Unsupervised Automated Keyphrase Extraction Approaches: A Literature Review

• Published in: 2023 International Conference on Information Technology, Applied Mathematics and Statistics (ICITAMS) pp. 269 - 274
• Main Authors: Kahdum, Alyaa Abdual, Al-Hameed, Wafaa
• Format: Conference Proceeding
• Language: English
• Published: IEEE 20.03.2023
• Subjects: Data mining; Feature extraction; Information retrieval; keyphrase; keyphrase extraction; Mathematics; Neural-based Approaches; review; Semantics; Surveys; Syntactics; Traditional Approaches; Unsupervised approach keyphrase extraction
• DOI: 10.1109/ICITAMS57610.2023.10525639

Because of technical improvements and the exponential growth of textual data and digital sources, extracting high-quality keywords at a high level in research is now more challenging. Keyphrase extraction characteristics, which are growing in popularity, are required for high-level keyword extraction. Automated keyphrase extraction (AKE) aims to find the significant subjects of a text document by automatically identifying a small group of single or multi-words from inside the text. AKE is crucial to many NLP and information retrieval tasks, including document summarizing and classification, article recommendation, and full-text indexing. This paper presents a survey of various techniques available for keyphrase extraction, discusses some important feature selection methods utilized by researchers to rank candidate keyphrases based on their importance, and finally debates various benefits and drawbacks for each category.