This course is intended for language technology students who do not have linguistic background. The purpose of the course is to give an overview of the structure of Icelandic, with a special consideration to features which can be problematic for natural language processing. The main topics that will be covered are the sound system of Icelandic and phonetic transcription (IPA and SAMPA); the inflectional and derivational morphology of Icelandic with a special consideration to Part-of-Speech tagging and tagsets; and the syntactic structure of Icelandic with emphasis on both phrase structure and dependency parsing.

The course is first and foremost organized for students in language technology that have a background in linguistics (or humanities) but are not experienced in computer science. This course is most often taken in the same semester as the course “Computer Science 1a”. If someone with a different background is interested in the course, please contact the teacher for further information.  The course is taught alongside ÍSL333G Programming for the humanities at the BA-level and all students attend the same lectures but MA students get longer assignments than BA students.

The main goal of this course is to support students in taking their first step toward learning programming, help them to knack the basis and train them in solving simple but diverse assignments in language technology using Python. Besides, students will be introduced to a few text processing tools that can be used for natural language processing.

Programming in Python (for computations in engineering and science): Main commands and statements (computations, control statements, in- and output), definition and execution of functions, datatypes (numbers, matrices, strings, logical values, records), operations and built-in functions, array and matrix computation, file processing, statistics, graphics. Object-oriented programming: classes, objects, constructors and methods. Concepts associated with design and construction of program systems: Programming environment and practices, design and documentation of function and subroutine libraries, debugging and testing of programmes.

Directed Study

Directed Study

Training in various aspects of the writing and editing of scientific texts. Various kinds of texts (non-fiction) examined and evaluated. Training in reviewing and commenting on scientific texts and in other aspects of editorial work. The main emphasis will be on the writing of articles, but other kinds of texts will also be considered, both shorter (conference abstracts, reviews) and longer (theses, books), as well as research proposals. Discussion of guidelines for the preparation of manuscripts. Types of plagiarism and how to avoid them and find them. Texts on different subjects will be used as examples, especially writings in linguistics, literature and history. The book Skrifaðu bæði skýrt og rétt will be used as a textbook (Höskuldur Þráinsson 2015).

This course is open to students of many MA programmes in the School of Humanities, cf. the regulations of the individual subjects. Students in the MA programmes in Icelandic literature, Icelandic linguistics, Icelandic studies and Icelandic teaching can take the course as part of the MA course requirements in Icelandic literature or Icelandic linguistics. Students in the MA programme in Icelandic teaching can, however, not have this course as the only linguistics or literature course in their MA.

An overview of some of the main concepts, techniques and algorithms in machine learning. Supervised learning and unsupervised learning. Data preprocessing and data visualization. Model evaluation and model selection. Linear regression, nearest neighbors, support vector machines, decision trees and ensemble methods. Deep learning. Cluster analysis and the k-means algorithm. The students implement simple algorithms in Python and learn how to use specialized software packages. At the end of the course the students work on a practical machine learning project.

In this course we cover deep neural networks and methods related to them. We study networks and methods for image, sound and text analysis. The focus will be on applications and students will present either a project or a recent paper in this field.

This is a course for people who want to be able to analyze datasets stastically to better understand them, for example through visualization with graphs. Recent years have seen an increased focus on data collection and statistical analysis within the humanities. This is particularly apparent in growing branches such as computational linguistics and psycholinguistics, cognitive literary studies and experimental philosophy, to name a few. The push towards quantitative methods occurs at a time where the validity and reliability of well-established statistical methods are called into question in other fields, with increased demands of replicability and open access as well as data protection and responsibility. In this course, students explore the value of quantitative methods in their field while getting training in the collection and analysis of data. A diverse set of research methods will be introduced, ranging from surveys to corpus analysis and experiments in which participants’ response to stimuli (such as words, texts or audio-visual materials) is quantified. Basic concepts in statistics will be reviewed, enabling students to know the difference between descriptive and inferential statistics, understand statistical significance and interpret visual representations of data in graphs. The course will be largely practical and students are expected to apply their knowledge of data collection and analysis under the instructor’s guidance. Students will work on a project within their own discipline but will also explore the possibility of cross-disciplinary work. Open source tools such as R Studio will be used for all assignments but no prior knowledge of the software or statistics in general is required. The course is suitable for all students within the humanities who want to collect quantitative data to answer interesting questions and could therefore be a useful preparation for a BA or MA project.

This course will be dedicated to the Computer Assisted Translation-technology available to translators. Students get an insight into the importance of translation memories, how humans and machines use these memories, and learn how to align text corpora to create language data and dictionaries. How to use online dictionaries, data bases and other online means. We will consider language policy, technical terms and neologisms. The translators working environment will be considered as well as skills that help freelancers get by in the gig-economy. It is hoped that experienced translators will contribute to the seminar. Students work on projects during class to prepare them for the home assignments.

The aim of the course is to create a venue in which graduate students can access an overview of the current landscape in Icelandic language technology and work on a project consistent with its latest challenges. The course is organized as a seminar series with weekly lectures sponsored by Máltæknisetur (the Icelandic Center for Language Technology, ICLT). Before each lecture, registered students meet and discuss the course readings with the instructor. The lectures will mostly be by researchers affiliated with the institutions of the ICLT (UI, RU and The Árni Magnússon Institute) but representatives from the private sector will also be invited.  

Faroeese is the language that has the strongest similarity to Icelandic among the Nordic languages but it has changed more than Icelandic with respect to phonology, inflections and syntax. Investigating Faroese is important for Icelandic linguistics because Faroese provides a unique perspective on how Icelandic could have changed or may change in the next centuries.

This course will give an overview of the grammar of Faroese (phonology, inflections, word-formation and syntax) in comparison to Icelandic and the other Nordic languages. Language changes, dialects and foreign influence on Faroese will also be discussed. Moreover, students will get some training in listening to spoken Faroese.

This seminar deals with the history of the inflectional system of Icelandic from Proto-Germanic to modern times with special emphasis on selected problems. Recent writings on Icelandic historical morphology will be discussed. We will study text examples and their value as sources of information on the development of Icelandic morphology. The development of Icelandic word formation and different types of compounds will also be discussed.

Assignments: Students will give presentations on text samples and/or particular morphological problems.

Directed Study

Directed Study

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. Machine translation I does not require programming skills as the objective is to lead together people working on language technology and people working on traditional translations.

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. A pre-requisite for Machine translation II is that students also take Machine Translation I and have taken Programming for Language technology or an equivalent course.

The course will introduce and discuss topics and methods in etymological research. Different types of etymological dictionaries will be compared. Examples from Icelandic will be discussed, i.e., the history of particular words and the information that etymological dictionaries provide on their development.

In this course we will discuss selected topics in linguistics, with a focus on the origin of language and its influence on thought. Most of the course will be devoted to the origin and evolution of language and speech, seen from a broad perspective. Classic theories and research in the field will be discussed, including hypotheses on the role of gesture (Corballis) and grooming (Dunbar), the “single mutation” theory (Chomsky), and research on the evolution of speech (Fitch). We will also discuss more recent research that provides insights into the origin and nature of speech and the language capacity, such as research on songbirds, musicality and interaction. Did human language originate in gesture or vocal calls of animals? Did it evolve out of the need for gossip and grooming? Did music have any role in the evolution of language? What can genetic studies tell us about the evolution of language? Do biological biases or the environment influence the evolution of languages?  In the course we will also discuss the relationship between language and thought. Categorization of various phenomena and objects in languages of the world will be discussed, for example in relation to color vocabulary. How does the language we speak influence the way we think and perceive the world around us?

In this course, we study the AI lifecycle, i.e. the productionisation of AI methods.
We explore the following parts of the lifecycle:
- Data collection and preparation
- Feature engineering
- Model training
- Model evaluation
- Model deployment
- Model serving
- Model monitoring
- Model maintenance
Three large projects will be handed out during the semester where students compete to solve AI problems.

The main purpose of the course is to give an overview of the syntax of the modern Scandinavian languages from a generative perspective. The emphasis is on the comparison of the Insular Scandinavian languages (Icelandic and Faroese) on the one hand and the Mainland Scandinavian languages (Danish, Norwegian and Swedish) on the other. Aspects of the syntax of some lesser-known Scandinavian varieties is also included for comparison, including Övdalian (Swe. Älvalsmålet), for instance, which preserves certain inflectional and syntactic features of Old Norse that have disappeared from the Mainland Scandinavian standard languages. Selected topics in recent research on variation in Scandinavian syntax are covered and the students will be trained in designing and administering syntactic questionnaires.

In this seminar some Eddic poems will be read and their language examined. Features which cast light on the age of the poems will be given particular attention. The evidence of the Eddic poems will be compared with that from other linguistic sources. Various methods of dating the Eddic poems will be discussed.

Do AI tools work in Icelandic? Do they work as well as in languages such as English? In this course we explore these two questions in the context of Large Language Models (LLMs) such as the ones underlying the ChatGPT and Claude AI assistants. We will examine the methods used to assess the language comprehension and production of LLMs in languages such as Icelandic and discuss whether various potential risks of increased LLM use (e.g. disinformation and bias propagation) are exacerbated in lower-resource language communities. We will place these discussions in the context of current theoretical debates, asking what AI performance in Icelandic tells us about the nature of LLMs and human language, e.g. regarding questions about how children and machines learn language.

Directed Study

Directed Study

Training in various aspects of the writing and editing of scientific texts. Various kinds of texts (non-fiction) examined and evaluated. Training in reviewing and commenting on scientific texts and in other aspects of editorial work. The main emphasis will be on the writing of articles, but other kinds of texts will also be considered, both shorter (conference abstracts, reviews) and longer (theses, books), as well as research proposals. Discussion of guidelines for the preparation of manuscripts. Types of plagiarism and how to avoid them and find them. Texts on different subjects will be used as examples, especially writings in linguistics, literature and history. The book Skrifaðu bæði skýrt og rétt will be used as a textbook (Höskuldur Þráinsson 2015).

This course is open to students of many MA programmes in the School of Humanities, cf. the regulations of the individual subjects. Students in the MA programmes in Icelandic literature, Icelandic linguistics, Icelandic studies and Icelandic teaching can take the course as part of the MA course requirements in Icelandic literature or Icelandic linguistics. Students in the MA programme in Icelandic teaching can, however, not have this course as the only linguistics or literature course in their MA.

An overview of some of the main concepts, techniques and algorithms in machine learning. Supervised learning and unsupervised learning. Data preprocessing and data visualization. Model evaluation and model selection. Linear regression, nearest neighbors, support vector machines, decision trees and ensemble methods. Deep learning. Cluster analysis and the k-means algorithm. The students implement simple algorithms in Python and learn how to use specialized software packages. At the end of the course the students work on a practical machine learning project.

In this course we cover deep neural networks and methods related to them. We study networks and methods for image, sound and text analysis. The focus will be on applications and students will present either a project or a recent paper in this field.

This is a course for people who want to be able to analyze datasets stastically to better understand them, for example through visualization with graphs. Recent years have seen an increased focus on data collection and statistical analysis within the humanities. This is particularly apparent in growing branches such as computational linguistics and psycholinguistics, cognitive literary studies and experimental philosophy, to name a few. The push towards quantitative methods occurs at a time where the validity and reliability of well-established statistical methods are called into question in other fields, with increased demands of replicability and open access as well as data protection and responsibility. In this course, students explore the value of quantitative methods in their field while getting training in the collection and analysis of data. A diverse set of research methods will be introduced, ranging from surveys to corpus analysis and experiments in which participants’ response to stimuli (such as words, texts or audio-visual materials) is quantified. Basic concepts in statistics will be reviewed, enabling students to know the difference between descriptive and inferential statistics, understand statistical significance and interpret visual representations of data in graphs. The course will be largely practical and students are expected to apply their knowledge of data collection and analysis under the instructor’s guidance. Students will work on a project within their own discipline but will also explore the possibility of cross-disciplinary work. Open source tools such as R Studio will be used for all assignments but no prior knowledge of the software or statistics in general is required. The course is suitable for all students within the humanities who want to collect quantitative data to answer interesting questions and could therefore be a useful preparation for a BA or MA project.

This course will be dedicated to the Computer Assisted Translation-technology available to translators. Students get an insight into the importance of translation memories, how humans and machines use these memories, and learn how to align text corpora to create language data and dictionaries. How to use online dictionaries, data bases and other online means. We will consider language policy, technical terms and neologisms. The translators working environment will be considered as well as skills that help freelancers get by in the gig-economy. It is hoped that experienced translators will contribute to the seminar. Students work on projects during class to prepare them for the home assignments.

The aim of the course is to create a venue in which graduate students can access an overview of the current landscape in Icelandic language technology and work on a project consistent with its latest challenges. The course is organized as a seminar series with weekly lectures sponsored by Máltæknisetur (the Icelandic Center for Language Technology, ICLT). Before each lecture, registered students meet and discuss the course readings with the instructor. The lectures will mostly be by researchers affiliated with the institutions of the ICLT (UI, RU and The Árni Magnússon Institute) but representatives from the private sector will also be invited.  

Faroeese is the language that has the strongest similarity to Icelandic among the Nordic languages but it has changed more than Icelandic with respect to phonology, inflections and syntax. Investigating Faroese is important for Icelandic linguistics because Faroese provides a unique perspective on how Icelandic could have changed or may change in the next centuries.

This course will give an overview of the grammar of Faroese (phonology, inflections, word-formation and syntax) in comparison to Icelandic and the other Nordic languages. Language changes, dialects and foreign influence on Faroese will also be discussed. Moreover, students will get some training in listening to spoken Faroese.

This seminar deals with the history of the inflectional system of Icelandic from Proto-Germanic to modern times with special emphasis on selected problems. Recent writings on Icelandic historical morphology will be discussed. We will study text examples and their value as sources of information on the development of Icelandic morphology. The development of Icelandic word formation and different types of compounds will also be discussed.

Assignments: Students will give presentations on text samples and/or particular morphological problems.

Final project

Directed Study

Directed Study

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. Machine translation I does not require programming skills as the objective is to lead together people working on language technology and people working on traditional translations.

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. A pre-requisite for Machine translation II is that students also take Machine Translation I and have taken Programming for Language technology or an equivalent course.

The course will introduce and discuss topics and methods in etymological research. Different types of etymological dictionaries will be compared. Examples from Icelandic will be discussed, i.e., the history of particular words and the information that etymological dictionaries provide on their development.

In this course we will discuss selected topics in linguistics, with a focus on the origin of language and its influence on thought. Most of the course will be devoted to the origin and evolution of language and speech, seen from a broad perspective. Classic theories and research in the field will be discussed, including hypotheses on the role of gesture (Corballis) and grooming (Dunbar), the “single mutation” theory (Chomsky), and research on the evolution of speech (Fitch). We will also discuss more recent research that provides insights into the origin and nature of speech and the language capacity, such as research on songbirds, musicality and interaction. Did human language originate in gesture or vocal calls of animals? Did it evolve out of the need for gossip and grooming? Did music have any role in the evolution of language? What can genetic studies tell us about the evolution of language? Do biological biases or the environment influence the evolution of languages?  In the course we will also discuss the relationship between language and thought. Categorization of various phenomena and objects in languages of the world will be discussed, for example in relation to color vocabulary. How does the language we speak influence the way we think and perceive the world around us?

In this course, we study the AI lifecycle, i.e. the productionisation of AI methods.
We explore the following parts of the lifecycle:
- Data collection and preparation
- Feature engineering
- Model training
- Model evaluation
- Model deployment
- Model serving
- Model monitoring
- Model maintenance
Three large projects will be handed out during the semester where students compete to solve AI problems.

The main purpose of the course is to give an overview of the syntax of the modern Scandinavian languages from a generative perspective. The emphasis is on the comparison of the Insular Scandinavian languages (Icelandic and Faroese) on the one hand and the Mainland Scandinavian languages (Danish, Norwegian and Swedish) on the other. Aspects of the syntax of some lesser-known Scandinavian varieties is also included for comparison, including Övdalian (Swe. Älvalsmålet), for instance, which preserves certain inflectional and syntactic features of Old Norse that have disappeared from the Mainland Scandinavian standard languages. Selected topics in recent research on variation in Scandinavian syntax are covered and the students will be trained in designing and administering syntactic questionnaires.

In this seminar some Eddic poems will be read and their language examined. Features which cast light on the age of the poems will be given particular attention. The evidence of the Eddic poems will be compared with that from other linguistic sources. Various methods of dating the Eddic poems will be discussed.

Do AI tools work in Icelandic? Do they work as well as in languages such as English? In this course we explore these two questions in the context of Large Language Models (LLMs) such as the ones underlying the ChatGPT and Claude AI assistants. We will examine the methods used to assess the language comprehension and production of LLMs in languages such as Icelandic and discuss whether various potential risks of increased LLM use (e.g. disinformation and bias propagation) are exacerbated in lower-resource language communities. We will place these discussions in the context of current theoretical debates, asking what AI performance in Icelandic tells us about the nature of LLMs and human language, e.g. regarding questions about how children and machines learn language.

Final project

This is a foundational course in single variable calculus. The prerequisites are high school courses on algebra, trigonometry. derivatives, and integrals. The course aims to create a foundation for understanding of subjects such as natural and physical sciences, engineering, economics, and computer science. Topics of the course include the following:

• Real numbers.
• Limits and continuous functions.
• Differentiable functions, rules for derivatives, derivatives of higher order, applications of differential calculus (extremal value problems, linear approximation).
• Transcendental functions.
• Mean value theorem, theorems of l'Hôpital and Taylor.
• Integration, the definite integral and rules/techniques of integration, primitives, improper integrals.
• Fundamental theorem of calculus.
• Applications of integral calculus: Arc length, area, volume, centroids.
• Ordinary differential equations: First-order separable and homogeneous differential equations, first-order linear equations, second-order linear equations with constant coefficients.
• Sequences and series, convergence tests.
• Power series, Taylor series.

In this course, we study several concepts related to digital literacy. The goal of the course is to introduce the students to a broad range of topics without necessarily diving deep into each one.

The Unix operating system is introduced. The file system organization, often used command-line programs, the window system, command-line environment, and shell scripting. We cover editors and data wrangling in the shell. We present version control systems (git), debugging methods, and methods to build software. Common concepts in the field of cryptography are introduced as well as concepts related to virtualization and containers.

Basics of linear algebra over the reals.  

Subject matter: Systems of linear equations, matrices, Gauss-Jordan reduction.  Vector spaces and their subspaces.  Linearly independent sets, bases and dimension.  Linear maps, range space and nullk space.  The dot product, length and angle measures.  Volumes in higher dimension and the cross product in threedimensional space.  Flats, parametric descriptions and descriptions by equations.  Orthogonal projections and orthonormal bases.  Gram-Schmidt orthogonalization.  Determinants and inverses of matrices.  Eigenvalues, eigenvectors and diagonalization.

The course provides an introduction to the methods at the heart of data science and introduces widely used software tools such as numpy, pandas, matplotlib and scikit-learn.

The course consists of 6 modules:

1. Introduction to the Python programming language.
2. Data wrangling and data preprocessing.
3. Exploratory data analysis and visualization.
4. Optimization.
5. Clustering and dimensionality reduction.
6. Regression and classification.

Each module concludes with a student project.

Note that there is an academic overlap with REI201G Mathematics and Scientific Computing and both courses cannot be valid for the same degree.

In this course, we study several concepts related to digital literacy. The goal of the course is to introduce the students to a broad range of topics without necessarily diving deep into each one.

The Unix operating system is introduced. The file system organization, often used command-line programs, the window system, command-line environment, and shell scripting. We cover editors and data wrangling in the shell. We present version control systems (git), debugging methods, and methods to build software. Common concepts in the field of cryptography are introduced as well as concepts related to virtualization and containers.

Open and closed sets. Mappings, limits and continuity. Differentiable mappings, partial derivatives and the chain rule. Jacobi matrices. Gradients and directional derivatives. Mixed partial derivatives. Curves. Vector fields and flow. Cylindrical and spherical coordinates. Taylor polynomials. Extreme values and the classification of stationary points. Extreme value problems with constraints. Implicit functions and local inverses. Line integrals, primitive functions and exact differential equations. Double integrals. Improper integrals. Green's theorem. Simply connected domains. Change of variables in double integrals. Multiple integrals. Change of variables in multiple integrals. Surface integrals. Integration of vector fields. The theorems of Stokes and Gauss.

Basic concepts in probability and statistics based on univariate calculus. 

Topics: 
Sample space, events, probability, equal probability, independent events, conditional probability, Bayes rule, random variables, distribution, density, joint distribution, independent random variables, condistional distribution, mean, variance, covariance, correlation, law of large numbers, Bernoulli, binomial, Poisson, uniform, exponential and normal random variables. Central limit theorem. Poisson process. Random sample, statistics, the distribution of the sample mean and the sample variance. Point estimate, maximum likelihood estimator, mean square error, bias. Interval estimates and hypotheses testing form normal, binomial and exponential samples. Simple linear regression. Goodness of fit tests, test of independence.

This course is intended for language technology students who do not have linguistic background. The purpose of the course is to give an overview of the structure of Icelandic, with a special consideration to features which can be problematic for natural language processing. The main topics that will be covered are the sound system of Icelandic and phonetic transcription (IPA and SAMPA); the inflectional and derivational morphology of Icelandic with a special consideration to Part-of-Speech tagging and tagsets; and the syntactic structure of Icelandic with emphasis on both phrase structure and dependency parsing.

The course is first and foremost organized for students in language technology that have a background in linguistics (or humanities) but are not experienced in computer science. This course is most often taken in the same semester as the course “Computer Science 1a”. If someone with a different background is interested in the course, please contact the teacher for further information.  The course is taught alongside ÍSL333G Programming for the humanities at the BA-level and all students attend the same lectures but MA students get longer assignments than BA students.

The main goal of this course is to support students in taking their first step toward learning programming, help them to knack the basis and train them in solving simple but diverse assignments in language technology using Python. Besides, students will be introduced to a few text processing tools that can be used for natural language processing.

Programming in Python (for computations in engineering and science): Main commands and statements (computations, control statements, in- and output), definition and execution of functions, datatypes (numbers, matrices, strings, logical values, records), operations and built-in functions, array and matrix computation, file processing, statistics, graphics. Object-oriented programming: classes, objects, constructors and methods. Concepts associated with design and construction of program systems: Programming environment and practices, design and documentation of function and subroutine libraries, debugging and testing of programmes.

Directed Study

Directed Study

Training in various aspects of the writing and editing of scientific texts. Various kinds of texts (non-fiction) examined and evaluated. Training in reviewing and commenting on scientific texts and in other aspects of editorial work. The main emphasis will be on the writing of articles, but other kinds of texts will also be considered, both shorter (conference abstracts, reviews) and longer (theses, books), as well as research proposals. Discussion of guidelines for the preparation of manuscripts. Types of plagiarism and how to avoid them and find them. Texts on different subjects will be used as examples, especially writings in linguistics, literature and history. The book Skrifaðu bæði skýrt og rétt will be used as a textbook (Höskuldur Þráinsson 2015).

This course is open to students of many MA programmes in the School of Humanities, cf. the regulations of the individual subjects. Students in the MA programmes in Icelandic literature, Icelandic linguistics, Icelandic studies and Icelandic teaching can take the course as part of the MA course requirements in Icelandic literature or Icelandic linguistics. Students in the MA programme in Icelandic teaching can, however, not have this course as the only linguistics or literature course in their MA.

An overview of some of the main concepts, techniques and algorithms in machine learning. Supervised learning and unsupervised learning. Data preprocessing and data visualization. Model evaluation and model selection. Linear regression, nearest neighbors, support vector machines, decision trees and ensemble methods. Deep learning. Cluster analysis and the k-means algorithm. The students implement simple algorithms in Python and learn how to use specialized software packages. At the end of the course the students work on a practical machine learning project.

In this course we cover deep neural networks and methods related to them. We study networks and methods for image, sound and text analysis. The focus will be on applications and students will present either a project or a recent paper in this field.

This is a course for people who want to be able to analyze datasets stastically to better understand them, for example through visualization with graphs. Recent years have seen an increased focus on data collection and statistical analysis within the humanities. This is particularly apparent in growing branches such as computational linguistics and psycholinguistics, cognitive literary studies and experimental philosophy, to name a few. The push towards quantitative methods occurs at a time where the validity and reliability of well-established statistical methods are called into question in other fields, with increased demands of replicability and open access as well as data protection and responsibility. In this course, students explore the value of quantitative methods in their field while getting training in the collection and analysis of data. A diverse set of research methods will be introduced, ranging from surveys to corpus analysis and experiments in which participants’ response to stimuli (such as words, texts or audio-visual materials) is quantified. Basic concepts in statistics will be reviewed, enabling students to know the difference between descriptive and inferential statistics, understand statistical significance and interpret visual representations of data in graphs. The course will be largely practical and students are expected to apply their knowledge of data collection and analysis under the instructor’s guidance. Students will work on a project within their own discipline but will also explore the possibility of cross-disciplinary work. Open source tools such as R Studio will be used for all assignments but no prior knowledge of the software or statistics in general is required. The course is suitable for all students within the humanities who want to collect quantitative data to answer interesting questions and could therefore be a useful preparation for a BA or MA project.

This course will be dedicated to the Computer Assisted Translation-technology available to translators. Students get an insight into the importance of translation memories, how humans and machines use these memories, and learn how to align text corpora to create language data and dictionaries. How to use online dictionaries, data bases and other online means. We will consider language policy, technical terms and neologisms. The translators working environment will be considered as well as skills that help freelancers get by in the gig-economy. It is hoped that experienced translators will contribute to the seminar. Students work on projects during class to prepare them for the home assignments.

The aim of the course is to create a venue in which graduate students can access an overview of the current landscape in Icelandic language technology and work on a project consistent with its latest challenges. The course is organized as a seminar series with weekly lectures sponsored by Máltæknisetur (the Icelandic Center for Language Technology, ICLT). Before each lecture, registered students meet and discuss the course readings with the instructor. The lectures will mostly be by researchers affiliated with the institutions of the ICLT (UI, RU and The Árni Magnússon Institute) but representatives from the private sector will also be invited.  

Faroeese is the language that has the strongest similarity to Icelandic among the Nordic languages but it has changed more than Icelandic with respect to phonology, inflections and syntax. Investigating Faroese is important for Icelandic linguistics because Faroese provides a unique perspective on how Icelandic could have changed or may change in the next centuries.

This course will give an overview of the grammar of Faroese (phonology, inflections, word-formation and syntax) in comparison to Icelandic and the other Nordic languages. Language changes, dialects and foreign influence on Faroese will also be discussed. Moreover, students will get some training in listening to spoken Faroese.

This seminar deals with the history of the inflectional system of Icelandic from Proto-Germanic to modern times with special emphasis on selected problems. Recent writings on Icelandic historical morphology will be discussed. We will study text examples and their value as sources of information on the development of Icelandic morphology. The development of Icelandic word formation and different types of compounds will also be discussed.

Assignments: Students will give presentations on text samples and/or particular morphological problems.

Directed Study

Directed Study

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. Machine translation I does not require programming skills as the objective is to lead together people working on language technology and people working on traditional translations.

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. A pre-requisite for Machine translation II is that students also take Machine Translation I and have taken Programming for Language technology or an equivalent course.

The course will introduce and discuss topics and methods in etymological research. Different types of etymological dictionaries will be compared. Examples from Icelandic will be discussed, i.e., the history of particular words and the information that etymological dictionaries provide on their development.

In this course we will discuss selected topics in linguistics, with a focus on the origin of language and its influence on thought. Most of the course will be devoted to the origin and evolution of language and speech, seen from a broad perspective. Classic theories and research in the field will be discussed, including hypotheses on the role of gesture (Corballis) and grooming (Dunbar), the “single mutation” theory (Chomsky), and research on the evolution of speech (Fitch). We will also discuss more recent research that provides insights into the origin and nature of speech and the language capacity, such as research on songbirds, musicality and interaction. Did human language originate in gesture or vocal calls of animals? Did it evolve out of the need for gossip and grooming? Did music have any role in the evolution of language? What can genetic studies tell us about the evolution of language? Do biological biases or the environment influence the evolution of languages?  In the course we will also discuss the relationship between language and thought. Categorization of various phenomena and objects in languages of the world will be discussed, for example in relation to color vocabulary. How does the language we speak influence the way we think and perceive the world around us?

In this course, we study the AI lifecycle, i.e. the productionisation of AI methods.
We explore the following parts of the lifecycle:
- Data collection and preparation
- Feature engineering
- Model training
- Model evaluation
- Model deployment
- Model serving
- Model monitoring
- Model maintenance
Three large projects will be handed out during the semester where students compete to solve AI problems.

The main purpose of the course is to give an overview of the syntax of the modern Scandinavian languages from a generative perspective. The emphasis is on the comparison of the Insular Scandinavian languages (Icelandic and Faroese) on the one hand and the Mainland Scandinavian languages (Danish, Norwegian and Swedish) on the other. Aspects of the syntax of some lesser-known Scandinavian varieties is also included for comparison, including Övdalian (Swe. Älvalsmålet), for instance, which preserves certain inflectional and syntactic features of Old Norse that have disappeared from the Mainland Scandinavian standard languages. Selected topics in recent research on variation in Scandinavian syntax are covered and the students will be trained in designing and administering syntactic questionnaires.

In this seminar some Eddic poems will be read and their language examined. Features which cast light on the age of the poems will be given particular attention. The evidence of the Eddic poems will be compared with that from other linguistic sources. Various methods of dating the Eddic poems will be discussed.

Do AI tools work in Icelandic? Do they work as well as in languages such as English? In this course we explore these two questions in the context of Large Language Models (LLMs) such as the ones underlying the ChatGPT and Claude AI assistants. We will examine the methods used to assess the language comprehension and production of LLMs in languages such as Icelandic and discuss whether various potential risks of increased LLM use (e.g. disinformation and bias propagation) are exacerbated in lower-resource language communities. We will place these discussions in the context of current theoretical debates, asking what AI performance in Icelandic tells us about the nature of LLMs and human language, e.g. regarding questions about how children and machines learn language.

Directed Study

Directed Study

Training in various aspects of the writing and editing of scientific texts. Various kinds of texts (non-fiction) examined and evaluated. Training in reviewing and commenting on scientific texts and in other aspects of editorial work. The main emphasis will be on the writing of articles, but other kinds of texts will also be considered, both shorter (conference abstracts, reviews) and longer (theses, books), as well as research proposals. Discussion of guidelines for the preparation of manuscripts. Types of plagiarism and how to avoid them and find them. Texts on different subjects will be used as examples, especially writings in linguistics, literature and history. The book Skrifaðu bæði skýrt og rétt will be used as a textbook (Höskuldur Þráinsson 2015).

This course is open to students of many MA programmes in the School of Humanities, cf. the regulations of the individual subjects. Students in the MA programmes in Icelandic literature, Icelandic linguistics, Icelandic studies and Icelandic teaching can take the course as part of the MA course requirements in Icelandic literature or Icelandic linguistics. Students in the MA programme in Icelandic teaching can, however, not have this course as the only linguistics or literature course in their MA.

An overview of some of the main concepts, techniques and algorithms in machine learning. Supervised learning and unsupervised learning. Data preprocessing and data visualization. Model evaluation and model selection. Linear regression, nearest neighbors, support vector machines, decision trees and ensemble methods. Deep learning. Cluster analysis and the k-means algorithm. The students implement simple algorithms in Python and learn how to use specialized software packages. At the end of the course the students work on a practical machine learning project.

In this course we cover deep neural networks and methods related to them. We study networks and methods for image, sound and text analysis. The focus will be on applications and students will present either a project or a recent paper in this field.

This is a course for people who want to be able to analyze datasets stastically to better understand them, for example through visualization with graphs. Recent years have seen an increased focus on data collection and statistical analysis within the humanities. This is particularly apparent in growing branches such as computational linguistics and psycholinguistics, cognitive literary studies and experimental philosophy, to name a few. The push towards quantitative methods occurs at a time where the validity and reliability of well-established statistical methods are called into question in other fields, with increased demands of replicability and open access as well as data protection and responsibility. In this course, students explore the value of quantitative methods in their field while getting training in the collection and analysis of data. A diverse set of research methods will be introduced, ranging from surveys to corpus analysis and experiments in which participants’ response to stimuli (such as words, texts or audio-visual materials) is quantified. Basic concepts in statistics will be reviewed, enabling students to know the difference between descriptive and inferential statistics, understand statistical significance and interpret visual representations of data in graphs. The course will be largely practical and students are expected to apply their knowledge of data collection and analysis under the instructor’s guidance. Students will work on a project within their own discipline but will also explore the possibility of cross-disciplinary work. Open source tools such as R Studio will be used for all assignments but no prior knowledge of the software or statistics in general is required. The course is suitable for all students within the humanities who want to collect quantitative data to answer interesting questions and could therefore be a useful preparation for a BA or MA project.

This course will be dedicated to the Computer Assisted Translation-technology available to translators. Students get an insight into the importance of translation memories, how humans and machines use these memories, and learn how to align text corpora to create language data and dictionaries. How to use online dictionaries, data bases and other online means. We will consider language policy, technical terms and neologisms. The translators working environment will be considered as well as skills that help freelancers get by in the gig-economy. It is hoped that experienced translators will contribute to the seminar. Students work on projects during class to prepare them for the home assignments.

The aim of the course is to create a venue in which graduate students can access an overview of the current landscape in Icelandic language technology and work on a project consistent with its latest challenges. The course is organized as a seminar series with weekly lectures sponsored by Máltæknisetur (the Icelandic Center for Language Technology, ICLT). Before each lecture, registered students meet and discuss the course readings with the instructor. The lectures will mostly be by researchers affiliated with the institutions of the ICLT (UI, RU and The Árni Magnússon Institute) but representatives from the private sector will also be invited.  

Faroeese is the language that has the strongest similarity to Icelandic among the Nordic languages but it has changed more than Icelandic with respect to phonology, inflections and syntax. Investigating Faroese is important for Icelandic linguistics because Faroese provides a unique perspective on how Icelandic could have changed or may change in the next centuries.

This course will give an overview of the grammar of Faroese (phonology, inflections, word-formation and syntax) in comparison to Icelandic and the other Nordic languages. Language changes, dialects and foreign influence on Faroese will also be discussed. Moreover, students will get some training in listening to spoken Faroese.

This seminar deals with the history of the inflectional system of Icelandic from Proto-Germanic to modern times with special emphasis on selected problems. Recent writings on Icelandic historical morphology will be discussed. We will study text examples and their value as sources of information on the development of Icelandic morphology. The development of Icelandic word formation and different types of compounds will also be discussed.

Assignments: Students will give presentations on text samples and/or particular morphological problems.

Final project

Directed Study

Directed Study

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. Machine translation I does not require programming skills as the objective is to lead together people working on language technology and people working on traditional translations.

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. A pre-requisite for Machine translation II is that students also take Machine Translation I and have taken Programming for Language technology or an equivalent course.

The course will introduce and discuss topics and methods in etymological research. Different types of etymological dictionaries will be compared. Examples from Icelandic will be discussed, i.e., the history of particular words and the information that etymological dictionaries provide on their development.

In this course we will discuss selected topics in linguistics, with a focus on the origin of language and its influence on thought. Most of the course will be devoted to the origin and evolution of language and speech, seen from a broad perspective. Classic theories and research in the field will be discussed, including hypotheses on the role of gesture (Corballis) and grooming (Dunbar), the “single mutation” theory (Chomsky), and research on the evolution of speech (Fitch). We will also discuss more recent research that provides insights into the origin and nature of speech and the language capacity, such as research on songbirds, musicality and interaction. Did human language originate in gesture or vocal calls of animals? Did it evolve out of the need for gossip and grooming? Did music have any role in the evolution of language? What can genetic studies tell us about the evolution of language? Do biological biases or the environment influence the evolution of languages?  In the course we will also discuss the relationship between language and thought. Categorization of various phenomena and objects in languages of the world will be discussed, for example in relation to color vocabulary. How does the language we speak influence the way we think and perceive the world around us?

In this course, we study the AI lifecycle, i.e. the productionisation of AI methods.
We explore the following parts of the lifecycle:
- Data collection and preparation
- Feature engineering
- Model training
- Model evaluation
- Model deployment
- Model serving
- Model monitoring
- Model maintenance
Three large projects will be handed out during the semester where students compete to solve AI problems.

The main purpose of the course is to give an overview of the syntax of the modern Scandinavian languages from a generative perspective. The emphasis is on the comparison of the Insular Scandinavian languages (Icelandic and Faroese) on the one hand and the Mainland Scandinavian languages (Danish, Norwegian and Swedish) on the other. Aspects of the syntax of some lesser-known Scandinavian varieties is also included for comparison, including Övdalian (Swe. Älvalsmålet), for instance, which preserves certain inflectional and syntactic features of Old Norse that have disappeared from the Mainland Scandinavian standard languages. Selected topics in recent research on variation in Scandinavian syntax are covered and the students will be trained in designing and administering syntactic questionnaires.

In this seminar some Eddic poems will be read and their language examined. Features which cast light on the age of the poems will be given particular attention. The evidence of the Eddic poems will be compared with that from other linguistic sources. Various methods of dating the Eddic poems will be discussed.

Do AI tools work in Icelandic? Do they work as well as in languages such as English? In this course we explore these two questions in the context of Large Language Models (LLMs) such as the ones underlying the ChatGPT and Claude AI assistants. We will examine the methods used to assess the language comprehension and production of LLMs in languages such as Icelandic and discuss whether various potential risks of increased LLM use (e.g. disinformation and bias propagation) are exacerbated in lower-resource language communities. We will place these discussions in the context of current theoretical debates, asking what AI performance in Icelandic tells us about the nature of LLMs and human language, e.g. regarding questions about how children and machines learn language.

Final project

This is a foundational course in single variable calculus. The prerequisites are high school courses on algebra, trigonometry. derivatives, and integrals. The course aims to create a foundation for understanding of subjects such as natural and physical sciences, engineering, economics, and computer science. Topics of the course include the following:

• Real numbers.
• Limits and continuous functions.
• Differentiable functions, rules for derivatives, derivatives of higher order, applications of differential calculus (extremal value problems, linear approximation).
• Transcendental functions.
• Mean value theorem, theorems of l'Hôpital and Taylor.
• Integration, the definite integral and rules/techniques of integration, primitives, improper integrals.
• Fundamental theorem of calculus.
• Applications of integral calculus: Arc length, area, volume, centroids.
• Ordinary differential equations: First-order separable and homogeneous differential equations, first-order linear equations, second-order linear equations with constant coefficients.
• Sequences and series, convergence tests.
• Power series, Taylor series.

In this course, we study several concepts related to digital literacy. The goal of the course is to introduce the students to a broad range of topics without necessarily diving deep into each one.

The Unix operating system is introduced. The file system organization, often used command-line programs, the window system, command-line environment, and shell scripting. We cover editors and data wrangling in the shell. We present version control systems (git), debugging methods, and methods to build software. Common concepts in the field of cryptography are introduced as well as concepts related to virtualization and containers.

Basics of linear algebra over the reals.  

Subject matter: Systems of linear equations, matrices, Gauss-Jordan reduction.  Vector spaces and their subspaces.  Linearly independent sets, bases and dimension.  Linear maps, range space and nullk space.  The dot product, length and angle measures.  Volumes in higher dimension and the cross product in threedimensional space.  Flats, parametric descriptions and descriptions by equations.  Orthogonal projections and orthonormal bases.  Gram-Schmidt orthogonalization.  Determinants and inverses of matrices.  Eigenvalues, eigenvectors and diagonalization.

The course provides an introduction to the methods at the heart of data science and introduces widely used software tools such as numpy, pandas, matplotlib and scikit-learn.

The course consists of 6 modules:

1. Introduction to the Python programming language.
2. Data wrangling and data preprocessing.
3. Exploratory data analysis and visualization.
4. Optimization.
5. Clustering and dimensionality reduction.
6. Regression and classification.

Each module concludes with a student project.

Note that there is an academic overlap with REI201G Mathematics and Scientific Computing and both courses cannot be valid for the same degree.

In this course, we study several concepts related to digital literacy. The goal of the course is to introduce the students to a broad range of topics without necessarily diving deep into each one.

The Unix operating system is introduced. The file system organization, often used command-line programs, the window system, command-line environment, and shell scripting. We cover editors and data wrangling in the shell. We present version control systems (git), debugging methods, and methods to build software. Common concepts in the field of cryptography are introduced as well as concepts related to virtualization and containers.

Open and closed sets. Mappings, limits and continuity. Differentiable mappings, partial derivatives and the chain rule. Jacobi matrices. Gradients and directional derivatives. Mixed partial derivatives. Curves. Vector fields and flow. Cylindrical and spherical coordinates. Taylor polynomials. Extreme values and the classification of stationary points. Extreme value problems with constraints. Implicit functions and local inverses. Line integrals, primitive functions and exact differential equations. Double integrals. Improper integrals. Green's theorem. Simply connected domains. Change of variables in double integrals. Multiple integrals. Change of variables in multiple integrals. Surface integrals. Integration of vector fields. The theorems of Stokes and Gauss.

Basic concepts in probability and statistics based on univariate calculus. 

Topics: 
Sample space, events, probability, equal probability, independent events, conditional probability, Bayes rule, random variables, distribution, density, joint distribution, independent random variables, condistional distribution, mean, variance, covariance, correlation, law of large numbers, Bernoulli, binomial, Poisson, uniform, exponential and normal random variables. Central limit theorem. Poisson process. Random sample, statistics, the distribution of the sample mean and the sample variance. Point estimate, maximum likelihood estimator, mean square error, bias. Interval estimates and hypotheses testing form normal, binomial and exponential samples. Simple linear regression. Goodness of fit tests, test of independence.

This course is intended for language technology students who do not have linguistic background. The purpose of the course is to give an overview of the structure of Icelandic, with a special consideration to features which can be problematic for natural language processing. The main topics that will be covered are the sound system of Icelandic and phonetic transcription (IPA and SAMPA); the inflectional and derivational morphology of Icelandic with a special consideration to Part-of-Speech tagging and tagsets; and the syntactic structure of Icelandic with emphasis on both phrase structure and dependency parsing.

The course is first and foremost organized for students in language technology that have a background in linguistics (or humanities) but are not experienced in computer science. This course is most often taken in the same semester as the course “Computer Science 1a”. If someone with a different background is interested in the course, please contact the teacher for further information.  The course is taught alongside ÍSL333G Programming for the humanities at the BA-level and all students attend the same lectures but MA students get longer assignments than BA students.

The main goal of this course is to support students in taking their first step toward learning programming, help them to knack the basis and train them in solving simple but diverse assignments in language technology using Python. Besides, students will be introduced to a few text processing tools that can be used for natural language processing.

Programming in Python (for computations in engineering and science): Main commands and statements (computations, control statements, in- and output), definition and execution of functions, datatypes (numbers, matrices, strings, logical values, records), operations and built-in functions, array and matrix computation, file processing, statistics, graphics. Object-oriented programming: classes, objects, constructors and methods. Concepts associated with design and construction of program systems: Programming environment and practices, design and documentation of function and subroutine libraries, debugging and testing of programmes.

Directed Study

Directed Study

Training in various aspects of the writing and editing of scientific texts. Various kinds of texts (non-fiction) examined and evaluated. Training in reviewing and commenting on scientific texts and in other aspects of editorial work. The main emphasis will be on the writing of articles, but other kinds of texts will also be considered, both shorter (conference abstracts, reviews) and longer (theses, books), as well as research proposals. Discussion of guidelines for the preparation of manuscripts. Types of plagiarism and how to avoid them and find them. Texts on different subjects will be used as examples, especially writings in linguistics, literature and history. The book Skrifaðu bæði skýrt og rétt will be used as a textbook (Höskuldur Þráinsson 2015).

This course is open to students of many MA programmes in the School of Humanities, cf. the regulations of the individual subjects. Students in the MA programmes in Icelandic literature, Icelandic linguistics, Icelandic studies and Icelandic teaching can take the course as part of the MA course requirements in Icelandic literature or Icelandic linguistics. Students in the MA programme in Icelandic teaching can, however, not have this course as the only linguistics or literature course in their MA.

An overview of some of the main concepts, techniques and algorithms in machine learning. Supervised learning and unsupervised learning. Data preprocessing and data visualization. Model evaluation and model selection. Linear regression, nearest neighbors, support vector machines, decision trees and ensemble methods. Deep learning. Cluster analysis and the k-means algorithm. The students implement simple algorithms in Python and learn how to use specialized software packages. At the end of the course the students work on a practical machine learning project.

In this course we cover deep neural networks and methods related to them. We study networks and methods for image, sound and text analysis. The focus will be on applications and students will present either a project or a recent paper in this field.

This is a course for people who want to be able to analyze datasets stastically to better understand them, for example through visualization with graphs. Recent years have seen an increased focus on data collection and statistical analysis within the humanities. This is particularly apparent in growing branches such as computational linguistics and psycholinguistics, cognitive literary studies and experimental philosophy, to name a few. The push towards quantitative methods occurs at a time where the validity and reliability of well-established statistical methods are called into question in other fields, with increased demands of replicability and open access as well as data protection and responsibility. In this course, students explore the value of quantitative methods in their field while getting training in the collection and analysis of data. A diverse set of research methods will be introduced, ranging from surveys to corpus analysis and experiments in which participants’ response to stimuli (such as words, texts or audio-visual materials) is quantified. Basic concepts in statistics will be reviewed, enabling students to know the difference between descriptive and inferential statistics, understand statistical significance and interpret visual representations of data in graphs. The course will be largely practical and students are expected to apply their knowledge of data collection and analysis under the instructor’s guidance. Students will work on a project within their own discipline but will also explore the possibility of cross-disciplinary work. Open source tools such as R Studio will be used for all assignments but no prior knowledge of the software or statistics in general is required. The course is suitable for all students within the humanities who want to collect quantitative data to answer interesting questions and could therefore be a useful preparation for a BA or MA project.

This course will be dedicated to the Computer Assisted Translation-technology available to translators. Students get an insight into the importance of translation memories, how humans and machines use these memories, and learn how to align text corpora to create language data and dictionaries. How to use online dictionaries, data bases and other online means. We will consider language policy, technical terms and neologisms. The translators working environment will be considered as well as skills that help freelancers get by in the gig-economy. It is hoped that experienced translators will contribute to the seminar. Students work on projects during class to prepare them for the home assignments.

The aim of the course is to create a venue in which graduate students can access an overview of the current landscape in Icelandic language technology and work on a project consistent with its latest challenges. The course is organized as a seminar series with weekly lectures sponsored by Máltæknisetur (the Icelandic Center for Language Technology, ICLT). Before each lecture, registered students meet and discuss the course readings with the instructor. The lectures will mostly be by researchers affiliated with the institutions of the ICLT (UI, RU and The Árni Magnússon Institute) but representatives from the private sector will also be invited.  

Faroeese is the language that has the strongest similarity to Icelandic among the Nordic languages but it has changed more than Icelandic with respect to phonology, inflections and syntax. Investigating Faroese is important for Icelandic linguistics because Faroese provides a unique perspective on how Icelandic could have changed or may change in the next centuries.

This course will give an overview of the grammar of Faroese (phonology, inflections, word-formation and syntax) in comparison to Icelandic and the other Nordic languages. Language changes, dialects and foreign influence on Faroese will also be discussed. Moreover, students will get some training in listening to spoken Faroese.

This seminar deals with the history of the inflectional system of Icelandic from Proto-Germanic to modern times with special emphasis on selected problems. Recent writings on Icelandic historical morphology will be discussed. We will study text examples and their value as sources of information on the development of Icelandic morphology. The development of Icelandic word formation and different types of compounds will also be discussed.

Assignments: Students will give presentations on text samples and/or particular morphological problems.

Directed Study

Directed Study

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. Machine translation I does not require programming skills as the objective is to lead together people working on language technology and people working on traditional translations.

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. A pre-requisite for Machine translation II is that students also take Machine Translation I and have taken Programming for Language technology or an equivalent course.

The course will introduce and discuss topics and methods in etymological research. Different types of etymological dictionaries will be compared. Examples from Icelandic will be discussed, i.e., the history of particular words and the information that etymological dictionaries provide on their development.

In this course we will discuss selected topics in linguistics, with a focus on the origin of language and its influence on thought. Most of the course will be devoted to the origin and evolution of language and speech, seen from a broad perspective. Classic theories and research in the field will be discussed, including hypotheses on the role of gesture (Corballis) and grooming (Dunbar), the “single mutation” theory (Chomsky), and research on the evolution of speech (Fitch). We will also discuss more recent research that provides insights into the origin and nature of speech and the language capacity, such as research on songbirds, musicality and interaction. Did human language originate in gesture or vocal calls of animals? Did it evolve out of the need for gossip and grooming? Did music have any role in the evolution of language? What can genetic studies tell us about the evolution of language? Do biological biases or the environment influence the evolution of languages?  In the course we will also discuss the relationship between language and thought. Categorization of various phenomena and objects in languages of the world will be discussed, for example in relation to color vocabulary. How does the language we speak influence the way we think and perceive the world around us?

In this course, we study the AI lifecycle, i.e. the productionisation of AI methods.
We explore the following parts of the lifecycle:
- Data collection and preparation
- Feature engineering
- Model training
- Model evaluation
- Model deployment
- Model serving
- Model monitoring
- Model maintenance
Three large projects will be handed out during the semester where students compete to solve AI problems.

The main purpose of the course is to give an overview of the syntax of the modern Scandinavian languages from a generative perspective. The emphasis is on the comparison of the Insular Scandinavian languages (Icelandic and Faroese) on the one hand and the Mainland Scandinavian languages (Danish, Norwegian and Swedish) on the other. Aspects of the syntax of some lesser-known Scandinavian varieties is also included for comparison, including Övdalian (Swe. Älvalsmålet), for instance, which preserves certain inflectional and syntactic features of Old Norse that have disappeared from the Mainland Scandinavian standard languages. Selected topics in recent research on variation in Scandinavian syntax are covered and the students will be trained in designing and administering syntactic questionnaires.

In this seminar some Eddic poems will be read and their language examined. Features which cast light on the age of the poems will be given particular attention. The evidence of the Eddic poems will be compared with that from other linguistic sources. Various methods of dating the Eddic poems will be discussed.

Do AI tools work in Icelandic? Do they work as well as in languages such as English? In this course we explore these two questions in the context of Large Language Models (LLMs) such as the ones underlying the ChatGPT and Claude AI assistants. We will examine the methods used to assess the language comprehension and production of LLMs in languages such as Icelandic and discuss whether various potential risks of increased LLM use (e.g. disinformation and bias propagation) are exacerbated in lower-resource language communities. We will place these discussions in the context of current theoretical debates, asking what AI performance in Icelandic tells us about the nature of LLMs and human language, e.g. regarding questions about how children and machines learn language.

Directed Study

Directed Study

Training in various aspects of the writing and editing of scientific texts. Various kinds of texts (non-fiction) examined and evaluated. Training in reviewing and commenting on scientific texts and in other aspects of editorial work. The main emphasis will be on the writing of articles, but other kinds of texts will also be considered, both shorter (conference abstracts, reviews) and longer (theses, books), as well as research proposals. Discussion of guidelines for the preparation of manuscripts. Types of plagiarism and how to avoid them and find them. Texts on different subjects will be used as examples, especially writings in linguistics, literature and history. The book Skrifaðu bæði skýrt og rétt will be used as a textbook (Höskuldur Þráinsson 2015).

This course is open to students of many MA programmes in the School of Humanities, cf. the regulations of the individual subjects. Students in the MA programmes in Icelandic literature, Icelandic linguistics, Icelandic studies and Icelandic teaching can take the course as part of the MA course requirements in Icelandic literature or Icelandic linguistics. Students in the MA programme in Icelandic teaching can, however, not have this course as the only linguistics or literature course in their MA.

An overview of some of the main concepts, techniques and algorithms in machine learning. Supervised learning and unsupervised learning. Data preprocessing and data visualization. Model evaluation and model selection. Linear regression, nearest neighbors, support vector machines, decision trees and ensemble methods. Deep learning. Cluster analysis and the k-means algorithm. The students implement simple algorithms in Python and learn how to use specialized software packages. At the end of the course the students work on a practical machine learning project.

In this course we cover deep neural networks and methods related to them. We study networks and methods for image, sound and text analysis. The focus will be on applications and students will present either a project or a recent paper in this field.

This is a course for people who want to be able to analyze datasets stastically to better understand them, for example through visualization with graphs. Recent years have seen an increased focus on data collection and statistical analysis within the humanities. This is particularly apparent in growing branches such as computational linguistics and psycholinguistics, cognitive literary studies and experimental philosophy, to name a few. The push towards quantitative methods occurs at a time where the validity and reliability of well-established statistical methods are called into question in other fields, with increased demands of replicability and open access as well as data protection and responsibility. In this course, students explore the value of quantitative methods in their field while getting training in the collection and analysis of data. A diverse set of research methods will be introduced, ranging from surveys to corpus analysis and experiments in which participants’ response to stimuli (such as words, texts or audio-visual materials) is quantified. Basic concepts in statistics will be reviewed, enabling students to know the difference between descriptive and inferential statistics, understand statistical significance and interpret visual representations of data in graphs. The course will be largely practical and students are expected to apply their knowledge of data collection and analysis under the instructor’s guidance. Students will work on a project within their own discipline but will also explore the possibility of cross-disciplinary work. Open source tools such as R Studio will be used for all assignments but no prior knowledge of the software or statistics in general is required. The course is suitable for all students within the humanities who want to collect quantitative data to answer interesting questions and could therefore be a useful preparation for a BA or MA project.

This course will be dedicated to the Computer Assisted Translation-technology available to translators. Students get an insight into the importance of translation memories, how humans and machines use these memories, and learn how to align text corpora to create language data and dictionaries. How to use online dictionaries, data bases and other online means. We will consider language policy, technical terms and neologisms. The translators working environment will be considered as well as skills that help freelancers get by in the gig-economy. It is hoped that experienced translators will contribute to the seminar. Students work on projects during class to prepare them for the home assignments.

The aim of the course is to create a venue in which graduate students can access an overview of the current landscape in Icelandic language technology and work on a project consistent with its latest challenges. The course is organized as a seminar series with weekly lectures sponsored by Máltæknisetur (the Icelandic Center for Language Technology, ICLT). Before each lecture, registered students meet and discuss the course readings with the instructor. The lectures will mostly be by researchers affiliated with the institutions of the ICLT (UI, RU and The Árni Magnússon Institute) but representatives from the private sector will also be invited.  

Faroeese is the language that has the strongest similarity to Icelandic among the Nordic languages but it has changed more than Icelandic with respect to phonology, inflections and syntax. Investigating Faroese is important for Icelandic linguistics because Faroese provides a unique perspective on how Icelandic could have changed or may change in the next centuries.

This course will give an overview of the grammar of Faroese (phonology, inflections, word-formation and syntax) in comparison to Icelandic and the other Nordic languages. Language changes, dialects and foreign influence on Faroese will also be discussed. Moreover, students will get some training in listening to spoken Faroese.

This seminar deals with the history of the inflectional system of Icelandic from Proto-Germanic to modern times with special emphasis on selected problems. Recent writings on Icelandic historical morphology will be discussed. We will study text examples and their value as sources of information on the development of Icelandic morphology. The development of Icelandic word formation and different types of compounds will also be discussed.

Assignments: Students will give presentations on text samples and/or particular morphological problems.

Final project

Directed Study

Directed Study

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. Machine translation I does not require programming skills as the objective is to lead together people working on language technology and people working on traditional translations.

The course is designed for master’s students in language technology and translation studies but also open to master’s students in other disciplines. It is possible to take the course for 5 as well as 10 ECTS, Machine translation I (5 ECTS) is taught before the project week and Machine translation II (5 ECTS) is taught after. A pre-requisite for Machine translation II is that students also take Machine Translation I and have taken Programming for Language technology or an equivalent course.

The course will introduce and discuss topics and methods in etymological research. Different types of etymological dictionaries will be compared. Examples from Icelandic will be discussed, i.e., the history of particular words and the information that etymological dictionaries provide on their development.

In this course we will discuss selected topics in linguistics, with a focus on the origin of language and its influence on thought. Most of the course will be devoted to the origin and evolution of language and speech, seen from a broad perspective. Classic theories and research in the field will be discussed, including hypotheses on the role of gesture (Corballis) and grooming (Dunbar), the “single mutation” theory (Chomsky), and research on the evolution of speech (Fitch). We will also discuss more recent research that provides insights into the origin and nature of speech and the language capacity, such as research on songbirds, musicality and interaction. Did human language originate in gesture or vocal calls of animals? Did it evolve out of the need for gossip and grooming? Did music have any role in the evolution of language? What can genetic studies tell us about the evolution of language? Do biological biases or the environment influence the evolution of languages?  In the course we will also discuss the relationship between language and thought. Categorization of various phenomena and objects in languages of the world will be discussed, for example in relation to color vocabulary. How does the language we speak influence the way we think and perceive the world around us?

In this course, we study the AI lifecycle, i.e. the productionisation of AI methods.
We explore the following parts of the lifecycle:
- Data collection and preparation
- Feature engineering
- Model training
- Model evaluation
- Model deployment
- Model serving
- Model monitoring
- Model maintenance
Three large projects will be handed out during the semester where students compete to solve AI problems.

The main purpose of the course is to give an overview of the syntax of the modern Scandinavian languages from a generative perspective. The emphasis is on the comparison of the Insular Scandinavian languages (Icelandic and Faroese) on the one hand and the Mainland Scandinavian languages (Danish, Norwegian and Swedish) on the other. Aspects of the syntax of some lesser-known Scandinavian varieties is also included for comparison, including Övdalian (Swe. Älvalsmålet), for instance, which preserves certain inflectional and syntactic features of Old Norse that have disappeared from the Mainland Scandinavian standard languages. Selected topics in recent research on variation in Scandinavian syntax are covered and the students will be trained in designing and administering syntactic questionnaires.

In this seminar some Eddic poems will be read and their language examined. Features which cast light on the age of the poems will be given particular attention. The evidence of the Eddic poems will be compared with that from other linguistic sources. Various methods of dating the Eddic poems will be discussed.

Do AI tools work in Icelandic? Do they work as well as in languages such as English? In this course we explore these two questions in the context of Large Language Models (LLMs) such as the ones underlying the ChatGPT and Claude AI assistants. We will examine the methods used to assess the language comprehension and production of LLMs in languages such as Icelandic and discuss whether various potential risks of increased LLM use (e.g. disinformation and bias propagation) are exacerbated in lower-resource language communities. We will place these discussions in the context of current theoretical debates, asking what AI performance in Icelandic tells us about the nature of LLMs and human language, e.g. regarding questions about how children and machines learn language.

Final project

This is a foundational course in single variable calculus. The prerequisites are high school courses on algebra, trigonometry. derivatives, and integrals. The course aims to create a foundation for understanding of subjects such as natural and physical sciences, engineering, economics, and computer science. Topics of the course include the following:

• Real numbers.
• Limits and continuous functions.
• Differentiable functions, rules for derivatives, derivatives of higher order, applications of differential calculus (extremal value problems, linear approximation).
• Transcendental functions.
• Mean value theorem, theorems of l'Hôpital and Taylor.
• Integration, the definite integral and rules/techniques of integration, primitives, improper integrals.
• Fundamental theorem of calculus.
• Applications of integral calculus: Arc length, area, volume, centroids.
• Ordinary differential equations: First-order separable and homogeneous differential equations, first-order linear equations, second-order linear equations with constant coefficients.
• Sequences and series, convergence tests.
• Power series, Taylor series.

In this course, we study several concepts related to digital literacy. The goal of the course is to introduce the students to a broad range of topics without necessarily diving deep into each one.

The Unix operating system is introduced. The file system organization, often used command-line programs, the window system, command-line environment, and shell scripting. We cover editors and data wrangling in the shell. We present version control systems (git), debugging methods, and methods to build software. Common concepts in the field of cryptography are introduced as well as concepts related to virtualization and containers.

Basics of linear algebra over the reals.  

Subject matter: Systems of linear equations, matrices, Gauss-Jordan reduction.  Vector spaces and their subspaces.  Linearly independent sets, bases and dimension.  Linear maps, range space and nullk space.  The dot product, length and angle measures.  Volumes in higher dimension and the cross product in threedimensional space.  Flats, parametric descriptions and descriptions by equations.  Orthogonal projections and orthonormal bases.  Gram-Schmidt orthogonalization.  Determinants and inverses of matrices.  Eigenvalues, eigenvectors and diagonalization.

The course provides an introduction to the methods at the heart of data science and introduces widely used software tools such as numpy, pandas, matplotlib and scikit-learn.

The course consists of 6 modules:

1. Introduction to the Python programming language.
2. Data wrangling and data preprocessing.
3. Exploratory data analysis and visualization.
4. Optimization.
5. Clustering and dimensionality reduction.
6. Regression and classification.

Each module concludes with a student project.

Note that there is an academic overlap with REI201G Mathematics and Scientific Computing and both courses cannot be valid for the same degree.

In this course, we study several concepts related to digital literacy. The goal of the course is to introduce the students to a broad range of topics without necessarily diving deep into each one.

The Unix operating system is introduced. The file system organization, often used command-line programs, the window system, command-line environment, and shell scripting. We cover editors and data wrangling in the shell. We present version control systems (git), debugging methods, and methods to build software. Common concepts in the field of cryptography are introduced as well as concepts related to virtualization and containers.

Open and closed sets. Mappings, limits and continuity. Differentiable mappings, partial derivatives and the chain rule. Jacobi matrices. Gradients and directional derivatives. Mixed partial derivatives. Curves. Vector fields and flow. Cylindrical and spherical coordinates. Taylor polynomials. Extreme values and the classification of stationary points. Extreme value problems with constraints. Implicit functions and local inverses. Line integrals, primitive functions and exact differential equations. Double integrals. Improper integrals. Green's theorem. Simply connected domains. Change of variables in double integrals. Multiple integrals. Change of variables in multiple integrals. Surface integrals. Integration of vector fields. The theorems of Stokes and Gauss.

Basic concepts in probability and statistics based on univariate calculus. 

Topics: 
Sample space, events, probability, equal probability, independent events, conditional probability, Bayes rule, random variables, distribution, density, joint distribution, independent random variables, condistional distribution, mean, variance, covariance, correlation, law of large numbers, Bernoulli, binomial, Poisson, uniform, exponential and normal random variables. Central limit theorem. Poisson process. Random sample, statistics, the distribution of the sample mean and the sample variance. Point estimate, maximum likelihood estimator, mean square error, bias. Interval estimates and hypotheses testing form normal, binomial and exponential samples. Simple linear regression. Goodness of fit tests, test of independence.