\n The GRAPH-SEARCH application was designed and developed by Francesco\n Invernici, Prof. Anna Bernasconi,\n and Prof. Stefano Ceri.\n
\n\n \n Deparment of Electronics, Information,\n and Bioengineering (DEIB)
\n Politecnico di Milano
\n Via Ponzio 34/5 Milano
\n 20133 Milano
\n Italy\n
\n The authors would like to acknowledge with thanks\n Luca Minotti for implementing the\n front-end of the GRAPH-SEARCH web application.\n
\n\n Our paper is in preparation. In the meanwhile, if you use GRAPH-SEARCH, please cite it as follows:
\n \n \n Francesco Invernici, Anna Bernasconi, and Stefano Ceri. 2023.
\n \"GRAPH-SEARCH: Searching COVID-19 clinical research using graph queries\".
\n \n http://gmql.eu/graph-search\n \n
\n {{ data.title }}\n
\n\n Searching for information over graphs is rather intuitive, as users can express queries in the form of graph\n patterns. We consider the use of \"graph queries\" (small graphs of concepts) as a means for expressing\n graph searches over existing biomedical literature, providing an interesting new application for exploring the\n supporting literature of given research findings.\n
\n\n We experiment our approach on the broad spectrum of COVID-19 research, which has many aspects yet to be\n explored. Extensive text material is available, especially in the COVID-19 Open Research Dataset (CORD-19), a corpus of more than one million\n scientific studies'\n abstracts created to accelerate the research against the disease.
\n To make CORD-19 searchable with graph patterns, we summarize its content as a graph of ontological terms (from\n UMLS terminology and CIDO ontology) that co-occur in the\n scientific abstracts.\n
\n GRAPH-SEARCH, a search engine for the exploration of scientific literature using the “graph query”\n paradigm. GRAPH-SEARCH outputs a list of ranked publications relevant for the user query.
\n GRAPH-SEARCH is built on top of state-of-the-art graph and data technologies: Neo4j, MariaDB,\n sciSpaCy, and NetworkX.\n