Before managing engineers, I spent three years turning research into software.
I joined OEG as a student and later worked as a Research Assistant and Software Engineer, between semantic web research and tools another person could run.
RESEARCH QUESTIONS→OPEN-SOURCE IMPLEMENTATIONS→DEMOS & PAPERS
Research only becomes engineering when someone makes it run.
CSV data hides constraints and relationships in ordinary tables. Morph-CSV combines metadata and declarative mappings to prepare a richer representation for virtual knowledge graph queries. It became the implementation I spent the most time on at OEG.
Morph-CSV repository ↗Lower the interface cost without lowering the power.
Mapping languages are powerful, but specialised syntax limits who can author rules. Mapeathor explored a spreadsheet-based interface translated into declarative mappings. I co-authored its ISWC demonstration paper with the OEG team.
Mapeathor repository ↗Make relationships in scientific literature navigable.
The wider collaborative initiative used NLP and semantic technologies to connect drugs, diseases and publications. I worked on the search-engine side; its source was published through my GitHub account.
Search-engine repository ↗Four publications, one common thread.
Make complex data easier to use. These papers represent different collaborations across my OEG years.
Morph-CSV: Virtual Knowledge Graph Access for Tabular Data
Read paper ↗Mapeathor: Simplifying the Specification of Declarative Rules for Knowledge Graph Construction
Read paper ↗Drugs4Covid: Drug-driven Knowledge Exploitation based on Scientific Publications
Read paper ↗Understand the model, make assumptions explicit, and turn abstract ideas into something someone else can test and question.