Siqi Wu🏃🏻‍♂️

If you are interested in studying good old human activity on pre-Musk Twitter, we are now sharing the complete 375M dataset of our 24h Twitter data collection with researchers. Get in touch! Please forward to interested scholars.

How H|O|T is social media this week? 🌡️ A new measurement system devised by UMSI researchers displays the levels of hateful, offensive and toxic (H|O|T) comments from users across platforms about top news stories: myumi.ch/VGgzW @presnick @avalanchesiqi

Why do Instagram, TikTok and other platforms delete millions of posts every day? UMSI's Paul Resnick presents the issues of an unfiltered social media experience, clarifying why content may be moderated for users' benefit. youtube.com/watch?v=BTCrPY… @presnick @Teleperformance

@manoelribeiro @PrincetonCS @PrincetonCITP Congrats!

I'm thrilled to announce that I'll join @PrincetonCS/@PrincetonCITP as an assistant professor in Spring 2025 — can't wait to join this amazing group of researchers. I will continue to work on online platforms / the societal impact of AI. If you’d like to collaborate, reach out!

Also excited to run my first in-person Boston Marathon next Monday! If you see me run past with a Michigan hat, yell at me🦄〽️ @bostonmarathon

I will join Indiana University Bloomington Luddy School Department of Information and Library Science as a tenure track assistant professor this fall. Very excited about my new life chapter❗️🧲 @IULuddy @iuils

We call the new approach counterfactual bots. It needs two ingredients: 1: real user traces (here YT histories) 2: bots—computer programs simulating user behavior For each YT history, we instantiate various bots. One is a "control bot" that watches the same videos as the user.

Are you a fresh Masters degree graduate interested in #datascience for #socialgood? 👉Apply to the @ISI_Fondazione Lagrange #Fellowship to lead 1-year research project, collaborate with humanitarian and academic data scientists, and contribute to society! #research #jobs #italy

@m_starnini @MaxFalken @TheWebConf @gdfm7 @CentaiInstitute @Federico_Cinus @e_colacrai Great work, look forward to reading it

"Navigating Multidimensional Ideologies with Reddit's Political Compass: Economic Conflict and Social Affinity" accepted at @TheWebConf! arxiv.org/abs/2401.13656 Why do we need multiple dimensions to understand opinion dynamics and polarization?

I am recruiting PhD students! If you are interested in working on studying misinformation, hate speech and political polarization with a focus in non WEIRD contexts, please consider applying and mention my name. The deadline for our department @Rutgers is January 5.

@jaybaxter Oh I get what you meant, the mode != the majority of user polarity scores

@avalanchesiqi While the mode is ~0, the majority of raters are still away from 0 so f_u * f_n != 0

Really excited about this expansion. I’m curious to see how the latent viewpoint spaces of these new expansion areas compare to the existing ones. Any guesses?

@jaybaxter If the mode (f_u or f_n, in this case it does not matter) is 0, f_u * f_n will be 0. The partisanship part contributes nothing to the final prediction. All the variances will be explained by i_u or i_n. Am I missing anything here?

@avalanchesiqi I believe much of the variance is still explained by partisanship even when the mode is 0

@jaybaxter Most Asian countries do not have a bi-partisan divide, but ppl in those countries still disagree on many issues. Unimodal is interesting, if the mode is close to 0, this means all variance in the rating matrix can be explained by the general user friendliness or majority voting

@avalanchesiqi Which countries are you curious about? Certainly in the data so far you see many countries have bimodal rater factor distributions, but some less-polarized ones have unimodal distributions

[Please RT] I’m recruiting PhD students to work with me at @UW! I’m looking for students passionate about using computational methods to study how social platforms can be reimagined to enable better conversations, bridge political divides, and reduce the spread of misinfo. >>>

We are excited to launch ICWSM-Global Initiative, 2 pg. applications due 11/30 for: Fully funded trip to the conference in 2024 (up to $5K) Mentorship support from a senior academic in the field Details: #global_initiative" target="_blank" rel="nofollow noopener">icwsm.org/2024/index.htm… 1/2

Want to deep dive into our sock puppet experiment and more? Check out the full paper 📜here: arxiv.org/abs/2307.14551 (8/8)

🎆New Paper🎆 Ever wonder how to get rid of unwanted YouTube recommendations? My ICWSM paper with @avalanchesiqi and @presnick investigates the efficacy of several YouTube features to do just that. We also surveyed real users on whether they're aware of these features. (1/8)

We started an awesome-list for computational social science @gesis_org. Everyone can contribute and help to make this list truly awesome! github.com/gesiscss/aweso…

Accepted @icwsm! Re-hydrating Tweets is deprecated -> collaborate with us on raw data! arxiv.org/abs/2301.11429 w/ @feedkoko @onurvarol @janalasser @_kenny_joseph @Diyi_Yang @avalanchesiqi @fredmorstatter @Dede198904 @DanielMRomero @dgarcia_eu @c_schwemmer @JahnaO @CorBrantner