N 55° 39' 33.588''
E  12° 35' 26.771''
H
H
C
H
O
A
S
H
E
E
I
O
L
G
A
O
J
L
A
I
Á
O
L
L
!
L
O
!
!
E
U
Á
O
!
I
T
!
!
E
!

NLPnorth Group Photo

Figure 1: The NLPnorth research unit in the Data Science section of the IT University of Copenhagen is a team of faculty, postdocs, PhD students, assistants and students who work together on fundamental and applied problems for computational understanding of human languages and speech.

Members

Full-time Faculty

Postdocs

PhD Fellows

External Affiliated Members

Visiting Scholars

PhD Graduates

Past Visiting Scholars

Other Alumni

Publications

Our most recent research outputs, with more over ITU's PURE.

MORE

Resources

Awesome Neural Adaptation

Survey of awesome work on neural unsupervised domain adaptation.
[repo] [paper]

BeeSL

Biomedical event extraction as sequence labelling.
[tool] [paper]

BiLSTM-Aux

BiLSTM sequence tagger in DyNet with multi-task learning / CRF / lexicon + contextual embedding support.
[tool] [paper]

CiteTracked

A longitudinal dataset of peer reviews and citations.
[paper]

DaCoref

Danish Coreference Dataset.
[data]

DaN+ Corpus

Danish Nested Named Entities and Lexical Normalization.
[data] [paper]

DaNewsroom

A Large-scale Danish Summarisation Dataset.
[data] [paper]

Danish NER

Evaluation Dataset for Danish Named Entity Recognition.
[data] [paper]

DDT

Danish Universal Dependencies Treebank.
[data] [paper]

DeepSig

Significance testing implementations, including Almost Stochastic Order, bootstrap and permutation.
[tool]

EMNLP 2021 PhD Course

Communicating state-of-the-art NLP research to a broader audience.
[blogs] [course]

EWT-NNER

Nested Named Entities for the English Web Treebank.
[data] [paper]

UD Frisian FAME

Dependency annotations for a subset of the Frisian FAME! speech corpus.
[data] [paper]

FTSpeech

A Danish speech corpus created from meetings of the Danish Parliament.
[data] [paper]

JobStack

Annotations for entities in job postings.
[data] [paper]

MaChAmp

Multi-task learning toolkit for sequence labeling, classification, generation, masked LM, dependency parsing and more.
[tool] [paper] [demo]

UniParse

Universal graph-based modular parsing framework for quick prototyping and comparison of parsers and parser components.
[tool] [paper]

xSID

Multilingual slot and intent detection.
[tool] [paper]

Jobs

All our funded positions will be advertised on the ITU website. There are also other oppurtunities by applying for your own project hosted by NLPnorth. More information can be found on our opportunities page.

PhD Life at NLPnorth

What would it be like to do a PhD in Denmark and ITU? Here's some experiences shared by NLPnorth PhD students.

Reach Out

ITU building

Corridor 3E, Rued Langgaards Vej 7, 2300 København S

ITU people

Contact: Head of research group Stefan Heinrich (stehe@itu.dk); our individual e-mails can be found on our personal webpages