AI Paper Atlas

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

CEMA – Cost-Efficient Machine-Assisted Document Annotations

Guowen Yuan, Ben Kao, Tien-Hsuan Wu

AAAI 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Multilingual language representation pretraining · 2 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
Use CEMA's cost- and knowledge-aware document selection with paragraph filtering instead of random selectionMAA · DTJhuman cost to reach MA F1 at least 0.8: 2,884 -> 887Improved
Use CEMA's cost- and knowledge-aware document selection with paragraph filtering instead of random selectionMAA · GLhuman cost to reach MA F1 at least 0.8: 22,384 -> 12,037Improved
Select the shortest document rather than randomly selecting a documentMAA · DTJhuman cost to reach MA F1 at least 0.8: 2,884 -> 3,537Hurt
Select the shortest document rather than randomly selecting a documentMAA · GLhuman cost to reach MA F1 at least 0.8: 22,384 -> 40,413Hurt
Select documents using Least Confidence rather than randomlyMAA · DTJhuman cost to reach MA F1 at least 0.8: 2,884 -> 2,731Improved
Select documents using Least Confidence rather than randomlyMAA · GLhuman cost to reach MA F1 at least 0.8: 22,384 -> 22,961Hurt
Select documents using Dream rather than randomlyMAA · DTJhuman cost to reach MA F1 at least 0.8: 2,884 -> 4,899Hurt
Select documents using Dream rather than randomlyMAA · GLhuman cost to reach MA F1 at least 0.8: 22,384 -> 22,307No clear effect
Select documents using MNLP rather than randomlyMAA · DTJhuman cost to reach MA F1 at least 0.8: 2,884 -> 2,700Improved
Select documents using MNLP rather than randomlyMAA · GLhuman cost to reach MA F1 at least 0.8: 22,384 -> 26,423Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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