AI Paper Atlas

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

Modelling Semantic Categories Using Conceptual Neighborhood

Zied Bouraoui, Jose Camacho-Collados, Luis Espinosa-Anke, Steven Schockaert

AAAI 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Multilingual language representation pretraining · 10 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
Incorporate predicted conceptual neighbors into category representationsSECOND-WEA₄ vs. Gauss · BabelNet category inductionF1: 22.3 -> 54.2Improved
Select predicted conceptual neighbors rather than assuming all siblings are neighborsSECOND-WEA₄ vs. Multi · BabelNet category inductionF1: 44.2 -> 54.2Improved
Select siblings by predicted conceptual neighborhood rather than cosine similaritySECOND-WEA₄ vs. Similarity₄ · BabelNet category inductionF1: 38.2 -> 54.2Improved
Use four rather than one predicted neighbor (ablation)SECOND-WEA · BabelNet category inductionF1: 41.9 -> 54.2Improved
Use four rather than two predicted neighbors (ablation)SECOND-WEA · BabelNet category inductionF1: 49.3 -> 54.2Improved
Use four rather than three predicted neighbors (ablation)SECOND-WEA · BabelNet category inductionF1: 50.4 -> 54.2Improved
Use five rather than four predicted neighbors (ablation)SECOND-WEA · BabelNet category inductionF1: 54.2 -> 51.1Hurt
Use BERT rather than averaged word embeddings to identify neighbors when selecting one neighbor (ablation)SECOND · BabelNet category inductionF1: 41.9 -> 47.0Improved
Use averaged word embeddings rather than BERT representations for text classification (ablation)Text classifier · Text classifier cross-validationF1: 65.8 -> 69.0Improved
Use a text representation classifier rather than co-occurrence sentence countsText classifier · Text classifier cross-validationF1: 46.6 -> 69.0Improved
Use four rather than one predicted neighbor (ablation)SECOND-BERT · BabelNet category inductionF1: 47.0 -> 53.3Improved
Use five rather than four predicted neighbors (ablation)SECOND-BERT · BabelNet category inductionF1: 53.3 -> 51.3Hurt

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