C2: Classic and Advanced IR Models
This module contains two sub-modules: Classic IR Models and Advanced IR Models.
It takes around 90 minutes per sub-module to go through the class notes and practice exercises. It is highly recommended that students read Chapter 2 of the textbook Modern Information Retrieval, written by Ricardo Baeza-Yates and Berthier Ribeiro-Neto.
Sub-Module: Classic IR Models
Objective
After learning this sub-module, students will have a general understanding about three classic information retrieval models: The Boolean model, vector model, and probabilistic model.
Description
This sub-module introduces concepts related to classic IR models. Major topics include:
- Brief description of operational modes related to IR models.
- A Precise characterization of an IR model;
- Basic Concepts and definitions;
- Description of classic IR models:
- Boolean
- Vector
- Probabilistic
- Brief Comparison of classic models
Class Notes
(Required)
- Basic Concepts of Classic IR Models [flash] | [Windows Media] | [.ppt] | [.doc] | [.pdf]
- Classic IR Models [flash] | [Windows Media] | [.ppt] | [.doc] | [.pdf]
Exercises
- Exercise: Intro to Classic IR Models
[Online Exercise] [.doc] [.pdf] - Exercise: Boolean Model Quiz
[Online Exercise] [.doc] [.pdf] - Exercise: Vector Model Quiz
[Online Exercise] [.doc] [.pdf]
Suggested Readings and Resources
In the Book
- Chapter 2 of textbook Modern Information Retrieval, written by Ricardo Baeza-Yates and Berthier Ribeiro-Neto, ISBN: 0-201-39829-X, Publisher: Addison-Wesley, 1999.
On the Web
- Boolean algebra
- Boolean Model-pdf
- Vector Model Information Retrieval by Rich Ackerman
- Probability-Bayes' Theorem
- Probabilistic Models in Information Retireval by Norbert Fuhr
- Open Source Search Engines in Java
Sub-Module: Advanced IR Models
Objective
To provide an overview and brief explanations highlighting the major points of advanced IR models. After learning this module, students should have a basic understanding of advanced IR models.
Description
This sub-module introduces concepts related to classic IR models. Major topics include:
- Description of advanced IR models:
- Extended Boolean
- Generalized Vector Space
- Latent Semantic Indexing
- Neural Network Model
Class Notes
(Required)
- Advanced IR Models [flash] | [Windows Media] | [.ppt] | [.doc] | [.pdf]
Suggested Readings and Resources
In the Books- Chapter 2 of textbook Modern Information Retrieval, written by Ricardo Baeza-Yates and Berthier Ribeiro-Neto, ISBN: 0-201-39829-X, Publisher: Addison-Wesley, 1999.
