Novel sparse matrix algorithm expands the feasible size of a self-organizing map of the knowledge indexed by a database of peer-reviewed medical literature

TL;DR

This paper introduces a new sparse matrix algorithm enabling the application of self-organizing maps to the entire Medline database, facilitating comprehensive and updatable mapping of medical knowledge.

Contribution

A novel sparse matrix multiplication algorithm that expands the feasible size of self-organizing maps for large biomedical datasets.

Findings

Mapped the entire Medline database using the new algorithm

Enhanced the ability to update the map over time

Demonstrated improved scalability of self-organizing maps

Abstract

Past efforts to map the Medline database have been limited to small subsets of the available data because of the exponentially increasing memory and processing demands of existing algorithms. We designed a novel algorithm for sparse matrix multiplication that allowed us to apply a self-organizing map to the entire Medline dataset, allowing for a more complete map of existing medical knowledge. The algorithm also increases the feasibility of refining the self-organizing map to account for changes in the dataset over time.