An Analysis of Web-scale Discovery Services From the Perspective of User's Relevance Judgment

Although web-scale discovery services have been implemented increasingly worldwide, there is a need for the exploration of search effectiveness from users' perspectives. In this sense, this study examined web-scale discovery services in the view of users' relevance judgment comparing to individual databases in the fields of Education and Library and Information Science. Using four search topics for the EBSCO Discovery Service (EDS), ERIC, ERC, LISA, and LISTA, their search effectiveness were measured in terms of modified precision, recall, and reciprocal rank based on the relevance judgments of four participants. Comparison of the measurements showed that the web-scale discovery service was less effective than individual databases. In particular, EDS's effectiveness was lower than that of ERIC and ERC in terms of modified precision and recall. The modified reciprocal rank of EDS was lower than those of most individual databases in the fields of Education and LIS. Moreover, the relationship between the rankings from EDS and those from four participants was weak, as indicated by Spearman rank-order correlation coefficients (0.141 in Education and 0.170 in LIS). In fact, the effectiveness in the fields of LIS and Education of EDS was lower than those individual databases to a degree dependent on the field.