Privacy and Security Issues in Data Mining and Machine Learning
International ECML/PKDD Workshop, PSDML 2010, Barcelona, Spain, September 24, 2010. Revised Selected Papers
Published on 16. April 2011
VII, 141 pages
978-3-642-19896-0 (ISBN)
Description
This book constitutes the refereed proceedings of the International ECML/PKDD Workshop on Privacy and Security Issues in Data Mining and Machine Learning, PSDML 2010, held in Barcelona, Spain, in September 2010.
The 11 revised full papers presented were carefully reviewed and selected from 21 submissions. The papers range from data privacy to security applications, focusing on detecting malicious behavior in
computer systems.
More details
Other editions
Additional editions
Christos Dimitrakakis | Aris Gkoulalas-Divanis | Aikaterini Mitrokotsa
Privacy and Security Issues in Data Mining and Machine Learning
International ECML/PKDD Workshop, PSDML 2010, Barcelona, Spain, September 24, 2010. Revised Selected Papers
Book
03/2011
1st Edition
Springer
€53.49
Shipment within 7-9 days
Content
- 6549
- Preface
- Organization
- Table of Contents
- Edit Constraints on Microaggregation and Additive Noise
- Introduction
- Data Editing
- Constrained Microaggregation
- An Overview of Microaggregation
- Edit Constraints and Microaggregation
- Additive Noise
- Edit Constraints and Additive Noise
- Noise Swapping
- Experiments
- Conclusions
- References
- Preserving Privacy in Data Mining via Importance Weighting
- Introduction and Framework
- Preserving Privacy via Importance Weighting
- How to Compute Importance Weights
- Research Questions
- Quadratic Error Minimization in a Distributed Environment with Privacy Preserving
- Introduction
- Related Work
- Quadratic Error Minimization
- Quadratic Error
- Gradient Descent Algorithm
- Experiments
- Cryptographic Tools
- The Protocols
- Protocol PQEM
- Protocol PDEM
- Complexity
- Discussion and Future Work
- References
- Secure Top-k Subgroup Discovery
- Introduction
- Preliminaries
- Privacy-Preserving Data-Mining and Secure Multi-party Computation
- Subgroup Discovery
- Distributed Secure Subgroup Discovery
- Computing the Maximum Quality
- Computing a Maximum Quality Subgroup
- The Protocol
- Top-k Subgroup Discovery and the Weighted Covering Scheme
- Prototypical Implementation
- Conclusions
- References
- ASAP: Automatic Semantics-Aware Analysis of Network Payloads
- Introduction
- The ASAP Framework
- Alphabet Extraction for Network Payloads
- Matrix Factorization
- Construction of Communication Templates
- Experiments and Applications
- A Showcase Analysis
- Analysis of Honeypot Data
- Analysis of Malware Communication
- Anomaly Detection
- Related Work
- Conclusions
- References
- Temporal Defenses for Robust Recommendations
- Introduction
- From One-Shot to Temporal Attacks
- Temporal Attack Model
- A Temporal Defence
- Global Monitoring - Many Sybils/Many Ratings Scenario
- User Monitoring - Few Sybils/Many Ratings Scenario
- Item Monitoring - Many Sybils/Few Ratings Scenario
- Related Work
- Conclusion and Future Work
- References
- SBAD: Sequence Based Attack Detection via Sequence Comparison
- Introduction
- Related Work
- Graph Based Dissimilarity Measure for Attack Detection
- Construction of Correlation Graph
- Sequence Coding
- Sequence Comparison and Dissimilarity Measure
- Experiment Result
- Datasets
- Evaluation Metrics
- Sensitivity Analysis and Robustness
- Effectiveness Analysis
- Real Case for Dissimilarity
- Conclusion
- References
- Classifier Evasion: Models and Open Problems
- Introduction
- The Near-Optimal Evasion Problem
- Security and the Near-Optimal Evasion Problem
- Previous Work
- Open Problems in the Theory of Near-Optimal Evasion
- Alternative Models for Evasion
- Additional Information about Training Data Distribution
- Beyond the Membership Oracle
- Evading Randomized Classifiers
- Querying with Real-World Objects
- Evading an Adaptive Classifier
- Conclusion
- References
- Large Margin Multiclass Gaussian Classification with Differential Privacy
- Introduction
- Differential Privacy
- Related Work
- Large Margin Gaussian Classifiers
- Differentially Private Large Margin Gaussian Classifiers
- Theoretical Analysis
- Proof of Differential Privacy
- Analysis of Excess Error
- Conclusion
- References
- Privacy Preserving Protocols for Eigenvector Computation
- Introduction
- Preliminaries
- Power Iteration Method
- Homomorphic Encryption
- Privacy Preserving Protocol
- Data Setup and Privacy Conditions
- The Basic Protocol
- Making the Protocol More Secure
- Extension to Multiple Parties
- Analysis
- Correctness
- Security
- Efficiency
- Conclusion
- References
- Content-Based Filtering in On-Line Social Networks
- Introduction
- Related Work
- Filtered Wall Conceptual Architecture
- Short Text Classifier
- Text Representation
- Machine Learning-Based Classification
- Content-Based Filtering with Blacklist
- A Case Study: DicomFW
- Problem and Dataset Description
- Demo Application
- Short Text Classifier Evaluation
- Overall Performance and Discussion
- Conclusions
- References
- Author Index
