Taking into account semantic similarities in correspondence analysis
Mattia Egloff, François Bavaud · SERVAL (Université de Lausanne) · 2018
Term-document matrices feed most distributional approaches to quantitative textual studies, without consideration for the semantic similarities between terms, whose presence arguably reduce the content variety. This contribution presents a formalism remedying this omission, and makes an explicit use of the semantic similarities as extracted from WordNet. A case study in similarity-reduced correspondence analysis illustrates the proposal.