Download presentation
Presentation is loading. Please wait.
Published byAbner Glenn Modified over 8 years ago
1
Data Mining, Security and Privacy Dr. Bhavani Thuraisingham The University of Texas at Dallas bhavani.thuraisingham@utdallas.edu March 2008
2
What is Data Mining? Data Mining Knowledge Mining Knowledge Discovery in Databases Data Archaeology Data Dredging Database Mining Knowledge Extraction Data Pattern Processing Information Harvesting Siftware The process of discovering meaningful new correlations, patterns, and trends by sifting through large amounts of data, often previously unknown, using pattern recognition technologies and statistical and mathematical techniques (Thuraisingham, Data Mining, CRC Press 1998)
3
What’s going on in data mining? l What are the technologies for data mining? - Database management, data warehousing, machine learning, statistics, pattern recognition, visualization, parallel processing l What can data mining do for you? - Data mining outcomes: Classification, Clustering, Association, Anomaly detection, Prediction, Estimation,... l How do you carry out data mining? - Data mining techniques: Decision trees, Neural networks, Market-basket analysis, Link analysis, Genetic algorithms,... l What is the current status? - Many commercial products mine relational databases l What are some of the challenges? - Mining unstructured data, extracting useful patterns, web mining, Data mining, security and privacy
4
Data Mining for Counter-terrorism
5
Data Mining Needs for Counterterrorism: Non-real-time Data Mining l Gather data from multiple sources - Information on terrorist attacks: who, what, where, when, how - Personal and business data: place of birth, ethnic origin, religion, education, work history, finances, criminal record, relatives, friends and associates, travel history,... - Unstructured data: newspaper articles, video clips, speeches, emails, phone records,... l Integrate the data, build warehouses and federations l Develop profiles of terrorists, activities/threats l Mine the data to extract patterns of potential terrorists and predict future activities and targets l Find the “needle in the haystack” - suspicious needles? l Data integrity is important l Techniques have to SCALE
6
Data Mining for Non Real-time Threats Integrate data sources Clean/ modify data sources Build Profiles of Terrorists and Activities Examine results/ Prune results Report final results Data sources with information about terrorists and terrorist activities Mine the data
7
Data Mining Needs for Counterterrorism: Real-time Data Mining l Nature of data - Data arriving from sensors and other devices l Continuous data streams - Breaking news, video releases, satellite images - Some critical data may also reside in caches l Rapidly sift through the data and discard unwanted data for later use and analysis (non-real-time data mining) l Data mining techniques need to meet timing constraints l Quality of service (QoS) tradeoffs among timeliness, precision and accuracy l Presentation of results, visualization, real-time alerts and triggers
8
Data Mining for Real-time Threats Integrate data sources in real-time Build real-time models Examine Results in Real-time Report final results Data sources with information about terrorists and terrorist activities Mine the data Rapidly sift through data and discard irrelevant data
9
Data Mining Outcomes and Techniques for Counter-terrorism
10
Example Success Stories l COPLINK developed at University of Arizona - Research transferred to an operational system currently in use by Law Enforcement Agencies l What does COPLINK do? - Provides integrated system for law enforcement; integrating law enforcement databases - If a crime occurs in one state, this information is linked to similar cases in other states l Some UTD Prototypes - Police Blotter with Raytheon; Carried out crime analysis using Google maps - Suspicious Event Detection
11
Data Mining for Intrusion Detection: Problem l An intrusion can be defined as “any set of actions that attempt to compromise the integrity, confidentiality, or availability of a resource”. l Attacks are: - Host-based attacks - Network-based attacks l Intrusion detection systems are split into two groups: - Anomaly detection systems - Misuse detection systems l Use audit logs - Capture all activities in network and hosts. - But the amount of data is huge!
12
Our Approach: Overview Training Data Class Hierarchical Clustering (DGSOT) Testing Testing Data SVM Class Training DGSOT: Dynamically growing self organizing tree
13
Results Training Time, FP and FN Rates of Various Methods Methods Average Accuracy Total Training Time Average FP Rate (%) Average FN Rate (%) Random Selection 52%0.44 hours4047 Pure SVM57.6%17.34 hours35.542 SVM+Rocchio Bundling 51.6%26.7 hours44.248 SVM + DGSOT69.8%13.18 hours37.829.8
14
Where are we now? l We have some tools for - building data warehouses from structured data - integrating structured heterogeneous databases - mining structured data - forming some links and associations - information retrieval tools - image processing and analysis - pattern recognition - video information processing - visualizing data - managing metadata
15
What are our challenges? l Do the tools scale for large heterogeneous databases and petabyte sized databases? l Building models in real-time; need training data l Extracting metadata from unstructured data l Mining unstructured data l Extracting useful patterns from knowledge-directed data mining l Rapidly forming links and associations; get the big picture for real- time data mining l Detecting/preventing cyber attacks l Mining the web l Evaluating data mining algorithms l Conducting risks analysis / economic impact l Building testbeds
16
IN SUMMARY: l Data Mining is very useful to solve Security Problems - Data mining tools could be used to examine audit data and flag abnormal behavior - Much recent work in Intrusion detection (unit #18) l e.g., Neural networks to detect abnormal patterns - Tools are being examined to determine abnormal patterns for national security l Classification techniques, Link analysis - Fraud detection l Credit cards, calling cards, identity theft etc. BUT CONCERNS FOR PRIVACY
17
Some Privacy concerns l Medical and Healthcare - Employers, marketers, or others knowing of private medical concerns l Security - Allowing access to individual’s travel and spending data - Allowing access to web surfing behavior l Marketing, Sales, and Finance - Allowing access to individual’s purchases
18
Data Mining as a Threat to Privacy l Data mining gives us “facts” that are not obvious to human analysts of the data l Can general trends across individuals be determined without revealing information about individuals? l Possible threats: - Combine collections of data and infer information that is private l Disease information from prescription data l Military Action from Pizza delivery to pentagon l Need to protect the associations and correlations between the data that are sensitive or private
19
Some Privacy Problems and Potential Solutions l Problem: Privacy violations that result due to data mining - Potential solution: Privacy-preserving data mining l Problem: Privacy violations that result due to the Inference problem - Inference is the process of deducing sensitive information from the legitimate responses received to user queries - Potential solution: Privacy Constraint Processing l Problem: Privacy violations due to un-encrypted data - Potential solution: Encryption at different levels l Problem: Privacy violation due to poor system design - Potential solution: Develop methodology for designing privacy- enhanced systems
20
Some Directions: Privacy Preserving Data Mining l Prevent useful results from mining - Introduce “cover stories” to give “false” results - Only make a sample of data available so that an adversary is unable to come up with useful rules and predictive functions l Randomization - Introduce random values into the data and/or results - Challenge is to introduce random values without significantly affecting the data mining results - Give range of values for results instead of exact values l Secure Multi-party Computation - Each party knows its own inputs; encryption techniques used to compute final results - Rules, predictive functions l Approach: Only make a sample of data available - Limits ability to learn good classifier
21
Privacy Constraint Processing l Privacy constraints processing - Based on prior research in security constraint processing - Simple Constraint: an attribute of a document is private - Content-based constraint: If document contains information about X, then it is private - Association-based Constraint: Two or more documents taken together is private; individually each document is public - Release constraint: After X is released Y becomes private l Augment a database system with a privacy controller for constraint processing
22
Architecture for Privacy Constraint Processing User Interface Manager Constraint Manager Privacy Constraints Query Processor: Constraints during query and release operations Update Processor: Constraints during update operation Database Design Tool Constraints during database design operation Database DBMS
23
Semantic Model for Privacy Control Patient John Cancer Influenza Has disease Travels frequently England address John’s address Dark lines/boxes contain private information
24
Data Mining and Privacy: Friends or Foes? l They are neither friends nor foes l Need advances in both data mining and privacy l Need to design flexible systems - For some applications one may have to focus entirely on “pure” data mining while for some others there may be a need for “privacy-preserving” data mining - Need flexible data mining techniques that can adapt to the changing environments l Technologists, legal specialists, social scientists, policy makers and privacy advocates MUST work together
25
Platform for Privacy Preferences (P3P): What is it? l P3P is an emerging industry standard that enables web sites t9o express their privacy practices in a standard format l The format of the policies can be automatically retrieved and understood by user agents l It is a product of W3C; World wide web consortium www.w3c.org l When a user enters a web site, the privacy policies of the web site is conveyed to the user l If the privacy policies are different from user preferences, the user is notified l User can then decide how to proceed
26
P3P and Legal Issues l P3P does not replace laws l P3P work together with the law l What happens if the web sites do no honor their P3P policies - Then appropriate legal actions will have to be taken l XML is the technology to specify P3P policies l Policy experts will have to specify the policies l Technologies will have to develop the specifications l Legal experts will have to take actions if the policies are violated
27
Challenges and Discussion l Technology alone is not sufficient for privacy l We need technologists, Policy expert, Legal experts and Social scientists to work on Privacy l Some well known people have said ‘Forget about privacy” l Should we pursue working on Privacy? - Interesting research problems - Interdisciplinary research - Something is better than nothing - Try to prevent privacy violations - If violations occur then prosecute l Discussion?
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.