About the Course

Information retrieval -aka “search”- plays a central role in our modern digital lives. In this course we cover the fundamental concepts of information retrieval as well as some of the recent advances in the field such as the use of knowledge graphs for retrieval, neural methods for retrieval tasks, issues of fairness and fake news, and the use of succinct data-structures in building efficient search systems.

Objectives

Understand and be able to discuss concepts such as document representation methods, information needs, search result effectiveness metrics and web search engine architectures. Implement and use retrieval algorithms; test them on standard and large-scale data collections. Apply information retrieval and web search methods to solve real-world problems, appreciate their impact on modern everyday life.

Sparse models for retrieval (Boolean, vector-space, probabilistic, language-model, Markov random fields, diversity-aware);
Design of test collections (TREC, crowd-sourcing) and retrieval effectiveness measures (micro-/macro-F measure, nDCG, BPref);
Collection models (multinomial repr.; topic mixtures) and topic modeling (LSA/LSI, LDA);
Search engine architecture (crawling, indexing, and web-page ranking);
Learning to rank including neural ranking;
Dense retrieval models;
Use of LLMs in retrieval and evaluation;
Agentic IR

NOTE: The course will involve a significant level of programming to process large datasets with focus on efficiency as well as quality of results.

Prerequisites:

datastructures and algorithms (COL106)
probability and statistics (MTL106 or equivalent)
comfort with programming in Java/C++/Python, and with linear algebra.
background in machine learning and/or NLP is not mandatory (although it will help if you have)

About the Course

Objectives

Contents

Prerequisites: