Smart impact-driven discovery service
This service aims to provide accurate estimates of the impact of research objects and detect emerging trends in research topics.
The service has the following components
Extraction of scientific fields from publications
This component consists of classification tools for automatically assigning publications to scientific fields of study (FoS). The classifier employs metadata included in publications, such as references, citations, titles, and abstracts. The existing OpenAIRE system will be expanded to support SciLake’s pilot case studies by leveraging domain-specific classification schemes and ontologies.
Contact person:
Multi-perspective analysis of scientific impact
This component combines multiple indicators to measure research impact from a multidimensional perspective. We will use AI to analyze the citation network of a research work and related scholarly metadata. This will enable us to assess research impact from different perspectives, such as the knowledge of the fields of the publications, the context of the citations considered, and the links to other research objects. Computed impact measures will be used to identify valuable research and topic trends.
Contact person:
Impact propagation among research objects
This component consists of a set of tools for measuring impact propagation among research objects of different types, whose direct impact is difficult to determine. To this end, and using as a starting point estimations of the direct impact of objects that are feasible (e.g., citation-based influence or publication popularity), SciLake will utilize models and techniques from the fields of network science and statistical mechanics. These include spreading processes on networks and belief propagation on graphs. As a result of these processes, impact scores are assigned to properly connected research objects within the relevant SKGs.
Contact person:
Trend identification analysis
This component tracks structural changes in research topics over time (e.g., topic splitting or merging). It uses aggregated impact scores for all topic-related papers to estimate the total attention each topic receives. It offers unique trend monitoring services for researchers and other interested parties.
Contact person: