Ghent University Academic Bibliography

Advanced

Learning representations for tweets through word embeddings

Cedric De Boom UGent, Steven Van Canneyt UGent, Thomas Demeester UGent and Bart Dhoedt UGent (2016)
abstract
In event detection and opinion mining on social media it is important to grasp the semantic meaning of a text post. In this abstract, we present a method to learn effective representations for Twitter posts through a combination of word embeddings and word frequency information. We design a semantic similarity task between tweet couples and a novel loss function to train our model. We test it on a manually crafted dataset of tweets, and we find that our method outperforms the traditional baselines.
Please use this url to cite or link to this publication:
author
organization
year
type
conference
publication status
published
subject
keyword
IBCN, artificial intelligence, word embeddings, information retrieval, natural language processing, representation learning
pages
3 pages
conference name
Benelearn
conference location
Kortrijk, Belgium
conference start
2016-09-12
conference end
2016-09-13
language
English
UGent publication?
yes
classification
C1
copyright statement
I have retained and own the full copyright for this publication
id
8057704
handle
http://hdl.handle.net/1854/LU-8057704
date created
2016-08-30 16:55:43
date last changed
2017-03-02 13:15:10
@inproceedings{8057704,
  abstract     = {In event detection and opinion mining on social media it is important to grasp the semantic meaning of a text post. In this abstract, we present a method to learn effective representations for Twitter posts through a combination of word embeddings and word frequency information. We design a semantic similarity task between tweet couples and a novel loss function to train our model. We test it on a manually crafted dataset of tweets, and we find that our method outperforms the traditional baselines.},
  author       = {De Boom, Cedric and Van Canneyt, Steven and Demeester, Thomas and Dhoedt, Bart},
  keyword      = {IBCN,artificial intelligence,word embeddings,information retrieval,natural language processing,representation learning},
  language     = {eng},
  location     = {Kortrijk, Belgium},
  pages        = {3},
  title        = {Learning representations for tweets through word embeddings},
  year         = {2016},
}

Chicago
De Boom, Cedric, Steven Van Canneyt, Thomas Demeester, and Bart Dhoedt. 2016. “Learning Representations for Tweets Through Word Embeddings.” In .
APA
De Boom, C., Van Canneyt, S., Demeester, T., & Dhoedt, B. (2016). Learning representations for tweets through word embeddings. Presented at the Benelearn.
Vancouver
1.
De Boom C, Van Canneyt S, Demeester T, Dhoedt B. Learning representations for tweets through word embeddings. 2016.
MLA
De Boom, Cedric, Steven Van Canneyt, Thomas Demeester, et al. “Learning Representations for Tweets Through Word Embeddings.” 2016. Print.