The EOSC Future Project is an EU-funded H2020 project that is implementing the European Open Science Cloud (EOSC). EOSC will give European researchers access to a wide web of FAIR data and related services.
EOSC Future will build on the existing baseline for the European Open Science Cloud to deliver a platform with a durable set of user-friendly components that are designed for the long haul. Α system-of-systems approach is adopted to the EOSC platform, linking together other research portals, resources and services to respond to the data needs of a wide range of researchers.
One way to think about EOSC is as a fully operational web of data and related services founded on FAIR protocols, principles and standards for accessing interoperable datasets. In practice, stakeholders will be approached to ensure a smooth user experience, developing:
- EOSC core, the set of enabling services needed to operate the EOSC
- EOSC exchange registering resources and services from research infrastructures, other EOSC projects and science clusters to the EOSC and integrating them with the EOSC core functionalities
- The EOSC interoperability framework will provide guidelines for providers that want to integrate services or data into EOSC
Our involvement and contributions focus on improving both the user and the provider experience through our Recommender System and “neighboring” components like the User Profile and Provider RS Insights.
Recommender System Use cases:
- Recommending Similar Services to the User. When a user is viewing a service recommend similar services that are personalized based on the user history. We utilize both metadata and text attributes of the services.
- Project Completion to the User. A project is a collection of resources that a user creates usually aiming to solve a specific research task. Similar to “My Cart” of ecommerce shops. Inspired by the “Frequently bought together functionality” of e-shops, if project A has many common resources with project B, then recommend non common resources to project B.
- Auto-completion Suggestion on Service Onboarding for the Provider. Onboarding a new service, i.e., adding a new service to the catalogue, is done by filling in several forms. This is a tedious and time-consuming process that might discourage providers. We aim at generating auto-completion suggestions for list input values such as the Category of the service removing the burden of the provider having to look through lengthy lists.
The user profile component focuses on collecting in a centralised location the information that describes the user that is of value for other components like the RS and the user dashboard . So far, its data sources are the EOSC Portal and OpenAIRE Graph (OAG) although we expect in the future to include other external sources too.
Having a single point of querying information about the user could be useful for other components of the system like:
- The Recommender System
- The user profile dashboard
- Provider insights on RS performance
Insights on RS Performance for the Provider
The providers understand that the discoverability of their services depends on the performance on the Recommender System. So, we calculate and make available a set of insights on the dashboard of the provider.
- Most recommended services. The provider gets a better feeling of the RS impact on discoverability of their services.
- Recommendations of services over time. Check the impact on the RS of modifying a description/metadata of services.
- Which services are mostly recommended next to your services. Check which are your main “competitors” and attempt to better describe the selling points of your services.