Higher education in the era of artificial intelligence: academic freedom as a case study

Abstract

Higher education is crucial for the development of states and societies and improving the overall quality of life. However, entry into higher education is often influenced by factors beyond qualifications, and individuals in the field face suppression from the controlling parties. These challenges undermine the value of education and the integrity of democratic processes like elections. In this paper, we study academic freedom in Lebanon and propose a technique that dynamically extracts the factors that might affect academic freedom. This technique comprises multiple stages: data collection, data preprocessing, static extraction of factors, dynamic extraction of factors, and evaluation. In the data collection stage, data was obtained from 254 participants through a questionnaire that discusses various facets of academic freedom. The preprocessing stage enhances data quality through cleaning, normalizing, and transforming. For static extraction, factors impacting academic freedom are identified using naive K-means clustering. In dynamic extraction, the Apriori algorithm identifies key metrics. Finally, a customized K-means algorithm clusters data based on a specific metric. This algorithm was applied on both, the statically and dynamically extracted metrics, and comparison was done based on the accuracy of the resultant clustering. This comparison demonstrates the effectiveness of the proposed technique in identifying and analyzing factors impacting academic freedom.