Semantic structure from correspondence analysis

Abstract

A common problem for clustering techniques is that clusters overlap, which makes graphing the statistical structure in the data difficult. A related problem is that we often want to see the distribution of factors (variables) as well as classes (objects). Correspondence Analysis (CA) offers a solution to both these problems. The structure that CA discovers may be an important step in representing similarity. We have performed an analysis for Italian verbs and nouns, and confirmed that similar structures are found for English.