Entity Recognition Using Contextual Embeddings
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F19%3A00332765" target="_blank" >RIV/68407700:21230/19:00332765 - isvavai.cz</a>
Alternative codes found
RIV/68407700:21730/19:00332765
Result on the web
—
DOI - Digital Object Identifier
—
Alternative languages
Result language
angličtina
Original language name
Entity Recognition Using Contextual Embeddings
Original language description
In this paper, we present a Named entity recognition sequence labeling task using contextual embeddings such as ELMO or BERT. We compare the result using traditional BiLSTM or BiLSTM-CRF models using word embeddings with the approaches taking advantage of contextual embeddings. These embeddings are trained on large corpora which helps the model to understand the language even if the task-specific dataset is limited. Additionally, the contextual nature of the representation allows us to describe the same word with a different representation regarding the context. For that purpose, we test the models on a commonly used dataset CONLL 2003 and a relatively small in-house-labeled dataset of conversations between bot and a user.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
—
Continuities
S - Specificky vyzkum na vysokych skolach
Others
Publication year
2019
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Proceedings of the International Student Scientific Conference Poster – 23/2019
ISBN
978-80-01-06581-5
ISSN
—
e-ISSN
—
Number of pages
2
Pages from-to
183-184
Publisher name
ČVUT FEL, Středisko vědecko-technických informací
Place of publication
Praha
Event location
ČVUT FEL, Technická 2, Praha 6
Event date
May 23, 2019
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—