Problem
Research questions and friction points this paper is trying to address.
Self-Organizing Maps
Database Abstraction
Topology-Driven Exploration
Queryable Data Map
Relational Data
Queryable Self-Organizing Maps: A Database Abstraction for Topology-Driven Data Exploration
🤖 AI Summary
Traditional self-organizing maps (SOMs) are ill-suited for direct use in topological exploration of relational data, as they typically operate outside database training and querying workflows. This work proposes a queryable data mapping abstraction that, for the first time, integrates SOMs into database systems as materialized topological constructs. The resulting topological artifact encapsulates representative prototypes, adjacency relationships, object assignments, and aggregate statistics, enabling topology-driven data exploration through standard SQL interfaces. A lightweight prototype system, MapDB, demonstrates that SOMs can be efficiently trained on moderate-scale datasets, support interactive queries once materialized, and yield topologically coherent regions that serve as meaningful targets for SQL-based exploratory analysis.
📝 Abstract
Self-Organizing Maps (SOMs) have long been used as exploratory tools for high-dimensional data: they organize objects into a two-dimensional topology that reveals clusters, gradients, sparse regions, dense regions, and boundaries. Yet, in modern data systems, SOMs are typically trained and visualized outside the DBMS, disconnected from the relational data they summarize. We introduce the abstraction of a queryable data map: a learned topological artifact consisting of representatives, neighborhood relations, object assignments, and derived summaries. We instantiate this idea with MapDB, a lightweight prototype that makes SOM artifacts queryable so users can explore data topology without leaving the database. Experimental study shows that SOM training is feasible at moderate analytical scale, that map queries are interactive after materialization, and that SOM regions provide meaningful targets for exploratory SQL.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Queryable Self-Organizing Maps
Database Abstraction
Topology-Driven Exploration
MapDB
Interactive Data Exploration
🔎 Similar Papers
No similar papers found.
