Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Striking a Balance between Classical and Deep Learning Approaches in Natural Language Processing Pedagogy

Aditya Joshi, Jake Renzella, Pushpak Bhattacharyya, Saurav Jha, Xiangyu Zhang

TL;DR

The paper investigates the role of classical NLP in modern curricula by comparing two introductory NLP courses (NLP-UNSW and NLP-IITB) taught in Australia and India. It analyzes textbooks, course outlines, lecture plans, and assessments to propose a balanced, interleaved approach that preserves intuition-building and annotation while incorporating deep learning like transformers. Key contributions include empirical observations from courses, detailed examples of assignments and projects, and a reasoning framework grounded in cognitive load theory and CS1 analogies. The findings suggest classical methods support understanding of NLP problems and even DL models themselves, making a case for their inclusion in today’s NLP pedagogy. This work provides practical guidance for educators designing curricula that bridge traditional and neural approaches.

Abstract

While deep learning approaches represent the state-of-the-art of natural language processing (NLP) today, classical algorithms and approaches still find a place in NLP textbooks and courses of recent years. This paper discusses the perspectives of conveners of two introductory NLP courses taught in Australia and India, and examines how classical and deep learning approaches can be balanced within the lecture plan and assessments of the courses. We also draw parallels with the objects-first and objects-later debate in CS1 education. We observe that teaching classical approaches adds value to student learning by building an intuitive understanding of NLP problems, potential solutions, and even deep learning models themselves. Despite classical approaches not being state-of-the-art, the paper makes a case for their inclusion in NLP courses today.

Striking a Balance between Classical and Deep Learning Approaches in Natural Language Processing Pedagogy

TL;DR

The paper investigates the role of classical NLP in modern curricula by comparing two introductory NLP courses (NLP-UNSW and NLP-IITB) taught in Australia and India. It analyzes textbooks, course outlines, lecture plans, and assessments to propose a balanced, interleaved approach that preserves intuition-building and annotation while incorporating deep learning like transformers. Key contributions include empirical observations from courses, detailed examples of assignments and projects, and a reasoning framework grounded in cognitive load theory and CS1 analogies. The findings suggest classical methods support understanding of NLP problems and even DL models themselves, making a case for their inclusion in today’s NLP pedagogy. This work provides practical guidance for educators designing curricula that bridge traditional and neural approaches.

Abstract

While deep learning approaches represent the state-of-the-art of natural language processing (NLP) today, classical algorithms and approaches still find a place in NLP textbooks and courses of recent years. This paper discusses the perspectives of conveners of two introductory NLP courses taught in Australia and India, and examines how classical and deep learning approaches can be balanced within the lecture plan and assessments of the courses. We also draw parallels with the objects-first and objects-later debate in CS1 education. We observe that teaching classical approaches adds value to student learning by building an intuitive understanding of NLP problems, potential solutions, and even deep learning models themselves. Despite classical approaches not being state-of-the-art, the paper makes a case for their inclusion in NLP courses today.
Paper Structure (18 sections, 6 figures, 2 tables)

This paper contains 18 sections, 6 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Context
  3. Parallels to Computer Science Education
  4. Observations from NLP textbooks and courses
  5. Lecture Plan
  6. Coding Assessments
  7. Course Assignment
  8. Course Project
  9. Project evaluation.
  10. Findings.
  11. Tutorials
  12. Making the Case for Classical Approaches
  13. Intuition-building
  14. Student Motivation
  15. Popular classical approaches
  16. ...and 3 more sections

Figures (6)

  • Figure 1: Course Structure for NLP-UNSW.
  • Figure 2: Course Structure for NLP-IITB.
  • Figure 3: Lecture Plan for NLP-UNSW.
  • Figure 4: Lecture Plan for NLP-IITB.
  • Figure 5: Word clouds showing the key deep learning and classical methods explored in the group projects.
  • ...and 1 more figures