English
 
User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
 
 
DownloadE-Mail
  Using Multi-Sense Vector Embeddings for Reverse Dictionaries

Hedderich, M. A., Yates, A., Klakow, D., & de Melo, G. (2019). Using Multi-Sense Vector Embeddings for Reverse Dictionaries. Retrieved from http://arxiv.org/abs/1904.01451.

Item is

Basic

show hide
Item Permalink: http://hdl.handle.net/21.11116/0000-0004-02B4-E Version Permalink: http://hdl.handle.net/21.11116/0000-0004-02B5-D
Genre: Paper

Files

show Files
hide Files
:
arXiv:1904.01451.pdf (Preprint), 309KB
Name:
arXiv:1904.01451.pdf
Description:
File downloaded from arXiv at 2019-07-10 10:53 Accepted as long paper at the 13th International Conference on Computational Semantics (IWCS 2019)
Visibility:
Public
MIME-Type / Checksum:
application/pdf / [MD5]
Technical Metadata:
Copyright Date:
-
Copyright Info:
-

Locators

show

Creators

show
hide
 Creators:
Hedderich, Michael A.1, Author
Yates, Andrew2, Author              
Klakow, Dietrich1, Author
de Melo, Gerard1, Author              
Affiliations:
1External Organizations, ou_persistent22              
2Databases and Information Systems, MPI for Informatics, Max Planck Society, ou_24018              

Content

show
hide
Free keywords: Computer Science, Computation and Language, cs.CL,Computer Science, Learning, cs.LG
 Abstract: Popular word embedding methods such as word2vec and GloVe assign a single vector representation to each word, even if a word has multiple distinct meanings. Multi-sense embeddings instead provide different vectors for each sense of a word. However, they typically cannot serve as a drop-in replacement for conventional single-sense embeddings, because the correct sense vector needs to be selected for each word. In this work, we study the effect of multi-sense embeddings on the task of reverse dictionaries. We propose a technique to easily integrate them into an existing neural network architecture using an attention mechanism. Our experiments demonstrate that large improvements can be obtained when employing multi-sense embeddings both in the input sequence as well as for the target representation. An analysis of the sense distributions and of the learned attention is provided as well.

Details

show
hide
Language(s): eng - English
 Dates: 2019-04-022019
 Publication Status: Published online
 Pages: 12 p.
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Identifiers: arXiv: 1904.01451
URI: http://arxiv.org/abs/1904.01451
BibTex Citekey: Hedderich_arXiv1904.01451
 Degree: -

Event

show

Legal Case

show

Project information

show

Source

show