Yaoyao Dai
@yaoyao_dai
Followers
876
Following
367
Media
9
Statuses
103
Assistant Professor in PoliSci @unccharlotte . Authoritarian politics, information manipulation, populism, ML, NLP.
Joined June 2018
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
#TolakPolitikDinasti
• 2588423 Tweets
FB JOIN OPENING BTV HOUSE
• 744569 Tweets
BATALKAN BUKAN TUNDA
• 565123 Tweets
Polisi
• 505877 Tweets
Mulyono
• 319697 Tweets
#NHK_MUSICEXPO
• 306542 Tweets
FOTFOT NEW SINGLE
• 286645 Tweets
RIIZE
• 166983 Tweets
Gus Walz
• 141270 Tweets
WayV
• 141093 Tweets
BOTO PARA SA SB19
• 99605 Tweets
Kompas
• 88433 Tweets
#GOATなINZM
• 60949 Tweets
チャイルドシート
• 43796 Tweets
Barron
• 38362 Tweets
Phantom
• 35135 Tweets
Pretender
• 21144 Tweets
ちいかわ寿司
• 19588 Tweets
Siren
• 15848 Tweets
ラストマイル
• 14336 Tweets
チッケム
• 11039 Tweets
ウォンビン
• 10212 Tweets
ボムギュ
• 10181 Tweets
Pinned Tweet
New Book! So excited to share my work with
@roseluqiu
on China's Wolf Warrior Diplomacy!
We trace the evolution of China's diplomacy since Mao and offer a large-scale analysis of China's diplomatic language and media reactions to WW diplomacy in the Global South.
33
113
514
Concluded 3-week Text Analysis course
@ICPSR
@ICPSRSummer
today with fine-tune LLMs! The most enjoyable teaching experience with extremely motivated bright and diverse students❤️ from 10+ countries across social sciences disciplines! Can't wait to see your projects using NLP!
2
5
41
Thanks to my awesome committee, I really enjoyed my grad school time! Now it’s time to move forward :)
Congratulations to Dr.
@yaoyao_dai
! As usual the dissertation committee members thoroughly enjoyed ourselves discussing her work. I love getting to work with wonderful scholars (and people) like Yaoyao - I mean, Dr. Dai.
1
0
27
7
1
39
Li Chen's comic is one of the few things keep me happy in the past two crazy years. So Ben commissioned a drawing for me featuring us and my favorite stuff including the buns ate my plants 😍 Thanks Li and Ben❤️ check out
@Exocomics
for the most heartwarming drawings!
2
0
29
Wow glad to see our paper with
@akoustov
on populism
@PSRMJournal
is trending. It is
#OpenAccess
at . We find NO effect of populist rhetoric alone on VOTE CHOICE. But populism might still be effective at attracting media attention & mobilizing voters.
Trending in
#PoliticalScience
:
1) (In)effectiveness of populist rhetoric: an experiment of campaign messaging (
@psrmjournal
)
2) Government’s biases against left-behind places (
@jepp_journal
)
3) “Rally around the flag” effects in the Russian–Ukrainian
0
22
71
1
6
27
It's great to see this conversation on measuring populism! Plz read both papers. They are relevant to measuring complex concepts in text and great examples of open science. I want to take the chance to talk about some differences between our paper w
@akoustov
and the PA paper.
In a new working paper (),
@Robert_A_Huber
and I discuss a recently proposed method for deriving populism scores from texts. We argue that a more comprehensive validation raises concerns about the validity of the suggested approach.
4
18
73
1
9
25
Happy to join this vigorous discussion on the myths about populism! A big thank you to
@PetraGuasti
and
@Zulianello_M
for organizing this series
@ECPR_TheLoop
. Populism might not be as effective in attracting votes as we think. Programmatic policies matter more:
🔮
#FutureOfPopulism
No.33
@yaoyao_dai
cautions readers that the electoral success of populists is not necessarily due to populism, and urges scholars to disentangle populism from host ideologies. Empirically, populism alone has limited effect.
0
4
7
2
8
22
Check it out if you are interested in China, diplomacy, and text analysis! 30% off discount with code "LXFANDF30" at
2
4
21
Political actors have been buying space in independent media outlets to publish advertisements that look like standard news stories. Find out more about this type of soft propaganda in this new article on
@Res_Pol
by
@roseluqiu
and me:
2
10
20
If you are interested in populism, populism in China and measurement using text... =)
Measuring Populism in Context: A Supervised Approach With Word Embedding Models
Yaoyao Dai (
@yaoyao_dai
)
0
1
4
0
4
19
Thanks to
@akoustov
for a great thread of our new paper
@polcommjournal
! Though the analysis focuses on the U.S., our theoretical model and measurement method are inspired by studies across multiple disciplines and should generalize well to cases outside of the U.S.
How do politicians decide whether to appear more populist? In our new
@polcommjournal
paper w.
@yaoyao_dai
, we show the initially less popular US presidential candidates are more likely to use populist rhetoric in a gamble to have some chance of winning
1
17
77
1
4
11
So sad. Becky was one of the kindest most honest and supportive people I ever met. She has helped so many young scholars around the world. I was lucky to have spent a year with her and got to know this wonderful person I aspire to be. I am going to miss her very much.
Polisci folks: with a heavy heart, I regret to inform you of the passing of Professor Rebecca Morton, Associate Dean of Social Science and Program Head of Political Science at NYU Abu Dhabi. She was a first-rate scholar, mentor, colleague, and friend. She will be deeply missed.
32
53
405
1
0
8
Wow! Check out this fantastic book trailer! Having read parts of it, it is the book I am most excited about this year! A must-read for anyone interested in migration, social identities, and politics of ethnicity and nationalism.
0
0
7
Had such a great panel with amazing people! Looking forward to experimenting with all the cool suggestions and to our next project on "Structured" Word Embedding
@ben_j_radford
Really cool work on training multilingual word embeddings from
@yaoyao_dai
. Python package coming soon!
#apsa2018
0
2
4
0
2
7
Sharing some findings from a new working paper with
@roseluqiu
. Wolf warrior diplomacy may allow Chinese diplomats to placate domestic nationalists and uphold Xi's demand for the fighting spirit while still leave room for China to control escalation.
1
3
7
@PJDunleavy
The paper finds a null effect of populist rhetoric alone in attracting votes or influencing voters' vote choice, as
@akoustov
said, not that populism is ineffective. It can be effective in mobilizing voters to vote, but we need more empirical studies to find out than assume.
2
0
7
@tldrpapers
summary for 2nd grader =) "Both Democratic and Republican US presidential candidates and incumbents use more populist rhetoric when they are less popular in public opinion polls."
Limited open access copy for full paper:
0
3
6
Can't survive this intense 3 weeks teaching a hybrid class of 129 students without my wonderful TAs
@caseyrandazzo
, Irene Morse, and Patrick Kaminski! They are more than capable of teaching their own text analysis class!
0
1
6
What to do with biased sources & missing data on conflict losses? Excited to share our new paper in
@PNASNews
: we construct an original data set on reported losses during
#RussiaUkraineWar
and develop a Bayesian model to estimate the losses & biases of sources.
@womenalsoknow
Out in
@PNASNews
: we are estimating conflict losses from cumulative and absolute fatality numbers reported in multiple, possibly biased reporting sources.
Joint work with
@yaoyao_dai
,
@niklas_stoehr
,
@AaronSchein
, Mya Fernandez and Hanif Sajid.
1
3
11
1
0
5
Congratulations! We are so lucky to have
@akoustov
joining us in the summer
@unccharlotte
!!!🎉🎉🎉🤩🤩🤩
Exciting news! This summer I'll be joining the University of North Carolina at Charlotte as an Assistant Professor of Political Science and Public Administration. Looking forward to continuing my work on migration and ethnic conflict with my great new colleagues
@unccharlotte
!
66
1
569
1
0
5
@PJDunleavy
@akoustov
Thanks Patrick. It's possible. That's why it is important to replicate experiments in different settings. Ours is the third experiment finding the small to null effects of populism alone on candidate choice in two different countries. We'd love to see more replications!
0
1
5
@akoustov
@PSRMJournal
Also, want to point out two other conjoint experiments with similar null effects of populist rhetoric in different contexts by
@FabianNeuner
@chriswratil
&
@b_castanho
: in Germany (…) and the US (…)
0
2
5
The deadline is Feb.4th! Still have time to apply! EITM 2018 was the best experience I had in grad school. It's like a 3 weeks summer camp with some of the kindest and smartest scholars. You DON'T need to have a formal model in your work to apply!
More information about the application process and the 2019 EITM Summer Institute at Emory can be found here:
0
0
1
0
3
4
@_shengche
@miket_32
@dasmiq
@arthur_spirling
@MatthewJDenny
@SKDreier24
@LauraK_Nelson
@GretaFranzini
@vanstriendaniel
I learned a lot from reading this thread. Adding another awesome paper on OCR here,
@ben_j_radford
and coauthors have this working paper on digitalizing COVID19 policy document with practical suggestions and novel evaluation methods:
0
0
4
We also use doc2vec to vectorize texts. Similar models can represent multiple languages in one shared vector space so we don't have to train a separate model for each language. It allows the transfer of knowledge from one language to another—just something for future research.
1
0
3
@b_castanho
@cmbosancianu
@littvay
Super cool! The cover says, "The most prestige and cutting-edge guide for quantitative methods." 🤩 It's a really good publisher in China.
1
0
3
@dtchimp
My working paper on measuring populism using party manifestos and news articles provides a further clarified conceptualization and coding criteria (plus a new computer-assisted method to PoliSci): . Hope it helps.
0
0
3
0
0
3
The blog post is largely based on my and
@akoustov
conjoint experiment. We embed populist and non-populist rhetoric with various hosting policies in campaign messages. More details:
0
0
3
Links to the three papers:
PA paper by
@JessicaDiCocco2
&
@BernardoMonechi
:
pre-print by
@michaelj505
&
@Robert_A_Huber
:
Our paper:
1
1
3
Because most docs are not populist, we also use active learning to populate the training data. For active learning see: by
@bapmiller
1
1
2
She made so many efforts to ensure a happy and collaborative intellectual environment for the postdocs in AD and made sure that we have the resources we need for our career. Hard to imagine a social science division without her.
0
0
2
1
0
2
0
0
1
@dtchimp
@team_populism
also has several versions of data and coding rules on measuring populism in political speeches: . They are going to release the latest version soon.
0
0
1
0
0
1
@jandrewharris
Some of the original papers on embeddings are pretty intuitive and I highly recommend: w2v: ; d2v: ; bert: . I also have a step by step explanation of word2vec in appendixB of
0
0
1