Special Topics In Artificial Intelligence
Public syllabus for 2025-2026
Academic overview
Teaching team
Learning time distribution
| Total | ||||||
|---|---|---|---|---|---|---|
| Curriculum | Lecture | Practice | Total Weekly | Lecture | Practice | |
| 42 | 28 | 14 | 3 | 2 | 1 | |
| Exam hours | ||||||
| 2 | ||||||
| Individual Study | Bibliography study | Field study | Homework | Tutoring | Others | |
| 106 | 46 | 19 | 40 | 1 | 0 | |
| Overall | ||||||
| 150 |
Learning outcomes
Knowledge
- Presentation and understanding of (1) the importance of NLP, with (2) application to use-cases using emerging technologies.
Skills
- The ability of communicating knowledge about the usage of NLP for different problems.
Responsibility
(none)
Online platform
Course content
| Content | Methods | Obs |
|---|---|---|
| C1-2 (4h). Introducere, prezentarea domeniului, a temelor generale, a principalelor scoli si conferinte, a principalilor actori din domeniu Introduction, history, overview of main topics, main conferences, significants researchers | Lecture, conversation, illustration | References:slides |
| C3-C4 (4h). Preprocesarea textului: normalizare, segmentareText Processing: Normalization, Tokenization (Segmentation) | Same as above | Same as above |
| C5-6 (2h). Lematization, StemmingLematizare, Stemming | Same as above | Same as above |
| C7 (2h). Etichetarea partilor de vorbire in propozitiePOS tagging | Same as above | Same as above |
| C8 (2h). Aspecte cantitative si cognitive ale limbajului natural: legi de tip minim efort, legea lu Zipf, principiul lui Menzerath, principiul economiei cognitive, analize lexicale, etc. Quantitative and cognitive aspects of natural language: minimum effort type laws, Zipf’s law, Menzerath’s principle , lexical analyses, etc. | Same as above | Same as above |
| C9 (4h). Probleme de similaritate in procesarea limbajului natural Similarity problems of natural language processing | Same as above | Same as above |
| C10 (2h). Stilistica computationala Computational stylistics. | Same as above | Same as above |
| C11-12 (2h). Limbajele figurative, si identificarea mesajelor inselatoare si a stirilor false. Figurative languages and deception detection | Same as above | Same as above |
| C13-14 (4h). Lingvistica istorica. Computational Historical Linguistics | Same as above | Same as above |
Course bibliography
Daniel Jurafsky and James H. Martin. 2009. Speech and Language Processing. An Introduction to Natural Language Processing, Computational Linguistics, and Speech Recognition (2nd Edition). Prentice Hall. The Oxford Handbook of Computational Linguistics 2nd edition (2 ed.), 2017, Oxford Manning, C., H. Schutze. Foundations of statistical natural language processing, MIT Press, 1999 G. Altmann (ed). Handbook of Quantitative Linguistics, 2003 Arhivele revistelor: Computational Linguistics, Literary and Linguistic Computing, Quantitative Linguistics, PNAS, etc Volumele principalelor conferinte din domeniu (ACL, EMNLP, EACL, NAACL, COLING, etc). Disponibile on-line la http://aclweb.org/anthology-new/
Seminar content
| Content | Methods | Obs |
|---|---|---|
| L1-7 (2h). Exercises on the topics presented as well as formalization of problems and application of NLP for different practical problems. | Questioning, dialogue, collaborative learning | Each lab will be available online. The students will have time until the next to solve it. At the next lab meeting they will present their work and receive feedback. |
| Bibliography:same as for the lecture | ||
| Bibliography: The content of the lecture is similar to others, on the same topic, from other universities. It covers the fundamental notions for understanding of NLP techniques. |
Seminar bibliography
(none)
Corroboration
(none)
AI tools guidance
Evaluation and delivery
| Activity | Criteria | Methods | Percentage |
|---|---|---|---|
| C |
|
|
|
| S |
|
|
|
Performance standards
Knowledge of the main concepts used in the NLP Knowledge of problems and solutions associated with NLP applications. Ability to identify the solution for an NLP problem The final mark is computed as a weighted average of the marks corresponding to the components specified at 10.4 and 10.5. The exam is considered passed if the average is at least 5 (it is not required that each mark is at least 5). In each session of exams (including re-examinations) the mark is computed using the same rule. The student can be re-examined only for the components for which the current mark is smaller than 5, excepting the cases when the student asks to be re-examined .
Additional info
(none)