Seostatavad

Seostatavad is a term used in Estonian information science and linguistics to describe data elements or concepts that can be linked to other elements through explicit relations. The form is derived from seos, meaning relation or link, with the participial suffix -atav, yielding a sense of “relational” or “capable of being connected.” In many projects the exact definition of seostatav depends on the chosen relation vocabulary, and it is not a universally standardized category.

In practical terms, seostatavad elements are nodes in a relational dataset or graph that can be connected via edges representing semantic similarity, co-occurrence, functional association, or other domain-specific relations. Their status is determined by the data model and the relation set defined by researchers or engineers, rather than by intrinsic properties of the item alone. This makes seostatav a flexible concept that can adapt to different domains and datasets.

Applications include knowledge graphs, recommender systems, and data integration, where identifying seostatavad items enables linking disparate datasets and inferring new connections. Methods to identify seostatavad elements include measuring semantic similarity, analyzing co-occurrence in text, or performing schema alignment across datasets. The concept supports modular data modeling, allowing scalable graph construction and efficient querying.

Example: in a retail catalog, products A and B may be seostatavad under an “often bought together” relation. In a linguistic corpus, the words “engine” and “car” might be seostatavad under a broader “domain of machinery” relation. See also: linked data, knowledge graph, semantic similarity, graph database, data integration.