Semantic Indexing of Open Scientific Literature to Help Users Discover and Navigate through Publications Networks

In recent years, several evolutions have drastically transformed the way researchers as well scientific and technical information (STI) services interact with literature. The amount pace of publications are skyrocketing, whether in journals conferences or through pre-publication repositories (e.g., arxiv.org), such that it is increasingly difficult to keep up, find make sense relevant articles. Furthermore, specialization research communities makes discover cross-disciplinary knowledge, which essential meet growing demand funding agencies for interdisciplinary projects. Scientific open archives central this landscape, however keyword-based search they usually provide fail grasp semantic relationships between Therefore, necessary develop new tools allow users their mass knowledge. talk, we wish present methods, implemented ISSA*2 project address these needs, discuss how could fit be deployed biodiversity area. Guided by science goals embracing FAIR*1 principles, aims to: a generic, transferable extensible pipeline analysis processing articles an archive; turn results into index stored published public RDF knowledge graph; innovative visualization leverage researchers, decision makers STI professionals explore thematic association rules, networks co-publications, co-occurring topics, etc. construction process involves artificial intelligence techniques: natural language processing, engineering Semantic Web. These techniques used publications’ metadata text automatically extract descriptors named entities. entities linked bases Wikidata, DBpedia GeoNames, domain-specific terminological resources suited archive's domain. third-party serves keystone support development rich aimed at and/or professionals. We demonstrated effectiveness solution use case Agritrop, institutional archive 110,000+ among 12,000 access articles, specialized fields agronomy, sustainable development. context, Agrovoc multilingual thesaurus was reference vocabulary. Fig. 1 illustrates concepts mentioned can visualize rules. example, mentioning COVID-19 food security (a) frequently mention concept pandemics (b). 2 shows other help health any its sub-concepts (a b), often co-mentioned climate change (c), get list related (d) time distribution (e). Being designed solution, delivered ISSA easily adapted Typically, references Darwin Core Terms, Access Biological Collection Data (ABCD), Digital Specimens (openDS), Audubon Metadata Schema various taxonomic registries, considered description article's linking From there, proposed reconfigured from answer competency questions, instance: what taxon child taxa? What museums/institutions more together certain groups? topics co-occur change, do evolve years? policies occur endangered species? extended including existing carry out e.g., extraction reconciliation authors' names.