Data Collection
A bibliometric analysis is a systematic approach that employs computer-assisted techniques to identify the core research or authors, and their relationships by scrutinizing publications pertaining to a particular topic or field (De Bellis, 2009). The two primary databases utilized for bibliometric analysis are Web of Science (WOS) and Scopus (Singh et al., 2021). Although both databases are widely used, WOS is known for its more stringent standards and has a 99.11% overlap with Scopus in terms of indexed journals (Singh et al., 2021). Consequently, the published papers on inference processing in discourse comprehension during 2001-2021 were collected from the WoS Core Collection, consisting of Science Citation Index Expanded (SCI-EXPANDED), Social Sciences Citation Index (SSCI), Arts and Humanities Citation Index (A and HCI), Conference Proceedings Citation Index-Science (CPCI-S), Conference Proceedings Citation Index Social Science and Humanities (CPCI-SSH) as well as Emerging Sources Citation Index (ESCI). All collected bibliographic records were written in English. The search strategy and the inclusion criteria were detailed in Figure 1. The selection of search terms for this study is informed by the bibliometric analysis of lexical inferencing conducted by Yang et al. (2023). A total of 3,131 records were obtained from 1,306 journals distributed in 200 WoS categories. The current research attempted to focus on inference processing during text reading from the perspectives of linguistics, psychology, and neuroscience. As a result, 1,373 articles were extracted, and the categories involved are shown in Figure 2. Among the 16 categories, Linguistics, Psychology Experimental, Psychology Educational, and Language Linguistics involve the most articles, each accounting for more than 20 percent.
Figure 1. The flowchart of the search strategy and selection process in this study
Figure 2. The categories involved in this study
Instrument
The CiteSpace software program is a free Java application that analyzes and visualizes the literature of a specified scientific or knowledge domain, providing a visual gateway to the literature of scholarly publications. The program generates interactive visualizations from bibliographic information, particularly citation data from the WoS, which allows users to navigate and explore patterns and trends uncovered in scientific publications (Chen, 2006).
In this study, the knowledge domain associated with inference processing during text reading in the framework of linguistics, psychology, and neuroscience was explored to reveal critical references, identify research patterns and hotspots to predict emerging trends in the literature.