1、 1 外文翻译 原文 An Optimized System Dynamics Approach for a Hotel Chain Management Material source: Hospitality Management Author:Valerio Lacagnina and Davide Provenzano 3.1 Introduction The proposed model consists of an integrated system dynamics-data envelopment analysis approach to value, in a dynamic
2、 framework, the effects over time of the policies implemented according to the relative efciency analysis. Roomsprice and competing facilities (the hedonics) are the decision variables to move in order to push the hotels towards a higher relative efciency at the end of the observation periods. In fa
3、ct, in competitive markets as tourism, hotels compete for money offering differentiated quality. Moreover, according to the microeconomic theory, a producer of differentiated goods is not a price taker but a price maker. Therefore, we assume that the decision maker of the hotel chain can freely set
4、the rooms price and the hedonics that will increase the relative economic efciency of all the hotels of the chain. The proposed model treats the rooms pricing and the hedonic setting problem in an environment characterized by uncertainty of the customers preferences. The relative efciency analysis i
5、s carried out by making use of data envelopment analysis that identies the peer group and targets for the inefcient units. The dynamic analysis of the effects over time of the policies implemented is carried out using system dynamics methodology. This combined approach will help the decision maker i
6、n answering the following questions: which hotels of the chain will be attractive, and which ones will be efcient? What adjustments on prices and hedonics will attract more tourism demand? What are the dynamic effects of the DEA policies? The remaining sections of this paper are organized as follows
7、. Section 3.2 is devoted to a brief survey of the theoretical background with particular attention to data envelopment analysis and system dynamics. Section 3.3 describes our model 2 both from the customer side and the hotel management side. Section 3.4 shows the parameters of the model and the resu
8、lts of the conducted experiments. Finally, Sect. 3.5 concludes the paper. 3.2 State of the Art and Theoretical Background The analysis of the hotels efciency up to now represents a neglected area. There is just a small number of studies among which the earliest ones make use of ratios to evaluate th
9、e performance of the lodging industry (Baker and Riley 1994), or use the break even analysis to discern the effectiveness of tourism management (Wijeysinghe 1993) or the yield management (Brotherton and Mooney 1992; Donaghy et al. 1995) to enhance hotels protability. In the efciency context, DEA tec
10、hniques have been applied by Anderson et al. (1999), Hwang and Chang (2003), Barros (2004, 2005), and Barros and Mascarenhas (2005). System dynamics methodology is also very new in this research eld. The only recent example we know is Georgantas (2003) who uses a system dynamics simulation model to
11、test Cyprus hotel value chain and protability. An example of integration of system dynamics with an economic model is illustrated by Smith and van Ackere (2002). Two papers, Png, (1988) and Scott et al. (1995) share a common theme with our study: the optimal hotel pricing cum capacity-utilization pr
12、oblem within an uncertain demand environment. They both deduce that optimal strategy leaves some unused capacity. 3.2.1 Data Envelopment Analysis Data envelopment analysis is a linear programming based technique for measuring the relative performance of organizational units where the presence of mul
13、tiple inputs and outputs makes comparisons difcult. Indeed, when the activity of measuring and comparing the efciency of relatively homogeneous units (authority departments, schools, hospitals, shops, banks, and so on) involves multiple inputs and outputs related to different resources, activities a
14、nd environmental factors, the usual measure of efciency is often inadequate. Output efciency = input In fact, the initial assumption is that the measure of technical efciency requires a common set of weights to be applied across all decision making units 3 (DMUs). This immediately raises the problem
15、 of how such a set of weights can be obtained. Such a problem was addressed the rst time by Farrell (1957) and developed by Farrell and Fieldhouse (1962), focusing on the construction of a hypothetical efcient unit, as a weighted average of efcient units, to act as a comparator for an inefcient one.
16、 Later, Charnes et al. (1978) recognized the legitimacy of the proposal that each unit might value inputs and outputs differently and therefore it should be allowed to adopt a set of weights which shows it in the most favourable light in comparison to the other units. Therefore, by making a generali
17、zation of the Farrells technical efciency measure, Charnes, Cooper, and Rhodes first introduced the data envelopment analysis to describe what is a mathematical programming approach to the construction of production frontiers and the measurement of efciency of developed frontiers. In their general-p
18、urpose DEA they assumes constant return-to-scale (CRS) and an input orientation. A few years later Banker et al. (1984) developed a variable returns-to-scale (VRS) model. Other models are the additive model (Charnes et al. 1985), the multiplicative model (Charnes et al. 1982), the cone-ratio DEA mod
19、el (Charnes et al.1990, the assurance region DEA model (Thompson et al. 1990), and Thompsonet al. 1986), and the super-efciency model (Anderson and Peterson 1993). Since 1978 over 4000 articles, books and dissertation have been published and DEA has rapidly extended to dummy or categorical variables
20、, discretionary and non-discretionary variables, incorporating value judgments, longitudinal analysis, weight restrictions, stochastic DEA, non-parametric Malmquist indices, technical change in DEA and many other topics. Up to now the DEA measure has been used to evaluate and compare educationaldepa
21、rtments (schools, colleges and universities), health care (hospitals, clinics), prisons, agricultural production, banking, armed forces, sports, market research, transportation (highway maintenance), courts, benchmarking, index number construction and many other applications. One of the main charact
22、eristics of DEA is its exibility in the choice of weights for the different inputs and outputs. Therefore, DEA may be appropriate where units can properly value inputs or outputs differently, or where there is a high uncertainty or disagreement over the value of some inputs or outputs. The heart of
23、the analysislies in nding the “best” virtual producer (a single peer or a peer group) for 4 each real producer (virtual because this producer does not necessarily exist and two or more DMUs can be combined to form a composite producer). If the virtual producer is better than the original producer by
24、 either making more output with the same input or making the same output with less input then the original producer is inefcient.By identifying the efciency score of each DMU in the sample, the slack variables in inputs and outputs of the inefcient DMUs and the peer group of efcient ones, DEA is one
25、 of the most promising techniques for the improvement of efciency. Yet, the same characteristics that make DEA a powerful tool can also create problems. An analyst should keep these limitations in mind when choosing whether or not to use DEA. First of all, since DEA is an extreme point technique, no
26、ise (even symmetrical noise with zero mean) such as measurement error can cause signicant problems. Second, DEA is good at estimating “relative” efciency of a DMU but it converges very slowly to “absolute” efciency. In other words, it can tell you how well you are doing compared to your peers but no
27、t compared to a “theoretical maximum”. Thirdly, DEA is a nonparametric technique and, therefore, statistical hypothesis tests are difcult. Finally, since a standard formulation of DEA creates a separate linear program for each DMU, large problems can be computationally intensive. More detailed revie
28、ws of the methodology are presented by Seiford and Thrall (1990), Ali and Seiford (1993), Lovell (1994), Charnes et al (1994), Seiford (1988, 1996), Thanassoulis and Dyson (1988), Dyson and Thanassoulis (1988), and Thanassoulis et al. (1987). 译文 应用于连锁酒店管理的一种优化的系统动力学方法 资料来源： Hospitality Management 作者
29、： Valerio Lacagnina and Davide Provenzano 3.1 介绍 该模型包括一个集成的系统动力学数据包络分析方法，根据相关效率分析在一个动态的价值框架内，超过政策实施时间的影响。为了在观察期结束时将酒店推向一个更高的相对效率的酒店。客房价格和竞争性设施是决定价值的变量因素。 事实上，在旅游业竞争激烈的市场，酒店为了赚钱而提供有区别的产品质5 量。此外，根据微观经济学理论，商品生产的区别不是一个价格接受者，而是价格的制定者。因此，我们假定连锁酒店的决策者能够自由设定价格和将增加的竞争设备，那么，就能提升所有连锁酒店的相关经济效益。 这个提议的模型认为客房的价格和享
30、 乐的设置问题具有客户偏好的不确定性这个环境特点。 通过使用数据环境分析，由相关效率进行分析得出为低效率单位的同水平团体和目标。由对超过实施政策时间所产生的影响所进行的动态分析得出运用系统动力学方法。 这种结合的方法将有助于决策者在回答下列问题：其中的连锁酒店将成为有吸引力的，哪些是有效的？什么价格和设施调整将吸引更多的旅游需求呢？什么是数据包络分析政策的动态效果？ 本文的其余部分组织如下： 3.2节是专门用来与特别注意数据包络分析和系统动力学的理论背景简要调查。 3.3节是从客户方和酒店管理方面介绍我们的模型。 3.4节是该模型的参数和所进行的实验结果。最后， 3.5节总结了所有观点。 3.
31、2 国家的艺术和理论背景 该酒店的效率可达分析，现在是一个被忽视的领域。现在只是研究其中最早的利用比例来评价的住宿业（贝克和赖利 1994年）的性能，或者使用盈亏平衡分析，以辨别旅游管理专业（ Wijeysinghe 1993）或效力少数产量管理（布拉泽和 Mooney 1992;。多纳吉等 1995），以提高酒店的盈利能力。在效率方面，数据包络分析技术已被由安德森（ 1999年），黄禹锡和 Chang（ 2003），巴罗斯（ 2004年， 2005年），巴罗斯和奥马斯卡雷（ 2005年）等人所应用。 系统动力学的方法在这一研究领域也是比较新的。最近的例子中，我们唯一知道的是 Georgant
32、as（ 2003）使用系统动力学仿真模型，以测试塞浦路斯酒店价值链和盈利能力。史密斯和 van Ackere（ 2002年）利用集成的系统动力学模型的一个例子也被证实了。 巴布亚新几内亚，（ 1988年）和斯科特等人的两篇论文，（ 1995）共享了与我们研究的共同主题：最佳酒店价格暨能力利用率在一个不确定的需求环境问题，他们都能用最佳的策略推断出留下一些未使用的容量。 3.2.1 数据包络分析 数据包络分析是一种存在多个输入和输出的存在使得组织单位相比困难而设计出的测了组织单元相关业绩的线性规划的基础技术。的确，在衡量和比较6 效益相对均匀单位活动（权威部门，学校，医院，银行，商店等）涉及多个
33、不同的投入和相关资源，活动和环境因素，通常衡量产出效率是不足的。 输 出 效率 = 输 入 事实上 ,初步设想它是技术效率的措施，需要一个共同的重量被应用在 所有包含决策单元。这立即引起了这样的权重如何设置可以得到的问题。这样的问题是解决是由（ 1957年）法瑞尔和菲尔德豪斯（ 1962年）首次开发的，在一个假设的高效机组对于低效为重点建设单位作为一种有效的加权平均。后来，Charnes（ 1978年）等认识到该建议的合法性，每个单位的投入与产出可能有所不同，因此，应允许采取的权数，显示在最有利的光线与其他单位比起来。 因此，通过对法瑞尔技术效率措施的概括， Charnes、库珀和罗兹首先介绍
34、了数据包络分析法来描述什么是数学规划的方法来对生产领域的建设和发展领域进行测量 。他们在通用数据包络分析评估出固定回报至量表（ CRS）的和输入方向。后来 Banker（ 1984）开发了一个变量返回到量表（ VRS）模型。其他型号是加模型（ Charnes， 1985年），乘法模型（ Charnes， 1982年），锥比率 DEA模型（ Charnes， 1990），保证地区的 DEA模型（汤普森等人， 1990）和 Thompsonet（ 1986年），以及超效率模型（安德森和 Peterson1993）。 自 1978年以来已经出版的 4000多篇文章、书籍和论文和 DEA已经迅速扩展到
35、虚拟或分类变量，任意性和非任 意变量，结合价值判断，纵向分析，权重限制，随机数据包络分析，非参数指数，技术改变 DEA和许多其他议题。 到现在， DEA的措施已经被用来评价和比较教育（学校，学院和大学），医疗（医院，诊所），监狱，农业生产，银行，军队，体育，市场调查，运输（公路养护），法院，基准指数的构建以及许多其他应用。 对 DEA的主要特点之一是它在为不同的输入和输出的权重选择的灵活性。因此， DEA可以适用于简单的价值不同的输入或输出，或有很高的不确定性或对一些输入或输出值的分歧的地方。在寻找每一个真正的生产者（虚拟的，因为这名制片人，并 不一定存在两个或更多的内燃动车组可结合形成复合生
36、产者）的 “最佳 “虚拟生产商（或一个单一的对等体组）。如果具有相同的输入输出或使原有的少的虚拟生产比原来的生产或制造更多更好的，那么生产者通过确定样本中每个决策单元的效率得分，松弛变量在相同的输出输入投入和效率差和有效率的对等组的产出， DEA是为改善效率最有前途的技术之一。 然而，相同的特性使 DEA成为一个强大的工具，同时也产生问题。一位分析师应该记住这些限制时，选择是否或不使用 DEA方法。首先，由于 DEA是一个极端的点技术，噪声等测量误差（甚至为零均值对称噪声 ）会造成重大7 问题。二， DEA是在估算 “亲属 ”的决策单元的效率非常好，但慢慢地收敛到 “绝对 “的效率。换句话说，它可以告诉你如何以及你正在做的比你的同龄人相比，但不是 “理论最大 “。第三， DEA是一种非参数方法，因此，统计假设检验的困难。最后，由于标准制定的 DEA为各 DMU创建一个单独的线性规划问题，大问题都可以密集计算。 更详细的观点是由 Seiford和萨尔（ 1990年），阿里和 Seiford（ 1993），洛威尔（ 1994年）， Charnes（ 1994年）， Seiford（ 1988年， 1996年）， Thanassoulis和戴森（ 1988），戴森和 Thanassoulis（ 1988）， Thanassoulis（ 1987年）提出的。
