Insights from a text mining survey on Expert Systems research
from 2000 to 2016
ALGORITMI Research Centre, University of
Minho, Braga, Portugal
ISTAR‐IUL, Instituto Universitário de Lisboa
(ISCTE‐IUL), Lisbon, Portugal
CIS‐IUL, Instituto Universitário de Lisboa
(ISCTE‐IUL), Lisbon, Portugal
NOVA Information Management School
(NOVA IMS), Universidade Nova de Lisboa,
School of Computing and Communications,
The Open University, Milton Keynes, UK
Paulo Cortez, ALGORITMI Research Centre,
University of Minho, 4710‐057 Braga,
This study presents a literature analysis using a semiautomated text mining and topic
modelling approach of the body of knowledge encompassed in 17 years (2000–2016)
of literature published in the Wiley's Expert Systems journal, a key reference in Expert
Systems (ESs) research, in a total of 488 research articles.
The methodological approach included analysing countries from authors' affiliations,
with results emphasizing the relevance of both U.S. and U.K. researchers, with
Chinese, Turkish, and Spanish holding also a significant relevance. As a result of the
sparsity found on the keywords, one of our goals became to devise a taxonomy for
future submissions under 2 core dimensions: ESs' methods and ESs' applications.
Finally, through topic modelling, data‐driven methods were unveiled as the most
relevant, pairing with evaluation methods in its application to managerial sciences,
arts, and humanities. Findings also show that most of the application domains are well
represented, including health, engineering, energy, and social sciences.
Expert Systems, literature analysis, research categorization, research evolution, text mining
Expert Systems (ESs) have been at the centre of decision support for managerial decision making. In the 70s and 80s, ESs were focused on
mimicking human experts and separating explicit knowledge (stored in a knowledge base) from the artificial intelligence inference machine
(Buchanan, 1986). Yet, since then, and in particular after the 2000s, much has changed due to the explosion of data, evolution of the Internet
(e.g., Internet of things), mobile, and social stances. Thus, there has been a pressure to extract as useful knowledge from past data and
incorporate such knowledge in ESs, leading to an increase in data related fields, such as Business Intelligence, data mining, big data, and data
science (Cortez & Santos, 2015).
Due to such ESs‐related research and technological evolution, it is relevant to perform a review of what has been recently published in the ESs
domain. As such, this paper focuses on analysing literature published in Wiley's Expert Systems journal (EXSY), emphasizing its relevance and
evolution through a recent timeline of 17 years, from 2000–2016. This journal, established more than 30 years ago, has always been at the
forefront of investigation on ESs, thus constituting one of the main sources for research in this major area. Single‐source literature analyses provide
an historical picture of the main topics addressed by that source, helping to guide the board of editors' future strategies and at the same time
providing a thorough perspective over the addressed topics (e.g., Moro, Rita, & Cortez, 2017). With this in mind, we set out to perform this task
by analysing all EXSY research articles published within the 17‐year timeline, in a total of 488 articles. Given the sparsity of keywords used by
authors to classify articles, we aimed to develop a taxonomy aggregating the main methods and applications of EXSY research. Finally, due to
the large volume of research, we conducted a semiautomated literature analysis using text mining (Moro, Cortez, & Rita, 2015) to assess the main
research trends. In particular, the text analysis includes an assessment of authors' affiliations and the analysis of research topics based on both ESs'
methods and ESs' application areas.
Received: 17 July 2017 Revised: 25 January 2018 Accepted: 5 March 2018
Expert Systems. 2018;35:e12280.
Copyright © 2018 John Wiley & Sons, Ltdwileyonlinelibrary.com/journal/exsy 1of10