Computational Linguistics

Academic Year 2020/2021

Learning outcomes

The student will learn the basic theoretical aspects of computational linguistics/natural language processing and will acquire practical skills to perform from tokenization and vectorization to the computation of similarities and supervised models (e.g., for topic identification, structural analysis, meaning analysis).

Course contents

Whereas the contents could be (slightly) adapted according to the students skills and interests, the general structure of the course will be as follows.

0. A gentle introduction to Python

This topic wont be covered in class. If you didn’t attend to the crash course, visit it before the first session

1. Introduction to Computational Linguistics

[25/02/21] Slides

Intermezzo

7. Intro to LSA

[08/04/21] Slides LSA and SVD
[08/03/21] Notebook

8. Intro to NN

[09/04/21] Slides on the perceptron
[09/04/21] Notebook on the perceptron
[15/04/21] Slides introducing neural networks and keras
[15/04/21] Notebook introducing neural networks and keras

9. Word Embeddings

[16/04/21] Slides on word2vec
[22/04/21] Slides on word embeddings
[22/04/21] Notebook

10. Visualisation

[23/04/21] Slides on visualization
[23/04/21] Notebook

11. From document representations, towards sequences

[23/04/21] Slides

12. Convolutions for text

[29/04/21] Slides
[29/04/21] Notebook

13. Text is Sequential

[30/04/21] Slides on RNN
[30/04/21] Notebook on RNN
[06/05/21] Slides on BiRNN
[06/05/21] Notebook on BiRNN

14. Long Short-Term Memory Networks

[06/05/21] Slides
[06/05/21] Notebook
[07/05/21] Slides on characters and generation
[07/05/21] Notebook on characters
[07/05/21] Notebook on generation

Projects

The evaluation is based on a project. If you want some inspiration, look at the projects presented last year

Some project ideas

Whereas you are supposed to apply the acquired knowledge on a problem of your own interest, here are some ideas, in case you find yourself lost

Standard research

Performing research on propaganda identification in other languages than English. For inspiration, see this IPM paper , this EMNLP paper , or this SemEval shared task
Estimating the complexity of a text for a non-native speaker. For inspiration, see READ-IT
Analysing the quality of Wikipedia articles across languages

Shared Tasks

The ADoBO shared task on automatic detection of borrowings (deadline: 17th May 2021).
The FakeDeS fake news detection challenge (in Spanish): (deadline: 30th April 2021).
Some of the problems proposed in EVALITA 2020 . There is age and gender profiling , misogyny identification , and complexity evaluation , among many others.

Readings/Bibliography

Core

Hobson Lane, Cole Howard, Hannes Hapke (2019). Natural Language Processing in Action Understanding, analyzing, and generating text with Python . Manning Publications.
Dan Jurafsky and James H. Martin. Speech and Language Processing . 3rd ed. draft. December 30, 2020

Optional

Steven Bird, Ewan Klein, and Edward Loper. Natural Language Processing with Python
Yoav Goldberg. (2017). Neural Network Methods for Natural Language Processing (G. Hirst, ed.). Morgan & Claypool Publishers.
Emily M. Bender (2013). Linguistic Fundamentals for Natural Language Processing: 100 Essentials from Morphology and Syntax Synthesis . Lectures on Human Language Technologies. Morgan & Claypool Publishers.
Kenneth Ward Church. Unix for poets .

Teaching methods

The course is a combination of seminar and practical sessions. In either case, active participation of the students is expected. Assuming you know the basics of programming (e.g., by completing the python course in Topic 0 ) we will cover a (practical) description of diverse models and tasks.

Evaluation

The student will work on addressing a problem within her own research interests with the knowledge acquired during the course. Upon agreement of the topic, the student will work on solving the problem and will produce a written report. A poster session will be organized before at the end of the course (or before every appello )in which the students will present their research work.

The final evaluation will be computed as a combination of both report and poster presentation.

Important points

If the conditions (covid) are not favourable, we wont have a poster session. In that case, your project should be submitted 1 week before the appello to be considered valid.
Do you want to target 30L? Your final project should be submitted to a conference for publication. Talk to me well in advance if you aim this, as it would require my heavy involvement and a high quality will be necessary.

Teaching tools

Seminars will be carried out with slides and coding will be carried out with jupyter notebooks. Continuous exercises will be carried out.

Office hours

See my UniBO website