Παρουσίαση/Προβολή

Course Description

The goal of M149 is to enable students to understand the theory and the techniques behind modern database systems. Students will learn about data storage and management, query optimization and execution, data access methods, modern data exploration and search approaches, and dive into cutting-edge technologies in the intersection of deep learning and database technologies. They will be able to understand the data ecosystem, from the database core going up to the user-facing search and exploration interfaces. They will be able to transfer this knowledge to get easily familiarized with different database systems going forward. Core concepts, such as query cost optimization, are also valuable for learning to build efficient solutions and algorithms to any data problem and application.

Grading

1. Written tests (30%)
2. Labs (20%)
3. Project (50%)

Syllabus

1. Introduction to DBMS, Relational Algebra, SQL (Operators, Operations, Nested Queries)  
2. Storage (Pages, Buffers, Heap Files, Data representation of Tuples in Postgres, Indexes, Column Stores, Memory Networks)  
3. Execution Algorithms (Joins, Sorts, Aggregations)  
4. Query Optimization (Rewriting, Cardinality Estimation, Cost estimation, Join Orders)
5. Query Execution (Transactions, ACID Properties, Concurrency control, Locks, Parallel and Distributed Execution, Knob tuning) 
6. Recovery System (ARIES algorithm )  
7. Integration of Deep Learning in DBMS, ​
8. Natural Language Data Interfaces (NL-to-SQL, SQL-to-NL, Data-to-Text) ​
9. Data Exploration (Query Recommendations)  ​
10. NoSQL DBMSes 

Βιβλιογραφία

• Database System Concepts (Avi Silberschatz, Henry F. Korth, S. Sudarshan, 7th edition, ISBN 9780078022159)
• Mention Memory: incorporating textual knowledge into Transformers through entity mention attention (M. D. Jong, Yury Zemlyanskiy, Nicholas FitzGerald, Fei Sha, W. Cohen, ArXiv abs/2110.06176 (2021))

• AI Meets Database: AI4DB and DB4AI (Guoliang Li,  Xuanhe Zhou, Lei Cao, SIGMOD '21: Proceedings of the 2021 International Conference on Management of Data, June 2021)

• Bao: Making Learned Query Optimization Practical (Ryan Marcus, Parimarjan Negi, Hongzi Mao, Nesime Tatbul, Mohammad Alizadeh, Tim Kraska, SIGMOD '21: Proceedings of the 2021 International Conference on Management of Data)

• Zero-shot cost models for out-of-the-box learned cost prediction (Benjamin Hilprecht, Carsten Binning, Proceedings of the VLDB Endowment, Volume 15, Issue 11, July 2022)

• Data-to-Text Generation with Content Selection and Planning (Ratish Puduppully, Li Dong, Mirella Lapata, AAAI'19: AAAI Conference on Artificial Intelligence Honolulu Hawaii USA 27 January 2019- 1 February 2019)

• Text-to-Text Pre-Training for Data-to-Text Tasks (Mihir Kale, Abhinav Rastogi, Proceedings of the 13th International Conference on Natural Language Generation, December 2020, Dublin, Ireland)

• Explaining Natural Language Query Results (Daniel Deutch, Nave Frost, Amir Gilad, The VLDB Journal (2020) 29:485–508)

• Explaining Queries over Web Tables to Non-Experts (Jonathan Berant, Daniel Deutch, Amir Globerson, Tova Milo, Tomer Wolfson, 2019 IEEE 35th International Conference on Data Engineering (ICDE) (2018): 1570-157)

• A Survey on Deep Learning Approaches for Text-to-SQL (George Katsogiannis, Georgia Koutrika, The VLDB Journal (2023))

• Automating Exploratory Data Analysis via Machine Learning: An Overview. (Tova Milo, Amit Somech: SIGMOD Conference 2020: 2617-2622)

• Overview of Data Exploration Techniques (Stratos Idreos, Olga Papaemmanouil, Surajit Chaudhuri:. SIGMOD Conference 2015)

RELATED SOURCES:

• Proceedings of VLDB

• The VLDB Journal

• ACM Transactions on Database Systems

• IEEE Transactions on Knowledge and Data Engineering

• ACL Anthology

• ACM Transactions on Interactive Intelligent Systems

• The Journal of Personalization Research