TaBIIC2: Interactive Ontology Building with Weighted Self-Organizing Maps
A new tool, TaBIIC2, enables progressive and interactive construction of ontological taxonomies from tabular data. It uses weighted self-organizing maps for clustering, allowing users to identify concepts and their intentional definitions. The approach addresses challenges in manual taxonomy building and purely automatic methods like agglomerative clustering or large language models, which can overwhelm users with little control. The tool is described in a paper on arXiv (2605.24899).
Key facts
- TaBIIC2 is a tool for interactive taxonomy construction.
- It uses weighted self-organizing maps for clustering.
- The tool processes tabular data to identify concepts.
- It provides intentional definitions of clusters.
- The approach is progressive and user-controlled.
- It addresses limitations of manual and automatic methods.
- The paper is available on arXiv with ID 2605.24899.
- The tool aims to simplify ontology building from data.
Entities
Institutions
- arXiv