Citations analysis of selected journals using DEA models

This thesis is focused on analysing citations of scientific journals from the field of operational research using data envelopment analysis models (DEA). At present, the quality evaluation of journals, from the citations point of view, is performed in different and often ineffective ways by the global public. This issue becomes the basis for the formulation of a number of suitable DEA optimisation models helping to draft a new approach in order to gain a new quality indicator for these journals. Firstly, the reader is acquainted with the current practice of evaluating scientific journals, especially the use of journal impact factor and article influence score, to show that both indicators suffer numerous drawbacks. Thus, an alternative approach to practise this kind of evaluation is proposed and further researched. This part is followed by a brief summary of the DEA models theory, which is used in subsequent parts of the thesis. A thorough description of how to obtain and process the data necessary for the application of mathematical models follows in the next part, accompanied by a series of attached computational experiments based on different approaches, including the analysis of results. The thesis confirms the appropriateness of the adoption of a journal quality evaluation using the efficiency scores which are the output values of DEA models. The idea is based on the relationship between the efficiency scores and the values of selected bibliometric indicators. The recommended approach is then supported by a correlation analysis of the results obtained. The reader of this thesis will not only able to read the results of operational research journals ranking, but to apply this approach within a wider field of science.