Hauptregion der Seite anspringen

Freitagsseminar

Das Freitagsseminar findet Freitags 13:00 Uhr im Raum APB/3027 statt. Es dient der Präsentation und Diskussion forschungsrelevanter Themen durch Student:innen und Mitarbeiter:innen. Die Veranstaltung ist offen für alle Interessierten.

Zukünftige VorträgeKalender importieren (iCal)

Datum Beschreibung

Vergangene Vorträge

2022

Datum Beschreibung
08.09. 10:00 Uhr
Max Friedrich: Parsing discontinuous syntactical structures with neural models.

Verteidigung Diplomarbeit.

We present our idea for a new method of parsing discontinuous constituency trees and its current implementation progress. It consists of a new method of representing discontinuous constituency trees as matrices, an algorithm each for conversion from trees to matrices and vice versa and a neural network architecture that can produce those matrix tree represetations in a fully differentiable manner.

20.05. 11:00 Uhr
Max Friedrich: Parsing discontinuous syntactical structures with neural models.

Zwischenvortrag Diplomarbeit.

We present our idea for a new method of parsing discontinuous constituency trees and its current implementation progress. It consists of a new method of representing discontinuous constituency trees as matrices, an algorithm each for conversion from trees to matrices and vice versa and a neural network architecture that can produce those matrix tree represetations in a fully differentiable manner.

Der Vortrag findet online via BigBlueButton statt.

22.04. 13:00 Uhr
Jonas Knobloch: Implementierung von Paraphrasierung als Übersetzungsproblem.

Kolloquium Forschungsprojekt Grundlagen

2021

Datum Beschreibung
17.12. 13:00 Uhr
Jonas Knobloch: Implementierung von Paraphrasierung als Übersetzungsproblem.

Zwischenvortrag Forschungsprojekt Grundlagen

26.02. 10:00 Uhr
Celina Pohl: Normal forms of weighted context-free grammars.

Verteidigung Bachelorarbeit

22.01. 13:00 Uhr
Franziska von Albedyll: Dijkstra-Algorithmus und dessen Komplexität.

Zweite Prüfungsleistung AuD

2020

Datum Beschreibung
20.11. 13:00 Uhr
Max Friedrich: Parsing von diskontinuierlichen syntaktischen Strukturen mittels neuronaler Modelle.

Analyse eines Forschungsthemas

12.06. 12:00 Uhr
Alexander Jenke: Extraktion von nichtdeterministischen Streaming-String-Transducern aus einem bilingualen Korpus.

Abschlussvortrag Praktikum

13:30 Uhr
Richard Mörbitz & Thomas Ruprecht: Lexicalization of Probabilistic Linear Context-free Rewriting Systems.

Probevortrag

26.05. 09:00 Uhr
Alex Ivliev: Parsing of lexicalised linear context-free rewriting systems via supertagging.

Kolloquium Profilprojekt Grundlagenforschung

10:30 Uhr
Max Friedrich: Morphosyntactic tagging with a meta-BiLSTM model over context-sensitive token encodings.

Kolloquium Profilprojekt Grundlagenforschung

12:00 Uhr
Frederic Dörband: Sequentiality of Group-Weighted Tree Automata.
24.04. 13:00 Uhr
Richard Mörbitz und Thomas Ruprecht: Lexikalisierung von Linear context-free rewriting systems.
14.02. 13:30 Uhr
Alexander Jenke: Extraktion von nichtdeterministischen Streaming-String-Transducern aus einem bilingualen Korpus.

Zwischenvortrag Praktikum

14:00 Uhr
Alex Ivliev: Parsing of lexicalised linear context-free rewriting systems via supertagging.

Zwischenvortrag Forschungsprojekt

24.01. 13:00 Uhr
Johanness Hayeß: k-best Decoding of Conditional Random Fields in Sequence Analysis.

Bachelorverteidigung

2019

Datum Beschreibung
20.12. 13:00 Uhr
Max Fiedrich: Moprhosyntactic tigging with a meta-BiLSTM model over context-sensitive token encodings.

Zwischenvortrag Forschungsprojekt

29.11. 13:00 Uhr
Johannes Hayeß: k-best Decoding of Conditional Random Fields in Sequence Analysis.

Zwischenvortrag Bachelorarbeit

08.11. 13:00 Uhr
Tilo Werdin: Evaluation of coarse-to-fine recognition for weighted tree-stack automata.

Forschungsprojekt Grundlagen

06.09. 09:00 Uhr
Pius Friedrich Meinert: The Problem of Computing the Most Probable Tree of a Probabilistic Tree Automaton.

Kolloquium Masterarbeit

02.09. 09:00 Uhr
Marcel Legler: Implementierung eines gewichteten Lattice-based Parsers für Multiple Context-free Grammars.

Zwischenvortrag Forschungsprojekt Grundlagen

19.07. 13:00 Uhr
Tilo Werdin: Evaluation of coarse-to-fine recognition for weighted tree-stack automata.

Zwischenvortrag Forschungsprojekt Grundlagen

15.07. 13:00 Uhr
Kilian Gebhardt: Data-driven Parsing of Natural Languages with Mildly Context-sensitive Grammars.

Statusvortrag zum Promotionsvorhaben

Several languages (and in consequence also syntactic treebanks for these languages) such as Dutch (Lassy), German (NeGra, TiGer), but also English (Penn Treebank) contain so-called discontinuous structures that cannot be adequately modelled by context-free grammars. In consequence, a class of more power grammar formalisms called mildly context-sensitive has been studied. Although the parsing complexity of these grammar formalisms is polymomial, it is usally of higher degree than that of CFG. Hence, also other methods based on graph reductions and oracle-driven transition systems with more favourable complexity have been proposed.

In this talk I give an overview on different methods proposed for discontinuous parsing. Furthermore, I present my own contributions to the area of grammar-based discontinuous parsing. In particular, I recall linear context-free rewriting systems and hybrid grammars and present a method to learn powerful probabilistic extensions of these grammars. I discuss empirical results when applying said models to data from multiple languages.

05.07. 13:00 Uhr
Edgar Dorauasch: The smallest grammar problem.
13:30 Uhr
Lena Jurkschat: Parsing as reduction.
14:00 Uhr
Klaus-Rudolph Kladny: Weighting finite-state transductions with neural context.
14:30 Uhr
Alexander van Le: Coarse-to-fine parsing for expressive grammar formalisms.
15:00 Uhr
Tilo Werdin: An efficient best-trees algorithm for weighted tree automata over the tropical semiring.
24.05. 13:00 Uhr (45min)
Max Friedrich: Parsing of lexicalised linear context-free rewriting systems via supertagging.

Kolloquium zum Großen Beleg

17.05. 13:00 Uhr
Pius Friedrich Meinert: The Problem of Computing the Most Probable Tree of a Probabilistic Tree Automaton.

Zwischenvortrag Masterarbeit

10.05. 13:00 Uhr
Thomas Ruprecht: Implementation of a Chomsky-Schützenberger n-best parser for weighted multiple context-free grammars.
03.05. 13:00 Uhr
Heiko Vogler: Abstract families of weighted tree languages.
26.04. 13:00 Uhr
Franz Pöschel: Kompositionseigenschaften von Baumübersetzungssystemen.
13:45 Uhr
Tilo Werdin: Neuerungen an j-Algo (Bericht über WHK-Tätigkeit).
12.04. 13:30 Uhr
Philipp Langen: Advances in discontinuous latent variable parsing.

Bericht über die WHK-Tätigkeit

05.04. 13:00 Uhr (25min)
Max Friedrich: Parsing of lexicalised linear context-free rewriting systems via supertagging.
13:30 Uhr
Hauke Mewes: Cones for weighted tree languages.
15.03. 13:00 Uhr
Max Korn: Neuerungen an Vanda-Studio (Demonstration zur SHK-Tätigkeit).
01.02. 13:00 Uhr
Richard Mörbitz: Shortest distances in graphs and hypergraphs.
25.01. 13:00 Uhr
[notespaper]

Kilian Gebhardt: On the NP-hardness of finding the best tree generated by a probabilistic regular tree grammar.
11.01. 13:00 Uhr
Tobias Denkinger: CYK parsing on GPUs (after a paper by Yi, Lai, Petrov, and Keutzer, 2011).

2018

Datum Beschreibung
21.12. 13:00 Uhr
Olaf Gröscho: Conditional Random Fields für Part-of-Speech Tagging.
13:30 Uhr
Franz Pöschel: Kompositionseigenschaften von Baumübersetzungssystemen.
07.12. 12:30 Uhr
Pius Friedrich Meinert: Transitionsbasierte Generierung von Sätzen aus Abstract Meaning Representations.
30.11. 13:00 Uhr
[slides]

Thomas Ruprecht: Extraction of Dyck words from finite automata.
23.11. 13:00 Uhr
Richard Mörbitz: Dynamic programming is parsing (and vice versa).
16.11. 13:00 Uhr
Kilian Gebhardt: On the internals of disco-dop.

I give an overview of the discontinuous parsing framework disco-dop. In particular, I discuss which high-level and low-level optimizations were chosen in order to achieve state-of-the-art LCFRS parsing.

09.11. 13:00 Uhr
[slides]

Tobias Denkinger: Some notes on data structures used in search algorithms.

We talk about pushdowns, logarithmic probabilities, and double-ended priority queues (with example code in Rust).

02.11. 13:00 Uhr
[slides]

Thomas Ruprecht: Ansätze zum parallelen Suchen und Parsen.
26.10. 13:00 Uhr
[slides]

Richard Mörbitz: A comparison of parsing approaches for linear context-free rewriting systems.
19.10. 13:00 Uhr
[slides]

Kilian Gebhardt: Constraining of parse charts.
12.10. 13:00 Uhr
[script]

Tobias Denkinger: An introduction to Goldstine-automata.
28.09. 13:00 Uhr
Johanna Björklund: Tree-to-graph transducers.
27.09. 10:30 Uhr
Richard Mörbitz: M-monoid parsing problems </br> (Diplomverteidigung).
13:00 Uhr
Toni Dietze: A Formal View on Training of Weighted Tree Automata by Likelihood-Driven State Splitting and Merging
(Promotionsverteidigung im Ratszimmer APB/1004).

18.09. 14:00 Uhr
Marcel Legler: Evaluation of coarse-to-fine recognition for weighted tree-stack automata (Zwischenvortrag).
04.09. 09:00 Uhr
Toni Dietze: A Formal View on Training of Weighted Tree Automata by Likelihood-Driven State Splitting and Merging (Probevortrag).
27.07. 10:00 Uhr
Laura Kallmeyer: Formalizing Role and Reference Grammar as a mildly context-sensitive tree-rewriting grammar.

Role and Reference Grammar (RRG) is a theory of the grammar of natural language which has been developed as a descriptive tool for the analysis of typologically distinct languages, and which takes into account the interaction between syntax, semantics, and pragmatics. While RRG already contains a number of formal elements such as syntactic templates and the logical structures, it lacks a thorough formalization, which makes it difficult to study its formal properties or to develop a computational implementation.

In the talk, we present an outline of how a formal framework for RRG could look like. Our approach adopts and extends concepts from Lexicalized Tree Adjoining Grammars (LTAG) for the specification and composition of syntactic templates. We introduce two general operations, sister adjunction and wrapping substitution, for the composition of elementary syntactic trees. The resulting formalism, Tree Wrapping Grammar (TWG), is more powerful than LTAG. However, if the number of wrapping substitutions that can stretch across a node in the derived tree is limited to some constant k, it is possible to construct a weakly equivalent simple context-free tree grammars of rank k, which is in turn weakly equivalent to a well-nested linear context-free rewriting systems (LCFRS) of fan-out k + 1. Hence, with this linguistically plausible restriction, our formalization of RRG leads to a mildly context-sensitive grammar formalism.

We furthermore show how the operator projection of RRG, which reflects the scopal properties of functional operators, can be integrated into a single tree representation by adding appropriate node and edge features. This extends our formalism to a feature-based TWG.

The work presented in this talk is to a large extent joint work with Rainer Osswald.

12:30 Uhr (45min)
[slides]

Tobias Denkinger: Two characterisation results of multiple context-free grammars and their application to parsing (Statusvortrag).

We present two parsing algorithms for weighted multiple context-free grammars (short: wMCFGs). The first one is based on an automata characterisation of wMCFGs and a coarse-to-fine parsing algorithm for weighted automata with data storage. The second one is based on a Chomsky-Schützenberger characterisation of wMCFGs.

20.07. 13:00 Uhr
[slides]

Niklas Wünsche: Implementierung und Evaluierung eines inkrementellen Parsers in Vanda.
14:00 Uhr
[slides]

Andy Püschel: Learning to prune LCFRS parse charts
(Forschungsprojekt Grundlagen).

13.07. 13:00 Uhr
[paper]

Felix Völker: Incremental Discontinuous Phrase Structure Parsing with the GAP Transition
(Proseminar).

14:00 Uhr
Pius Friedrich Meinert: Training Deterministic Parsers with Non-Deterministic Oracles
(Seminar).

06.07. 13:00 Uhr
Tilman Hinnerichs: Minimalisierung gewichteter Baumautomaten.
29.06. 13:00 Uhr
Richard Mörbitz: M-monoid parsing problems.
13:30 Uhr
Felix Völker: Implementierung eines NeGra-Subkommandos für Vanda-Haskell.
22.06. 13:00 Uhr
[book]

Antonia Gerdes: Syntax and semantics of popular recurrent neural network architectures.
15.06. 13:00 Uhr
Tobias Denkinger: Automata & transducers with weights & data storage (based on a paper by Jonathan Goldstine).

We develop a definition of automata with data storage by stripping down the classic definition. We observe that automata with data storage can be extended to weighted transducers elegantly, using the free semiring.

14:00 Uhr
Heiko Vogler: Principle abstract families of weighted tree languages.
08.06. 13:00 Uhr
[slides]

Niklas Wünsche: Zwischenvortrag: Implementierung und Evaluierung eines inkrementellen Parsers in Vanda.
01.06. 13:00 Uhr
[slides]

Thomas Ruprecht: Chomsky-Schützenberger parsing is chart parsing.

We have a close look at the constructions of C-S parsing and find strong similarities to the well-established chart parsing approach.

18.05. 13:00 Uhr
[slides]

Anja Reusch: Chatbots zur Unterstützung des Spracherwerbs.
04.05. 13:15 Uhr
Johannes Osterholzer: New Results on Context-Free Tree Languages
(Promotionsverteidigung im Ratszimmer).

27.04. 13:00 Uhr
Tobias Denkinger: Practical problems with Chomsky-Schützenberger parsing for weighted multiple context-free grammars (in English).
20.04. 13:00 Uhr
Tobias Denkinger: Underlying Principles and Recurring Ideas of Formal Grammars (nach einem gleichnamigen Paper von Alexander Okhotin).
13.04. 13:00 Uhr
Johannes Osterholzer: New Results on Context-Free Tree Languages.
14:00 Uhr
[slides]

Pius Meinert: Parsing of natural language sentences to syntactic and semantic graph representations.
21.03. 14:00 Uhr
[slides]

Thomas Ruprecht: Implementation and evaluation of k-best Chomsky-Schützenberger parsing for weighted multiple context-free grammars.
13.03. 10:00 Uhr
[slides]

Richard Mörbitz: M-monoid parsing und Reduktbildung.
02.03. 12:00 Uhr
Andy Püschel: Learning to prune LCFRS parse charts.
02.02. 13:00 Uhr
Anja Reusch: Chatbots zur Unterstützung des Spracherwerbs.
26.01. 13:00 Uhr
[slides]

Christian Lewe: A brief view on Rust and rustomata.
19.01. 13:00 Uhr
[slides]

Till Köhler und Andy Püschel: Implementierung eines Part-Of-Speech-Taggers in Haskell.
12.01. 13:00 Uhr
[slides]

Thomas Ruprecht: Implementation and evaluation of k-best Chomsky-Schützenberger parsing for weighted multiple context-free grammars.
05.01. 13:00 Uhr
[slides]

Heiko Vogler: Attribute Grammars.

2017

Datum Beschreibung
15.12. 13:00 Uhr
[slides]

Richard Mörbitz: M-monoid parsing und Reduktbildung.
08.12. 13:00 Uhr
[slides]

Thomas Ruprecht: Parsing linear context-free rewriting systems in Vanda.
01.12. 13:00 Uhr
Pius Meinert: Parsing of natural language sentences to syntactic and semantic graph representations.
24.11. 13:00 Uhr
Kilian Gebhardt: Word2Vec – The skip-gram model.

I give a short introduction to word embeddings and, in particular, to the skip-gram model implemented in word2vec.

10.11. 13:00 Uhr
[slides]

Heiko Vogler: Dynamic programming algorithms for transition-based dependency parsing.
03.11. 13:00 Uhr
Kilian Gebhardt: The Inside-Outside Algorithm is just Backpropagation.

The inside-outside algorithm computes so-called inside weights and outside weights for a weighted (intersection-)grammar. I illustrate that each outside weight can also be obtained as a certain partial derivative of inside weights. This property can be exploited to obtain an EM algorithm from automatic differentiation present in tools such as dynet or Tensorflow.

27.10. 13:00 Uhr
[slides]

Max Korn: Coarse-to-fine recognition for weighted tree-stack automata.

We apply the idea of coarse-to-fine parsing to recognition with weighted automata with storage by introducing the concept, and implementing an algorithm in the framework rustomata.

20.10. 13:00 Uhr
[slides]

Thomas Ruprecht: Parsing linear context-free rewriting systems.

We use weighted deduction systems to describe three different types of parsers for weighted linear context-free rewriting systems. Furthermore, we introduce heuristic weights for these parsers to enable an informed best-first search among parsing results. We implemented all described formalisms in Haskell using the Vanda framework and tested the implementation with grammars extracted from a natural language corpus.

13.10. 13:00 Uhr
Tobias Denkinger: Chomsky-Schützenberger parsing for weighted multiple context-free languages.

We present an n-best parsing algorithm based on the Chomsky-Schützenberger theorem for weighted multiple context-free languages. The algorithm will be implemented and evaluated in the future.

Kontakt

Stand: 19.10.2020 15:25 Uhr