Mova0.5 pre-release⌕
☰
home / docs / papers / authors / Roberto Navigli

Roberto Navigli

5 paper cards in Mova's library · 2017–2024

Roberto Navigli wrote 5 of the papers Mova reads, in 2017–2024, most of them on Word sense disambiguation, Semantic role labeling, Machine translation; often with Simone Conia, Alessandro Raganato, Andrea Di Fabio.

OpenAlex · ORCID · h-index 61 · 19197 citations · 555 works — openalex, 2026-10-18

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

Elsewhere: Google Scholar · Semantic Scholar · dblp · ACL Anthology · ORCID

Topics: Word sense disambiguation 2 · Semantic role labeling 2 · Machine translation 1

Papers

  1. Beyond Correlation: Interpretable Evaluation of Machine Translation Metrics (2024)
    Correlation with human judgement says little about what a metric score means. The paper evaluates metrics in two proxy scenarios: as binary classifiers for data filtering (good against bad, and perfect against other, labelled from WMT23 MQM with cut-offs set by MQM error weights), and as re-rankers, where the…
  2. Exploring Non-Verbal Predicates in Semantic Role Labeling: Challenges and Opportunities (2023)
    Most SRL research and benchmarks assume that predicates are verbs. The authors test a state-of-the-art PropBank system (CN-22) trained on OntoNotes. Its argument F1 falls from 84.7 on verbs to 16.4 on nouns, and a noun-trained model falls from 72.8 to 11.2 on verbs, so it does not transfer knowledge across predicate…
  3. Generationary or «How We Went beyond Word Sense Inventories and Learned to Gloss» (2020)
    The paper replaces the choice from a sense inventory with generation of a definition for a word in its context.
  4. VerbAtlas: a Novel Large-Scale Verbal Semantic Resource and Its Application to Semantic Role Labeling (2019)
    VerbAtlas clusters all 13,767 WordNet verb synsets by hand into 466 frames (EAT, KILL, GIVE, …; about 30 synsets each), each with one prototypical argument structure over 25 cross-frame roles taken from VerbNet (14 rare VerbNet roles merged into coarser ones, e.g.
  5. Word Sense Disambiguation: A Unified Evaluation Framework and Empirical Comparison (2017)
    Results for all-words word sense disambiguation (WSD) are hard to compare, because the datasets, sense-inventory versions and preprocessing differ.

Co-authors (11)

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/authors/roberto-navigli/