Single and Multi-Documents Text Summarization Technologies for Natural Language Processing: a Systematic Review on Method and Dataset

The summarization technologies have been increasingly used in recent decades. Those technologies are a very important part of emerging topics in computer science and engineering, such that Natural Language Processing (NLP). Several methods have been used for analysis to get good summary results. There are two types of document summaries: single document summaries and multi-document summaries. Single document summaries aim to extract information from a single document to get new and relevant summary information, while multi-document summaries extract information from multiple documents. This study focuses on the activities of the semantic literature in previous studies to obtain the basis of the widely used base methods and data sets used in this study. Data was collected from Scopus publication sources from 2019 until 2022 Q2 for analysis. Researchers use guidelines from the semantic literature method by using the basis of Kitchenham and Charters as a reference in its design. In this study, there were forty-eight articles obtained from the filtering results of several criteria used, including exclusion and quality assessment.