Two Papers Accepted at ESWC 2026
EthereumKG: Building a Knowledge Graph of the Ethereum Blockchain
by Juan Cano de Benito, Andrea Cimmino Arriaga, Sven Hertling, Heiko Paulheim and Raúl García-Castro
This paper presents a method for transforming raw data from the Ethereum blockchain (block and smart contract data) into a knowledge graph. First, we extend current ontologies (Solidity ontology and EthOn ontology) by creating two new ontologies that model Ethereum blocks and the Application Binary Interfaces (ABI), covering all their properties. Based on these new and existing ontologies, we design and implement a method to generate a knowledge graph based on semantic web standards that integrates Ethereum block, ABI and Solidity data. The result is stored in a triple store that is organised into three interconnected knowledge graphs for blocks, ABIs, and Solidity contracts, containing over 27 million triples. Finally, to demonstrate the applications of the Ethereum knowledge graph, three scenarios are presented. In the first scenario, we explore the graph using SPARQL queries. In the second scenario, we detect similar behaviour wallets that may reflect a single entity. Finally, at the contract level, we use the ABI semantic data to group contracts by functional similarity and detect identical contracts.
Integrating Meta-Features with Knowledge Graph Embeddings for Meta-Learning
by Antonis Klironomos, Ioannis Dasoulas, Mohamed H. Gad-Elrab, Anastasia Dimou, Heiko Paulheim, Evgeny Kharlamov and Francesco Periti
The vast collection of machine learning records available on the web presents a significant opportunity for meta-learning, where past experiments are leveraged to improve performance. Two crucial meta-learning tasks are pipeline performance estimation (PPE), which predicts pipeline performance on target datasets, and dataset performance-based similarity estimation (DPSE), which identifies datasets with similar performance patterns. Existing approaches primarily rely on dataset meta-features (e.g., number of instances, class entropy, etc.) to represent datasets numerically and approximate these meta-learning tasks. However, these approaches often overlook the wealth of past experimental results and pipeline metadata available. This limits their ability to capture dataset – pipeline interactions that reveal performance similarity patterns. In this work, we propose KGmetaSP, a knowledge-graph-embeddings approach that leverages existing experiment data to capture these interactions and improve both PPE and DPSE. We represent datasets and pipelines within a unified knowledge graph (KG) and derive embeddings that support pipeline-agnostic meta-models for PPE and distance-based retrieval for DPSE. To validate our approach, we construct a large-scale benchmark comprising 144,177 OpenML experiments, enabling a rich cross-dataset evaluation. KGmetaSP enables accurate PPE using a single pipeline-agnostic meta-model and improves DPSE over baselines. The proposed KGmetaSP, KG, and benchmark are released, establishing a new reference point for meta-learning and demonstrating how consolidating open experiment data into a unified KG advances the field.