Research Fronts in Data Envelopment Analysis
1. Introduction
Data envelopment analysis (DEA) is a methodology for performance evaluation and benchmarking
where multiple performance measures are present [1, 2]. Charnes, et al. [3] in 1978 establish an
efficient frontier formed by the best performing decision making units (DMUs) and assign an
efficiency index to each non-frontier units according to their distances to the efficient frontier.
After more than thirty-five years of development, research activities relating to DEA are still
growing at a very fast rate. Our search in the Web of Science (WOS) database indicates that during
the period 2010 to 2014, around 2,000 more DEA-related papers have been published, which is in
addition to the 4,500 existing collections prior to 2010 as reported by Liu, et al. [4].
Amidst such a large amount of literature, it is getting more and more difficult to comprehend the
development of the field without the guidance of survey type studies. Several previous studies have
reviewed the general DEA literature. They include, in the sequence of their publication year,
Seiford and Thrall [5] (1990), Seiford [6] (1996), Gattoufi, et al. [7] (2004), Gattoufi, et al. [8]
(2004), Cooper, et al. [9] (2007), Emrouznejad, et al. [10] (2008), Cook and Seiford [11] (2009),
Liu, et al. [4] (2013), Liu, et al. [12] (2013), etc. Seiford and Thrall [5] review DEA development
when the field was in its early stage. Seiford [6] and Gattoufi, et al. [7] extensively list the DEA
literature up to the time when their articles was published. Cooper, et al. [9] review DEA models
and measures from a theoretical perspective. Gattoufi, et al. [8] and Emrouznejad, et al. [10]
conduct bibliometric style surveys and present the DEA publication statistics. Cook and Seiford
[11] present a comprehensive review on the methodological developments since 1978. Liu, et al. [4]
(2013) and Liu, et al. [12] (2013) survey the general DEA literature and examine how DEA is
applied through a citation-based methodology.
There is also no lack of survey studies in specific DEA subfields. Studies that survey methods to
further discriminate DEA results include Adler, et al. [13], Angulo-Meza and Lins [14], and
Hosseinzadeh Lotfi, et al. [15]. Zhu [16] and Hatami-Marbini, et al. [17] review the approaches to
handle imprecise and fuzzy data in DEA. Zhou, et al. [18] and Song, et al. [19] survey energy and
environmental studies that apply DEA. For financial institution and banking applications, Berger
and Humphrey [20], Fethi and Pasiouras [21], and Paradi and Zhu [22] conduct comprehensive
surveys. Recently, Cook, et al. [23], Castelli, et al. [24], and Kao [25] go over the developments in
network DEA and two-stage process model.
