Latifis Konstantinos (Phd Candidate)

Thesis title:   «Parametric collecting and classifying scientometric data & evaluation of research entities

Supervisor: Sidiropoulos Antonis

Advisory Committee Members:

Antoniou Efstathios, Professor, Dept. of Information and Electronic Engineering, IHU
Ougiaroglou Stefanos, Assistant Professor, Dept. of Information and Electronic Engineering, IHU

Abstract:

Scientometrics constitutes a branch of information science that focuses on the measurement of scientific output, employing statistical tools to understand the impact and evolution of research. However, the continuous increase in scientific publications has led to complex and large-scale datasets, rendering analysis more challenging.

To address these challenges, techniques such as multidimensional data analysis are applied, with core methodologies including skyline analysis and rainbow ranking. The former identifies points that are not dominated by others, offering candidate solutions without predefined preferences, whereas the latter proposes a fair ranking by balancing all evaluation criteria. In cases of high-dimensional data, the phenomenon of the “curse of dimensionality” emerges. For this reason, dimensionality reduction techniques such as PCA, UMAP, or other weighted variants are employed.

The proposed doctoral dissertation aims to combine the aforementioned techniques to establish a coherent and reliable framework for ranking research entities such as universities, researchers, and journals, guided by principles of fair, multi-criteria evaluation.

At the final stage, the development of an application is envisaged, which will not only present ranking results but also allow users to experiment with different weighting schemes, filter data according to their interests, and extract interpretable and customizable outcomes. This functionality positions the proposed system not only as a research analysis tool but also as a decision-support system for funding policies, academic administration, and collaboration planning.

Compared to existing ranking methodologies, which often rely on a limited number of criteria, the proposed approach seeks to distinguish itself through transparency, flexibility, and its capacity to integrate diverse evaluation perspectives. Ultimately, the dissertation aspires to provide an alternative, methodologically rigorous approach to the multidimensional parametric evaluation of the academic system, grounded in modern statistical tools.