AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Predicting and Analyzing Language Specificity in Social Media Posts

Yifan Gao, Yang Zhong, Daniel Preoţiuc-Pietro, Junyi Jessy Li

AAAI 2019 · Large language models › Language, safety and interpretability · approach: Controlled text generation

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add surface & lexical features (ablation)all-features SVR · Tweet specificity test setMAE: 0.3856 -> 0.3583Improved
add distributional word representations (ablation)all-features SVR · Tweet specificity test setMAE: 0.3965 -> 0.3583Improved
add emotion features (ablation)all-features SVR · Tweet specificity test setMAE: 0.3617 -> 0.3583Improved
add social media content features (URL, reply) (ablation)all-features SVR · Tweet specificity test setMAE: 0.3578 -> 0.3583No clear effect
add social media features (ablation)all-features SVR · Tweet specificity test setPearson correlation: 0.735 -> 0.734No clear effect
use trained feature model instead of average predictor (ablation)SVR with all features · Tweet specificity test setMAE: 0.5318 -> 0.3583Improved
use all features instead of length aloneSVR · Tweet specificity test setMAE: 0.4726 -> 0.3583Improved
use all features instead of Speciteller trained on newsSVR · Tweet specificity test setMAE: 1.0614 -> 0.3583Improved
use word representations alone vs best baseline (ablation)SVR · Tweet specificity test setMAE: 0.4726 -> 0.3885Improved
use social media content alone vs length baseline (ablation)SVR · Tweet specificity test setMAE: 0.4726 -> 0.5180Hurt
use emotion features alone vs average baseline (ablation)SVR · Tweet specificity test setMAE: 0.5318 -> 0.5324No clear effect
use all features vs word representations alone (ablation)SVR · Tweet specificity test setPearson correlation: 0.661 -> 0.734Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy