Data credit distribution through lineage

Data are a fundamental asset in the current world of research. Data citation is becoming more common and supported by research databases, but it still presents many research challenges. This paper describes Data Credit, a new measure of value for data derived from data citation, that enables us to annotate databases with real values representing their importance. Credit, computed through the citations, can be used alongside them to better understand the importance of data. We introduce the task of Data Credit Distribution, the process by which credit produced by a citation is and assigned to the data in a database responsible for producing the output information being cited. We describe how this process can be performed and, through experiments, we show that credit can serve, among other things, to highlight “hotspots” in the database.