Machine Learning and Expert Judgement: Analyzing Emerging Topics in Accounting and Finance Research in the Asia–Pacific

Abstract

In this paper, we focus on the question to what extent machine learning (ML) tools can be used to support systematic literature reviews. We apply a ML approach for topic detection to analyze emerging topics in the literature-our context is accounting and finance research in the Asia-Pacific region. To evaluate the robustness of the approach, we compare findings from the automated ML approach with the results from a manual analysis of the literature. The automated approach uses a keyword algorithm detection mechanism whereby the manual analysis uses common techniques for qualitative data analysis, that is, triangulation between researchers (expert judgement). From our paper, we conclude that both methods have strengths and weaknesses. The automated analysis works well for large corpora of text and provides a very standardized and non-biased way of analyzing the literature. However, the human researcher is potentially better equipped to evaluate current issues and future trends in the literature. Overall, the best results might be achieved when a variety of tools are used together.

Research of globalization influence to paradigm change in financial services

“…According to derived data following preliminary results of the research was obtained after application of mean average formula (1) as it is shown in table 3. It was 0 values for the category "V" for 2017 [11][12][13], 2018 [14][15][16][17][18] and 2019 [19][20][21][22] publishing years, but in only one case in 2020 [23] all categories had their respective values.…”

“…According to that formula (1) all the values from data set (see table 4), in this case values (total number of words per category) derived from each article, were added up and then divided to the number of articles within the one-year range. In this way data from articles within the range of 2017 (3 articles) [11][12][13] 2018 (5 articles) [14][15][16][17][18], 2019 (4 articles) [19][20][21][22] and 2020 (4 articles) [23][24][25][26] are analyzed separately from the rest publishing years. See table 3.…”

Sixty years of Accounting & Finance: a bibliometric analysis of major research themes and contributions

“…To implement the burst detection algorithm, we used the Python library called burst_detection (Version 0.1.3). The parameters used were the resolution of state jumps set to 2 and the gamma set to 0.5 (see Cai et al ., 2019, for further explanation of these parameters). Topics had to appear at least five times to be considered for our analysis.…”

“…In a recent study, Cai et al . (2019) already examined emerging trends in 11 accounting and finance journals (including 597 articles published in Accounting & Finance ) between 2009 and 2018. This study identified several emergent themes such as China‐related research, integrated reporting, as well as Fintech and environmentally focused research.…”

Interdisciplinary Research Maps: A new technique for visualizing research topics

“…The approach introduced here offers an improved approach to topic modeling. Topic modelling typically focuses on identifying a group of words (i.e., topic) from a collection of textual data [12][13][14], but has a number of limitations. In topic modelling, the user establishes the number of topics that will be extracted, which is based on an arbitrary selection [15].…”