|
Authors |
Ma脙卤a, Nathaniel Jay; Babiera, Johniel; Bayloces, Kriziah Lynn; Palmer, Xavier-Lewis; Potter, Lucas; Lavilles, Rabby; Velasco, Lemuel Clark |
| Abstract |
The progress of Information Retrieval Systems (IRS) research has been influenced by available technologies. With the rapid development of technologies associated with the IRS, there is a need to examine the extant literature related to its advancement. This study reviewed the literature on the IRS, particularly on the methods and gaps in developing the IRS. Specifically, it aimed to synthesize IRS methods implementation, including frameworks and recommendations for future studies from articles ranging from 1997 to 2022 using a methodological review. The methodological review followed the 2020 Systematic Reviews and Meta-Analyses (PRISMA) guidelines and only considered indexed articles from Scopus and Web of Science (WoS). The publications showed momentum in 2018 onwards. The authors working in IRS were mainly from China, the United States of America, and the United Kingdom. Besides the IRS, the recent leading theme is its application in the medical field. However, it should be noted that most articles mentioned that the IRS鈥檚 applications cut across other areas, including business and social sciences. Regarding the query presentation and indexing methodology, though the Vector Space was the most frequently used method, the probabilistic approach was the most widely used method for different applications. Meanwhile, for the matching process, the leading techniques were ranking-based, semantic approach, and ontology-based. For future works, researchers can work on checking assumptions, model or method accuracy, and parameter tuning for enhancement of IRS. Integrating user information, domain knowledge, rich thesauri, and other contextual information was strongly recommended by existing research. |
Mindanao State
University - Iligan Institute of Technology