Teacher

Prof. Lam Wai
wlam@se.cuhk.edu.hk
 

Annoucement

• Quiz 3 will be held on April 14 (Thur) 15:30 as originally planned. The scope will cover the topics "Latent Semantic Model", "Link Analysis", "Vector Semantics and Embeddings", "Sequence Processing with Recurrent Networks and Transformers", "Encoder-Decoder and Attention", "Transfer Learning with Pretrained Models and Contextual Embeddings".
• Quiz 2 will be held on Mar 10 (Thur) 3:30pm. It covers from the topic "Evaluation of IR systems" up to the topic "Text Classification - kNN and Linear Classifier".
• Quiz 1 will be held on Feb 17 (Thur) 4:15pm. It covers from the first topic to "Term Weighting and Vector Space Model".
• Since your mobile phone will be used as a proctoring device for at least 2 hours while taking video, make sure it is fully charged. You must prepare enough sheets of blank paper. You must get ready 5 minutes before the quiz start time.
• Since the quiz is changed to online, the quiz date and arrangement have been changed. There will be an additional Quiz0 on Feb10 and the usual Quiz1 on Feb17. Quiz0 is a short quiz testing only your familiarity with the instructions of the online quiz. Quiz0 will carry some points towards the total quiz score. The instructions can be found right below the Announcement section. On Feb 10, I will conduct lecture 15:30-16:15, then it is followed by Quiz0 at 16:30. (If you intend to join Quiz0 only, you can join at 16:30.)

Instructions for Online Quiz

Course Schedule and Content (ongoing updates)

Course Outline

1. basic information retrieval (IR)
2. vector space model
3. evaluation in IR
4. Naive Bayes text classification
5. text classification - linear classifiers and SVM
6. link analysis
7. vector semantics
8. word embeddings
9. neural models
10. recurrent networks, encoder-decoder
11. neural attention
12. neural IR

Reference Books

• "Introduction to Information Retrieval", C. Manning, P. Raghavan, H. Schutze, Cambridge University Press. (A copy has been reserved at the CUHK main library)
• "Speech and Language Processing", 3rd Edition, D. Jurafsky and J. Martin.
• "An Introduction to Neural Information Retrieval", B. Mitra and N. Craswell, Foundations and Trends in Information Retrieval, 13(1), pp.1-126, 2018.
• "Mining Text Data", C.C. Aggarwal and C.X. Zhai, Springer Science.

Lecture Notes

• Introduction to Information Retrieval and Boolean Models Ref: Manning's book Chapter 1 and 2.4
• Text Preprocessing Ref: Manning's book Chapter 2.1 and 2.2
• Parametric Search and Machine Learning Based Scoring Ref: Manning's book Chapter 6.1
• Term Weighting and Vector Space Model Ref: Manning's book Chapter 6.2, 6.3, and 6.4
• Evaluation of IR systems Ref: Manning's book Chapter 8
• Basic Probabilistic IR Ref: Manning's book Chapter 11
• Text Classification - Naive Bayes Ref: Manning's book Chapter 13
• Sample Application - Bayesian spam filtering
• Text Classification - kNN and Linear Classifier Ref: Manning's book Chapter 14
• Text Classification - basic SVM Ref: "The Elements of Statistical Learning", by T. Hastie, R. Tibshirani, J. Friedman, Springer Section 4.5, 12.1-3
  Section 4.5
  Section 12.1-3
• A case study on application of SVM for aviation report classification
• Introduction to Lucene IR Package
  demos for Lucene usage
• Latent Semantic Model Ref: Manning's book Chapter 18
• Latent Semantic Indexing - An Example
• Latent Semantic Analysis Application
• Text Classification Demo
  demos for text classification
• Link Analysis Ref: Manning's book Chapter 21
• Vector Semantics and Embeddings Ref: Jurafsky's book Chapter 6
• Neural Networks and Natural Language Model Ref: Jurafsky's book Chapter 7
• Sequence Processing with Recurrent Networks and Transformers Ref: Jurafsky's book Chapter 9
• Encoder-Decoder and Attention Ref: Jurafsky's book Chapter 10
• Transfer Learning with Pretrained Models and Contextual Embeddings Ref: Jurafsky's book Chapter 11
• Demo on neural models for text
  Link for the code

Assessment Scheme

• Quiz (in-class) 65%
• Presentation 15%
• Project Report 20%

Project

Project description is given here
The project presentation time of each student is less than 10 minutes. There is no live presentation. You should record your project presentation via Zoom with camera on showing your face, and save to the cloud. Then submit the link to the cloud to the Blackboard system before Apr 29 (Fri) 23:59. The deadline of submitting an electronic version of the project report via the Blackboard system is May 4 17:00.