Introduction

Sharing research data alongside journal articles is becoming more and more common. Sharing data provides a recognized set of benefits for science, such as allowing for easier reproduction of the published results, benefiting future research and in general providing a more solid ground for credibility of the published work. Authors may benefit because of increased citation of articles with associated research data\cite{Piwowar2013}. Still, while researchers are encouraged to share more data, the practice is still limited \cite{Borgman_2012}.
The most likely reason for this disconnect is that authors and researchers are not finding the right incentives to share data (see, for example\cite{Mongeon_2017}): while most people see the overall benefit of data sharing, there is no clear reward system that permits researchers to directly use their data for career advancement or funding requests. 
A growing number of research funders have data sharing policies in place, including data management requirements for grant applications \cite{sparc}, but data is rarely - if ever - a decisive part of the research quality assessment. This is because it is very challenging to evaluate the quality and impact of research data (see, for example, \cite{Piwowar_2008}). 
Despite being much debated, the scientific community has an evaluation process for journal articles. Well known and recognized metrics, both qualitative and quantitative in nature, are commonly used. The paramount example is citations. Counting article citations is a relatively streamlined procedure nowadays, with a number of major players applying similar systems to count citation in different databases, such as ScopusWeb of Science, and CrossRef
In Fig. \ref{284706} a simplified version of the Scopus data model is displayed. It represents the baseline for identifying article citations, assigning them to the right author and affiliation. The entire Scopus database is built around article records, that are analysed to extract and normalize author, affiliation and journal information \cite{j2017}.