Content-Based Citation Analysis of Open Access and Non-Open Access Medical Articles Using Opinion Mining of Citances

Abstract

Introduction: Scientific communities have always been concerned about validity of open-access articles. Given the challenges of quantitative citation analysis in evaluating scientific articles, content-based citation analysis, including opinion mining of citances, can bring about more transparent results about their validity. In view of this, the present study compared the opinions contained in citances about open-access and non-open-access articles. Methods: We used a quantitative content analysis method with citation and opinion analysis approaches. The citances, bibliographic, and bibliometric data were extracted from Colil and PubMed databases. Opinion scores were assigned to the citances through SentiWords. After processing the titles, abstracts, and citances, Cosine similarity of Term Frequency-Inverse Document Frequency (TF-IDF) values were calculated. The open-access and non-open-access articles were then paired by their similarities in abstracts, titles, and citances. The data were analyzed using Friedman test and Spearman correlation. Results: There was no significant difference between the open-access and non-open-access articles in terms of their opinion scores, despite a significant difference in citation advantages. The pairs’ citance and textual similarities had no significant correlation with their opinion distance. Conclusion: Although the open-access studies had citation advantage over their similar non-open-access peers, they showed no significant opinion distance. Besides, similar texts did not necessarily follow the same opinion patterns. Consequently, to complete the results of quantitative citation analysis, the content-based citation analysis is emphasized.

Keywords

• citances
• citation analysis
• open access to information
• opinion mining