Unlocking Scholarly Insights Using Deep Learning-Based Evaluation Method for Research Impact Assessment in Academic Institutions

Background: This paper addresses the academic institutions, especially the recently founded Indian Institutes of Technology (IITs), who struggle to appropriately evaluate and communicate their research potential. We call attention to the shortcomings in the present evaluation techniques, especially the dependence on bibliometric data, which would not fairly represent the spectrum of scholarly contributions. Objectives: Using data analytics and altmetrics, our study suggests a complete strategy to assess the research impact and exposure of these institutions with an eventual aim of raising their positions in the National Institutional Ranking Framework (NIRF). We demonstrate how Scopus publishing data analysis and comparison with NIRF rankings can provide real-time insights into academic influence, hence augmenting conventional metrics. Methodology and Findings: Together with Firefly optimization, we also offer a new deep learning technique called Deep Belief Networks (DBN), which improves the precision of data analytics process. By use of DBN-Firefly optimization, we hope to find links and patterns in academic data thereby enabling academic institutions and legislators to make better decisions.