# KGAP: An RDF knowledge graph of agricultural commodity prices

## TL;DR

KGAP is a knowledge graph that integrates agricultural price data from Brazil into a standardized format for easier analysis and comparison.

## Contribution

KGAP introduces a harmonized RDF knowledge graph of agricultural commodity prices using FAIR data principles and semantic alignment.

## Key findings

- KGAP integrates data from Cepea, Conab, and Ipea into a unified RDF format.
- The graph supports semantically-aware queries and reveals insights obscured by data heterogeneity.
- KGAP is publicly accessible via a SPARQL endpoint and can be used for policy analysis and machine learning.

## Abstract

This article presents the Knowledge Graph for Agricultural Prices (KGAP), which is a knowledge graph (KG) that integrates agricultural commodity prices data from three major Brazilian institutions: Cepea, Conab, and Ipea. The datasets, originally published in heterogeneous formats, were harmonized and converted into RDF/Turtle using the Almes Core metadata schema as the data model. Agricultural products were classified with the Agricultural Product Types Ontology (APTO), and geographic references were aligned with GeoNames identifiers, ensuring semantic consistency and adherence to the FAIR data principles. KGAP is archived on Zenodo and GitHub, and hosted on the Platform Linked Data Nederland (PLDN) with a public SPARQL endpoint. It contains metadata, price observations, product types, and location entities, allowing users to query and compare agricultural prices across institutions, regions, and time periods. The knowledge graph can potentially support applications in agricultural economics, policy analysis, journalism, data science, and machine learning. By explicitly modeling metadata such as reference quantities, KGAP enables semantically-aware queries that prevent common analytical errors and reveal insights previously obscured by data heterogeneity.