Retrievability in an integrated retrieval system: an extended study

Retrievability measures the influence a retrieval system has on access to information in given collection of items. This measure can help making an evaluation search based which insights be drawn. In this paper, we investigate retrievability integrated consisting items from various categories, particularly focussing datasets, publications \ijdl{and variables} real-life Digital Library (DL). The traditional metrics, that is, Lorenz curve and Gini coefficient, are employed visualize diversity scores \ijdl{three} retrievable document types (specifically publications, variables). Our results show significant popularity bias with certain being retrieved more often than others. Particularly, it been shown datasets likely other same category. contrast, variable or publication category evenly distributed. We have observed distribution is diverse for as compared variables.