The fourth annual ELG conference



Together Towards Digital Language Equality

Brussels, Belgium
Hybrid conference

Berlin Skyline
© Adobe Stock – Sergij Figurnyi

Project Expo

Project Profile

Project abbreviation: USeA

Project name: Universal Semantic Annotator

Project coordinator: Roberto Navigli

Project consortium: Sapienza University of Rome

Funding: ELG (European Language Grid) Pilot Projects Open Call 2

Project duration: 1 year

Main key words: Natural Language Understanding, REST API, Word Sense Disambiguation, Semantic Role Labeling, Semantic Parsing, Multilinguality

Background of the research topic: Although machines are nowadays able to process massive amounts of text and information, current approaches to artificial intelligence are still not able to fully capture the underlying meaning structure. Developing techniques that capture and represent semantic information is, arguably, key to enabling Natural Language Understanding (NLU). Over the years, researchers have made encouraging steps towards achieving this goal, but the barriers to entry are still high as current tools require expert knowledge of advanced NLU topics.

Goal of the project: The goal of this project is to create the first unified API for three major tasks in Natural Language Understanding (NLU):

  • Word Sense Disambiguation (WSD)
  • Semantic Role Labeling (SRL)
  • Semantic Parsing (Abstract Meaning Representation, AMR)
We aim at offering a simple yet efficient way to use state-of-the-art multilingual models within a single framework accessible via REST API, browsers, and programmatically. This will ease the integration of NLU models in NLP pipelines (also for low-resource languages), allowing it to exploit the information obtained from one semantic task to improve the performance of another one. In addition, WSD, SRL, and AMR, when used jointly, can help the resolution of each other, e.g. WSD can help the disambiguation of predicates in SRL, and predicate-argument structures can be used in AMR parsing.


Project abstract: Universal Semantic Annotator (USeA) is the first unified API for high-performance Natural Language Understanding (NLU), with the aim of enabling easy integration of semantic knowledge into real-world applications. The API integrates state-of-the-art multilingual models developed recently at Sapienza NLP. With a simple HTTP request, they will allow to annotate texts with multiple semantic annotations. Specifically, we will release the APIs for the following models:

  • WSD: Simone Conia and Roberto Navigli. EACL 2021. Framing Word Sense Disambiguation as a multi-label problem for model-agnostic knowledge integration.
  • SRL: Simone Conia, Andrea Bacciu, Roberto Navigli. NAACL 2021. Unifying cross-lingual Semantic Role Labeling with heterogeneous linguistic resources.
  • AMR parsing: Michele Bevilacqua, Rexhina Blloshmi, Roberto Navigli. AAAI 2021. One SPRING to Rule Them Both: Symmetric AMR Semantic Parsing and Generation without a Complex Pipeline.



  • Riccardo Orlando, Simone Conia, Stefano Faralli, Roberto Navigli. 2022. Universal Semantic Annotator: the First Unified API for WSD, SRL and Semantic Parsing. In Proceedings of the 13th Language Resources and Evaluation Conference (LREC 2022). European Language Resources Association.
  • Riccardo Orlando, Simone Conia, Fabrizio Brignone, Francesco Cecconi, Roberto Navigli. 2021. AMuSE-WSD: An All-in-one Multilingual System for Easy Word Sense Disambiguation. In Proceedings of EMNLP 2021: System Demonstrations. Association for Computational Linguistics.
  • Simone Conia, Riccardo Orlando, Fabrizio Brignone, Francesco Cecconi, Roberto Navigli. 2021. InVeRo-XL: Making Cross-Lingual Semantic Role Labeling Accessible with Intelligible Verbs and Roles. In Proceedings of EMNLP 2021: System Demonstrations. Association for Computational Linguistics.
  • Rexhina Blloshimi, Michele Bevilacqua, Edoardo Fabiano, Valentina Caruso, Roberto Navigli. 2021. SPRING Goes Online: End-to-End AMR Parsing and Generation. In Proceedings of EMNLP 2021: System Demonstrations. Association for Computational Linguistics.