RELATIONAL DATA WATERMARKING: ROBUST SCHEMES

This course provides essential knowledge and skills that directly support the PhD program in Computer Science, with a particular emphasis on cybersecurity.
It equips students with both theoretical and practical foundations in information hiding techniques, focusing specifically on digital watermarking for relational databases.
The course primarily addresses robust schemes designed to ensure ownership protection and enable copy tracing. Its objectives are to analyze, apply, and extend robust watermarking techniques for relational databases, with particular attention to adversary models.
Emphasis is placed on classifying schemes in terms of robustness versus fragility, and on examining the different types of attacks that may compromise watermark synchronization.
By meeting these objectives, students will be able to design data protection solutions that safeguard access and distribution rights, thereby broadening the potential applications of digital data.

At the end of this course, the students will be able to:
- Identify how security and robustness differ and how they can be combined independently, with or without affecting data quality.
- Identify threats at different levels that may compromise the preservation of the watermark.
- Design and implement advanced watermark-based protection strategies that will boost watermark preservation, contributing not only to robustness but also to security.

Among the main prerequisites of the course, it is required:
- The capacity to understand (read and listen) the content presented in English and the ability to communicate (write and speak) using this language.
- To know the basics of digital watermarking and relational data watermarking.

Note: Although a basic understanding of relational database theory is desirable, this is not mandatory for students, considering a brief introduction to this topic is included in the course.

1. Introduction to digital watermarking (the adversary model)
2. Robust and fragile techniques
3. Side-effects of relational data protection
4. Lab. Implementing basic watermark detectability rules
5. Virtual Primary Key (VPK) based synchronization
6. Lab. Implementation and evaluation of VPK schemes
7. First-degree attacks
8. Lab. Facing first-degree attacks (pseudo-random selection and majority voting)
9. Second-degree attacks (robustness vs. security)
10. Lab. Facing second-degree attacks (false ownership claims, collusion, & brute force attacks).

- Agrawal, R., & Kiernan, J. (2002, January). Watermarking relational databases. In VLDB'02: Proceedings of the 28th International Conference on Very Large Databases (pp. 155-166). Morgan Kaufmann.
- Halder, R., Pal, S., & Cortesi, A. (2010). Watermarking techniques for relational databases: Survey, classification and comparison. Journal of universal computer science, 16(21), 3164-3190.
- Rani, S., & Halder, R. (2022). Comparative analysis of relational database watermarking techniques: An empirical study. IEEE Access, 10, 27970-27989.
- Cox, I., Miller, M., Bloom, J., Fridrich, J., and Kalker, I., Digital Watermarking and Steganography. Morgan kaufmann, 2007.
- Barni, M. and Bartolini, F., 2004. Watermarking systems engineering: enabling digital assets security and other applications. Crc Press.

The main evaluation of the course takes place at the end of the program and consists of a 45-minute written exam in the form of a test, including matching questions, true/false items, completion tasks, and other formats. The exam is designed to assess the student’s ability to identify appropriate solutions to situations that may compromise the ownership of a relational database. It covers both theoretical and practical aspects, with a focus on the core concepts of robust watermarking for relational data and the adversary model.
During the exam, the use of printed materials (such as books, articles, or class notes) is not permitted. The use of cell phones or other electronic devices, including tablets and laptops, is also strictly prohibited.
In addition to the written exam, students will be evaluated on their performance in labs and lessons. This performance will serve as an individual criterion considered in the professor’s final assessment. Thus, while the written exam constitutes the main evaluation, student performance throughout the course also contributes to the overall assessment, which, together with the exam, determines the final grade.
The written exam carries a maximum of 25 points out of 30. The remaining 5 points are awarded on the basis of individual performance throughout the course.
To encourage participation and support learning, each lesson includes examples followed by questions, with exercises first completed collectively and then individually. This approach ensures that any uncertainties regarding the evaluation process are clarified in advance. In addition, students have access to supplementary materials on the university’s e-learning platform (moodle.unive.it).

written

Regarding the grading scale, scores will be assigned according to the following scheme:
A. Scores in the 18–22 range will be awarded when the student demonstrates sufficient knowledge of the content delivered in class and an adequate ability to analyze and investigate watermarking methods for relational databases.
B. Scores in the 23–26 range will be awarded in the presence of a fair knowledge of the content taught in class, as well as a satisfactory performance in the analysis and investigation of watermarking methods for relational databases, and the ability to consider solutions related to other disciplines or security methods.
C. Scores in the 27–30 range will be awarded in the presence of good or excellent knowledge of the content offered in the lessons and strong analytical and investigative skills regarding the security concepts and methods presented in the cases. At this level, the student must demonstrate greater independence in conducting research, as well as a high ability to analyze and integrate watermarking methods with other solutions or disciplines.
D. “Cum laude” will be awarded in the presence of excellent knowledge of the content and outstanding performance in participation and implementation of solutions discussed in class.

This in-person course incorporates structured learning modules available on the university’s online learning platform, moodle.unive.it. These modules host materials considered essential to the course and serve as a reference for students throughout their studies. They are designed to engage students in solving real-world practical cases, fostering critical thinking and problem-solving skills. Additionally, the platform allows students to compare their solutions with those proposed by the instructor or applied in professional practice, thereby reinforcing understanding and promoting the integration of theoretical knowledge with practical application.

This subject deals with topics related to the macro-area "Cities, infrastructure and social capital" and contributes to the achievement of one or more goals of U. N. Agenda for Sustainable Development