More on Projects for Comp 8390

*******

Note that this information is also posted on the announcement page

of the course web page.

http://cs.uwindsor.ca/~cezeife/courses/60-539/

Remember that project presentations will be made on

Mar. 31.

Each project (group or individual) is presented in roughly 5 to 10

minutes giving information on:

Project topic, work involved and importance

New things learnt for the project

Relevance to data warehousing or mining

Parts of projects completed and parts left to complete

Note that for group projects, each group member's contribution has to

be stated clearly. It is better if each member of the group

actually participates during the presentation. Note that

all members of a project group might not necessarily get the

same mark if they do not contribute the same amount of work

and do not show the same level of understanding in the work

involved.

Note that research projects should clearly present the topic

with an example discussing a few techniques in the literature

already for solving this topic. For example, a research in

Data Warehousing Cleaning, should clearly summarize what the

topic is about and present with an example about 2 or 3 methods

they already researched for handling data warehouse cleaning.

A comprehensive list of techniques must be included in the final

report that must have not less than 10 current and most

relevant references with at least 5 papers as the core of the report.

Question time takes between 3 and 5 minutes.

By your project presentation date, I expect that you have

completed at least 50% of the work involved in the project.

The complete project report and demonstration (if a system is built

or an algorithm implemented) has to be done by project demonstration date of April 15th.

Final Project demonstration (different from project presentation) time for each project should be booked

for Monday, April 7th. Project demo time slots

are: 2:30 - 3:00pm, 3:30 - 4:00pm, 4:00 - 4:30, 4:30 - 5:00pm, 5:00-5:30pm.

Reports should be submitted at least a day or two before demo.

Project Report

-----

Report should not be more than 20 pages on 12 point, 1.5 or double line

spaced font and should include the following points:

(a) purpose of application,

(b) system designers and their contributions,

(d) summary of interface,

(e) features of the system built or studied

(f) installation guide,

(g) listing of all files involved in the application and their functions,

and relationships to each other.

(h) user manual and any other pertinent information.

1. More on Seminars

********

Remember that research seminars begin on Monday, March 11. Refer back to the seminar schedule list I gave to you for information on who

is presenting when and which papers to read in preparation.

Students are expected to grade seminar based on clarity, organization,

quality of talk and technical content. Clarity judges how you

understand what the presenter is giving from the way they present it

(not based on your lack of concentration on what they are talking

about). Organization judges how cohesive (the flow or togetherness) of

the material taught (e.g., do they clearly tell you the topic, explain the purpose before discussing the solutions and results or do they

jump in and talk of results when you do not even know what the

topic is about. Efforts put into slide presentations also come in

here). Quality of talk judges the correctness of the presentation

from the paper presented (that is, how well does the presenter

understand materials in this paper or are they providing wrong

information?). Technical content judges level of difficulty of

the paper presented and efforts put in by the student. Some

students avoid actually trying to understand and explain the

technical solution (algorithms) presented in the paper, but

rather spend too much time discussing purpose and performance

analysis.

Recall that seminar reports should be a maximum of 5 pages

in 12-point double line format for your research paper you are

supposed to summarize. That is, a maximum of 5 pages for one

paper. Hand in your research report on or after the day

your research paper is presented. Last reports should be submitted

on the last day of seminar presentations.

On seminars, the course information documents 1 and 2 with their links also found through the course announcement page, provide some details such as:

Seminar report should clearly state title of paper, authors, proceedings

or journal it came from, year of publication, name of student and

seminar group. Then, include clearly, problem addressed by paper,

contributions of this paper, solution provided with clear algorithms and

running example, limitations and advantages of solution and your opinion of work. Seminar reports are due on the day of presentation of the paper or latest the last day of seminar presentations.

For seminar presentation, the goal is to focus on making clear:

Then, include clearly, problem addressed by paper,

contributions of this paper, solution provided with clear algorithms and

running example, limitations and advantages of solution and your opinion of work. Check for example, how I tried to present the Apriori and the FP-tree growth algorithms for mining frequent patterns. Follow that approach as much as possible. Think of the Input data, Output data, then the algorithm steps for the solution presented by paper. Summarize these first and try to use an example to make yourself clear.

Check a sample presentation power-point slides I left for you in the Comp 8390 (60-539) web page, database tools. The link is:

http://cezeife.myweb.cs.uwindsor.ca/courses/60-539/databases/index.htm.

Please, find a sample seminar presentation grading sheet I will hand out to class during each seminar or keep in Brightspace folder, for their own assessment of your presentation. Focus on grading a seminar is on clarity (2 marks), organization (2 marks), Quality of presentation (3 marks), Technical content (3 marks) for 10 marks total.

You should also start to book the time slot for your project

demo.

2. More on Test

****

Test covers materials taught in the lecture part of course and distributed into 3 sections for i) database foundation, ii) data warehouse, XML and NOSQL databases and iii) data mining techniques. The sample test handed out in class is a good model with a fairly good coverage of test materials (note that XML and NOSQL Databases are new addition and not included in the sample test).

For section A of test, Know how to design source data bases that are normalized, querying through Sql, relational algebra and calculus; disk anatomy (e.g., what is a sector, cylinder,
platter, track and give the size of a sector, what is the size of
a track, disk etc.). File organizations and Indexes etc.
For section B, know how to design data warehouses from source databases and multi-dimensional modelling aspects of the data warehouse, data warehouse querying. Also know the differences between components of other non-traditional databases discussed in class (XML and NOSQL databases) and the traditional databases.

For section C, know how to mine association rules using the basic Apriori and FP-growth techniques, olap querying and operations such as drill-down and roll-up etc.