Language Modelling Makes Sense - WSD (and more) with Contextual Embeddings

Incorporating Dictionary Knowledge into Neural Word Sense Disambiguation(ACL 2018)

A Word Sense Disambiguation system integrating implicit and explicit external knowledge.

Disambiguate is a tool for training and using state of the art neural WSD models

A pipeline for transliteration, spell correction, POS tagging and word sense disambiguation of Hinglish code mixed data to Hindi Devanagari script.

Word Sense Induction with BERT MLM

Sequence Labeling in Tensorflow

Data and code for "Nibbling at the Hard Core of Word Sense Disambiguation" (ACL 2022).

Watasense: an Unsupervised WSD System for Under-Resourced Languages.

Repository containing code for the paper "Learning to Learn to Disambiguate: Meta-Learning for Few-Shot Word Sense Disambiguation", published at Findings of EMNLP.

NAACL'2021: Non-Parametric Few-Shot Learning for Word Sense Disambiguation

This project is a text summarization system that leverages word sense disambiguation and pre-trained T5 models. It identifies and summarizes repeating words with different meanings, using spaCy for NLP and an n-gram model for candidate word selection. It compares summaries using Rouge scores and outputs the best one. Written in Python with Flask.

Exploiting the global WordNet graph to perform WSD

WSD for Word-in-Context (WiC) disambiguation, experimenting with BERT feature-based and fine-tuning approaches (GlossBERT)

A 78.5% word sense disambiguator based on Transformers and RoBERTa (PyTorch)

[🏆 Azure Prize at TreeHacks] readAR -- 🌲 TreeHacks 2020 Backend

Russian Word Sense Induction system (for the RUSSE'18 shared task)

SIR is a sense-enhanced Information Retrieval system for multiple languages (EMNLP2021).

FUSE: Multi-Faceted Set Expansion by Coherent Clustering of Skip-grams (ECML-PKDD 2020)

A multi-labeling model for knowledge integration into Word Sense Disambiguation (EACL 2021).