Natural Language Processing 1

Course manual 2021/2022

Course content

This course aims at providing the student with the background that is needed for studying statistical models that are used in the field of Computational Linguistics. We will mostly depart from shallow labeling tasks and consider tasks that involve hierarchical structure (e.g., syntactic trees) and/or hidden structure (alignment of word and their translations in machine translation). For these tasks the course will concentrate on the fundamentals of probabilistic modeling and statistical learning from data by supervised and unsupervised statistical learning algorithms.

Study materials

Literature

• Daniel Jurafsky & James H. Martin, 'Speech and Language Processing' (3rd Edition)  [freely and legally available online]

Other

• Articles to be distributed during the course.

Objectives

• describe computational modelling methods for several levels of language analysis (morphology, syntax, semantics and discourse)
• discuss the strengths and limitations of these methods
• construct language processing models for several tasks, such as word representation learning, sentence representation learning, text classification etc.
• implement supervised (and some unsupervised) estimation procedures for these models
• evaluate these models experimentally and analyse their performance
• use the above techniques in NLP applications, such as sentiment analysis

Teaching methods

• Lecture
• Computer lab session/practical training
• Self-study

Semi-flipped classroom lectures and lab sessions

Learning activities

Attendance

This programme does not have requirements concerning attendance (OER part B).

Assessment

• exam 40%
• practical 1 (group work) 20%
• practical 2 (group work) 30%
• pen-and-paper exercises 10%

Assignments

Practical 1

• Sentiment analysis practical (part 1)

Practical 2

• Sentiment analysis practical (part 2)

Pen and paper exercises

Fraud and plagiarism

The 'Regulations governing fraud and plagiarism for UvA students' applies to this course. This will be monitored carefully. Upon suspicion of fraud or plagiarism the Examinations Board of the programme will be informed. For the 'Regulations governing fraud and plagiarism for UvA students' see: www.student.uva.nl

Course structure

Timetable

The schedule for this course is published on DataNose.

Additional information

recommended prior knowledge:

• probability theory
• basics of machine learning
• programming experience
• prior exposure to natural language processing / computational linguistics may be helpful 

 Maximum number of students: 40

Contact information

Coordinator

• dr. S. Pezzelle

Staff

• W. Ferreira Aziz
• Mario Giulianelli