“Everything I count, I count it for you”. How usage statistics can measure research, beyond views and downloads
- 26 September 2023 |
- 16:00 |
- Session 3 |
- Sala Nouvel - Reina Sofia Museum
Usage statistics are based on the processing of usage events, triggered by metadata views or full text downloads of digital objects (eg. articles, datasets, books, etc.) in content providers like repositories, e-journals and CRISs. In this talk, we will discuss OpenAIRE’s UsageCounts service, the usage statistics service developed by OpenAIRE. UsageCounts service collects usage events and publishes COUNTER CoP standardized usage statistics, like the number of metadata views and full-text downloads for research items. However, it goes beyond this typical deployment of usage statistics and provides more qualitative metrics and adds value on how the usage statistics are exploited by a variety of stakeholders. This is realized via two paths. The first path is the enhancement of usage statistics with information from OpenAIRE Research Graph to build indicators that could be used to discover, analyze and evaluate the impact of Open Science. For example, how open access routes, Green, Gold, Hybrid, etc. are used, or how FAIR principles are accommodated in user behavior. UsageCounts service, also allows usage statistics, combined with topics from Field-of-Science (FoS) and Sustainable Development Goals (SDGs) to reflect relevance of a particular research output, over the course of time and up to the present. The second path of adding value using the UsageCounts service, is the exploitation of a hybrid content-based and collaborative filtering approach, using research product topics and usage events. This approach allows the construction of user communities, i.e., group of users sharing common interests and preferences. Following these two paths, UsageCounts provide for researchers, authors, funders and policy makers, a better understanding of the usage patterns of research products and useful insights for research uptake. They can be more recent than bibliometric indicators and could show the usage, via user communities, of a published work by non-authors, for example students, academics, and non-academic users who do not publish but may read scholarly publications and providing an important indicator to analyze trends. Usage statistics enhanced information, can also be exploited by stakeholders like repository managers to drive bibliographic instruction, internal marketing, and the development of the provider’s datasource. Finally, user communities can be used to develop a recommendation service that offers similar research products to the members of the community, based not only on the similarity of the content, but also on the similarity of the usage behaviour.