Mixing syntagmatic and paradigmatic information for concept detection

TL;DR

This paper improves concept detection in textual data by combining syntagmatic and paradigmatic information through a novel topic model based on word embeddings, leading to better performance and flexibility.

Contribution

It introduces a topic model that integrates paradigmatic relations via word embeddings, enhancing concept detection accuracy over traditional methods.

Findings

Significant increase in concept detection performance.

Enhanced flexibility in expressing target concepts.

Effective integration of syntagmatic and paradigmatic information.

Abstract

In the last decades, philosophers have begun using empirical data for conceptual analysis, but corpus-based conceptual analysis has so far failed to develop, in part because of the absence of reliable methods to automatically detect concepts in textual data. Previous attempts have shown that topic models can constitute efficient concept detection heuristics, but while they leverage the syntagmatic relations in a corpus, they fail to exploit paradigmatic relations, and thus probably fail to model concepts accurately. In this article, we show that using a topic model that models concepts on a space of word embeddings (Hu and Tsujii, 2016) can lead to significant increases in concept detection performance, as well as enable the target concept to be expressed in more flexible ways using word vectors.