Selected work

Representative Papers

The Agentic Automation Canvas: a structured framework for agentic AI project design

Feb 16, 2026

This work addresses the lack of forward-looking, structured design and governance methodologies in current agentic AI systems, where existing documentation is often retrospective, non-machine-readable, and lacking interoperability. To bridge this gap, the paper proposes the Agentic Automation Canvas (AAC)—a structured framework encompassing six dimensions: scope definition, user expectations, development feasibility, governance stages, data sensitivity, and outcomes. Integrating semantic web technologies—including controlled vocabularies, Schema.org, and the DCAT ontology—and aligned with FAIR principles, AAC enables the generation of machine-readable, versioned, and privacy-aware project charters. Implemented as an open-source web tool, it supports export to RO-Crate format. Empirical validation across research, clinical, and institutional settings demonstrates that AAC significantly enhances design transparency and cross-role collaboration in autonomous AI systems.

0 citationsRead paper

OxO2 -- A SSSOM mapping browser for logically sound crosswalks

Jun 04, 2025

To address the fidelity loss, provenance gaps, and service instability of the original OxO ontology mapping system, we designed and implemented OxO2. Methodologically, we first deeply integrated the SSSOM standard into the ontology mapping browser; proposed sound mapping generation conditions and verification mechanisms based on Nemo Datalog; and established a provenance-driven, trustworthy mapping management paradigm—unifying provenance modeling, incremental reasoning, and memory-aware optimization. Contributions include: achieving zero timeouts and zero crashes across all requests, markedly enhancing service robustness; and enabling auditable, reproducible, and high-confidence cross-ontology mapping discovery and browsing. OxO2 thus provides a logically consistent, source-traceable infrastructure for integrating heterogeneous biomedical data.

0 citationsRead paper

OLS4: A new Ontology Lookup Service for a growing interdisciplinary knowledge ecosystem

Jan 22, 2025

Existing ontology lookup services (e.g., BioPortal, OntoBee) struggle to cope with the exponential growth of biomedical knowledge and increasing ontology complexity, exhibiting weak multilingual support, poor user experience, and low retrieval efficiency—particularly in emerging domains such as spatial transcriptomics. To address these limitations, we present OLS4, a next-generation ontology lookup service. OLS4 is the first to fully implement the OWL 2 specification; it integrates internationalization for multilingual support, an enhanced UX interface, and direct connectivity to external databases. Built on Spring Boot and Elasticsearch, it employs OWL API 4.x for ontology parsing and delivers a backward-compatible RESTful API. Deployed operationally at EMBL-EBI, OLS4 has completely replaced OLS3, serving global users with zero-API migration cost and achieving a 40% average improvement in response latency.

0 citationsRead paper

Recent publications

Latest Papers

The Agentic Automation Canvas: a structured framework for agentic AI project design

Feb 16, 2026

This work addresses the lack of forward-looking, structured design and governance methodologies in current agentic AI systems, where existing documentation is often retrospective, non-machine-readable, and lacking interoperability. To bridge this gap, the paper proposes the Agentic Automation Canvas (AAC)—a structured framework encompassing six dimensions: scope definition, user expectations, development feasibility, governance stages, data sensitivity, and outcomes. Integrating semantic web technologies—including controlled vocabularies, Schema.org, and the DCAT ontology—and aligned with FAIR principles, AAC enables the generation of machine-readable, versioned, and privacy-aware project charters. Implemented as an open-source web tool, it supports export to RO-Crate format. Empirical validation across research, clinical, and institutional settings demonstrates that AAC significantly enhances design transparency and cross-role collaboration in autonomous AI systems.

0 citationsRead paper

OxO2 -- A SSSOM mapping browser for logically sound crosswalks

Jun 04, 2025

To address the fidelity loss, provenance gaps, and service instability of the original OxO ontology mapping system, we designed and implemented OxO2. Methodologically, we first deeply integrated the SSSOM standard into the ontology mapping browser; proposed sound mapping generation conditions and verification mechanisms based on Nemo Datalog; and established a provenance-driven, trustworthy mapping management paradigm—unifying provenance modeling, incremental reasoning, and memory-aware optimization. Contributions include: achieving zero timeouts and zero crashes across all requests, markedly enhancing service robustness; and enabling auditable, reproducible, and high-confidence cross-ontology mapping discovery and browsing. OxO2 thus provides a logically consistent, source-traceable infrastructure for integrating heterogeneous biomedical data.

0 citationsRead paper

OLS4: A new Ontology Lookup Service for a growing interdisciplinary knowledge ecosystem

Jan 22, 2025

Existing ontology lookup services (e.g., BioPortal, OntoBee) struggle to cope with the exponential growth of biomedical knowledge and increasing ontology complexity, exhibiting weak multilingual support, poor user experience, and low retrieval efficiency—particularly in emerging domains such as spatial transcriptomics. To address these limitations, we present OLS4, a next-generation ontology lookup service. OLS4 is the first to fully implement the OWL 2 specification; it integrates internationalization for multilingual support, an enhanced UX interface, and direct connectivity to external databases. Built on Spring Boot and Elasticsearch, it employs OWL API 4.x for ontology parsing and delivers a backward-compatible RESTful API. Deployed operationally at EMBL-EBI, OLS4 has completely replaced OLS3, serving global users with zero-API migration cost and achieving a 40% average improvement in response latency.

0 citationsRead paper