Sense Cluster

Sense clustering aims to group different meanings (senses) of a word based on contextual usage, enabling analysis of semantic change over time and across languages. Current research focuses on improving clustering algorithms, often employing graph-based methods and incorporating multimodal data like gaze and scene information to enhance accuracy and contextual understanding. These advancements are crucial for various NLP tasks, including knowledge graph completion, human motion prediction, and diachronic word meaning analysis, ultimately leading to more robust and nuanced natural language processing systems.