Unlocking the Potential of Keyword Extraction: The Need for Access to High-Quality Datasets

Amur, Z.H. and Hooi, Y.K. and Soomro, G.M. and Bhanbhro, H. and Karyem, S. and Sohu, N. (2023) Unlocking the Potential of Keyword Extraction: The Need for Access to High-Quality Datasets. Applied Sciences (Switzerland), 13 (12).

Full text not available from this repository.
Official URL: https://www.scopus.com/inward/record.uri?eid=2-s2....

Abstract

Keyword extraction is a critical task that enables various applications, including text classification, sentiment analysis, and information retrieval. However, the lack of a suitable dataset for semantic analysis of keyword extraction remains a serious problem that hinders progress in this field. Although some datasets exist for this task, they may not be representative, diverse, or of high quality, leading to suboptimal performance, inaccurate results, and reduced efficiency. To address this issue, we conducted a study to identify a suitable dataset for keyword extraction based on three key factors: dataset structure, complexity, and quality. The structure of a dataset should contain real-time data that is easily accessible and readable. The complexity should also reflect the diversity of sentences and their distribution in real-world scenarios. Finally, the quality of the dataset is a crucial factor in selecting a suitable dataset for keyword extraction. The quality depends on its accuracy, consistency, and completeness. The dataset should be annotated with high-quality labels that accurately reflect the keywords in the text. It should also be complete, with enough examples to accurately evaluate the performance of keyword extraction algorithms. Consistency in annotations is also essential, ensuring that the dataset is reliable and useful for further research. © 2023 by the authors.

Item Type: Article
Additional Information: cited By 2
Depositing User: Mr Ahmad Suhairi UTP
Date Deposited: 04 Jun 2024 14:10
Last Modified: 04 Jun 2024 14:10
URI: https://khub.utp.edu.my/scholars/id/eprint/18490

Actions (login required)

View Item
View Item