Analyzing Semantic Change through Lexical Replacements
This work addresses semantic change detection for natural language processing applications, representing an incremental advancement with a novel interpretable approach.
The paper tackles the problem of semantic change in language models by modeling it through lexical replacements, proposing a replacement schema and an interpretable model for semantic change detection. It also evaluates LLaMa for this task, achieving competitive results on benchmark datasets.
Modern language models are capable of contextualizing words based on their surrounding context. However, this capability is often compromised due to semantic change that leads to words being used in new, unexpected contexts not encountered during pre-training. In this paper, we model \textit{semantic change} by studying the effect of unexpected contexts introduced by \textit{lexical replacements}. We propose a \textit{replacement schema} where a target word is substituted with lexical replacements of varying relatedness, thus simulating different kinds of semantic change. Furthermore, we leverage the replacement schema as a basis for a novel \textit{interpretable} model for semantic change. We are also the first to evaluate the use of LLaMa for semantic change detection.