About
AMuSE-WSD is an All-in-one Multilingual System for Easy Word Sense Disambiguation (WSD).
Why AMuSE-WSD?
WSD is often considered a fundamental step towards Natural Language Understanding as it provides information about the meaning of a word in context. While current readily-available WSD systems are often based on graph-based pattern matching algorithms, Natural Language Processing (NLP) has recently seen tremendous growth, especially thanks to neural networks. Following this trend, AMuSE-WSD is the first all-in-one neural-based multilingual system for WSD which provides word-level semantic information in an end-to-end fashion. In AMuSE-WSD, users do not need to worry about data preprocessing (document splitting, tokenization, lemmatization, part-of-speech tagging), finding and implementing a state-of-the-art model, and data post-processing (recomposing a long text, choosing an inventory, etc.).
AMuSE-WSD is the perfect choice if you want to extract meanings from text or integrate semantic information in a downstream application!
How does AMuSE-WSD work?
Under the hood, AMuSE-WSD uses spaCy and/or Stanza NLP for transparent preprocessing. The core of AMuSE-WSD is, however, a recently proposed WSD model (Conia et al., 2021) which takes advantage of neural language models to achieve state-of-the-art results in multilingual WSD.
Read more details about how AMuSE-WSD works:
- Riccardo Orlando, Simone Conia, Fabrizio Brignone, Francesco Cecconi, and Roberto Navigli. AMuSE-WSD: An All-in-one Multilingual System for Easy Word Sense Disambiguation . In Proceedings of EMNLP 2021: System Demonstrations.
