Paragraphic Query-Based Summarization Approach for Arabic Extracts

The internet's exponential increase in information availability imposes a significant cost on individuals seeking such content. This challenge is being addressed by numerous researchers striving to overcome its complexity. The current study presents an alternative approach to query-based automatic text summarization. Unlike earlier methods, this approach generates the query based on the first paragraph sentences of the material being summarized, eliminating the need for user input during query submission. Applying the generated query to the document results in a final summary. Additionally, the study investigates the impact of different query lengths and similarity measures. The evaluation utilized the ROUGE metric and the EASC dataset. The experimental findings show that the suggested approach, which makes use of the Russel similarity measure and a longer query length, performs better than alternative scenarios that use the cosine, Forbes, and Jaccard similarity measures and shorter query lengths.