CS-563

Advanced Topics in Information Retrieval Systems

Postgraduate

This course focuses on modern and advanced topics in Information Retrieval, specifically on:
(A)  Background- Exploratory Search
(B) Natural Language Processing ○ Connection to Linguistics, Pre-processing, morphological, syntactical, semantic and pragmatic analysis ○ Entity Linking
(C) Semantic Similarity and Machine Learning Techniques in Information Retrieval ○ Classification, Regression, Word Embeddings, Document Embeddings, Word Similarity, Document Similarity
(D) Question Answering Systems ○ Typology of QA systems, techniques for question answering and dialogue systems over documents and data
(E) Open Information Extraction ○ Entity and Relation Extraction, Knowledge bases
(F) Other issues ○ ChatGPT ○ Machine Translation ○ Summarization ○ Techniques for Sentiment and Opinion Mining, Bias and Fairness in Information Access

LEARNING OUTCOMES
1. Knowledge: Having attended and succeeded in the course, the student is able to understand new technologies related to the latest developments in the scientific field of information retrieval (Embeddings, Large Language Models) and how selected information retrieval problems are addressed in the literature. .
2. Understanding: Having attended and succeeded in the course, the student will have
● the basic notions of natural language processing ● how similarity methods and embeddings work and how to use them ● how question answering systems work ● how machine learning can be used in the context of IR and techniques and algorithms for opinion mining
3. Application: Having attended and succeeded in the course, the student is able to  Carry out various basic NLP tasks through python ● Use and evaluate Named Entity Recognition tools ● Perform Document and Entity Similarity through several models ● Perform Machine learning tasks for several IR problems ● Design and implement a QA system ● Evaluate an interactive information retrieval system. ● Use existing models for any IR problem.
4. Analysis: Having attended and succeeded in the course, the student is able to analyze the steps of specific problems and treat them as subproblems.
5. Synthesis: Having attended and succeeded in the course, the student is able to use tools and methodologies from different sub-fields of information retrieval, e.g. Entity Recognition, Large Language Models, and others, so that he can achieve the best possible solution.
6. Evaluation: Having attended and succeeded in the course, the student is able to measure/quantify the quality of solutions of information retrieval problems and to compare them to other existing solutions.

STUDENT PERFORMANCE EVALUATION
Class attendance/participation
2 Exercises
1 project
2 presentations of papers


 

6

CS-463

http://www.csd.uoc.gr/~hy563