Document Type : Innovative Paper

Authors

• Tahereh Abbasi
• Hadi Saboohi
• Javad Mohammadzadeh
• Amineh Amini
• Azam Bastanfard

Department of Computer Engineering, Ka.C., Islamic Azad University, Karaj, Iran.

Abstract

Background and Objectives: Keyword extraction from specialised texts is challenging because the importance of terms depends on both their semantic relationships and structural roles within a document. This study introduces the Distributed Adaptive Self-Learner (DASL), which combines a dynamic semantic network with reinforcement learning for adaptive keyword selection.
Methods: DASL represents candidate keywords as nodes connected according to semantic similarity. Semantic and structural characteristics are jointly considered, and keyword extraction is formulated as a sequential decision-making process with Select and Discard actions. An action-dependent reward function guides the learning process, allowing the model to adapt keyword selection to different textual contexts. DASL was evaluated on Scientific Papers, Technical Reports, and News Texts and compared with TF-IDF, BERT, GPT-based, INSPECT, and Attention-Seeker methods.
Results: DASL achieved Precision values of 0.88, 0.86, and 0.90 and Recall values of 0.81, 0.79, and 0.84 across the three datasets, resulting in F1-Scores of 0.84, 0.82, and 0.87, respectively. The extracted keywords also showed high frequency alignment. In the information retrieval experiment, DASL achieved a Precision of 0.84 and an MRR of 0.81, corresponding to a 35% improvement over the TF-IDF baseline in retrieval Precision.
Conclusion: The results indicate that combining dynamic semantic representation with reinforcement-based decision-making provides an effective approach to keyword extraction. By jointly considering semantic relevance and structural importance, DASL improves keyword selection and contributes to more effective downstream information retrieval. Further evaluation on larger and multilingual datasets is warranted.

Keywords

• Extraction
• Semantic
• Relationships
• Distributed Adaptive Self-Learner
• Natural Language Processing,

Main Subjects

• Text Classification

Open Access

This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit: http://creativecommons.org/licenses/by/4.0/

 

Publisher’s Note

JECEI Publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

 

Publisher

Shahid Rajaee Teacher Training University