|
Course Description Database systems have come along a long way since their
inception in the 1970s. Database Management Systems (DBMSs) have been widely successful
and are the heart of most information management system. However, there are a
number of significant challenges that future DBMSs must meet if they are to
continue playing the center role in information processing and management. We
are on the verge of a new revolution in ubiquitous computing in which
zillions of devices, ranging from small personal digital assistants (PDAs) to
“invisible” embedded sensor devices, will demand answers to
queries under a wide range of system conditions. These devices will rely on a
distributed backend infrastructure to deliver the query results. The data
sets in the back-end systems are growing at astonishing rates, demanding
scalable distributed data management techniques. Furthermore, the data sets
are increasingly complex, and are not limited to simple alphanumeric data
types (which traditional relational DBMS manage very effectively). Database
query processing and database storage techniques that exist today fall far
short of meeting the demands of these future systems. What then are the
techniques that will deliver this new world to us? This is the question that
we will explore in this course. The course will focus primarily on query
processing and query evaluation techniques that are likely to be applicable
in mobile, distributed, and sensor database environments of the future. Since
most of the questions in this area are unanswered, this course will be very
exploratory. The syllabus for this course is a list
of paper readings. Some of these papers are also included in the popular
“red book”, which is on reserve at the library. For roughly the
first half of the semester, I will present papers to bring you up-to-speed
with the database query processing and data management techniques that are
currently deployed in DBMSs, and also cover a few papers describing
techniques that are likely to be successful in the environment outlined
above. The second half of the semester will include paper presentation by the
students. We will cover ~2-3 papers a week and every student in the class
must read all these papers (including the papers that your fellow class-mates
present). Before each class you will be required to hand in a brief summary
(~10 sentences total) of the paper that will be discussed in the class. The
summary should not a facsimile of the abstract of the paper, but should be
your assessment of the key contributions and limitations of the paper.
The reviews will be graded on a scale of 0-4 (4 being the highest grade). Note for Masters Students: EECS 684 is approved for
500-level Masters credit. |
|
Course Project A class project, in which you pick a topic in this area
and explore it in detail, is a big component of this course. I will provide a
list of suggested project topics, though you are free to select a project
outside of this list provided you get prior approval. You will have to meet
with me every third week throughout the semester updating me on the progress
of your project. I will help with the direction, but unless you take the
initiative to actively explore the topic you choose, you are unlikely to
accomplish much in the project. I will expect that at the end of the course,
your project report should be at a level of a workshop/conference submission. For the projects you may work in a group. The maximum group size is 2, and I encourage most projects to be individual projects. My expectations will be scaled based on the group size. |
|
Time and Place TTH 9:00-10:30, 3437 EECS. |
|
Office Hours T 10:30-12:30, 4717 CSE, or by prior appointment. [Note OH are in the new CSE building next
to the DOW building] |
|
Prerequisites EECS 484 or equivalent. Note EECS 584 is not a pre-requisite for this course. |
|
Text No formal text. The reading list is a collection of papers,
which is posted on the course web page. Reference text: We will occasionally refer to the
following sources. Note you don’t need to purchase these textbooks; two
copies of each of these references are on reserve at the Media Union Library. ·
Red Book: ·
Cow Book: Database Management Systems (3rd
edition) - by Raghu Ramakrishnan and Johannes Gehrke, McGraw Hill 1999. |
|
Grading · One midterm: 35%. Date: TBD. ·
Course
Project: 50%. ·
Student
Paper Presentation: 5%. ·
Paper
Summaries and Class Participation: 10%. |