A Physical Metaphor to Study Semantic Drift

TL;DR

This paper introduces a novel approach using a physical metaphor of social mechanics to model and analyze semantic drift in digital preservation, leveraging large-scale machine learning and vector field analysis.

Contribution

It presents a new framework applying social mechanics metaphors to study semantic changes, integrating term importance and similarity in a dynamic model.

Findings

Semantic drifts can be modeled using a physical metaphor of gravitation.

Term importance influences semantic similarity over time.

The approach provides insights into knowledge dynamics in digital archives.

Abstract

In accessibility tests for digital preservation, over time we experience drifts of localized and labelled content in statistical models of evolving semantics represented as a vector field. This articulates the need to detect, measure, interpret and model outcomes of knowledge dynamics. To this end we employ a high-performance machine learning algorithm for the training of extremely large emergent self-organizing maps for exploratory data analysis. The working hypothesis we present here is that the dynamics of semantic drifts can be modeled on a relaxed version of Newtonian mechanics called social mechanics. By using term distances as a measure of semantic relatedness vs. their PageRank values indicating social importance and applied as variable `term mass', gravitation as a metaphor to express changes in the semantic content of a vector field lends a new perspective for experimentation. From `term gravitation' over time, one can compute its generating potential whose fluctuations manifest modifications in pairwise term similarity vs. social importance, thereby updating Osgood's semantic differential. The dataset examined is the public catalog metadata of Tate Galleries, London.