Note: in order to facilitate a stay abroad during the third semester, the electives in this semester can be taken at theuniversity abroad. The Research Seminar (as part of the Research Module) can be prepared with the thesis advisorbefore the stay abroad and is then conducted as an independent survey, communicating progress at regular intervalsand finalizing the Master thesis project plan after the return from abroad.
3
2 Vertiefungsmodule
2.1 Methods in Computational Linguistics
MODUL: Methods in Computational Linguistics Stand: XX.YY.ZZZZ1 Modulname Methods in Computational Linguistics2 Modulkurzel 0524003003 Leistungspunkte (LP) 94 Semesterwochenstunden
(SWS)6
5 Moduldauer 1 Semester6 Turnus jedes 2. Semester (WiSe)7 Sprache Englisch8 Modulverantwortliche(r) Jonas Kuhn
9 Dozenten Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Vertiefungsmodul; Pflicht; 1. Semester
11 Voraussetzungen Programming skills (in a scripting language); Undergraduate training inComputational Linguistics, Computer Science, and Formal Linguistics; stu-dents with an undergraduate background in only one of the relevant subdis-ciplines are expected to do extra readings and exercises to catch up
12 Lernziele Students become familiar with the main concepts, research questions andmethodological frameworks of computational linguistics; they know whatmethodological and practical tool basis to start from if they want to doresearch or technological development in a particular subarea.
13 Inhalt In the lectures, the most important concepts of computational lingui-stics are briefly introduced (or reviewed); standard methodologies arediscussed and put to use in practical exercises.The following topics are covered:- Levels of linguistic description- Main application areas of Computational Linguistics- Acoustic phonetics, Signal processing, speech recognition and synthe-sis- Finite state techniques, parsing techniques, unification grammars- Corpora, methodology of empirical evaluation- Language models, Hidden Markov models, probabilistic grammars- Machine learning techniques, supervised and unsupervised learning- Noisy channel model
14 Literatur/Lernmaterialien Daniel Jurafsky and James H. Martin, 2008, Speech and Language Pro-cessing, An Introduction to Natural Language Processing, ComputationalLinguistics and Speech Recognition, Prentice Hall.
15 Lehrveranstaltungen undLehrformen (Englisch)
Lecture with exercises (6 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 84h, Selbststudium: 180h
17a Studienleistungen (Englisch) assignments (requirement for exam)17b Prufungsleistungen (Englisch) written exam of 120 min.
4
18 Grundlage fur . . . CL-Vertiefungslinien19 Medienform Tafel, Folienprojektion20 Bezeichnung der zugehorigen
11 Voraussetzungen Knowledge of syntactic and semantic theory; standard parsing techniques12 Lernziele Advanced knowledge in at least two subareas of Computational Syntax and
Semantics; students are able to understand current scientific contributionsin this field and apply concepts and methods from Computational Syntaxand Semantics to new problem settings; they are able to relate insights fromtheoretical syntax and semantics to research questions in computational lin-guistics and language technology.
13 Inhalt Selection of courses comprising a total of 8 SWS from at least twosubareas of Computational Syntax and Semantics:- Grammar formalisms and grammar engineering (4 SWS)- Machine translation (2 SWS)- Statistical dependency parsing (2 SWS)- Philosophy of language (2 SWS)- Advanced Semantics (2 SWS)- Lexical Semantics (2 SWS)- further courses from the MCL catalogue that are announced for thisconcentration
6
14 Literatur/Lernmaterialien Joakim Nivre, 2005, Dependency grammar and dependency parsing. Tech-nical report, Vaxjo University.Arturo Trujillo, 1999, Translation Engines: Techniques for Machine Trans-lation. Springer.M. Butt, T. King, M.-E. Nino, and F. Segond, 1999, A Grammar Writer’sCookbook. CSLI Publications.
15 Lehrveranstaltungen undLehrformen (Englisch)
Selection of courses comprising 8 SWS (lectures, lectures with exerci-ses and/or seminar courses) from the following list:- Grammar formalisms and grammar engineering (4 SWS)- Machine Translation (2 SWS)- Statistical dependency parsing (2 SWS)- Philosophy of language (2 SWS)- Advanced Semantics (2 SWS)- Lexical Semantics (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 112h, Selbststudium: 240h
17a Studienleistungen (Englisch) ungraded course achievements in the included courses (e.g. assignments orpresentations): requirement for the final module exam
17b Prufungsleistungen (Englisch) final oral exam of 45 min.18 Grundlage fur . . .19 Medienform Tafel, Folienprojektion20 Bezeichnung der zugehorigen
11 Voraussetzungen Knowledge of phonetics and phonological theory, acoustic phonetics12 Lernziele Advanced knowledge in at least two subareas of laboratory phonology and
speech processing; students are able to understand current scientific contri-butions in the field and apply concepts and methods from laboratory phone-tics and speech processing to new problem settings; they are able to relateinsights from phonetics and phonology to research questions in computa-tional linguistics and speech technology.
13 Inhalt Selection of courses comprising a total of 8 SWS from at least twosubareas of laboratory phonology and speech processing:- Speech recognition (2 SWS)- Speech synthesis (2 SWS)- Experimental phonetics (2 SWS)- Laboratory phonology (2 SWS)- Language and speech in the human brain: Advanced methods inneurolinguistics and neurophonetics (2 SWS)- Brain computer interfaces (2 SWS)- further courses from the MCL catalogue that are announced for thisconcentration
14 Literatur/Lernmaterialien J. Clark, C. Yallop, J. Fletcher, 2007, An Introduction to Phonetics andPhonology, BlackwellDaniel Jurafsky and James H. Martin, 2008, Speech and Language Pro-cessing, An Introduction to Natural Language Processing, ComputationalLinguistics and Speech Recognition, Prentice HallK. Johnson, 2003, Acoustic and Auditory Phonetics, BlackwellPaul Taylor, 2009, Text-to-speech synthesis, Cambridge University Press
8
15 Lehrveranstaltungen undLehrformen (Englisch)
Selection of courses comprising 8 SWS (lectures, lectures with exerci-ses and/or seminar courses) from the following list:- Speech recognition (2 SWS)- Speech synthesis (2 SWS)- Experimental phonetics (2 SWS)- Laboratory Phonology (2 SWS)- Language and Speech in the Human Brain: Advanced methods inNeurolinguistics and Neurophonetics (2 SWS)- Brain Computer Interfaces (2 SWS)- Advanced Speech Perception (2 SWS)- Advanced Speech Production (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 112h, Selbststudium: 240h
17a Studienleistungen (Englisch) ungraded course achievements in the included courses (e.g. assignments orpresentations): requirement for the final module exam
17b Prufungsleistungen (Englisch) final oral exam of 45 min.18 Grundlage fur . . .19 Medienform Tafel, Folienprojektion20 Bezeichnung der zugehorigen
9 Dozenten Prof. Dr. Hinrich Schutze und weitere Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Vertiefungsmodul; Wahlpflicht; 1.-2. Se-mester
11 Voraussetzungen Familiarity with the foundations of statistical natural language processing12 Lernziele Students have gained a deep understanding of the methods used in statisti-
cal natural language processing to address computational tasks involvingwritten or spoken language. They have acquired in depth knowledge ofat least two subareas of statistical natural language processing; they un-derstand the strengths and weaknesses of different methods used in thesesubareas; they are familiar with the relevant literature; they know about exi-sting software tools relevant to the subarea and which problems they can beapplied to.
13 Inhalt Selection of courses comprising a total of 8 SWS from at least twosubareas of Statitsical Natural Language Processing- Advanced Statistical Natural Language Processing (2 SWS)- Statistical language models and smoothing (2 SWS)- Statistical constituent parsing (2 SWS)- Statistical machine translation (2 SWS)- Advanced information retrieval (2 SWS)- Machine learning for NLP (2 SWS)- Distributional and statistical approaches to semantics (2 SWS)- Statistical NLP applications (2 SWS)- further courses from the MCL catalogue that are announced for thisconcentration
14 Literatur/Lernmaterialien Manning, Christopher D., Schutze, Hinrich: Foundations of Statistical Na-tural Language Processing. MIT Press, 1999.
10
115 Lehrveranstaltungen undLehrformen (Englisch)
Selection of courses comprising 8 SWS (lectures, lectures with exerci-ses and/or seminar courses) from the following list:- Advanced Statistical Natural Language Processing (2 SWS)- Statistical language models and smoothing (2 SWS)- Statistical constituent parsing (2 SWS)- Statistical machine translation (2 SWS)- Advanced information retrieval (2 SWS)- Machine learning for NLP (2 SWS)- Distributional and statistical approaches to semantics (2 SWS)- Statistical NLP applications (2 SWS)- Probabilistic models of language and cognition (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 112h, Selbststudium: 240h
17a Studienleistungen (Englisch) ungraded course achievements in the included courses (assignments, pre-sentations, reports and/or written test); requirement for the final moduleexam
17b Prufungsleistungen (Englisch) final oral exam of 45 min.18 Grundlage fur . . .19 Medienform Tafel, Folienprojektion20 Bezeichnung der zugehorigen
9 Dozenten Dozenten, Mitarbeiter und Doktoranden des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Pflicht; 2. Seme-ster
11 Voraussetzungen Knowledge of Computational Linguistics methods, programming skills12 Lernziele Students gather practical experience in putting Computational Linguistics
methods to use in a team; they develop team skills and gather experience inthe presentation of results according to typical standards
13 Inhalt Project course (with preparatory meeting and regular team meetings)and team project work: Planning and implementation of a Computa-tional Linguistics project in a team of two-four participants; problemanalysis and specification; clarification of interfaces; implementation,testing and revision; evaluation; documentation.
14 Literatur/Lernmaterialien Daniel Jurafsky and James H. Martin, 2008, Speech and Language Pro-cessing, An Introduction to Natural Language Processing, ComputationalLinguistics and Speech Recognition, Prentice Hall.
15 Lehrveranstaltungen undLehrformen (Englisch)
Computational Linguistics Team Laboratory, project course (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 28h, Selbststudium: 150h
17a Studienleistungen (Deutsch) graded course achievement: system and report documenting the team pro-ject
17b Prufungsleistungen (Englisch) –18 Grundlage fur . . . Research Module, Master Thesis19 Medienform ggf. Kollaborationswerkzeuge (Wikis etc.)20 Bezeichnung der zugehorigen
9 Dozenten Prof. Dr. Jonas Kuhn und andere Dozenten und Gastprofessoren des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Knowledge of syntactic and semantic theory; standard parsing techniques12 Lernziele Students broaden their knowledge and skills in Computational Syntax; they
are able to understand and apply insights and methods from current workin two subareas.
13 Inhalt The module consists of two courses comprising 2 SWS each, specializingon a problem setting from the area of Computational Syntax, for instan-ce: Machine Translation, Statistical dependency parsing, Natural LanguageGeneration.
14 Literatur/Lernmaterialien Joakim Nivre, 2005, Dependency grammar and dependency parsing. Tech-nical report, Vaxjo University.Arturo Trujillo, 1999, Translation Engines: Techniques for Machine Trans-lation. Springer.Ehud Reiter, Robert Dale (2000): Building Natural Language GenerationSystems (Studies in Natural Language Processing). Cambridge UniversityPress.
15 Lehrveranstaltungen undLehrformen (Englisch)
Two of the following courses (with 2 SWS each)- Machine Translation (2 SWS)- Statistical dependency parsing (2 SWS)- Natural Language Generation (2 SWS)- Advanced Computational Syntax (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 56h, Selbststudium: 120h
13
17a Studienleistungen (Englisch) graded assignments and graded presentation/final project, if applicable;weighting of the two parts: 0.5 : 0.5
17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
9 Dozenten Prof. Dr. Jonas Kuhn und andere Dozenten und Gastprofessoren des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Knowledge of the fundamentals of semantic theory12 Lernziele Students broaden their knowledge and skills in Semantics; they are able
to understand and apply insights and methods from current work in twosubareas.
13 Inhalt The module consists of two courses comprising 2 SWS each, speciali-zing on a problem setting from the area of Theoretical and Computatio-nal Semantics, for instance: Lexical semantics, Tense, Plural, Presuppositi-ons, Propositional attitudes, Computational Semantics (Underspecification,Anaphora)
14 Literatur/Lernmaterialien Patrick Blackburn und Johan Bos. 2005. Representation and Inference forNatural Language. A First Course in Computational Semantics. CSLI Pu-blications.K. von Heusinger, C. Maienborn, P. Portner (eds.). Erscheint. Semantics:An International Handbook of Natural Language Meaning. Vol 2. Berlin:de Gruyter.Pustejovsky/Boguraev: Lexical Semantics. The Problem of Polysemy. Ox-ford University Press, 1996.
15 Lehrveranstaltungen undLehrformen (Englisch)
Two of the following courses (with 2 SWS each)- Philosophy of language (2 SWS)- Advanced Semantics (2 SWS)- Lexical Semantics (2 SWS)- Advanced Computational Semantics (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 56h, Selbststudium: 120h
17a Studienleistungen (Englisch) graded assignments and graded presentation/final project, if applicable;weighting of the two parts: 0.5 : 0.5
17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
11 Voraussetzungen Fundamental knowledge of phonetics and phonology12 Lernziele Students broaden their knowledge and skills in two subareas of Speech
Processing; they become familiar with current tools and methodologies forspeech processing and are able to apply them in their own work
13 Inhalt The module consists of two courses comprising 2 SWS each, specializingon a problem setting from Speech Processing, for instance Speech recogni-tion, Speech synthesis, Experimental phonetics, etc.
14 Literatur/Lernmaterialien as in the courses chosen15 Lehrveranstaltungen und
Lehrformen (Englisch)Two of the following courses (with 2 SWS each)- Speech recognition (2 SWS)- Speech synthesis (2 SWS)- Experimental phonetics (2 SWS)- Brain Computer Interfaces (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 56h, Selbststudium: 120h
17a Studienleistungen (Englisch) graded assignments and graded presentation/final project, if applicable;weighting of the two parts: 0.5 : 0.5
11 Voraussetzungen Fundamental knowledge of phonetics and phonology12 Lernziele Students broaden their knowledge and skills in two subareas Laboratory
Phonology; they are able to understand the current scientific literature fromthose areas and relate the insights to a specific problem setting
13 Inhalt The module consists of two courses comprising 2 SWS each, specializingon a problem setting from Laboratory Phonology, such as Advanced SpeechPerception, Advanced Speech Production, etc.
14 Literatur/Lernmaterialien as in the courses chosen15 Lehrveranstaltungen und
Lehrformen (Englisch)Two of the following courses (with 2 SWS each)- Laboratory Phonology (2 SWS)- Language and Speech in the Human Brain: Advanced methods inNeurolinguistics and Neurophonetics (2 SWS)- Advanced Speech Perception (2 SWS)- Advanced Speech Production (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 56h, Selbststudium: 120h
17a Studienleistungen (Englisch) graded assignments and graded presentation/final project, if applicable;weighting of the two parts: 0.5 : 0.5
11 Voraussetzungen Statistical natural language processing12 Lernziele Students are familiar with the main methods that are used in statistical na-
tural language processing, with particular emphasis on generative and dis-criminative models. They have acquired in depth knowledge of at least twoof these methods. They understand the strengths and weaknesses of thesemethods and can explain which criteria can be used to select an appropriatemethod for a problem. They are familiar with the relevant literature. Theyknow for a representative number of applications which methods are stan-dardly used to address them.
13 Inhalt The module consists of two courses comprising 2 SWS each, specializingon a set of methods in statistical natural language processing, such as Dis-tributional and statistical approaches to semantics, Unsupervised and semi-supervised learning, Distributional and statistical approaches to semantics,etc.
14 Literatur/Lernmaterialien as in the courses chosen15 Lehrveranstaltungen und
Lehrformen (Englisch)Two of the following courses (with 2 SWS each):- Advanced Statistical Natural Language Processing (2 SWS)- Machine learning for NLP (2 SWS)- Distributional and statistical approaches to semantics (2 SWS)- Unsupervised and semisupervised learning (2 SWS)- Evaluation and statistical testing (2 SWS)- Probabilistic models of language and cognition (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 56h, Selbststudium: 120h
17a Studienleistungen (Englisch) graded assignments and graded presentation/final project, if applicable;weighting of the two parts: 0.5 : 0.5
17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
Modulprufung(en)(wird vom Prufungsamt ausgefullt)
21 Import-Export Anbieter (Fakultat/Institut):
18
Nutzer (Studiengang):
19
Applications of Statistical Natural Language Processing
MODUL: Applications of Statistical Natural Language Processing Stand: XX.YY.ZZZZ1 Modulname Applications of Statistical Natural Language Processing2 Modulkurzel 0524006013 Leistungspunkte (LP) 64 Semesterwochenstunden
(SWS)4
5 Moduldauer 1-2 Semester6 Turnus unregelmaßig7 Sprache Englisch8 Modulverantwortliche(r) Prof. Dr. Hinrich Schutze
11 Voraussetzungen Statistical natural language processing12 Lernziele Students are familiar with the applications that statistical natural language
processing is used for. They have acquired in depth knowledge of at leasttwo of these applications. They understand the strengths and weaknessesof statistical versus symbolic approaches to these applications. They arefamiliar with the relevant literature. They know for a representative numberof applications which existing software tools can be used to address them.
13 Inhalt The module consists of two courses comprising 2 SWS each, specializingon a type of application of statistical natural language processing, such asStatistical constituent parsing, Statistical machine translation, Advanced in-formation retrieval, etc.
14 Literatur/Lernmaterialien as in the courses chosen15 Lehrveranstaltungen und
Lehrformen (Englisch)Two of the following courses (with 2 SWS each):- Statistical language models and smoothing (2 SWS)- Statistical NLP applications (2 SWS)- Statistical constituent parsing (2 SWS)- Statistical machine translation (2 SWS)- Advanced information retrieval (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 56h, Selbststudium: 120h
17a Studienleistungen (Englisch) graded assignments and graded presentation/final project, if applicable;weighting of the two parts: 0.5 : 0.5
17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
9 Dozenten Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen as in the course chosen12 Lernziele Students become familiar with an advanced topic area of computational
linguistics and at the same time develop their oral presentation skills andscientific writing skills; they practise self-organized work in an independentstudy and train their competence to put specific scientific contributions in abroader context and provide a critical discussion.
13 Inhalt This module consists of (1) a course in an advanced topic area of compu-tational linguistics (comprising 2 SWS), such as Machine Translation, Na-tural Language Generation, Advanced Semantics, Advanced Speech Per-ception, Advanced Speech Production, Statistical constituent parsing, Sta-tistical machine translation, etc., and (2) students’ independent studies of aspecific thematic complex from the area covered in the course; the investi-gations are conveyed in a long student presentation during the course andwritten up as a seminar paper of c. 20 pages.
14 Literatur/Lernmaterialien as in the course chosen
21
15 Lehrveranstaltungen undLehrformen (Englisch)
(1) One seminar course (2 SWS) from catalogue below, under thecondition (a) that the lecturer agrees that the student may take thecourse as a seminar with a long student presentation and a seminarpaper, and (b) that the course has not been chosen for any other module;(2) Long presentation and seminar paper.Course catalogue:- Machine Translation (2 SWS)- Statistical dependency parsing (2 SWS)- Natural Language Generation (2 SWS)- Philosophy of language (2 SWS)- Advanced Computational Syntax (2 SWS)- Advanced Semantics (2 SWS)- Advanced Computational Semantics (2 SWS)
- Speech recognition (2 SWS)- Speech synthesis (2 SWS)- Experimental phonetics (2 SWS)- Laboratory Phonology (2 SWS)- Language and Speech in the Human Brain: Advanced methods inNeurolinguistics and Neurophonetics (2 SWS)- Brain Computer Interfaces (2 SWS)- Advanced Speech Perception (2 SWS)- Advanced Speech Production (2 SWS)
- Advanced Statistical Natural Language Processing (2 SWS)- Statistical language models and smoothing (2 SWS)- Statistical NLP applications (2 SWS)- Statistical constituent parsing (2 SWS)- Statistical machine translation (2 SWS)- Advanced information retrieval (2 SWS)- Machine learning for NLP (2 SWS)- Distributional and statistical approaches to semantics (2 SWS)- Unsupervised and semisupervised learning (2 SWS)- Evaluation and statistical testing (2 SWS)- Probabilistic models of language and cognition (2 SWS)
9 Dozenten Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen as in the course chosen12 Lernziele Students become familiar with an advanced topic area of computational
linguistics and at the same time develop their oral presentation skills andscientific writing skills; they practise self-organized work in an independentstudy and train their competence to put specific scientific contributions in abroader context and provide a critical discussion.
13 Inhalt This module consists of (1) a course in an advanced topic area of compu-tational linguistics (comprising 2 SWS), such as Machine Translation, Na-tural Language Generation, Advanced Semantics, Advanced Speech Per-ception, Advanced Speech Production, Statistical constituent parsing, Sta-tistical machine translation, etc., and (2) students’ independent studies of aspecific thematic complex from the area covered in the course; the investi-gations are conveyed in a long student presentation during the course andwritten up as a seminar paper of c. 20 pages.
14 Literatur/Lernmaterialien as in the course chosen15 Lehrveranstaltungen und
Lehrformen (Englisch)(1) One seminar course (2 SWS) from catalogue below, under thecondition (a) that the lecturer agrees that the student may take thecourse as a seminar with a long student presentation and a seminarpaper, and (b) that the course has not been chosen for any other module;(2) Long presentation and seminar paper.Course catalogue:- Machine Translation (2 SWS)- Statistical dependency parsing (2 SWS)- Natural Language Generation (2 SWS)- Philosophy of language (2 SWS)- Advanced Computational Syntax (2 SWS)- Advanced Semantics (2 SWS)- Advanced Computational Semantics (2 SWS)
23
- Speech recognition (2 SWS)- Speech synthesis (2 SWS)- Experimental phonetics (2 SWS)- Laboratory Phonology (2 SWS)- Language and Speech in the Human Brain: Advanced methods inNeurolinguistics and Neurophonetics (2 SWS)- Brain Computer Interfaces (2 SWS)- Advanced Speech Perception (2 SWS)- Advanced Speech Production (2 SWS)
- Advanced Statistical Natural Language Processing (2 SWS)- Statistical language models and smoothing (2 SWS)- Statistical NLP applications (2 SWS)- Statistical constituent parsing (2 SWS)- Statistical machine translation (2 SWS)- Advanced information retrieval (2 SWS)- Machine learning for NLP (2 SWS)- Distributional and statistical approaches to semantics (2 SWS)- Unsupervised and semisupervised learning (2 SWS)- Evaluation and statistical testing (2 SWS)- Probabilistic models of language and cognition (2 SWS)
Module im Katalog MCL 3 umfassen 6 Leistungspunkte und werden aufgrund von Prufungsleistungen bewertet.
(derzeit keine Module)
24
3.2.4 MCL 4
Module im Katalog MCL 4 umfassen 6 Leistungspunkte und werden aufgrund von Prufungsleistungen bewertet. Furdie Prufungsleistung wird eine Studienleistung als Vorleistung vorausgesetzt.
Grammar Formalisms and Grammar Engineering
MODUL: Grammar Formalisms and Grammar Engineering Stand: XX.YY.ZZZZ1 Modulname Grammar Formalisms and Grammar Engineering2 Modulkurzel 0524004103 Leistungspunkte (LP) 64 Semesterwochenstunden
(SWS)4
5 Moduldauer 1 Semester6 Turnus unregelmaßig7 Sprache Englisch8 Modulverantwortliche(r) Prof. Dr. Jonas Kuhn
9 Dozenten Prof. Dr. Jonas Kuhn und andere Dozenten und Gastprofessoren des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Good knowledge of syntactic theory; parsing techniques12 Lernziele Students have an understanding of the theoretical and computational issues
in the representation and processing of grammatical and semantic knowled-ge; they are familiar with engineering methods for language resources andthe specification of linguistic knowledge; they know about possible waysof combining corpus-based methods and knowledge based manual specifi-cation techniques.
13 Inhalt Constraint-based grammar formalisms; algorithmic considerations; broad-coverage grammar writing; means of abstraction; disambiguation techni-ques; robust processing techniques.
14 Literatur/Lernmaterialien M. Butt, T. King, M.-E. Nino, and F. Segond, 1999, A Grammar Writer’sCookbook. CSLI Publications.
15 Lehrveranstaltungen undLehrformen (Englisch)
seminar course (4 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 56h, Selbststudium: 120h
17a Studienleistungen (Englisch) Homework assignments, as a requirement for the final exam17b Prufungsleistungen (Englisch) written exam of 90 min.18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
Module im Katalog MCL 5 umfassen 6 Leistungspunkte. Die erfolgreiche Teilnahme wird aufgrund von unbenotetenStudienleistungen festgestellt.
(derzeit keine Module)
3.2.6 MCL 6
Module im Katalog MCL 6 umfassen 3 Leistungspunkte und werden aufgrund von benoteten Studienleistungen be-wertet.
(Bitte beachten: Damit die Gesamtzahl der benoteten Einzelmodule im Studium nicht zu groß wird, sollte die Belegungvon Modulen aus MCL 6 der Ausnahmefall bleiben. Die entsprechenden Einzelveranstaltungen lassen sich auch imRahmen der kombinierten Module “Topics in Computational Syntax”, “Topics in Computational Semantics”, “Topicsin Laboratory Phonology”, “Topics in Speech Processing”, “Applications of Statistical Natural Language Processing”und “Methods in Statistical Natural Language Processing”) belegen.
9 Dozenten Prof. Dr. Jonas Kuhn10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Knowledge of computational syntax and fundamentals of statistical NaturalLanguage Processing
12 Lernziele Students are familiar with the most important approaches in classical sym-bolic and in statistical machine translation; they have learned to identify andclassify translation challenges and are able to put some standard translationmethods to use for a given dataset.
26
13 Inhalt Classical symbolic translation approaches:- direct translation- syntactic and semantic transfer- interlingual translation- example-based translationStatistical machine translation:- noisy channel model- word-based and phrase-based translation- syntactically informed statistical translationParallel corpus based techniques in Computational Linguistics
9 Dozenten Prof. Dr. Jonas Kuhn, Dr. Bernd Bohnet10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Knowledge of standard parsing techniques; statistical Natural LanguageProcessing
12 Lernziele Students are familiar with the different approaches to dependency parsingfrom the current literature in Computational Linguistics; they understandrepresentational choices made in dependency parsing and are able to applyavailable parsers to standard data sets and perform a comparative evaluation
13 Inhalt - constituent parsing vs. dependency parsing- classical dependency grammar, representational issues- Eisner’s algorithm- transition-based dependency parsing- graph-based dependency parsing (minimum spanning tree algorithms)- dealing with non-projectivity
9 Dozenten Dr. Bernd Bohnet, Prof. Dr. Jonas Kuhn10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Knowledge in computational syntax and semantics: parsing, grammar for-malisms
12 Lernziele Students are familiar with the field of Natural Language Generation and thevarious current generation approaches; they know about typical generationissues and the application contexts of Natural Generation Systems.
13 Inhalt The Architecture of a Natural Language Generation System; DocumentPlanning; Microplanning; Surface Realisation
14 Literatur/Lernmaterialien Ehud Reiter, Robert Dale (2000): Building Natural Language GenerationSystems (Studies in Natural Language Processing). Cambridge UniversityPress.
15 Lehrveranstaltungen undLehrformen (Englisch)
seminar course (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 28h, Selbststudium: 60h
17a Studienleistungen (Englisch) graded course achievements (assignments, presentation and/or final project)17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
9 Dozenten Jonas Kuhn und andere Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Thorough background in Computational Syntax12 Lernziele Students develop an understanding of state-of-the-art research in a parti-
cular subfield of Computational Syntax; they are able to assess the ad-vantages and disadvantages of particular approaches against a theoreticalbackground; they are able to put to use available systems for the subfieldcovered.
13 Inhalt Current original scientific contributions (mainly conference papers) from aparticular subfield of Computational Syntax are discussed and contextuali-zed, taking theoretical considerations into account and discussing practicalaspects and the evaluation methodology.
14 Literatur/Lernmaterialien Current conference papers from the respective subfield15 Lehrveranstaltungen und
Lehrformen (Englisch)seminar course (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 28h, Selbststudium: 60h
17a Studienleistungen (Englisch) graded course achievements (assignments, presentation and/or final project)17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Knowledge of Semantic Theory and the field of Pragmatics12 Lernziele Participants are familiar with the positions and problems relevant for the
shaping of the semantic and pragmatic concepts used in present-day formaland computational linguistics.
13 Inhalt The course discusses the two main trends in the philosophy of language:The “ideal language philosophy”, which propagates the application ofmathematical or logical methods in the semantic analysis of natural lan-guage, and the “ordinary language philosophy”, which takes the closeobservation of everyday usage to be fundamental for semantics. Morerecent work shows, that a synthesis between the two trends is possible.After these “conventional” theories of meaning, we discuss modern “na-turalistic” theories: i.e. Quine’s attempt to reduce meaning to sense sti-muli and dispositions to react, or the Gricean attempt at a reduction ofmeaning to speaker’s intentions (which in turn were to be explained asneurophysiological states).
14 Literatur/Lernmaterialien Peter Ludlow, Readings in the Philosophy of Language, MIT Press, 1997.15 Lehrveranstaltungen und
Lehrformen (Englisch)seminar course (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 28h, Selbststudium: 60h
17a Studienleistungen (Englisch) graded course achievements (assignments, presentation and/or final project)17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
9 Dozenten PD Dr. Antje Roßdeutscher10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Knowledge in syntax and basic knowledge in semantics (Discourse Repre-sentation Theory)
12 Lernziele Students are familiar with methods and frameworks for the formal represen-tation of lexical knowledge on different linguistic levels. They can designsemantic representations for lexical items.
13 Inhalt - compositional semantics of (complex) words, phrases or sentences- concepts for representing the semantics of basic categories, (verbs,nouns, adjectives, or prepositions), e.g. argument structure, temporalprofile, semantic classes for verbs)- overview of lexical-semantic resources (such as FrameNet, VerbNet,and WordNet) and role-semantically annotated corpora (such as Prop-Bank, NomBank) - lexical and structural ambiguity- syntax-semantics-interface- lexicon and text-representation- Space in Natural Language
14 Literatur/Lernmaterialien Dowty, David R.: Word Meaning and Montague Grammar, Kluwer Acade-mic Publishers, 1979.Pustejovsky/Boguraev: Lexical Semantics. The Problem of Polysemy. Ox-ford University Press, 1996.Geeraerts, Dirk: Theories of Lexical Semantics, Oxford University Press,2010
15 Lehrveranstaltungen undLehrformen (Englisch)
seminar course (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 28h, Selbststudium: 60h
17a Studienleistungen (Englisch) graded course achievements (assignments, presentation and/or final project)17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
11 Voraussetzungen Thorough background in Computational Syntax and Semantics12 Lernziele Students develop an understanding of state-of-the-art research in particular
subfields of semantics, among them are tense, plural, presuppositions, pro-positional attitudes, information theory, dialogue; they are able to assess theadvantages and disadvantages of particular theories and are able to checktheir predictions.
13 Inhalt Current original scientific contributions from a particular subfield of Se-mantics are discussed and contextualized.
14 Literatur/Lernmaterialien Shalom Lappin. 1995. The Handbook of Contemporary Semantic Theory.Oxford: Blackwell’s.K. von Heusinger, C. Maienborn, P. Portner (eds.). Erscheint. Semantics:An International Handbook of Natural Language Meaning. Vol 2. Berlin:de Gruyter.
15 Lehrveranstaltungen undLehrformen (Englisch)
seminar course (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 28h, Selbststudium: 60h
17a Studienleistungen (Englisch) graded course achievements (assignments, presentation and/or final project)17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
9 Dozenten Prof. Dr. Uwe Reyle und andere Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Thorough background in Computational Syntax and Semantics12 Lernziele Students develop an understanding of state-of-the-art research in particu-
lar subfields of Computational Semantics; they are able to assess the ad-vantages and disadvantages of particular approaches against a theoreticalbackground; they are able to put to use available systems for the subfieldcovered.
13 Inhalt Current original scientific contributions from particular subfields of Com-putational Semantics are discussed and contextualized.
14 Literatur/Lernmaterialien K. von Heusinger, C. Maienborn, P. Portner (eds.). Erscheint. Semantics:An International Handbook of Natural Language Meaning. Vol 2. Berlin:de Gruyter.J. van Genabith, H. Kamp und U. Reyle. Erscheint. Discourse Represen-tation Theory. In: Dov Gabbay (ed.): Handbook of Philosophical Logic.Kluwer.Patrick Blackburn und Johan Bos. 2005. Representation and Inference forNatural Language. A First Course in Computational Semantics. CSLI Pu-blications.
15 Lehrveranstaltungen undLehrformen (Deutsch)
Seminar (2 SWS)
Lehrveranstaltungen undLehrformen (Englisch)
seminar course (2 SWS)
16 Abschatzung des Arbeitsauf-wands
Prasenzzeit: 28h, Selbststudium: 60h
17a Studienleistungen (Englisch) graded course achievements (assignments, presentation and/or final project)17b Prufungsleistungen (Englisch) –18 Grundlage fur . . .19 Medienform20 Bezeichnung der zugehorigen
9 Dozenten Antje Schweitzer und weitere Dozenten und Gastprofessoren des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen acoustic phonetics12 Lernziele Students are familiar with formant synthesis and various approaches to con-
catenative synthesis; they are familiar with the typical architecture of text-to-speech systems and their components. They are able to apply synthesistools.
13 Inhalt applications of speech synthesis, formant synthesis, concatenative synthe-sis, components of text-to-speech systems
14 Literatur/Lernmaterialien Paul Taylor, Text-to-speech synthesis.15 Lehrveranstaltungen und
9 Dozenten Katrin Schneider und weitere Dozenten und Gastprofessoren des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen acoustic phonetics12 Lernziele Students are able to plan and to carry out their own phonetic experiments;
they are able to statistically analyse and interpret their results.13 Inhalt methods in experimental phonetics, statistical exploration of phonetic data14 Literatur/Lernmaterialien Ladefoged, 2005, Phonetic Data Analysis: An Introduction to Fieldwork
and Instrumental Techniques, Blackwell Publishing15 Lehrveranstaltungen und
11 Voraussetzungen knowledge of phonetics and phonology12 Lernziele experimental validation of phonological and general linguistic topics13 Inhalt Study of current phonological research issues, selection of experimental
methods for validation of phonological theories14 Literatur/Lernmaterialien C. Fougeron, B. Kuhnert, M. D’Imperio, N. Valle, Laboratory Phonology
10, 2010, De Gruyter Moutonand conference proceedings of recent Laboratory Phonology conferences
Language and Speech in the Human Brain: Advanced methods in Neurolinguistics and Neurophonetics
MODUL: Language and Speech in the Human Brain: Advanced methods inNeurolinguistics and Neurophonetics Stand: XX.YY.ZZZZ1 Modulname Language and Speech in the Human Brain: Advanced methods in Neuro-
linguistics and Neurophonetics2 Modulkurzel 0524005143 Leistungspunkte (LP) 34 Semesterwochenstunden
(SWS)2
5 Moduldauer 1 Semester6 Turnus SoSe7 Sprache Englisch8 Modulverantwortliche(r) Prof. Dr. Grzegorz Dogil
11 Voraussetzungen12 Lernziele Students are able to understand and appreciate literature in neurolinguistics
and neurophonetics13 Inhalt Basic methods and procedures of neurolinguistics and neurophonetics14 Literatur/Lernmaterialien Internet-Tutorial Sprache & Gehirn, http://www.ims.uni-
stuttgart.de/phonetik/joerg/sgtutorial/15 Lehrveranstaltungen und
11 Voraussetzungen12 Lernziele Students understand various research programs and possible applications
of brain computer interfaces13 Inhalt various research programs on brain computer interfaces14 Literatur/Lernmaterialien Sitaram et al., 2007, FMRI brain-computer interface: a tool for neuros-
cientific research and treatment, in Computational intelligence and neu-roscience, 25487, Hindawi Publishing Corporation. Dogil/Reiterer, 2009,Language Talent and Brain Activity, de Gruyter
9 Dozenten Grzegorz Dogil und weitere Dozenten und Gastprofessoren des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Thorough background in Phonetics and Phonology12 Lernziele Students develop an understanding of state-of-the-art research in Speech
Perception; they are able to assess the advantages and disadvantages ofparticular approaches against a theoretical background.
13 Inhalt Current original scientific contributions (mainly conference papers) fromthe field of Speech Perception are discussed and contextualized, takingtheoretical considerations into account and/or discussing practical aspects.
11 Voraussetzungen Thorough background in Phonetics and Phonology12 Lernziele Students develop an understanding of state-of-the-art research in Speech
Production; they are able to assess the advantages and disadvantages ofparticular approaches against a theoretical background.
13 Inhalt Current original scientific contributions (mainly conference papers) fromthe field of Speech Production are discussed and contextualized, takingtheoretical considerations into account and/or discussing practical aspects.
14 Literatur/Lernmaterialien W.J.M. Levelt, Speaking: From Intention to Articulation, 1989, MIT PressW.J.M. Levelt, A. Roelofs, A.S. Meyer, A theory of lexical access in speechproduction, Behavioral and Brain Sciences 22, 1999, Cambridge UniversityPressCurrent conference papers from the respective subfield
9 Dozenten Prof. Dr. Hinrich Schutze10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl; 1.-3. Seme-ster
11 Voraussetzungen Familiarity with the foundations of statistical natural language processing12 Lernziele Students have acquired in depth knowledge of one advanced subarea of
statistical natural language processing; they understand the strengths andweaknesses of different methods used in the subarea; they are familiar withthe relevant literature; they know about existing software tools relevant tothe subarea and which problems they can be applied to.
13 Inhalt Recent publications in a subarea of statistical natural language processingare presented, analyzed and discussed.
14 Literatur/Lernmaterialien Recent publications in a subarea of statistical natural language processing15 Lehrveranstaltungen und
9 Dozenten Prof. Dr. Hinrich Schutze10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen Statistical natural language processing (recommended)12 Lernziele Students have acquired in depth knowledge of several different language
models and smoothing techniques and are familiar with the relevant litera-ture.
13 Inhalt - Discounting models- Jelinek-Mercer models- Kneser-Ney models- Class-based models- Similarity-based models- The size-accuracy tradeoff in language modeling
14 Literatur/Lernmaterialien Manning, Christopher D., Schutze, Hinrich: Foundations of Statistical Na-tural Language Processing. MIT Press, 1999. Chen, Goodman: An Empi-rical Study of Smoothing Techniques for Language Modeling, TR-10-9,Microsoft, 1998. Jelinek: Statistical methods for speech recognition. 1997.MIT Press.
9 Dozenten Prof. Dr. Hinrich Schutze, PD Dr. Helmut Schmid10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen Statistical natural language processing (recommended)12 Lernziele Students have acquired in depth knowledge of at least one application area
of statistical natural language processing and are familiar with the relevantliterature.
13 Inhalt The seminar will cover one or two NLP applications in depth. Examplesinclude:- Part-of-speech tagging- Chunking- Sentiment detection- Coreference resolution- Named entity recognition- Summarization- Paraphrasing and textual entailment- Segmentation methods
14 Literatur/Lernmaterialien Manning, Christopher D., Schutze, Hinrich: Foundations of Statistical Na-tural Language Processing. MIT Press, 1999. Jurafsky/Martin: Speech andLanguage Processing, An Introduction to Natural Language Processing,Computational Linguistics and Speech Recognition, Prentice Hall, 2008.
9 Dozenten PD Dr. Helmut Schmid10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen Knowledge of parsing and statistical methods12 Lernziele Students have acquired in depth knowledge of statistical constituent parsing
methods and are familiar with the relevant literature.13 Inhalt - Probabilistic context-free grammars
9 Dozenten Dr. Alexander Fraser10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen knowledge of statistical methods for natural language processing12 Lernziele Students have acquired in depth knowledge of statistical machine translati-
on methods and are familiar with the relevant literature and an open sourcestatistical machine translation system.
13 Inhalt - Basic statistical modeling for machine translation- Automatic and manual evaluation of machine translation output- Bitext alignment of parallel sentence pairs- Basic phrase-based statistial machine translation models and decoding- Log-linear models and minimum error rate training- Advanced topics: discriminative word alignment, morphologicalmodeling, syntactic modeling
14 Literatur/Lernmaterialien Philipp Koehn. Statistical Machine Translation. Cambridge UniversityPress. 2010.
9 Dozenten Prof. Dr. Hinrich Schutze, Dr. Christina Lioma10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen Information Retrieval and Text Mining (recommended)12 Lernziele Students have acquired in depth knowledge of several advanced areas of
information retrieval and are familiar with the relevant literature.13 Inhalt - Question answering
- Probabilistic information retrieval models- Statistical language models for information retrieval- Latent semantic indexing- Text classification and support vector machines- Learning to rank- NLP methods for information retrieval
14 Literatur/Lernmaterialien Manning/Raghavan/Schutze, Introduction to Information Retrieval, Cam-bridge University Press, 2008.
9 Dozenten Prof. Dr. Hinrich Schutze, Florian Laws10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen Statistical natural language processing (recommended)12 Lernziele Students have acquired in depth knowledge of several machine learning
methods that are used in natural language processing and are familiar withthe relevant literature.
13 Inhalt - Maximum entropy models- Regression and regularized regression- Support vector machines- Sequence models- Generative models- Parameter estimation
14 Literatur/Lernmaterialien Abney, Semisupervised Learning for Computational Linguistics, Chapmanand Hall/CRC, 2007.Bishop, Pattern Recognition and Machine Learning, Springer, 2007.
9 Dozenten Prof. Dr. Hinrich Schutze10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen Statistical natural language processing (recommended)12 Lernziele Students have acquired in depth knowledge of distributional and statistical
approaches to semantics and are familiar with the relevant literature.13 Inhalt - Semantic vector spaces
- Statistical word sense disambiguation- Acquisition of lexical semantics and world knowledge- Semantic role labeling- Statistical representations of context- Semantic feature design and acquisition for NLP applications
14 Literatur/Lernmaterialien Manning, Christopher D., Schutze, Hinrich: Foundations of Statistical Na-tural Language Processing. MIT Press, 1999.Jurafsky/Martin: Speech and Language Processing, An Introduction to Na-tural Language Processing, Computational Linguistics and Speech Reco-gnition, Prentice Hall, 2008.
9 Dozenten Prof. Dr. Hinrich Schutze10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen Statistical natural language processing (recommended)12 Lernziele Students have acquired in depth knowledge of unsupervised and semisuper-
vised learning methods used in natural language processing and are familiarwith the relevant literature.
14 Literatur/Lernmaterialien Manning, Christopher D., Schutze, Hinrich: Foundations of Statistical Na-tural Language Processing. MIT Press, 1999.Abney, Steven: Semisupervised Learning for Computational Linguistics,Chapman and Hall/CRC, 2007.Manning/Raghavan/Schutze, Introduction to Information Retrieval, Cam-bridge University Press, 2008.
9 Dozenten Prof. Dr. Hinrich Schutze10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen Statistical natural language processing (recommended)12 Lernziele Students have acquired in depth knowledge of the evaluation methodology
used in statistical natural language processing and of statistical hypothe-sis testing, are familiar with the relevant literature and know how to use astatistical package such as R.
13 Inhalt - Evaluation methodology in StatNLP- Statistical hypothesis tests- The main distributions used in hypothesis testing- The statistical package R- Hypothesis tests used for NLP applications like collocations
14 Literatur/Lernmaterialien Manning, Christopher D., Schutze, Hinrich: Foundations of Statistical Na-tural Language Processing. MIT Press, 1999.Snedecor, Cochran: Statistical methods, Iowa State University Press, 1989.
9 Dozenten Prof. Dr. Hinrich Schutze10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computerlinguistik; Spezialisierungsmodul; Wahl; 1.-3. Semester
11 Voraussetzungen Statistical natural language processing (recommended)12 Lernziele Students have acquired in depth knowledge of probabilistic models of lan-
guage and cognition and are familiar with the relevant literature.13 Inhalt - Neural networks
14 Literatur/Lernmaterialien Rumelhart, McClelland, PDP research group (1986), Parallel DistributedProcessing. Explorations in the Microstructure of Cognition. Volume 1.MIT Press.
Module im Katalog MCL 7 umfassen 3 Leistungspunkte. Die erfolgreiche Teilnahme wird aufgrund von unbenotetenStudienleistungen festgestellt.
(Bitte beachten: Es darf nur eine begrenzte Anzahl von Spezialisierungsmodulen durch unbenotete Studienleistungerbracht werden. Diese konnen sinnvollerweise dafur genutzt werden, Lehrveranstaltungen außerhalb des eigenenHauptschwerpunkts zu belegen.)
9 Dozenten Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl
11 Voraussetzungen Fundamental knowledge in the broader area of the topic chosen.12 Lernziele Students become familiar with an additional subarea from Computational
Linguistics, typically distinct from their main focus area; they understandwhat the specific problem setting in this area is and are able to addressinterface issues with specialists; they get to know what standard tools andmethodologies are available for the area, so they can integrate insights fromthis area in their own work.
13 Inhalt This module type consists of a 2 SWS course, with ungraded course achie-vements, from a subarea of Computational Linguistics, such as such asMachine Translation, Natural Language Generation, Advanced Semanti-cs, Advanced Speech Perception, Advanced Speech Production, Statisticalconstituent parsing, Statistical machine translation, etc.
14 Literatur/Lernmaterialien as in the course chosen
55
15 Lehrveranstaltungen undLehrformen (Englisch)
One seminar course (2 SWS) from the following catalogue that has notbeen chosen for any other module:- Machine translation (2 SWS)- Statistical dependency parsing (2 SWS)- Natural language generation (2 SWS)- Advanced Computational Syntax (2 SWS)- Philosophy of language (2 SWS)- Advanced Semantics (2 SWS)- Lexical Semantics (2 SWS)- Advanced Computational Semantics (2 SWS)
- Speech recognition (2 SWS)- Speech synthesis (2 SWS)- Experimental phonetics (2 SWS)- Laboratory Phonology (2 SWS)- Language and Speech in the Human Brain: Advanced methods inNeurolinguistics and Neurophonetics (2 SWS)- Brain Computer Interfaces (2 SWS)- Advanced Speech Perception (2 SWS)- Advanced Speech Production (2 SWS)
- Advanced Statistical Natural Language Processing (2 SWS)- Statistical language models and smoothing (2 SWS)- Statistical NLP applications (2 SWS)- Statistical constituent parsing (2 SWS)- Statistical machine translation (2 SWS)- Advanced information retrieval (2 SWS)- Machine learning for NLP (2 SWS)- Distributional and statistical approaches to semantics (2 SWS)- Unsupervised and semisupervised learning (2 SWS)- Evaluation and statistical testing (2 SWS)- Probabilistic models of language and cognition (2 SWS)
9 Dozenten Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl
11 Voraussetzungen Fundamental knowledge in the broader area of the topic chosen.12 Lernziele Students become familiar with an additional subarea from Computational
Linguistics, typically distinct from their main focus area; they understandwhat the specific problem setting in this area is and are able to addressinterface issues with specialists; they get to know what standard tools andmethodologies are available for the area, so they can integrate insights fromthis area in their own work.
13 Inhalt This module type consists of a 2 SWS course, with ungraded course achie-vements, from a subarea of Computational Linguistics, such as such asMachine Translation, Natural Language Generation, Advanced Semanti-cs, Advanced Speech Perception, Advanced Speech Production, Statisticalconstituent parsing, Statistical machine translation, etc.
14 Literatur/Lernmaterialien as in the course chosen15 Lehrveranstaltungen und
Lehrformen (Englisch)One seminar course (2 SWS) from the following catalogue that has notbeen chosen for any other module:- Machine translation (2 SWS)- Statistical dependency parsing (2 SWS)- Natural language generation (2 SWS)- Advanced Computational Syntax (2 SWS)- Philosophy of language (2 SWS)- Advanced Semantics (2 SWS)- Lexical Semantics (2 SWS)- Advanced Computational Semantics (2 SWS)
- Speech recognition (2 SWS)- Speech synthesis (2 SWS)- Experimental phonetics (2 SWS)- Laboratory Phonology (2 SWS)- Language and Speech in the Human Brain: Advanced methods inNeurolinguistics and Neurophonetics (2 SWS)- Brain Computer Interfaces (2 SWS)- Advanced Speech Perception (2 SWS)- Advanced Speech Production (2 SWS)
57
- Advanced Statistical Natural Language Processing (2 SWS)- Statistical language models and smoothing (2 SWS)- Statistical NLP applications (2 SWS)- Statistical constituent parsing (2 SWS)- Statistical machine translation (2 SWS)- Advanced information retrieval (2 SWS)- Machine learning for NLP (2 SWS)- Distributional and statistical approaches to semantics (2 SWS)- Unsupervised and semisupervised learning (2 SWS)- Evaluation and statistical testing (2 SWS)- Probabilistic models of language and cognition (2 SWS)
9 Dozenten Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Wahl
11 Voraussetzungen Fundamental knowledge in the broader area of the topic chosen.12 Lernziele Students become familiar with an additional subarea from Computational
Linguistics, typically distinct from their main focus area; they understandwhat the specific problem setting in this area is and are able to addressinterface issues with specialists; they get to know what standard tools andmethodologies are available for the area, so they can integrate insights fromthis area in their own work.
13 Inhalt This module type consists of a 2 SWS course, with ungraded course achie-vements, from a subarea of Computational Linguistics, such as such asMachine Translation, Natural Language Generation, Advanced Semanti-cs, Advanced Speech Perception, Advanced Speech Production, Statisticalconstituent parsing, Statistical machine translation, etc.
14 Literatur/Lernmaterialien as in the course chosen15 Lehrveranstaltungen und
Lehrformen (Englisch)One seminar course (2 SWS) from the following catalogue that has notbeen chosen for any other module:- Machine translation (2 SWS)- Statistical dependency parsing (2 SWS)- Natural language generation (2 SWS)- Advanced Computational Syntax (2 SWS)- Philosophy of language (2 SWS)- Advanced Semantics (2 SWS)- Lexical Semantics (2 SWS)- Advanced Computational Semantics (2 SWS)
- Speech recognition (2 SWS)- Speech synthesis (2 SWS)- Experimental phonetics (2 SWS)- Laboratory Phonology (2 SWS)- Language and Speech in the Human Brain: Advanced methods inNeurolinguistics and Neurophonetics (2 SWS)- Brain Computer Interfaces (2 SWS)- Advanced Speech Perception (2 SWS)- Advanced Speech Production (2 SWS)
59
- Advanced Statistical Natural Language Processing (2 SWS)- Statistical language models and smoothing (2 SWS)- Statistical NLP applications (2 SWS)- Statistical constituent parsing (2 SWS)- Statistical machine translation (2 SWS)- Advanced information retrieval (2 SWS)- Machine learning for NLP (2 SWS)- Distributional and statistical approaches to semantics (2 SWS)- Unsupervised and semisupervised learning (2 SWS)- Evaluation and statistical testing (2 SWS)- Probabilistic models of language and cognition (2 SWS)
9 Dozenten Professoren des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Spezialisierungsmodul; Pflicht; 3. Seme-ster
11 Voraussetzungen Advanced knowledge in a subarea of Computational Linguistics12 Lernziele Students are able to perform an independent survey of relevant literature
(with guidance from an advisor); they practice the skills needed for iden-tifying a research topic and writing a project proposal; they gather experi-ence presenting ongoing scientific work and dealing with feedback
13 Inhalt (a) 3rd semester: Formulation of a Master thesis project plan, in consul-tation with the examiner of the thesis; (b) 4th semester: participation ina colloquium series and presentation of ongoing work on Master thesis.In order to facilitate a stay abroad during the 3rd semester, the rese-arch seminar (part (a)) can be arranged as follows: the candidate andthe advisor determine the thematic area for the independent survey andhave preparatory meetings before the stay abroad, taking into accountthe specialization of the hosting university. During the course of the 3rdsemester, the candidate and advisor communicate about progress on theproject plan at regular intervals. After the return from abroad the projectplan is finalized under the supervision of the advisor.
14 Literatur/Lernmaterialien Conference Proceedings of Association for Computational Linguistics andother international conferences.
15 Lehrveranstaltungen undLehrformen (Englisch)
(a) research seminar (2 SWS) and/or directed reading and planning (3rdsemester) and (b) colloquium (2 SWS), with a presentation (4th semester)
9 Dozenten Dozenten des IMS10 Verwendbarkeit/Zuordnung
zum CurriculumMSc Computational Linguistics; Masterarbeit; Pflicht; 4. Semester
11 Voraussetzungen The student must have successfully passed modules from the MSc programcomprising 60 LP; the Master thesis research plan from the research modulemust have been approved by the thesis examiner. If admission to the Masterprogram was conditional on the completion of certain additional modulesor the proof of certain skills, these conditions have to be satisfied at thepoint of registering the master thesis.
12 Lernziele The Master thesis shows that the student is able to independently completea defined research task in Computational Linguistics within a fixed period,following scientific methodology, and to present the results in an adequateway.
13 Inhalt The content depends on the thesis topic, which is set by an examiner fromthe area of Computational Linguistics, taking into account the Master thesisresearch plan developed by the student.
14 Literatur/Lernmaterialien Stiebels/Pinkal (2006): Merkblatt fur die Anfertigung von Seminararbeiten.Ms. ZAS Berlin, Uni Potsdam.Academic Writing. Ms. Dublin City University.http://www.computing.dcu.ie/˜john/writing guide.html
15 Lehrveranstaltungen undLehrformen (Deutsch)
Selbstandige Arbeit nach Aufgabenstellung
Lehrveranstaltungen undLehrformen (Englisch)
Independent work following the assignment of tasks by the examiner.
