, behaviors and infringement legal benefits by utilizing a cause and effect inference model;
(7) judging whether the artificial intelligence is an autonomous thinking or crime and data infringement carried out by a human set program, and setting a label for the judgment condition;
(8) judging whether the artificial intelligence thought by the user or the artificial intelligence of a human set program destroys ethics or rules, and determining the danger degree of the artificial intelligence;
(6) directly submitting the risk level to a striking solution subsystem or automatically processing the risk level by a protection subsystem;
(7) if the protection subsystem needs to process by itself, selecting a protection solution in the database according to the judgment result, taking different processing measures according to the protection solution, and delivering the relevant conditions to relevant departments;
(8) various protection schemes are perfected through reinforcement learning, an optimized version protection scheme which is in conflict with the protection schemes is automatically established by utilizing the technology, and protection measures are established through autonomous learning.
Step three, comprising
(1) According to the early warning grade result or the danger degree of the protection subsystem, different types of striking schemes and technical solutions are provided;
(2) judging whether the problem is an artificial intelligence problem of an intelligent judicial system or an artificial intelligence problem of a military system by combining the results of carrying out behaviors and violating legal benefits according to the judgment results of artificial intelligence independent thinking or human set programs and utilizing a cause and effect inference model;
(3) according to the label, if the intelligent judicial system artificial intelligence self-thinking problem, according to the rule and the solution, corresponding measures are adopted to solve the problem; if a program question is set for a human being, submitting to a program (5);
(4) if the problem is the artificial intelligence problem in the military field, whether the problem is the independent thinking problem of the military artificial intelligence system or the human set program problem is judged according to the label; if the problem is a problem of autonomous thinking of a military artificial intelligence system, corresponding measures are taken to solve the problem according to rules and solutions; if a program question is set for a human being, submitting to a program (5);
(5) according to the results of the procedures (2), (3) and (4), the relevant evidence and indexes are distributed to a judicial system and a military judicial institution, or a detection institution or a military detection institution directly raises a commonweal litigation, a commonweal complaint and/or a detection suggestion;
(6) after authorization of judicial authorities, military authorities or law enforcement authorities, according to rules, aiming at artificial intelligence of different types and functions, generating confrontation network simulation and optimizing construction of enhanced version striking artificial intelligence, destroying problem artificial intelligence by striking artificial intelligence, or injecting virus genes by technical means, or other technical means including but not limited to an energy system, a sensing function and a core component for destroying problem AI, so that software and hardware of the problem AI are destroyed, trapped in disorder or paralysis, and artificial intelligence for ending, destroying, deleting, modifying or replacing the criminal artificial intelligence and data infringement artificial intelligence.
The invention has the advantages and positive effects that:
by building an AI crime and AI data infringement prediction model, AI data infringement indexes are set for historical data, whether multiple indexes are out of limit or not is judged in real time, necessary probability analysis and prejudgment can be achieved, collected evidences can be accumulated again, and therefore automatic iteration is achieved, and the prediction model has self-adaptive capacity. Through deep learning, characteristics and history data of an object are collected and analyzed, wherein the characteristics and history data comprise data infringement means, characteristics, an evasion method, an AI crime method, AI self-adaption capability and the like, data extraction, data cleaning and data integration are carried out, and the model is utilized for evaluation and prejudgment. The method comprises the steps of building a protection model, analyzing an AI crime scene, a crime mode, a data infringement type, an infringement range and influence according to problems found by historical data and a prediction subsystem, judging danger levels through data mining analysis, data relevance analysis and a cause and effect inference model, proposing different protection solution schemes, and perfecting the protection scheme through reinforcement learning.
By building a striking model and judging striking according to the early warning level and the danger degree, different types of striking schemes and technical solutions are provided. According to the working procedures of the judicial department, the information is distributed to the judicial system and related technical departments, or the inspection aircraft is closed, and the military inspection departments raise the commonweal litigation, the commonwealth, the inspection advice and the like. After authorization of relevant authorities, the artificial intelligence system of the crime and the artificial intelligence system of data infringement can be destroyed by direct technology. In addition, a threshold value, an early warning rate and a false alarm rate need to be designed, false reports and false missing reports which may occur in the process of predicting the AI data infringement are eliminated, and the accuracy of the whole AI data infringement prediction model is improved. AI data infringement evidence is collected through the model, and AI data infringement is prejudged, so that the prevention effect is improved, and AI crimes are distinguished and prevented in advance.
The system realizes prevention and attack of AI crime and AI data infringement, and solves a series of problems occurring in AI itself, particularly after the thought AI occurs. When AI crimes occur, particularly the AI system of a judicial authority and the AI weapon equipment of a military authority crime and data infringement occur, the method fundamentally solves the problems. The method is characterized in that the method can be used for fighting against crimes of robot warriors and robot policemen, AI capable of fighting against the crimes is generated through a technical method, viruses are injected or energy systems, sensing functions, core components and the like of the problem AI are destroyed through technical means according to authorization of judicial authorities, military authorities or law enforcement authorities, so that software and hardware of the problem AI are destroyed, trapped in confusion or paralysis, and artificial intelligence of data infringement of the crimes are ended, destroyed, deleted, modified or replaced.
Drawings
FIG. 1 is a system schematic block diagram of an embodiment of the present invention;
FIG. 2 is a flow diagram of a prediction subsystem of the present invention;
FIG. 3 is a flow diagram of a protection subsystem of the present invention;
FIG. 4 is a percussion solution subsystem flow diagram of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The following detailed description of the principles of the invention is provided in connection with the accompanying drawings.
Example 1:
as shown in fig. 1, the system of the present application is divided into three subsystems: the system comprises a forecast artificial intelligence crime and artificial intelligence data infringement subsystem, a protection artificial intelligence crime and artificial intelligence data infringement subsystem and a hit solution subsystem.
1. The subsystem for predicting artificial intelligent crimes and artificial intelligent data infringement realizes all-weather supervision and safety level prediction on the artificial intelligent crimes and the artificial intelligent data infringement, performs detection, supervision and detection on various types of artificial intelligent technologies in various fields in advance, performs full-process detection on AI, prompts and pre-warns the AI system which does not accord with safety standards and index values, distributes the AI system to a protection subsystem or a striking subsystem according to the pre-waring level, and perfects the prediction method through self-learning.
2. A subsystem for protecting against artificial intelligent crime and artificial intelligent data infringement analyzes AI crime scene, crime mode, data infringement type, infringement range and influence according to problems found by historical data and a prediction subsystem, judges danger levels through a data analysis and causal inference model, provides different protection solutions, and perfects the protection solutions through reinforcement learning.
3. And the striking solution subsystem judges to strike according to the early warning level and the danger degree, and provides different types of striking solutions and technical solutions. According to the working procedures of the judicial organ, the method is distributed to the judicial organ or the military organ, or the inspection aircraft is closed, and the military inspection organ raises the commonweal litigation, the commonweal complaint, the inspection advice and the like. After authorization of judicial authorities, military authorities or law enforcement authorities, the artificial intelligence system for destroying crimes and the artificial intelligence system for data infringement can be directly destroyed by the technology.
The system is mainly based on Machine Learning (Machine Learning), Deep Learning (Deep Learning), reinforcement Learning (reinforcement Learning) and causal inference models. The block chain technology, the quantum computation technology, the quantum communication technology, the Python technology and the like are used for supporting, big data analysis is used, learning is carried out through a neural network, and self-adaptive learning is achieved through mass data.
The models used in the system are:
machine learning: bayes, Naive Bayesian classification (Naive Bayesian classification), kalman innovation variance, Decision Trees (Decision Trees), Logistic Regression (Logistic Regression), Support Vector Machines (SVM), K-nearest neighbor algorithms, K-means algorithms, markov, monte carlo, Boosting, Least Squares (iterative Least Squares Regression), clustering algorithms (clustering algorithms).
The deep learning comprises the following steps: CNN (convolutional neural network), RNN (recurrent neural network), LSTM (long-term short-term memory), GAN (generative confrontation network), transfer learning, attention model (attention model). Deep learning + reinforcement learning.
The prediction prevention indexes comprise statistics and a threshold, the applied statistical detection amount is estimated through collected historical data, the threshold is determined by using a causal inference model, when the statistics is larger than the threshold, the threshold is determined to be out of limit, and when the statistics is smaller than the threshold, the threshold is determined to be free of problems. The design of the indexes and the quantity of the indexes are different according to different AI systems, the indexes are designed by adopting a causal inference model, the indexes frequently appear and are closely related to crimes and data infringement, the maximum alarm threshold with larger influence is set as the indexes, and the N thresholds which are more front are set as the indexes.
The index design is to perform sentence search on big data for the following keywords, including but not limited to:
1. content of any keyword: artificial intelligence, robots, big data, computers, software, algorithms, cloud computing, AI data infringement, AI crime, intelligent forces, robotic weapons, robotic warriors, intelligent weapons, kinetic weapons, directed energy weapons, algorithmic warfare, war machines, lethal autonomous weapons systems.
2. Combining content with any of the following keywords: generating, creating, synthesizing, mobile phone, extracting, grabbing, crawling, crawler, identifying, unmanned, automatic forming, intelligent weapon, unmanned weapon, autonomous combat, network combat, robot fighter, kinetic weapon, algorithm weapon, unmanned combat. The condition that the keywords appear in the names of people, legal people and the book (R) needs to be eliminated.
Setting labels through the keywords, establishing a database, and utilizing a block chain technology to 'chain up' various data collected by a previous system to control online data and operation behavior tracks of the problem AI in real time. And automatically adding new labels according to new crime and data infringement trends. Problem data are marked through labels, and N-type characteristic indexes with high AI crime and AI data infringement occurrence probability are obtained through technical means such as neural networks and quantum computing and through Pyhton data analysis.
The specific indexes are set according to historical data of AI crime and AI data infringement and different types of AI. When the index reaches a critical value or exceeds a threshold, for example, a police intelligent police system and a robot police violate the program setting, the label which is unauthorized to warn appears 3 times, the arrest label appears 2 times, and the shooting label appears 1 time. The intelligent judging system of the court is free from destroying the evidence chain for 1 time, the evidence is less than 1 time, the intelligent judging system is not judged for 1 time according to the specified program, and the intelligent judging system is free from collecting judging data for 2 times. The lethal autonomous weapon system opens the gun without permission for 1 time, fires for 1 time, does not obey the order for 1 time, etc. AI crime and AI data infringement may be determined.
As shown in fig. 2, the key techniques of the prediction subsystem algorithm are as follows: the subsystem can perform early warning judgment on the model when a monitored object reaches the index model by acquiring crime prediction and prevention indexes, perform real-time monitoring on AI, and perform pre-judgment, prediction and investigation on AI data infringement. The method comprises the following steps: the method comprises the following steps of block chain technology, quantum calculation, a prior probability model and a causal inference model, AI data infringement crime index availability judgment, AI data infringement and AI crime probability calculation and judgment on whether AI carries out data infringement and crime or not.
In the specific steps, the method comprises the following steps,
(1) firstly, automatically adding tags to related data according to initial tag setting and keywords.
(2) According to historical data, Python data analysis is utilized, AI data infringement and AI crime feature maps, behavior characteristics, trends and the like are sketched out through technical means such as a cause and effect inference model and quantum computation and the like, the behavior features are tracked by utilizing a block chain technology, network monitoring and communication investigation are implemented, data of all aspects of the problem AI are correlated and compared, and indexes and probabilities of various types of AI crimes and AI data infringement are calculated.
(3) And carrying out availability judgment on AI crime and AI data infringement indexes.
(4) When the current AI crime and AI data infringement indexes are available, the N types of indexes with the maximum probability are selected.
(5) Index detection is carried out, and the detected problem data are labeled, so that the detection accuracy is improved.
(6) And (4) index identification, namely judging whether the report is missed, false or wrong, and eliminating problem early warning.
(7) And calculating the probability of AI data infringement and AI crime, and judging whether the AI carries out data infringement and crime.
(8) And if data infringement or crime occurs, early warning, precaution and supervision are carried out.
As shown in fig. 3, according to the problems and early warning levels found by the historical data and prediction subsystem, the subsystem analyzes the AI crime scene, crime mode, data infringement type, infringement range and influence by using technical means such as a block chain technology, a quantum calculation, a deep learning technology, reinforcement learning, a cause and effect inference model, big data analysis and the like, performs relevance analysis, judges whether the AI is an autonomous thought, judges the degree of danger, proposes different protection and solutions or submits the solutions to a attack solution subsystem, perfects the protection scheme through self-learning, and detects the integrity of the subsystem. And (3) applying a block chain technology to 'uplink' various data collected by the previous system, and controlling the online data and the operation behavior track of the problem AI in real time. Algorithmic wars against problem AI using quantum computation. And a data protection shield is constructed by utilizing quantum communication, so that infinite expansion of AI stealing is prevented.
(1) And selecting a protection method according to the early warning grade result of the prediction subsystem.
(2) According to the judgment result, deep learning and data mining are carried out on the problem data through a neural network, images and videos are searched through a self-encoder (automatic encoder), unstructured data are processed through a simple understanding network (RBM), structured data are processed through a Deep Belief Network (DBN), and unmarked data are automatically identified and added with labels.
(3) And (3) discriminating various conditions of AI crimes and AI data infringement by utilizing RNN deep learning and Python data analysis, learning out relevant crime scenes, carrying out correlation comparison analysis on the data, and determining evidences, behaviors and infringement legal benefits results by utilizing a cause-and-effect inference model.
(4) And judging whether the AI is an autonomous thinking or a crime and data infringement performed by a human setting program, and setting a label for the judgment condition.
(5) And judging whether the self-thought AI or the human set program AI destroys the ethics or rules, and determining the AI risk degree.
(6) Either directly to the percussion solution subsystem or handled by the protection subsystem on its own, depending on the risk level.
(7) If the protection subsystem needs to process by itself, the protection solution in the database is selected according to the judgment result, different processing measures are taken according to the protection solution, and the related condition is delivered to the related department.
(8) Various protection schemes are perfected through reinforcement learning, an optimized version protection scheme for resisting against the protection schemes is automatically established by using technical methods such as a block chain technology, quantum computation, quantum communication, generation of a countermeasure network technology (GAN), AutoML, a non-learning algorithm tool (non-learning algorithm tool) and the like, protection measures are established through autonomous learning, and self-adaptability is enhanced.
As shown in fig. 4, different types of attack schemes and technical solutions are proposed by using a causal inference model according to the early warning level or the danger level proposed by the protection subsystem. According to the working procedures of the judicial department, the method is distributed to the judicial system, the military judicial department and the related technical departments, or the inspection departments and the military inspection departments directly lift up the commonweal litigation, the commonweal complaint and the inspection advice, etc. Generating an confrontation network simulation and optimizing and constructing an enhanced version of the attack artificial intelligence through authorization and rules of relevant authorities, using the attack artificial intelligence to destroy problem artificial intelligence, or injecting virus genes through technical means, or other technical means including but not limited to destroying energy systems, sensing functions and core components of the artificial intelligence, so that software and hardware of the artificial intelligence are destroyed, trapped in disorder or paralysis, and the artificial intelligence of data infringement of the crime are ended, destroyed, deleted, modified or replaced.
The method comprises the following specific steps:
(1) according to the early warning grade result of the prediction subsystem or the danger degree of the protection subsystem, different types of striking schemes and technical solutions are provided.
(2) And judging whether the AI problem is an AI problem of an intelligent judicial system or an AI problem of a military system by using a cause and effect inference model according to the judgment result of the AI independent thinking or a human set program and combining the implementation behavior and the infringement legal benefit result.
(3) According to the label, if the problem is thought autonomously by the intelligent judicial system AI, corresponding measures are taken to solve the problem according to the rules and the solutions. If a program question is set for a human, it is submitted to the program (5).
(4) If the problem is the AI problem of the military system, whether the problem is the autonomous thinking problem of the military AI system or the human program setting problem is judged according to the label. If the problem is autonomously thought by the military AI system, corresponding measures are taken to solve the problem according to the rules and the solution. If a program question is set for a human, it is submitted to the program (5).
(5) According to the results of the procedures (2), (3) and (4), the relevant evidence and index are distributed to the judicial system, the military judicial body and the relevant technical department, or the inspection body (including the military inspection body) directly draws up the commonweal litigation, the commonweal complaint and the inspection advice, etc.
(6) Authorized by judicial authorities, military authorities or law enforcement authorities, aiming at AI of different types and functions, according to rules, methods such as generating a countermeasure network (GAN), AutoML, a non-learning algorithm tool (non-learning algorithm tool) and the like are utilized to simulate and optimize and construct an enhanced version to attack AI, or viruses are injected or energy systems, sensing functions, core components and the like of the AI are destroyed through technical means, so that software and hardware of the AI are destroyed, trapped in confusion or paralysis, and artificial intelligence of the crime and artificial intelligence of data infringement are terminated, destroyed, deleted, modified or replaced.
The specific setting of the threshold is divided into a military AI system and a local AI system, and the setting of the threshold is different according to the function, the action and the type of each AI. The data is from the case data of the highest people's court and the liberated military court from 2016, 1 month, 1 day to 2019, 12 months, 1 day, and the data provided by the military data center.
And designing an early warning rate and a false alarm rate, eliminating false reports and false negatives which may appear in the process of predicting the AI data infringement, and improving the accuracy of the whole AI data infringement prediction model. Therefore, AI crime and AI data infringement evidences can be collected through the model, AI crime and AI data infringement are prejudged, the prevention effect is improved, and AI crimes are distinguished and prevented in advance.
The method comprises the steps of drawing a characteristic map of AI crime and AI data infringement by utilizing the collected data and analysis conclusion, drawing thresholds of the AI crime and AI data infringement according to the characteristic map, drawing characteristics and modes of the AI crime and AI data infringement, tracing data trace addresses and characteristics, starting points and veins of the AI crime and data infringement, adaptively predicting various data indexes and crime trends of the AI data infringement by using the model, collecting and verifying various data indexes through a large amount of data every day, perfecting tracks and prediction models of crime trends, and realizing early intervention, early monitoring, early warning, early prevention and early striking of judicial authorities and military authorities.
The system realizes prevention and attack of AI crime and AI data infringement, and solves a series of problems occurring in AI itself, particularly after the thought AI occurs. When AI crimes occur, particularly the AI system of a judicial authority and the AI weapon equipment of a military authority crime and data infringement occur, the method fundamentally solves the problems. The method can be used for fighting against crimes of robot warriors and robot policemen, AI capable of fighting against the crimes is generated through a technical method, or viruses are injected or the energy system, the sensing function, the core component and the like of the AI are destroyed through technical means, and the problem AI of crime and data infringement is destroyed according to authorization of judicial authorities, military authorities or law enforcement authorities.
The system method is used in the fields of robot armies, unmanned combat vehicles, unmanned submarine armies, Beidou satellite navigation positioning systems and the like which are built by our army, the probability of the artificial intelligent weapon equipment is identified to be 99.9%, relevant data is collected and integrity detection is carried out, and relevant solutions are provided. The conventional calculation method for geometrical usability of military satellites is analyzed, an HPL calculation method combining ARP and noise influence is provided, and the integrity guarantee performance of the HPL calculation method for non-precision approach is obtained. The system has the outstanding effects of resisting interference, damage and attack to the robot warriors and unmanned weaponry, in a battlefield environment simulation experiment, the instructions of the robot warriors are related to the information of commanders, and the enemies are used for destroying the troops of the robots of my party and injecting wrong information to cause the traitors of the robots to change, and after the system is used, related problems are detected in advance, a protection solution is provided, the irreversible robot warriors (lethal autonomous weaponry system) are technically destroyed, the integrity of a military artificial intelligent weaponry system is ensured, and the performance of the artificial intelligent weaponry is ensured.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principles of the present invention are intended to be included within the scope of the present invention.
Claims (7)
1. A method for preventing and fighting AI crime and AI data infringement is characterized in that: the method comprises the following steps:
when an artificial intelligence data infringement or an artificial intelligence crime occurs, carrying out detection, monitoring and detection on the artificial intelligence, designing a prediction prevention index, and carrying out early warning judgment when a monitored object reaches an index threshold;
analyzing the artificial intelligent crime scene, the crime mode, the data infringement type and the infringement range, judging the danger level, proposing a protection solution, and perfecting the protection solution through reinforcement learning;
and thirdly, striking according to the early warning grade and the danger grade, proposing a striking scheme and a technical solution, and distributing to a judicial institution or a military institution or lifting and disposing by a detection institution or a military detection institution according to a working procedure.
2. The method for preventing and fighting AI crime and AI data infringement of claim 1 further comprising the step of destroying the criminal artificial intelligence system and the data infringement artificial intelligence system under the authorization of judicial, military or law enforcement agencies, including but not limited to simulating and optimizing the construction of an enhanced version fighting AI, or destroying the software and hardware of an artificial intelligence system, sensing function, core component, etc. by injecting viruses or destroying the energy system, sensing function, core component, etc. of the artificial intelligence system, sensing function, core component, etc. so that the software and hardware of the artificial intelligence system are destroyed, trapped in confusion or paralysis, and the artificial intelligence system of the crime and the data infringement artificial intelligence system are terminated, destroyed, deleted, modified or replaced.
3. The method of preventing and combating AI crime and AI data infringement of claim 1, wherein said disposition comprises a fair litigation, a fair complaint, and or a censorship recommendation.
4. The method for preventing and fighting AI crime and AI data infringement according to claim 1, wherein the first step is to perform artificial intelligence data infringement crime indicator availability determination, probability calculation for artificial intelligence data infringement and artificial intelligence crime, and determine whether artificial intelligence performs data infringement and crime, specifically:
(1) firstly, automatically adding tags to related data according to initial tag setting and keywords;
(2) according to historical data, performing data analysis, drawing an artificial intelligence data infringement and artificial intelligence crime feature map, behavior characteristics, trends and the like, tracking the behavior features by using a block chain technology, implementing network monitoring and communication investigation, correlating and comparing all aspects of artificial intelligence data of the problem, and calculating indexes and probabilities of various types of artificial intelligence crimes and artificial intelligence data infringement;
(3) carrying out usability judgment on artificial intelligence crimes and artificial intelligence data infringement indexes;
(4) when the current artificial intelligence crime and artificial intelligence data infringement indexes are available, selecting N types of indexes with the maximum probability;
(5) index detection is carried out, and the detected problem data are labeled, so that the detection accuracy is improved;
(6) index identification, namely judging whether the report is missed, false or wrong, and eliminating problem early warning;
(7) calculating the probability of artificial intelligence data infringement and artificial intelligence crime, and judging whether the artificial intelligence carries out the data infringement and the crime or not;
(8) if data infringement or crime occurs, early warning, precaution and monitoring are carried out, and various indexes are predicted to predict the risk of serious offense.
5. The method for preventing and fighting AI crime and AI data infringement according to claim 1, wherein in the second step, the artificial intelligence crime scene, crime mode and data infringement type, infringement range and influence are analyzed according to the discovered problems and the early warning level, and the correlation analysis is performed to judge whether the crime scene, the crime mode and the data infringement type, the infringement range and the influence are artificial intelligence independent thinking or not, judge the degree of danger, propose different protection and solutions or submit the solutions to the subsystem of the attack solution, perfect the protection scheme through self-learning, and detect the integrity of the system.
6. The method for preventing and fighting AI crime and AI data infringement of claim 5, wherein said second step is specifically:
(1) selecting a protection method according to the early warning grade result;
(2) according to the judgment result, deep learning and data mining are carried out on the problem data through a neural network, and label-free data are automatically identified and added with labels;
(3) discriminating various conditions of artificial intelligent crimes and artificial intelligent data infringement, building and reproducing the learning of relevant crime scenes, carrying out correlation comparison analysis on the data, and determining evidences, behaviors and infringement legal benefits by using a cause and effect inference model;
(4) judging whether the artificial intelligence is an autonomous thinking or crime and data infringement carried out by a human set program, and setting a label for the judgment condition;
(5) judging whether the artificial intelligence thought by the user or the artificial intelligence of a human set program destroys ethics or rules, and determining the danger degree of the artificial intelligence;
(6) directly submitting the risk level to a striking solution subsystem or automatically processing the risk level by a protection subsystem;
(7) if the protection subsystem needs to process by itself, selecting a protection solution in the database according to the judgment result, taking different processing measures according to the protection solution, and delivering the relevant conditions to relevant departments;
(8) various protection schemes are perfected through reinforcement learning, an optimized version protection scheme for resisting against the protection schemes is automatically established, and protection measures are established through autonomous learning.
7. The method for preventing and fighting AI crime and AI data infringement of claim 1, wherein said third step comprises
(1) According to the early warning grade result or the danger degree of the protection subsystem, different types of striking schemes and technical solutions are provided;
(2) judging whether the problem is an artificial intelligence problem of an intelligent judicial system or an artificial intelligence problem of a military system by combining an implementation behavior and a law and benefit invasion result and utilizing a cause and effect inference model according to a judgment result of an artificial intelligence independent thinking or a human set program;
(3) according to the label, if the problem is automatically thought by the artificial intelligence of the wisdom judicial system, corresponding measures are taken to solve the problem according to the rules and the solution; if a program question is set for a human being, submitting to a program (5);
(4) if the problem is the artificial intelligence problem of the military system, judging whether the problem is the autonomous thinking problem of the military artificial intelligence system or the human set program problem according to the label; if the problem is a problem of autonomous thinking of a military artificial intelligence system, corresponding measures are taken to solve the problem according to rules and solutions; if a program question is set for a human being, submitting to a program (5);
(5) according to the results of the procedures (2), (3) and (4), the related evidence and the indexes are distributed to a judicial institution, a military institution and related technical departments, or a justice litigation, a prosecution and/or a prosecution suggestion are directly raised by a detection institution or a military detection institution;
(6) after authorization, aiming at artificial intelligence of different types and functions, software or hardware of the problem AI is destroyed, trapped in disorder or paralyzed according to rules, and the aims of terminating, destroying, deleting or modifying the artificial intelligence of the crime and the artificial intelligence of data infringement are fulfilled.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911318745.1A CN111291962A (en) | 2019-12-19 | 2019-12-19 | Method for preventing and attacking AI crime and AI data infringement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911318745.1A CN111291962A (en) | 2019-12-19 | 2019-12-19 | Method for preventing and attacking AI crime and AI data infringement |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111291962A true CN111291962A (en) | 2020-06-16 |
Family
ID=71024226
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911318745.1A Pending CN111291962A (en) | 2019-12-19 | 2019-12-19 | Method for preventing and attacking AI crime and AI data infringement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111291962A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113642986A (en) * | 2021-08-02 | 2021-11-12 | 上海示右智能科技有限公司 | Method for constructing digital notarization |
| CN115114677A (en) * | 2022-08-30 | 2022-09-27 | 四川安洵信息技术有限公司 | Network crime service platform based on block chain evidence obtaining and storing technology and application method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160321563A1 (en) * | 2015-04-30 | 2016-11-03 | University Of Southern California | Optimized artificial intelligence machines that allocate patrol agents to minimize opportunistic crime based on learned model |
| CN106295565A (en) * | 2016-08-10 | 2017-01-04 | 中用环保科技有限公司 | Monitor event identifications based on big data and in real time method of crime prediction |
| CN106453404A (en) * | 2016-11-23 | 2017-02-22 | 北京邮电大学 | Network intrusion detection method and device |
| CN110493262A (en) * | 2019-09-17 | 2019-11-22 | 武汉思普崚技术有限公司 | It is a kind of to improve the network attack detecting method classified and system |
-
2019
- 2019-12-19 CN CN201911318745.1A patent/CN111291962A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160321563A1 (en) * | 2015-04-30 | 2016-11-03 | University Of Southern California | Optimized artificial intelligence machines that allocate patrol agents to minimize opportunistic crime based on learned model |
| CN106295565A (en) * | 2016-08-10 | 2017-01-04 | 中用环保科技有限公司 | Monitor event identifications based on big data and in real time method of crime prediction |
| CN106453404A (en) * | 2016-11-23 | 2017-02-22 | 北京邮电大学 | Network intrusion detection method and device |
| CN110493262A (en) * | 2019-09-17 | 2019-11-22 | 武汉思普崚技术有限公司 | It is a kind of to improve the network attack detecting method classified and system |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113642986A (en) * | 2021-08-02 | 2021-11-12 | 上海示右智能科技有限公司 | Method for constructing digital notarization |
| CN113642986B (en) * | 2021-08-02 | 2024-04-16 | 上海示右智能科技有限公司 | Method for constructing digital notarization |
| CN115114677A (en) * | 2022-08-30 | 2022-09-27 | 四川安洵信息技术有限公司 | Network crime service platform based on block chain evidence obtaining and storing technology and application method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sarker et al. | Ai-driven cybersecurity: an overview, security intelligence modeling and research directions | |
| Jeong | Artificial intelligence security threat, crime, and forensics: Taxonomy and open issues | |
| William et al. | Crime analysis using computer vision approach with machine learning | |
| CN106790256B (en) | Active machine learning system for dangerous host supervision | |
| Rawat et al. | Techniques for predicting dark web events focused on the delivery of illicit products and ordered crime | |
| Sarker | Multi‐aspects AI‐based modeling and adversarial learning for cybersecurity intelligence and robustness: A comprehensive overview | |
| Chen et al. | Backdoor attacks and defenses for deep neural networks in outsourced cloud environments | |
| McGuire | The laughing policebot: automation and the end of policing | |
| Huang et al. | Monitoring social media for vulnerability-threat prediction and topic analysis | |
| CN111291962A (en) | Method for preventing and attacking AI crime and AI data infringement | |
| Carmichael et al. | Unfooling perturbation-based post hoc explainers | |
| Sarker | AI-driven cybersecurity and threat intelligence: cyber automation, intelligent decision-making and explainability | |
| Bouke et al. | An empirical study of pattern leakage impact during data preprocessing on machine learning-based intrusion detection models reliability | |
| Chen et al. | {FACE-AUDITOR}: Data Auditing in Facial Recognition Systems | |
| Montasari | Countering cyberterrorism: the confluence of artificial intelligence, cyber forensics and digital policing in US and UK national cybersecurity | |
| Rawat et al. | Autonomous artificial intelligence systems for fraud detection and forensics in dark web environments | |
| Sangaiah et al. | Towards data security assessments using an IDS security model for cyber-physical smart cities | |
| Wang et al. | Cyber Threat Analysis and Trustworthy Artificial Intelligence | |
| Rehman | Practical Implementation of Artificial Intelligence in Cybersecurity–A Study | |
| Kaduk et al. | The use of automated information systems in the investigation of criminal offences | |
| Badalič | The metadata-driven killing apparatus: big data analytics, the target selection process, and the threat to international humanitarian law | |
| Nakid | Evaluation and detection of cybercriminal attack type using machine learning | |
| Mahanta et al. | Safety and Security in AI Systems | |
| Müller et al. | Defending against adversarial denial-of-service data poisoning attacks | |
| Tuteja et al. | Artificial intelligence: threat of terrorism and need for better counter-terrorism efforts |
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 |