A novel machine extraction algorithm for implicit and explicit keywords based on dynamic web metadata of scientific scholars' corpus. (25th May 2023)

• Record Type: Journal Article
• Authors: Mosbah, Mawloud
• Is Part Of: International journal of Web engineering and technology. Volume 18:Number 1(2023)
• Page Start: 29
• Page End: 44
• Publication Date: 2023-05-25
• Subjects: natural language processing -- keywords extraction -- automatic construction of metadata -- implicit keywords -- explicit keywords; World Wide Web -- Periodicals; Web site development -- Periodicals; Application software -- Development -- Periodicals; 006.7
• Languages: English
• ISSNs: 1476-1289

Abstract

Keywords extraction, as an operation to construct metadata, is an important pre-processing task considered by many natural language processing applications such as text summarisation, information retrieval, and clustering of documents. In this paper, we introduce a novel machine extraction algorithm for implicit and explicit keywords. The algorithm relies on a dynamic corpus of similar documents built by information retrieval engines. In addition to the direct utilisation of the keywords for similar documents, our algorithm combines some basic techniques. The given results, compared with some basic methods of the literature, seem to be very promising and we claim also the efficiency of our solution.