Comparative Analysis of the Performance of the Fasttext and Word2vec Methods on the Semantic Similarity Query of Sirah Nabawiyah Information Retrieval System: A systematic literature review

Abstract

In the field of text processing or Natural Language Processing, the increasing popularity of the use of words used in the field of Natural Language Processing can motivate the performance of each of the existing word embedding models to be compared. Semantic similarities have an important role in the field of language. Keywords are the most important thing in finding information. Seeing the benefits and studying the knowledge of Sirah Nabawiyah, especially from unreliable sources, a system to retrieve information was proposed for viewing and studying Sirah Nabawiyah. Based on the explanation above, the writer conducted a Systematic Literature Review (SLR) which was related to the semantic similarity comparison of the word embedding method, namely fast text and word2vec which contributed to the information retrieval system using the knowledge discovery in database method approach to documents in the form of Sirah nabawiyah text. Analysis and comparisons of several pieces of literature have been carried out. As a result, there are several criteria to complement the advantages and disadvantages of previous research.