This course homepage is accessible from https://www.cse.ust.hk/~dlee/csit6000I/
Fall 2019
Course and Instructor/TA Information
Instructor: |
Prof. Dik Lun Lee |
Email: |
dlee@cse.ust.hk |
Office: |
3534 (Lift 25/26) |
Office Hours: |
By email appointment |
|
|
TA: |
Xiang Li |
Email: |
xiang.li@connect.ust.hk |
Office: |
via email |
Office Hours: |
By email appointment |
|
|
Lectures Time: |
Saturday 10:30AM - 1:20PM |
Lecture Room: |
Rm 2464, Lift 25-26 |
|
Course Outline
Week 1 |
Sep 7 |
Introduction and course overview; Information Retrieval models: Boolean model |
Week 2 |
Sep 21 |
Information Retrieval models: vector-space model and term weighting |
Week 3 |
Sep 28 |
Indexing and query processing |
Week 4 |
Oct 5 |
Web-based information retrieval: Early projects, PageRank, |
Week 5 |
Oct 12 |
Hub and authority webs, applications beyond search |
Week 6 |
Oct 19 |
Search
engine architectures: Distributed search, metasearch |
Week 7 |
Oct 26 |
Relevance feedback: Implicit and explicit feedback (mid-term 1 in last hour of lecture) |
Week 8 |
Nov 2 |
Experimental Evaluation of ranking and relevance feedback algorithms |
Week 9 |
Nov 9 |
Implicit feedback: query log analysis, learning to rank, Personalization |
Week 10 |
Nov 16 |
Introduction to Recommendation Systems (RSs) |
Week 11 |
Nov 23 |
Curiosity based and industrial projects in RSs (mid-term 2 in last hour of lecture) |
Week 12 |
Nov 30 |
Summarization Systems |
Detailed topics and slides >>> (Use your student ID as username and paasswordd)
Homework |
30% |
3 homework assignments |
Homework 1 questions , Solution |
|
|
Homework 2 questionsHomework 3 questions Solution new! |
|
|
Exercise |
|
|
Mid TermMid Term 1 SolutionMid Term 2 Solution new! |
30% |
Two mid-term exams, 1 hour each, in lectureHigher of the two mid-term scores is taken- Open notes, bring a calculator |
Class participation |
5% |
0.5% for each lecture attended (5% max) |
Final Exam |
35% |
2-2.5 hours |
Bonus |
5% |
Participated in Q/A and discussion in lectures |
Course Description
Information retrieval techniques; traditional indexing, searching and ranking; search methods for web data, personalization, learning to rank; applications
At the end of the objective, the students will understand:
Homework assignments must be done individually. Collaboration between students is strictly forbidden. Any violation will be passed to the Department's Undergraduate/Postgraduate Studies Committee for assessment. The result may lead to dismissal from the University.
How is bonus considered?
Grades are first assigned to all students without considering bonus points. Thresholds between subgrades are set. Then, bonus points are added to students. A student’s grade will be re-assigned (moved up) according to his/her new score. The end result is that students who do not have bonus points will not be penalized by other students having bonus points.
Warning on Open Book/Note Exams
Both the mid-term and final exams are open book. You can bring your lecture notes (slides and notes) and any printouts to the exam venue. You can written on the notes. While you do not need to memorize everything (formula and pseudo code, etc.) by heart, the examinations are set assuming you know the materials well. That is, the notes/slides are there to help you with “is my cosine similarity formula correct?” and “if the PR formula 1-p… or p – 1 …” etc., but flipping through the slides page by page to find the answer of a question would waste too much time. At the end you do not have enough time to finish all of the questions. Bear in mind that you still need to study hard!