چکیده:
Google Scholar has recently attracted great attentions as an open access multidisciplinary citation database, and a tool for retrieving scientific works for scientometricians and researchers. The present research intended to highlight the limitations brought about by efficiency policies of the search engine and its impact on the results available to users. To do so, it examined the accessibility of the retrieval results, through conducting 54 searches in this database. The results showed that the estimation of the results on the top of the first page returned by Google Scholar did not match that of the accessible results. Therefore, these statistics could not be accounted for to precisely determine the number of documents on a topic. Moreover, the results showed that although the subjects selected for the searches were very specific, the number of results for each search was very wide and exceeded the upper limit of 1,000 records authorized in Google Scholar for display. By limiting the searches to the title field, the number of the results was dramatically reduced. Since title is one of the most important representations of document contents in scientific and technical fields, this strategy can increase the precision of the results and thus the effectiveness of the retrievals. The investigation of the accessibility of the search results for the title field also showed that some documents, though scarce in number, were still inaccessible despite the fact that they were within the 1000-record limits. In addition, in title field search, some rare cases of duplicate records, incompatibilities between queries and documents were observed regarding the language of the documents and exact phrase search. The lack of automatic truncation in field searches was one of the most important issues necessitating the use of sophisticated search strategies.
خلاصه ماشینی:
ir Abstract Google Scholar has recently attracted great attentions as an open access multidisciplinary citation database, and a tool for retrieving scientific works for scientometricians and researchers.
The investigation of the accessibility of the search results for the title field also showed that some documents, though scarce in number, were still inaccessible despite the fact that they were within the l000-record limits.
Although the results of the initial research were not satisfactory, recent research have suggested a dramatic improvement in the citation coverage of this database (Chen, 2010; Gehanno, Rollin Darmoni, 2013; Harzing, 2013a; Harzing, 2013b), the relevance of the results (Adriaanse & Rensleigh, 2011; Bar-Ilan, Levene, & Lin, 2007; Garcia-Perez, 2010; Kirkwood & Kirkwood 2011; Meier & Conkling 2008; Shariff et al.
Using GS's data retrieved for Sotudeh and Houshyar (2018), the present study tries to demonstrate how the system's efficiency observations affect the accessibility of GS search results in free searching (via full text keywords) and title field searching.
, 2013) in terms of the coverage of some subjects (Walters, 2007), and the relevance of the retrieved results; while evaluated to be parallel to some databases such as WoS, Scopus (Adriaanse & Rensleigh, 2011; Bar-Ilan, Leven, & Lin, 2007; Garcia-Perez, 2010), Compendex (Meier, & Conkling 2008) and BIOSIS Previews (Kirkwood & Kirkwood 2011).
For instance, due to the automatization of the process of citation identification, GS is subject to citation errors as well as non-academic resource coverage (Cathcart & Roberts, 2005; Donlan & Cooke, 2005; Jacso, 2005a; Vine, 2006; Wleklinski 2005; Lopez-Cozar, , Robinson-Garcia, & Torres-Salinas„ 2012).