A profile of information systems research published in expert systems with applications from 1995 to 2008 Expert Systems with Applications 38 (2011) 3999–4005 Contents lists available at ScienceDirect Expert Systems with Applications j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / e s w a A profile of information systems research published in expert systems with applications from 1995 to 2008 Wen-Lung Shiau ⇑ Ming Chuan University, Department of Information Management, Taiwan, ROC a r t i c l e i n f o Keywords: IS research Literature analysis Expert system Research profile Citation analysis 0957-4174/$ - see front matter � 2010 Elsevier Ltd. A doi:10.1016/j.eswa.2010.09.061 ⇑ Address: 12F.-3, No. 3, Lane 24, Wunhua St., Pingjh Taiwan, ROC. Tel.: +886 3 4948766; fax: +886 3 4663 E-mail address: mac@mail.mcu.edu.tw a b s t r a c t Expert systems with applications (ESWA) has been regarded as one of the highly qualified journals in the information system. This paper profiles research published in ESWA from 1995 to 2008. Based on the multidimensional analysis, we identified the most productive author and universities, research paper numbers per geographic region, and the most employed issues and methodologies used by the most highly published authors. Our results indicate that (1) ESWA is clearly an internationalized journal, (2) the most employed methodologies are fuzzy ESs and knowledge-based systems, and (3) the leading highly published authors always have diverse methodologies and applications. Furthermore, the implica- tions for researchers, journal editors, universities, and research institution are presented. � 2010 Elsevier Ltd. All rights reserved. 1. Introduction Expert system is a kind of information system, a subfield of arti- ficial intelligence, and a software that is used to reproduce the per- formance of one or more human experts. The simulation of performance of the expert is to help human workers solve real world problems by expertise, a specific domain of knowledge. There are diverse problems which need to be solved in the real world. Thus, the use of expert system becomes prolific in many fields (Liao, 2005). ESWA is regarded as one of the distinguished international journals which focus on expert and intelligent sys- tems applied to real world problems globally. As a journal of expert and intelligent systems, it addresses many problems, methods, and performance. Therefore, it is useful for IS researchers to examine the major research issues in expert and intelligent systems and their trends in order to continuously progress the research. This paper profiles the types of research that have been pub- lished in ESWA from 1995 to 2008. The overview of research pub- lished in the journal provides a broader understanding of IS for authors, reviewers, and journal editors, universities, and research institutions (Avison, Dwivedi, Fitzgerald & Powell, 2008). It is worthwhile to classify, categorize, and profile this paper into various dimensions, such as research issue (Alavi & Carlson, 1992; Claver, Gonzalez, & Llopis, 2000; Culnan & Swanson, 1986; Farhoomand & Drury, 1999; Vessey, Ramesh, & Glass, 2002), research methodology (Chen & Hirschheim, 2004; Grover, Lee, & ll rights reserved. en City, Taoyuan County 324, 107. Durand, 1993; Ives, Hamilton, & Davis, 1980; Palvia, 2004; Palvia, Mao, Salam & Soliman, 2003), productive authors and university (Athey & Plotnicki, 2000; Im, Kim, & Kim, 1998; Nath & Jackson, 1991), because the results could provide some frequently called for guidelines and suggestions on how to publish in highly rated journals. Thus, this paper analyzes publication trends in ESWA to realize those objectives as follows: 1. To identify the most productive authors. 2. To identify the universities associated with the most research publications. 3. To identify the research impact of the most productive authors and geographic location based on the AIS classification. 4. To determine the issues and methodologies most investigated and analyze their trends. 5. To discuss the issues and methodologies of the most productive authors. To achieve these objectives, a systematic and comprehensive review is needed. This review is constructed as follows. The next section describes the research methodology used in this paper. The third section presents the results of data analysis. The final sec- tion is the conclusion and implications of this research. 2. Research method Expert system with applications published their first issue in 1990 and had full online information from this journal and was opened to general users in 1995. Thus this paper surveys the devel- opment of ESWA from 1995 to 2008. To focus on research articles, http://dx.doi.org/10.1016/j.eswa.2010.09.061 mailto:mac@mail.mcu.edu.tw http://dx.doi.org/10.1016/j.eswa.2010.09.061 http://www.sciencedirect.com/science/journal/09574174 http://www.elsevier.com/locate/eswa Table 1 Top 11 productive authors. Name Affiliation AIS region Count Ingoo Han Korea Advanced Institute of Science and Technology 3 26 Tzung-Pei Hong National University of Kaohsiung 3 13 Elif Derya Ubeyli TOBB Ekonomi ve Teknoloji Üniversitesi 2 13 Dirk Van den Poel Ghent University 2 13 Salih Gune Selcuk University 2 12 So Young Sohn Yonsei University 3 11 Shyi-Ming Chen National Taiwan University of Science and Technology 3 11 Kemal Polat Selcuk University 2 11 Yueh-Min Huang National Cheng Kung University 3 10 Sang Chan Park Korea Advanced Institute of Science and Technology 3 10 Inan Guler Gazi University 2 10 4000 W.-L. Shiau / Expert Systems with Applications 38 (2011) 3999–4005 we follow Gallivan and Benbunan-Fich’s suggestions to exclude editorials, guest editorials and book reviews from the analysis (Gallivan & Benbunan-Fich, 2007). Research articles between 1995 and 2008 totaled 1836 articles. For a detailed analysis, each article was carefully examined to capture the relevant data. Relevant data include title, author, sub- ject terms, abstract, affiliation, publication, year, and citations. Various items were counted and recorded for each article includ- ing the productivity of authors, geographic regions, author’s back- ground, research issues, and the impact of the research by the most productive authors. We used self citation within ESWA, ISI citation counts, and Google’s scholar citation counts to assess the number of contributions per author. Institutional productivity was assessed by using each publication counted as one for all authors regardless of the number of co-authors from the same university (Avison et al., 2008; Yogesh & Jasna, 2008). The geo- graphic location variables were counted and categories were based on the AIS guidelines. The profile of methodologies em- ployed in this research is from a review of expert system method- ologies and applications (Liao, 2005). Liao (2005) collects 166 articles from 78 journals from the year 1995–2004, surveys and classifies ES methodologies using 11 categories: rule-based sys- tems, knowledge-based systems, neural network, fuzzy ESs, ob- ject-oriented methodology, case-based reasoning (CBR), system architecture development, intelligent agent (IA) systems, model- ing, ontology, and database methodology. We collected relevant subjects of 11 categories. Then those subjects were used to clas- sify the title, keywords, and abstract of the articles. The issues and methodologies of the most productive authors were also counted. It is essential to emphasize that the findings of this study including the most productive authors and universities should be regarded as ‘indicative and not an authoritative statement’(Palvia, Pinjani, & Sibley, 2007). This is because such profiling analysis was applied only within ESWA. Some important researchers and uni- versities might not have published in ESWA. They may have exper- tise and skills to publish in other important outlets. Therefore, similar to previous profiling studies (Claver et al., 2000; Palvia et al., 2007), it is best to interpret these results cautiously. Table 2 Top 10 productive universities. Rank School Counts 1 Korea Advanced Institute of Science and Technology 90 2 National Chiao-Tung University 59 3 National Cheng-Kung University 55 4 Yonsei University 32 5 National Yunlin University of Science and Technology 31 6 Hong Kong Polytechnic University 29 7 Yuan-Ze University 28 8 National Taiwan University of Science and Technology 27 9 National Central University 24 10 Firat University 23 10 National Kaohsiung First University of Science and Technology 23 10 Nanyang Technological University 23 3. Results 3.1. Productive authors For assessing research productivity, an analysis is made of the authors who have published the most during the period 1995– 2008 in the expert system with applications. All publications naming the researcher are counted equally (Palvia, Pinjani, et al., 2007; Yogesh & Jasna, 2008). For example, an article with two co-authors would provide one count for each. This approach results in the combined count of all authors being greater than the total number of articles. The findings of this study only include authors who have published 10 or more articles during the period 1995– 2008 for reporting purposes. Table 1 lists the 11 most productive authors, sorted by the number of publications, along with the asso- ciation for information systems (AIS) region and current affiliation. The top 11 authors with 10 or more publications are: Ingoo Han, Tzung-Pei Hong, Elif Derya Ubeyli, Dirk Van den Poel, Salih Gune, So Young Sohn, Shyi-Ming Chen, Kemal Polat, Yueh-Min Huang, Sang Chan Park, and Inan Guler. The geographical regions sug- gested by AIS are region 1 (America), region 2 (Europe, the Middle East and African), and region 3 (the Asia and Pacific). It is quite obvious that the most productive authors came from region 2 and region 3. This is because ESWA has attracted non-American authors and has become an internationalized IS journal. 3.2. Leading research university To assess the contribution of institutions and or universities, an analysis is made of institutions and or universities which have published the most during the period 1995–2008. As stated previ- ously, the results should be regarded as indicative and not as an authority’s statement of university research. It was misleading to have several researchers from the same university co-authoring an article (Palvia, Pinjani, et al., 2007). Therefore a university was counted only once when it had two or more authors on a single publication. Table 2 lists the top 10 universities having 23 or more articles published in ESWA during the period 1995–2008. A total of 12 are listed as the top 10 universities because there are three uni- versities having the same counts. The top place in the lists is Korea Advanced Institute of Science and Technology, with 90 contribu- tions. The next 11 universities are National Chiao-Tung University, National Cheng-Kung University, Yonsei University, National Yun- lin University of Science and Technology, Hong Kong Polytechnic University, Yuan-Ze University, National Taiwan University of Sci- ence and Technology, National Central University, Firat University, National Kaohsiung First University of Science and Technology, and Nanyang Technological University. It may show that Asian univer- sities and researchers are more likely to prefer this international- ized journal focusing on expert and knowledge domain. 3.3. Research numbers by AIS region The association of information system (AIS) suggests that the geographical regions are region 1 (America), region 2 (Europe, W.-L. Shiau / Expert Systems with Applications 38 (2011) 3999–4005 4001 the Middle East and African), and region 3 (the Asia and Pacific). An analysis is made of the regions. Fig. 1 lists the total number of con- tributions by AIS regions during the period 1995–2008. Before the year 2000, the total number of articles of region 1 (America) reached its highest in 1998. After the year 2001 region 2 (Europe, the Middle East and African) and region 3 (the Asia and Pacific) grew consistently. After the year 2006 all regions grew quickly be- cause the journal received more and more articles and also has more capacity to publish more articles. 3.4. Research methodologies The profile of methodologies employed in this research follows Liao’s (2005) 11 categories: rule-based systems, knowledge-based systems, neural network, fuzzy ESs, object-oriented methodology, case-based reasoning (CBR), system architecture development, intelligent agent (IA) systems, modeling, ontology, and database methodology (Liao, 2005). We collected the keywords from the Liao’s relevant 11 categories and listed them in Appendix A. We used the keywords in three stages to attribute a total of 1836 articles to 11 categories. The first stage was to compare the collected keywords to each title of research article. There are 757 identified articles and 1079 articles are unidentified. The second stage was to analyze the 1079 articles by comparing the collected keywords to each keyword of each research article. There are 392 identified leaving 687 articles unidentified. The third stage was to analyze the 687 articles by comparing the collected keywords to each abstract of each research article. There are 354 identified articles and 333 articles are unidentified. After three stage analysis Fig. 1. Research numb Fig. 2. The number o the totals are 1503 articles identified and 333 articles unidentified. The identified rate is 81.86%. Fig. 2 lists the number of 11 catego- ries published during the period 1995–2008. The most popular methodology is fuzzy ESs, followed by knowledge-based systems and neural network. From 1990, fuzzy became more and more an important issue worldwide because of the uncertainty in problem solving. It is not surprising that the fuzzy systems were of most concern to the expert and intelligence research community. 3.4.1. Trends in research methodologies To see the trends in research methodologies, we display 11 cat- egories year by year during the period 1995–2008. Fig. 3 lists the trends in research methodologies. The results show that fuzzy ESs started high in the mid 1990s, went down in 2001 and reached the top in 2008. The neural network started to grow quickly from 2003 and reached second in 2008. The knowledge-based systems are constantly in the high rate of publications and rise up to rank third in 2008. The rank fourth is rule-based systems. Other meth- odologies always show a low rate in the trend analysis. 3.5. Leading research profiles The profile of leading research provides important information for researchers to realize what leading authors researched. Junior researchers learn much from publications of leading authors and find some niches to continue their careers. Two analyses were made of issues and methodologies researched by leading authors. Table 3 lists the researched issues by top 11 authors. Table 4 pre- sents the preferred research methodologies by the top 11 authors. ers by AIS region. f 11 categories. Fig. 3. Trends in research methodologies. Table 3 Researched issue by top 11 authors. Name Affiliation Issue researched Ingoo Han Korea Advanced Institute of Science and Technology Action mechanism, virtual community recommender, information noise CRM, bankruptcy prediction, stock price index credit risk, mobile, recommendation, IDS behavior of physicians, activity-based costing prediction of interest rate, maintaining system bond rating, exchange-rate forecasting impact of measurement scale Tzung-Pei Hong National University of Kaohsiung Efficient sanitization, FP tree, bit-based feature selection, machine learning, data mining, parallelized indexing, classification, knowledge-integration, medical Elif Derya Ubeyli TOBB Ekonomi ve Teknoloji Üniversitesi Medical Dirk Van den Poel Ghent University CRM, marketing Salih Gune Selcuk University Medical, medical DSS So Young Sohn Yonsei University Mobile service, CRM, finance, dynamic scoring government fund, bipartite scorecard, selecting air force pilot trainee, commercialization, credit operations Shyi-Ming Chen National Taiwan University of Science and Technology Students’ evaluation, fuzzy system, adapting learning systems, fuzzy risk, temperature prediction, document retrieval, text categorization estimating null values Kemal Polat Selcuk University Medical, medical DSS Yueh-Min Huang National Cheng Kung University Net work, E-learning, adaptive learning, multiprocessor system, proportionate flow shop mining interesting patterns, intelligent, human- expert forum system, job-scheduling Sang Chan Park Korea Advanced Institute of Science and Technology Minimizing information gap, visualization of patent analysis, personalization CRM, electronic commerce, process control system evaluation, integrated yield management in semiconductor manufacturing, database marketing Inan Guler Gazi University Image restoration, medical 4002 W.-L. Shiau / Expert Systems with Applications 38 (2011) 3999–4005 Table 4 Preferred research methodology by top 11 authors. Name Affiliation Employed research methodology Ingoo Han Korea Advanced Institute of Scienceand Technology Rule-based Knowledge-based Neural network Fuzzy ESs Object-oriented Cased-based System architecture Intelligent agent system Modeling Ontology Tzung-Pei Hong National University of Kaohsiung Rule-based Knowledge-based Neural network Fuzzy ESs Object-oriented Cased-based Intelligent agent system Ontology Elif Derya Ubeyli TOBB Ekonomi ve Teknoloji Üniversitesi Rule-based Knowledge-based Neural network Fuzzy ESs Cased-based System architecture Intelligent agent system Database Modeling Dirk Van den Poel Ghent University Knowledge-based Neural network Fuzzy ESs Ontology Salih Gune Selcuk University Rule-based Knowledge-based Neural network Fuzzy ESs Cased-based System architecture Database Modeling So Young Sohn Yonsei University Knowledge-based Neural network Fuzzy ESs Object-oriented Shyi-Ming Chen National Taiwan University of Science and Technology Rule-based Fuzzy ESs System architecture Intelligent agent system Kemal Polat Selcuk University Rule-based Knowledge-based Neural network Fuzzy ES Cased-based System architecture Database Modeling Ontology Yueh-Min Huang National Cheng Kung University Rule-based Knowledge-based Neural network Fuzzy ESs Object-oriented Cased-based System architecture Modeling Intelligent agent system Ontology Table 4 (continued) Name Affiliation Employed research methodology Sang Chan Park Korea Advanced Institute of Scienceand Technology Rule-based Knowledge-based Neural network Fuzzy ESs Object-oriented Cased-based Intelligent agent system Modeling Ontology Inan Guler Gazi University Rule-based Knowledge-based Neural network Fuzzy ESs Cased-based System architecture Database Modeling Intelligent agent system W.-L. Shiau / Expert Systems with Applications 38 (2011) 3999–4005 4003 Most leading researchers have focused their attention on the industry or different methods. For example, professor Han focuses on financial field, professors Ubeyli and Gune focus on medical field, professor Poel focuses on marketing, and professors Hong and Chen use diverse methods to solve specific problems in a do- main. Even though top leading researchers seem to be dealing with different issues, they still focus on a few domains. When they pay more attention to a specific research stream, they are able to solve more core problems in a domain. 4. Conclusions and implications It is important for any field or any journal to do self-introspec- tion in order to realize what happened inside of a journal or a field. This paper profiles the researches that have been published in ESWA from 1995 to 2008. Some interesting results emerge by ana- lyzing the data during the period. First, ESWA is really an internationalized journal from the view- point of geography. There are no significant differences in the num- ber of publications between the three regions before the year 2001. After the year 2001 Asia and Pacific (region 3) published more arti- cles, followed by region 2 (the Europe, Middle East and African) and region 1 (America). Second, the results of the leading research universities show that Asian universities and researchers contrib- uted more to this journal, especially after the year 2001. Third, the more popular methodologies are fuzzy ESs, knowledge-based systems, and neural network. Finally, there are many researchers who develop mathematical models or specialize in analyzing prob- lems that are seen in the real world. Most productive authors deal with diverse issues. Similarly most productive authors employed diverse methodologies to solve problems because they confront different difficulties. There are several implications for researchers, journal editors, research institutions or universities. First, the analysis of this study shows that leading authors deal with different issues with diverse methods. But they still focus their interests and preferred method- ologies on certain research streams. Second, leading authors also develop sophisticated mathematical models or specialize in solv- ing problems that emerge in the real world. Upcoming researchers may apply the knowledge learned from these models to develop even better models. Third, the 11 categories of methodology in this paper are not complete. Other social science methodologies includ- ing expert system methodologies need to be taken into consider- ation, such as psychology and human behavior. Thus, prospects Appendix A (continued) Categories First keyword Second keyword Knowledge_base Medical Financ Planning 4004 W.-L. Shiau / Expert Systems with Applications 38 (2011) 3999–4005 from other social sciences methodologies are an important work in the near future. Finally, our recommendation to journal editors and research programs is to continue diverse research and to publish special issues in urgent problems, such as global recession and financial crisis, in order to meet the worldwide requirements of timely fashion. Appendix A Arranged keywords adapted from Liao (2005) Categories First keyword Second keyword Case_Based Case-base Case Base Manufac Process Knowledge management Ultrasonic Medical Appl Fault Diagn Knowledge Model e-Learning Database Geograph Traditional Chinese medicin Medical Expert Fuzzy_ESs Power load forecasting Scheduling Chemical Diagn Power Diagn Control sys Uncertain Reason Knowledge Integ Fault Diagn Power sys Fault detect Demand Wastewater treatment Data Select On-line Medical Diagn Job match Performance Index Secur Recognition Expert sys Intelligent_agent_systems Intelligent agent Tutor Sys Supply Chain System anal System desig Knowledge represent Adaptive Sys Air pollution Building Design Knowledge Internet Waste Production Decision DSS Knowledge KM Power Design Financial Tumor Business Agricult Steel composit Strateg Isokinetics Chemical Process Plant Control Concurrent Case Chip Power Trans Urban Robot Modeling Process Control Medical Software Estima Assembly Planning Project allocation Neural_networks Neural network Fault Diagn Optimal Power Decision Making Inference Diagnostic sys Machine learning Power load forecasting Process control Knowledge Learn Robot Parameter Waste Treat Biomedical Crude oil distillation Neural network Objective_oriented Object-orient Industr Diagn Knowledge Represent Knowledge Learn Knowledge Engineer Manufacturing information network Syntactic Appendix A (continued) Categories First keyword Second keyword Ontology Medical deci Knowledge Reuse Prevent Landscape assessment Knowledge Acquisition Knowledge Model Rule_based Transit Plann Product Psych Teach Advis Syst Devel Knowledge Veri Knowledge Vali Knowledge Base Scheduli Resource util Probab Diagn Sensor Tutor Knowledge represent Rule base Rule-base System_architecture System architect Material Evalu Material Select Computer aid Implement Corpora Decis Military Ferryboat W.-L. Shiau / Expert Systems with Applications 38 (2011) 3999–4005 4005 References Athey, M., & Plotnicki, J. (1992). An evaluation of research productivity in academic. Communications of the AIS, 3. Avison, D. E., Dwivedi, Y. K., Fitzgerald, G., & Powell, P. (2008). The beginnings of a new Era: Time to reflect on 17 years of the ISJ. Information Systems Journal, 15, 5–21. Chen, W., & Hirschheim, R. (2004). A paradigmatic and methodological examination of information systems research from 1991 to 2001. Information Systems Journal, 14, 197–235. Claver, E., Gonzalez, R., & Llopis, J. (2000). An analysis of research in information systems (1981–1997). Information & Management, 37, 181–195. Culnan, M. J., & Swanson, E. B. (1986). Research in management information systems, 1980–1984: Points of work and reference. MIS Quarterly, 10, 289–302. Farhoomand, A., & Drury, D. H. (1999). A historiographical examination of information systems. Communications of the AIS, 1, 1–27. Gallivan, M. J., & Benbunan-Fich, R. (2007). Analysing IS research productivity: An inclusive approach to global IS scholarship. European Journal of Information Systems, 16, 36–53. Grover, V., Lee, C. C., & Durand, D. (1993). Analyzing methodological rigor of MIS survey research from 1980–1989. Information & Management, 24, 305–317. Im, K. S., Kim, K. Y., & Kim, J. S. (1998). An assessment of individual and institutional research productivity in MIS. Decision Line, 8–11. Ives, B., Hamilton, S., & Davis, G. B. (1980). A framework for research in computer- based management information systems. Management Science, 26, 910–934. Liao, S.-H. (2005). Expert system methodologies and applications – A decade review from 1995 to 2004. Expert Systems with Applications, 28(1), 93–103. Nath, R., & Jackson, W. M. (1991). Productivity of management information systems researchers: Does Lotka’s Law apply? Information Processing & Management, 27, 203–209. Palvia, P. (2004). Research methodologies in MIS: An update. Communications of the Association for Information Systems, 14. Palvia, P., Mao, E., Salam, A. F., & Soliman, K. S. (2003). Management information systems research: what’s there in a methodology. Communications of AIS, 11. Palvia, P., Pinjani, P., & Sibley, E. H. (2007). A profile of information systems research. Information & Management, 44(1), 1–11. Vessey, I., Ramesh, V., & Glass, R. L. (2002). Research in information systems: An empirical study of diversity in the discipline and its journals. Journal of Management Information Systems, 19, 129–174. Yogesh, K. D., & Jasna, K. (2008). Profile of IS research. European Journal of Information Systems, 17, 678–693. A profile of information systems research published in expert systems with applications from 1995 to 2008 Introduction Research method Results Productive authors Leading research university Research numbers by AIS region Research methodologies 3.4.1 Leading research profiles Conclusions and implications Appendix A References