Natalia Vanetik

5 paper cards in Mova's library · 2022–2026

Natalia Vanetik wrote 5 of the papers Mova reads, in 2022–2026; often with Chaya Liebeskind, Marina Litvak, Gili Berger Hefetz.

The profile is the library's match of the name; two people with one name are not always told apart.

Papers

Annotation Matters: Resolving Cross-Corpus Performance Drops in Hebrew Offensive Language Detection (2026)
Hebrew offensive-language detectors trained on one forum corpus lose much of their accuracy on two external corpora, a Facebook set and a Twitter set.
Towards Safer Hebrew Communication: A Dataset for Offensive Language Detoxification (2025)
The paper builds HeDetox, 600 Hebrew sentence pairs in which an explicitly offensive forum comment is paired with a non-offensive rewrite.
From Linguistics to Practice: a Case Study of Offensive Language Taxonomy in Hebrew (2024)
The paper applies a Hebrew translation of the explicit-offence taxonomy of Lewandowska-Tomaszczyk et al. (2023) to Hebrew social-media text. Candidate offensive terms were found by co-occurrence (Dice coefficient) from 67 seed terms, and 683 were chosen by hand. Tweets were then extracted with those terms, and 450…
Detection of Negative Campaign in Israeli Municipal Elections (2022)
The paper introduces TONIC, a dataset of Hebrew Facebook posts by 27 mayoral candidates in 12 Israeli cities (2018 municipal elections).
Offensive language detection in Hebrew: can other languages help? (2022)
The authors build a binary offensive-language classifier for Hebrew, using a new set of 5217 annotated Facebook comments (two labels per comment, Kappa 0.82), and compare it with Arabic (9000 tweets) and English (OLID, 13 240 tweets).