AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Short Text Representation for Detecting Churn in Microblogs

Hadi Amiri, Hal Daume III

AAAI 2016 · Label-efficient and robust learning › Recognition and long-tail · approach: Multilingual language representation pretraining · 30 citations (OpenAlex)

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 neighboring-word features to unigrams (ablation)BOW churn classifier (hinge) · Churn detection, macro-avgF1: 65.30 -> 73.63Improved
add dependency-path features to unigrams (ablation)BOW churn classifier (hinge) · Churn detection, macro-avgF1: 65.30 -> 72.40Improved
use RNN vector for dependency path instead of unigram+Depchurn classifier (hinge) · Churn detection, macro-avgF1: 72.40 -> 75.66Improved
add context (thread) indicators to unigram+Nb (ablation)BOW churn classifier (hinge) · Churn detection, macro-avgF1: 73.63 -> 74.27Improved
combine Nb, Dep and Cntx features (ablation)BOW churn classifier (hinge) · Churn detection, macro-avgF1: 73.63 -> 77.03Improved
combine all RNN features (twRep+NbRep+DepRep+CntxRep) over twRep+NbRep (ablation)RNN representation (logistic) · Churn detection, macro-avgF1: 73.90 -> 77.56Improved
combine BOW and RNN representations (ablation)churn classifier (hinge) · Churn detection, macro-avgF1: 77.03 -> 78.30Improved
concatenate tRep and wRep (twRep) instead of unigrams (ablation)tweet representation · Churn detection, macro-avgF1: 65.30 -> 66.13Improved
tRep alone instead of unigrams (128 dims) (ablation)tweet representation · Churn detection, macro-avgF1No clear effect
train RNN on balanced instead of imbalanced development data (ablation)twRep+NbRep (logistic) · Churn detection, macro-avgF1: 75.89 vs 73.90 (BOW-comparison cell); balanced twRep 68.27Mixed
use sentiment classifiers instead of churn-specific featureschurn classification · Churn detection, VerizonF1: 40.06 -> 65.17 (TargetSnt vs Unigram)Hurt
concatenate neighboring-word embeddings instead of averaging (ablation)NbRep · Churn detectionF1Improved

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