The emergence and global adoption of social media has rendered possible the
real-time estimation of population-scale sentiment, bearing profound
implications for our understanding of human behavior. Given the growing
assortment of sentiment measuring instruments, comparisons between them are
evidently required. Here, we perform detailed tests of 6 dictionary-based
methods applied to 4 different corpora, and briefly examine a further 8
methods. We show that a dictionary-based method will only perform both reliably
and meaningfully if (1) the dictionary covers a sufficiently large enough
portion of a given text's lexicon when weighted by word usage frequency; and
(2) words are scored on a continuous scale.
%0 Generic
%1 reagan2015benchmarking
%A Reagan, Andrew J.
%A Tivnan, Brian
%A Williams, Jake Ryland
%A Danforth, Christopher M.
%A Dodds, Peter Sheridan
%D 2015
%K analysis comparison kallimachos sentiment survey toread
%T Benchmarking sentiment analysis methods for large-scale texts: A case
for using continuum-scored words and word shift graphs
%U http://arxiv.org/abs/1512.00531
%X The emergence and global adoption of social media has rendered possible the
real-time estimation of population-scale sentiment, bearing profound
implications for our understanding of human behavior. Given the growing
assortment of sentiment measuring instruments, comparisons between them are
evidently required. Here, we perform detailed tests of 6 dictionary-based
methods applied to 4 different corpora, and briefly examine a further 8
methods. We show that a dictionary-based method will only perform both reliably
and meaningfully if (1) the dictionary covers a sufficiently large enough
portion of a given text's lexicon when weighted by word usage frequency; and
(2) words are scored on a continuous scale.
@misc{reagan2015benchmarking,
abstract = {The emergence and global adoption of social media has rendered possible the
real-time estimation of population-scale sentiment, bearing profound
implications for our understanding of human behavior. Given the growing
assortment of sentiment measuring instruments, comparisons between them are
evidently required. Here, we perform detailed tests of 6 dictionary-based
methods applied to 4 different corpora, and briefly examine a further 8
methods. We show that a dictionary-based method will only perform both reliably
and meaningfully if (1) the dictionary covers a sufficiently large enough
portion of a given text's lexicon when weighted by word usage frequency; and
(2) words are scored on a continuous scale.},
added-at = {2016-07-13T09:44:42.000+0200},
author = {Reagan, Andrew J. and Tivnan, Brian and Williams, Jake Ryland and Danforth, Christopher M. and Dodds, Peter Sheridan},
biburl = {https://www.bibsonomy.org/bibtex/2898e615821a92a1cf681ea87b83a7ad5/hotho},
interhash = {d0c87b7102d87210aa68cc28596a04b8},
intrahash = {898e615821a92a1cf681ea87b83a7ad5},
keywords = {analysis comparison kallimachos sentiment survey toread},
note = {cite arxiv:1512.00531Comment: 34 pages, 30 figures. Minor edits and upgrades},
timestamp = {2016-07-13T09:44:42.000+0200},
title = {Benchmarking sentiment analysis methods for large-scale texts: A case
for using continuum-scored words and word shift graphs},
url = {http://arxiv.org/abs/1512.00531},
year = 2015
}