Downloads: 0 | Views: 87

Research Paper | Computer Science and Information Technology | Saudi Arabia | Volume 12 Issue 6, June 2023

Graph-based Model for Keyphrases Extraction from Arabic Text (GMKE)

Amirah Al Shammari | Abdullah Al Ghamdi

Abstract: Keyword extraction is an important step in several natural language processing and information retrieval applications, including text summarization and search engine optimization. Keywords include the most essential information characterizing the document's content. As the number of available documents increases, it is extremely difficult for a user to read each one in depth. Therefore, knowing the subject of the documents without performing an in-depth analysis is essential, and an automatic method of keyword extraction is required. Arabic research is still leaking in this area. In this study, we introduce a graph-based model for extracting the keyphrase using the K-mean clustering algorithm and TF-IDF ranking for single document text. Experiments were conducted using the ArabicKPE dataset. The experimental findings show that our model gives encouraging results compared to TF-IDF approaches in the Arabic KPE domain based on Recall, precision, and F-measure.

Keywords: Arabic NLP, keyphrase extraction, graph -based, clustering

Edition: Volume 12 Issue 6, June 2023,

Pages: 2755 - 2761