kattumus

Kattumus is a term used in Estonian- and other scholarly contexts to denote the overlap, convergence, or alignment between two sets of data, features, or phenomena. It generally describes the extent to which elements in one domain correspond to elements in another, rather than indicating a causal relationship.

Usage varies by discipline. In linguistics and anthropology, kattumus can refer to the degree to which features recur across languages, communities, or cultural practices. In information science, statistics, and data analysis, it is often used more loosely to signal similarity or shared coverage between datasets, model outputs, or representations, without committing to a specific formal definition.

Measurement and interpretation are context dependent. A common formalization treats kattumus as a proportion of shared items relative to a defined baseline, such as the size of the intersection divided by the size of the union of two sets, analogous to the Jaccard similarity. Some researchers use alternative measures that emphasize different aspects of overlap, such as shared attributes or conditional agreement. The choice of measure depends on the data structure, the research question, and the practical needs of comparison.

Applications and examples. Kattumus appears in cross-domain studies where researchers compare inventories, feature sets, or outcome patterns across groups or models. It is also used in collaborative filtering, taxonomy alignment, and consensus formation to describe how much two sources agree. Limitations include sensitivity to sample size, the definition of the underlying sets, and the fact that overlap alone does not reveal the direction or strength of underlying relationships.

Related concepts include overlap, similarity measures, and the Jaccard index.