Instructor: Yuhan Liu (Office: DC3301) & Mahdi Esmailoghli (Office: DC3620)
Lecture Room: MC 4059
Lecture Time: Tuesday & Thursday 8:30-9:50am
Office Hour Yuhan Liu (Tues 10:30-11:30am), Mahdi Esmailoghli ([out of office until Oct 19th] Wed 1:00-2:00pm)
Introduction to data engineering issues in data science. Data management technology objectives. Structured data management: Relational database technology, database workloads (OLTP vs OLAP). Big data issues: dealing with volume (geo-distributed, cluster parallel, and cloud-native data management), dealing with variety (data type-native systems, NoSQL database systems), dealing with velocity (streaming data management), and big data processing platforms (MapReduce, Spark). Data preparation pipeline: data acquisition, data integration (data warehouses, data lakes, lake houses), dataset selection, data quality and cleaning, data provenance management. Introduction to several current topics in database research, such as Large Language Models, vector databases.
Open to Master of Data Science and Artificial Intelligence students and others without an undergraduate course on database systems (instructor approval required).
This is a course that is specially designed for the data science program. It is an in-person course and no accommodations are made for remote attendance. Please make arrangements to attend lectures.
The course will use LEARN for dissemination of notes and for discussions.
We will be posting lecture slides on LEARN (look under Content/Course Slides). However, they may be posted shortly before lectures or sometimes even after a lecture. Some of them will be detailed, others just a skeleton. So, it is important to attend lectures to get the most from these.
There is no textbook for the course. Professor Özsu has started to write his notes and we’ll be posting them on LEARN (look under Content/Course Notes). We make no promises about the availability of these notes for every topic. We may assign reading from other textbooks and papers as appropriate.
We intend to have guest lecturers for some topics and will update the schedule as we get them confirmed. These guest lecture material are important and integral components of the course.
Some logistics:
Final exam schedule will be announced by the Registrar’s Office in due course and we cannot change the schedule. There will be no makeup for the final. You will need to take it the next time the course is offered (TBD). You have to pass the final exam to pass the course.
There will be four quizzes in the course. These will be 20-30 minute quizzes and will be taken online within LEARN.
There will be two paper reviews. The logistics of these will be revealed later.
Two 48 hour extensions per student are provided. They may be used on one of the two paper reviews (at most one may be used per paper review). Email us and the TA at least 24 hours before the deadline to let us know that you’re using it, and why. We will adjust the deadline on LEARN.
Students can use generative AI tools as aid, but have to write their own text in paper reviews. These will be checked using appropriate tools. Note that generative AI is known to hallucinate and may fabricate facts and inaccurately express ideas. They also commonly falsify references to other work.
In addition, you should be aware that the legal/copyright status of generative AI inputs and outputs is unclear. Exercise caution when using large portions of content from AI sources.
Bottom line: students are accountable for the content and accuracy of all work you submit in this class, including any supported by generative AI. You should be able to readily demonstrate your knowledge of your submissions. It is the students’ responsibility to check and use these tools responsibly.
Paper critiques (2): 40% Guidelines
Quizzes (4): 20%
Final: 40%