Search

Purpose: Relevance ranking algorithms rank retrieved documents based on the degrees of topical similarity (relevance) between search queries and documents. This paper aims to introduce a new relevance ranking method combining a probabilistic topic modeling algorithm with citation data. Data and Method: We applied this method to the iSearch corpus of c. 435,000 physics papers. We first ran the topic modeling algorithm on titles and summaries of all papers for 65 search queries and obtained the relevance ranking lists. We then used citation data with the existing relevance rankings, thereby incrementally refining the results. The outcome produced better relevance rankings with papers covering various aspects of the topic searched as well as the more marginal ones. Finally, we evaluated the retrieval performance of the proposed method. Findings: Findings suggest that the topic modeling algorithm might sometimes overlook the terms used in different contexts in the papers. However, the fusion of citation data to relevance ranking lists provides additional contextual information, thereby enriching the results further with various papers of higher relevance. Moreover, results can easily be re-ranked. Implications: We argue that once it is tested on dynamic corpora for computational load, robustness, replicability, and scalability, the proposed method can, in time, be used in both local and international information systems such as TR-Dizin, Web of Science, and Scopus. Originality: The proposed method is, as far as we know, the first one that shows that relevance rankings produced with a topic modeling algorithm can be incrementally refined using citation data. [ABSTRACT FROM AUTHOR]

Objective: This paper aims to find out the factors that are likely to increase the citation impact of Turkey-addressed international papers. Method: For this purpose, we downloaded bibliographic records of some 226,000 TR-addressed papers that were published between 2006 and 2015, there from selected a sample (2%) using stratified probabilistic sampling technique, and obtained relevant data from various sources such as the number of citations and coauthors, journal impact factors (JIF) and journal quartiles pertaining to each paper in the sample. We then identified 19 independent variables on article, journal, cooperation, support and access levels and analyzed the data using the hurdle model. Findings: The strongest factors increasing the number of citations of TR-addressed papers are five-year JIFs and journal quartiles (Q1) in which TR-addressed papers appeared and the number of references in papers' reference lists. These factors also reduce the number of papers with zero citations. Papers published in cooperation with other countries collect higher numbers of citations. Factors such as the subjects of papers, whether papers received monetary support or not, and methods by which papers can be accessed (open or closed access) have no impact on the number of citations received. Implications: Findings indicate that academic promotion and support policies encouraging publishing in high impact journals while forestalling publishing in low impact journals should be developed in order to increase the scientific impact of TR-addressed papers. Originality: This is the first study using hurdle model to investigate the direct impact of several article, journal, collaboration, monetary support and access level explanatory variables on the number of citations that TR-addressed papers receive. [ABSTRACT FROM AUTHOR]