(S²M \(-IQ):\) Semiotics Similarity Measurement and Information Quantification, Lecture Information Weightage Calculation

AI and ML-based smart applications will be the next best way towards essential services, like basic education, and quality optimization. In the educational delivery context, not only summarization but advanced stages of NLP require capabilities to extract information and quantify the delivered knowledge. Summarizing and evaluating summary quality alone is insufficient as it ignores whether the original lecture adequately covered relevant information about the topic—testing a summary against a perfect reference is meaningless if the lecture itself was incomplete. Hence, the perfect summarisation of educational delivery is insufficient, but information quantification is imperative, particularly for enhancing educational quality of service (E-QoS). This paper meticulously analyses the existing quality assessment metrics used in recent benchmark works and proposes/supports the need for information quantification by thoroughly examining some known metrics for generated summary quality assessment and their limitations concerning educational deliveries and respective summaries. The paper contributes in four key areas: first, a comprehensive study and evaluation of the sufficiency of available metrics for assessing the quality of candidate summaries in the context of educational deliveries; second, a proposal of the need for information/knowledge quantification (not only summary quality) in educational delivery summaries; third, mathematical modeling based on known metrics; and finally, experimentation and analysis of proposed mathematics on both benchmark and real-time novel datasets.