The SKM team aims to develop innovative approaches to generating value from scholarly data by leveraging artificial intelligence, large-scale data mining, semantic technologies, and visual analytics.

We are pursuing several research avenues, including:

• AI systems for automated literature reviews, hypothesis generation, citation prediction, and metascience.
• Generation of knowledge graphs, as well as their refinement through automatic methods and human-in-the-loop approaches.
• Large-scale classification of scientific publications. Our technologies also powered Stanford’s AI Index, the most well-known AI report.
• Analysis and integration of knowledge organization systems of research fields, including the automatic generation and evolution of taxonomies – see, for example, the Computer Science Ontology.
• AI tools for exploring and analysing the scientific landscape – see, for example, the AIDA Dashboard.
• Automatic forecasting of the emergence of new research fields.

We released a number of knowledge bases for exploring research dynamics, including:

• The Computer Science Ontology (CSO), the largest taxonomy of research topics in the field.
• The Academia/Industry DynAmics (AIDA) Knowledge Graph is an innovative resource that describes 14M publications and 8M patents according to their topics and industrial sectors.
• The Computer Science Knowledge Graph (CS-KG), a large-scale automatically generated knowledge graph that describes 25M entities (e.g., tasks, methods, metrics, materials, others) extracted from 15M research publications.

We collaborate with major publishers and universities to generate scalable applications, such as search engines, recommender systems, and analytics tools. In particular, we are currently working closely with Springer Nature in the development of several semantically enhanced solutions.

In 2019, we released the Computer Science Ontology (CSO), which is currently the largest taxonomy of research areas in the field and has been officially adopted by Springer Nature.  In the context of our collaboration with Springer Nature, I have also designed and co-developed the Smart Topic Miner, a tool is used by editors at Springer Nature to generate automatically the scholarly metadata for all their computer science proceedings, including flagship series, such as Lecture Notes in Computer Science (LNCS), Lecture Notes in Artificial Intelligence, and others.

In 2021, we generated the Academia/Industry DynAmics (AIDA) Knowledge Graph, an innovative resource for supporting large-scale analyses of research trends across academia and industry.  AIDA describes 14M publications and 8M patents according to the research topics drawn from CSO, the type of the author’s affiliations (e.g., academy, industry, collaborative), and 66 industrial sectors (e.g., automotive, financial, energy, electronics). In 2022, we also produced the Computer Science Knowledge Graph (CS-KG), a large-scale automatically generated knowledge graph that describes 25M entities (e.g., tasks, methods, metrics, materials, others) extracted from 15M articles. CS-KG was designed to support a large variety of intelligent services for analysing and making sense of research dynamics, assisting researchers, and informing the decisions of founding bodies and research policymakers.

Recently, our work has focused on the intersection of generative AI and knowledge graphs, aiming to create advanced systems for scientific exploration, horizon scanning, research hypothesis generation, and the automated production of literature reviews. As part of this effort, we published an in-depth survey on the transformative impact of AI systems for literature reviews, providing key recommendations for future research.