What papers do I need to read about measuring emotion in tweets. Fear, anger, shame etc., not valence (positive, negative, neutral.). #rstats #twitter #nlp #python #nltk.

Interested in learning about AI & public affairs? Ethics for using data in policy making? Then UCLA's Data Analytics in Public Affairs certificate is for you. Ioin our online information session Wednesday July 23 at 12 p.m. PST. Register here:

Generate videos in just a few seconds. Try Grok Imagine, free for a limited time.

The paper demonstrates applications to digital censorship, event detection, and national elections. The data are available here: explore, and enjoy!.

See new user registrations since Mar. '06; # of users, tweets, and tweets by language starting in Sept. '13; & images shared + still available. Covers 246 countries & territories, 81 languages.

Curious about how Twitter's popularity has changed over time and across countries? Wish you had your own data with which to study Twitter? The paper analyzes 4 datasets of metadata created from my corpus of 10 yrs. of geolocated tweets.

Check out my new paper, "THIS Was Twitter: Introducing the Twitter History and Image Sharing v1.0 Datasets" and its data:

May 13th 8 p.m. PST, come to an online info session to learn more. Registration here:

In defense of government services:

Government is good:

I'm a bit behind but everyone who cares about transparent government and research should bookmark this project:

Rejected from NSF MMM with good/fair, good, very good, good, very good/good, excellent, excellent, and fair ratings. How do I interpret that?.

RT @AmandaMurdie: Oh, boy, friends, I've got GREAT NEWS to share. Dr. @IdeanSalehyan - likely the world's leading peace science scholar stu….

RT @polanalysis: Now in FirstView: “Improving Computer Vision Interpretability: Transparent Two-Level Classification for Complex Scenes.” @….

Happy to see this formatted for print, a nice offset to several rejections last week. Don't forget to check out the API!.

RT @CUP_PoliSci: #OpenAccess from @polanalysis -. Improving Computer Vision Interpretability: Transparent Two-Level Classification for Comp….

My favorite part of the holiday season is that, with the holiday break looming, all the rejections come in at once.

RT @davidhogg111: It’s really not complicated. The answer to the housing crisis is to build housing.

Is there an easy source of data on #nba players points per game or total per season? I want to teach regression using LeBron's trends in career points. is ok but I'd prefer not to pay, surely someone else has done this.

RT @matmagnani: #toread "Recommendations for sharing network data and materials", by @zpneal,@zack_almquist,@bagrow,@aaronclauset,@janadie….

👇👇Current CV transparency rests on interpreting pixels, not v. useful. We show how to use actual objects (hats, people, etc.) to understand CV classification. Read the whole thread, check out the paper, test the demo, and use the API.

Wow, super cool dataset: New dataset of daily global geolocated sentiment from 2019-2023 at admin-2 level for 164 countries.