This course who want to understand more about the linguistic forces that hold a text together beyond sentences. Master-Students with an interest in NLP are welcome in this class. The ideal participant in this course already has some background in general NLP, but beginners in NLP are also welcome. The topic covered concerns a quite specific set of tasks that go beyond basic NLP. However, all background information that is required to understand the main concepts treated in this course will be summarised in the first set of lectures+activities.
Students will learn the theory of the core discourse phenomena (defined as all linguistic operations that add meaning beyond the single-sentence level), and NLP approaches to recognise these phenomena automatically in text. Exercises are an integral part of this intensive course. Students will experience the various analyses proposed in the literature themselves in first-hand in exercises, for instance by performing annotation exercises on real text, or by learning how to operationalise a theory into actual instructions. This way, they will encounter the level of ambiguity that is faced by automatic methods, rather than just seeing idealised examples chosen for illustrative purposes in the literature.
As a result, students who directly address discourse phenomena in their research should be able to put the new knowledge from this course into practice by creating more sophisticated automatic treatments of discourse phenomena. But students not directly working on discourse phenomena should also benefit. Background knowledge of discourse factors should help them
with the design of their own research work in related NLP applications (not treated here), such as information retrieval, sentiment detection, QA, summarisation, etc. All students, whatever their exact subjects, should hopefully profit by learning to design more meaningful evaluations of their NLP systems.
topic segmentation, lexical chains, pronoun resolution, rhetorical relations, argumentation mining, discourse structure
✔ Specialist skills | ✔ Intercultural skills | Communication skills | Critical thinking skills | Practical and/or problem-solving skills |
The course is provided as an intensive course in late August. We alternate lectures and activity.
Students will be required to read one paper (8-15 pages) per topic (6 papers in total). Because the course is intensive, they should ideally do this reading before the course starts, at least for the first three topics covered. Reading will be provided well ahead of schedule.
Course schedule | Required learning | |
---|---|---|
Class 1 | Lecture1: Basics of NLP (1.5 h) | Specified in the class |
Class 2 | Activity1: Basics of NLP (1.5 h) | |
Class 3 | Lecture2: Topics and Topic Segmentation (1.5 h) | |
Class 4 | Activity2: Topics and Topic Segmentation (1.5 h) | |
Class 5 | Lecture3: Lexical Chains (1.5 h) | |
Class 6 | Activity3: Lexical Chains (1.5 h) | |
Class 7 | Lecture4: Pronoun Resolution (1.5 h) | |
Class 8 | Activity4:Pronoun Resolution (1.5 h) | |
Class 9 | Lecture5: Definite Noun Phrase Resolution and Information Status (1.5 h) | |
Class 10 | Activity5: Definite Noun Phrase Resolution and Information Status (1.5 h) | |
Class 11 | Lecture6: Rhetorical Relations (1.5 h) | |
Class 12 | Activity6: Rhetorical Relations (1.5 h) | |
Class 13 | Lecture7: Argumentation Mining (1.5 h) | |
Class 14 | Activity7: Argumentation Mining (1.5 h) | |
Class 15 | Exam |
Not specified.
Not specified.
A final exam will be conducted on all topics treated in the course.
Those who want to understand more about the linguistic forces that hold a text together beyond sentences.
One week intensive course in late August.