SemAxis: A Lightweight Framework to Characterize Domain-Specific Word
Semantics Beyond Sentiment
J. An, H. Kwak, and Y. Ahn. (2018)cite arxiv:1806.05521Comment: Accepted in ACL 2018 as a full paper.
Abstract
Because word semantics can substantially change across communities and
contexts, capturing domain-specific word semantics is an important challenge.
Here, we propose SEMAXIS, a simple yet powerful framework to characterize word
semantics using many semantic axes in word- vector spaces beyond sentiment. We
demonstrate that SEMAXIS can capture nuanced semantic representations in
multiple online communities. We also show that, when the sentiment axis is
examined, SEMAXIS outperforms the state-of-the-art approaches in building
domain-specific sentiment lexicons.
Description
SemAxis: A Lightweight Framework to Characterize Domain-Specific Word
Semantics Beyond Sentiment
%0 Generic
%1 an2018semaxis
%A An, Jisun
%A Kwak, Haewoon
%A Ahn, Yong-Yeol
%D 2018
%K domain embeddings semaxis specific word
%T SemAxis: A Lightweight Framework to Characterize Domain-Specific Word
Semantics Beyond Sentiment
%U http://arxiv.org/abs/1806.05521
%X Because word semantics can substantially change across communities and
contexts, capturing domain-specific word semantics is an important challenge.
Here, we propose SEMAXIS, a simple yet powerful framework to characterize word
semantics using many semantic axes in word- vector spaces beyond sentiment. We
demonstrate that SEMAXIS can capture nuanced semantic representations in
multiple online communities. We also show that, when the sentiment axis is
examined, SEMAXIS outperforms the state-of-the-art approaches in building
domain-specific sentiment lexicons.
@misc{an2018semaxis,
abstract = {Because word semantics can substantially change across communities and
contexts, capturing domain-specific word semantics is an important challenge.
Here, we propose SEMAXIS, a simple yet powerful framework to characterize word
semantics using many semantic axes in word- vector spaces beyond sentiment. We
demonstrate that SEMAXIS can capture nuanced semantic representations in
multiple online communities. We also show that, when the sentiment axis is
examined, SEMAXIS outperforms the state-of-the-art approaches in building
domain-specific sentiment lexicons.},
added-at = {2018-06-28T09:48:55.000+0200},
author = {An, Jisun and Kwak, Haewoon and Ahn, Yong-Yeol},
biburl = {https://www.bibsonomy.org/bibtex/2818ee64cd0efa2ea71d94d0dff9b3a53/thoni},
description = {SemAxis: A Lightweight Framework to Characterize Domain-Specific Word
Semantics Beyond Sentiment},
interhash = {085b0598f0e7ad2502014fb5ddb2af15},
intrahash = {818ee64cd0efa2ea71d94d0dff9b3a53},
keywords = {domain embeddings semaxis specific word},
note = {cite arxiv:1806.05521Comment: Accepted in ACL 2018 as a full paper},
timestamp = {2018-06-28T09:48:55.000+0200},
title = {SemAxis: A Lightweight Framework to Characterize Domain-Specific Word
Semantics Beyond Sentiment},
url = {http://arxiv.org/abs/1806.05521},
year = 2018
}