The major results of the EPIQUE project are:
- Evaluation of topic extraction and alignment methods: We implemented, compared different approaches to topic extraction from scientific archives and alignment methods adapted to these models.
- New models of science evolution: Starting from the concept of phylomemies, we collaborated to explore the problem of representing the evolution of scientific domains and developed different models of evolution based on graphs, hierarchical clustering and spectral analysis.
- Phylomemy Quality Measures: The quality of generated phylomemies is influenced by many parameters (archive content, topic diversity, extraction, and alignment methods). We have proposed different quality measures and solutions that allow users to evaluate and control the quality of generated phylomemies.
- Phylomemy exploration interfaces: We have designed and developed some visualization and query interfaces for phylomemy exploration.
- Use cases and validation: Within the framework of the project, we have defined and studied various use cases with our partners of the IHPST on specific corpora. These use cases have been completed by other corpora for the study of evolution in different scientific domains.
- Prototypes: the EPIQUE project has allowed us to rethink and extend the Gargantext system of the ISC-PIF and to realize several prototypes for the construction and exploration of phylomemies.