Search

Abstract This paper proposes a methodology for conducting importance-performance analysis (IPA) through online reviews. The methodology is composed of three stages: (1) mining useful information from online reviews, (2) estimating each attribute's performance and importance, and (3) constructing IPA plot, where the latent dirichlet allocation (LDA), the improved one-vs-one strategy based support vector machine (IOVO-SVM) and the ensemble neural network based model (ENNM) are respectively used. A case study on two five-star hotels is given, and the results obtained by the proposed methodology through online reviews are compared with those obtained by the existing methods through questionnaires (or online ratings). The results indicate that the proposed methodology can obtain effective analysis results with lower cost and shorter time since online reviews are publicly available and easily collected. The proposed methodology can give managers or market analysts one more choice for conducting IPA or serve as a preparing process of large-scale survey. Highlights • A novel methodology for conducting IPA through online reviews is proposed. • The performance of each attribute is estimated according to the sentiment strengths of online reviews. • The importance of each attribute is estimated using ENNM. • Four types of IPA can be easily conducted through online reviews. • A case study of IPA for two five-star hotels is given. [ABSTRACT FROM AUTHOR]

Service quality evaluation is vital to identifying the strengths and weaknesses of the services provided by service organizations. In practical situations, because of customers’ inherent uncertainty and hesitancy, hesitant fuzzy linguistic information is often employed by customers to assess expected and perceived services. The purpose of this paper is to propose a novel method for evaluating service quality based on the use of hesitant fuzzy linguistic term sets (HFLTS). In this method, customers’ comparative linguistic expressions concerning the expected service, perceived service, and attribute weights are transformed into HFLTS. By using a transformation formula, HFLTS are expressed as trapezoidal fuzzy numbers. Then, based on Gap 5 of the Parasuraman Zeithaml Berry service quality model, calculation formulas and a related theoretical analysis of the discrepancy degree between fuzzy perceived service and expected service are given. A fuzzy evaluation result for each dimension is determined by aggregating the discrepancy degrees of customers with respect to all attributes in the same dimension. Furthermore, a linguistic evaluation result for each dimension is achieved by calculating and comparing the similarity degrees between the fuzzy evaluation result and the predefined linguistic variables. On this basis, an overall service quality evaluation result is determined by aggregating the evaluation results for all dimensions. Finally, an example is used to illustrate the feasibility and effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

This paper investigates a new screening alternative problem in which evaluation index sets and index requirements provided by multiple departments are different. To solve the problem, screening alternative method based on the soft set theory is proposed. Taking the screening alternative problem with two departments as a case, the three rules of screening alternatives and the calculation process of screening alternatives are given. In the proposed method, the soft set on the index requirements and the alternatives that reach the index requirements for each department is first constructed. Then, evaluation index subsets concerning the acceptable level and the satisfactory level provided by each department are set up, respectively. Further, the soft sets concerning the acceptable levels and the satisfactory levels are constructed based on the evaluation index subsets. Afterwards, according to a selected screening alternative rule, the constructed soft sets are integrated using the ∧-products and the uni  −  int operators to obtain the screening alternative set satisfying the rule. In addition, the screening alternatives processes considering multiple departments and the different importance of indexes are also given, respectively. Finally, an example is used to illustrate the feasibility and effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

This study analyzes the survival status of shared and non-shared listings in the peer-to-peer accommodation market. Using a large data set from Airbnb in Beijing, we identify 8640 shared listings and 50,741 non-shared listings. We then investigate the exit event and the identity transition event for both types of listings by applying a discrete-time hazard model. Our results suggest that, for the exit event, the two types of listings show significant differences in terms of survival determinants, including response time, tourism specialization, market volume, professionalization, and Covid-19. For the identity transition event, we find that internal flow exists in the market, mainly from shared listings to non-shared listings, and this flow is influenced by certain factors (i.e., capacity, facility, rating, reviews, minimum stay, service quality, tourism specialization, market volume, platform professionalization, and Covid-19). [ABSTRACT FROM AUTHOR]

Abstract: Customer response modeling is essential for a firm to allocate the marketing resources to active customers who have potential values. With the development of social media, customer response modeling in social media plays important roles in the firms’ marketing decisions. For customer response modeling in social media, the inputs involve multiple types of data and the purposes are to identify respondents to multiple items. In this study, a multi-task multi-kernel transfer learning (MT-MKTL) method is proposed to integrate shared, task-specific and transferred features in a framework for customer response modeling in social media. A two-phase algorithm is applied to solving the MT-MKTL problem. A computational experiment is conducted on microblog data. The experimental results show that the MT-MKTL method exhibits good performance. [Copyright &y& Elsevier]

Taking one grassland and 15-, 24-, and 30 years old Mongolian pine (Pinus sylvestris var. mongolica) plantations in the southeast of Keerqin Sandy Land as test objects, this paper studied the effects of grassland afforestation on the soil chemical and biological properties, including pH value, C, N, and P concentration, inorganic N concentration, C and N mineralization rates, microbial biomass C, and enzyme activities. At the early stage of afforestation, the concentration of soil C, N, and P decreased; but with increasing stand age, they recovered gradually. Compared with those in grassland, the soil C, N, and P concentration in 24 years old plantation decreased by 29%, 34%, and 33%, respectively, but the soil C and N concentration in 30 years old plantation had less difference. Afforestation led to an increase of soil NH4+-N concentration and a decrease of soil NO3-N concentration, but had less effects on the soil N mineralization and nitrification rates. The soil C mineralization rate was decreased in the order of 24 years old plantation >30 years old plantation > grassland >15 years old plantation. Soil microbial biomass C and invertase activity decreased at the early stage of afforestation, but recovered with increasing stand age. After afforestation, soil urease activity had less change, but soil catalase activity increased gradually. Our results indicated that in Keerqin Sandy Land, grassland afforestation with Mongolian pine had significant effects on the soil chemical and biological properties, and the effects differed with stand age. [ABSTRACT FROM AUTHOR]

Abstract: We investigated the seasonal patterns of soil phosphorus (P) fractions under five vegetation types –Ulmus macrocarpa savanna, grassland, Pinus sylvestris var. mongolica plantation, Pinus tabulaeformis plantation, and Populus simonii plantation – in the southeastern Keerqin Sandy Lands of China. The measured P fractions (0–20cm depth) included: soil total P (TP), total organic and inorganic P (TPo and TPi), bicarbonate extractable organic and inorganic P (BPo and BPi), microbial biomass P (MBP), and in situ resin-adsorbed P (resin-P). Soil TP and TPo concentrations in the savanna and grassland were significantly lower in summer than in spring and autumn. However, they were relatively stable in three forest plantations. Soil labile P fractions showed a significant seasonal pattern under all vegetation types with the peak in summer, except soil MBP that was constant in the savanna and grassland and BPo that decreased over time in the savanna. This pattern of labile P fractions was attributed to a combination of seasonal climatic changes, low P availability, as well as the biological controls of soil P transformation in the study area. Litter decomposition played a key role in soil P availability. The monthly resin-P released from litter decomposition in summer was 2.6–7.4 times greater than in other seasons, and was 1.7–3.4 times of that in the 10cm depth soil. Concentrations of soil P fractions were obviously affected by vegetation type. The savanna had the highest total P and MBP concentrations and the P. tabulaeformis plantation had the highest BPi and resin-P among all vegetation types. Among forest plantations, P. simonii plantation had the highest total P and MBP. These results suggest that U. macrocarpa savanna is the best system conserving soil nutrient (particularly P) stocks and microbial activity, followed by the grassland and P. simonii plantation, while the pine plantations are the worst. [Copyright &y& Elsevier]

Abstract: The assessment of the level of experts in group-decision analysis forms an important research topic. This article proposes a factor score method (FAM) for obtaining a ranking of the assessment levels of experts in group-decision analysis. The proposed FAM can be used to obtain a ranking of the assessment levels of experts from difference preference relations, where fuzzy preference relations, utility function, preference ordering of the alternatives, multiplicative preference relations and linguistic preference relations, if any, are transformed into fuzzy preference relations through transformation technique proposed. Theorems and algorithm related to the transformation technique are developed. A numerical example is examined to illustrate the applications of the FAM and its advantages. [Copyright &y& Elsevier]

To leverage computer vision technology to improve the accuracy of tourism demand forecasting, a model based on deep learning with time series imaging is proposed. The model consists of three parts: sequence image generation, image feature extraction, and model training. In the first part, the tourism demand data are encoded into images. In the second part, the convolution and pooling layers are used to extract features from the obtained images. In the final part, the extracted features are input into long short-term memory networks. Based on historical tourism demand data, the model for forecasting future tourism demand can be obtained. The performance of the proposed model is experimentally assessed through comparing against seven benchmark models. • A novel hybrid deep learning-based model is proposed for tourism demand forecasting. • A new idea to encode tourism time series data into image data is presented. • Convolution and pooling layers are used to extract features from the encoded images. • The validity of the new model is verified through a systematic experimental analysis. [ABSTRACT FROM AUTHOR]

Understanding the asymmetric effects of attribute performance (AP) on customer satisfaction (CS) is important for the managers in the hotel industry. Although several studies concerning this issue have been conducted, the varies of asymmetric effects across different market segments have not been revealed. To this end, this study aims to explore the asymmetric effects of AP on CS with respect to different market segments, including different types of hotels, different types of travelers and travelers from different regions. Four theories, i.e., expectation-disconfirmation paradigm, three-factor theory of CS, customer delight theory and prospect theory, are adopted to explain the formation of CS from the perspective of different market segments. The penalty–reward contrast analysis (PRCA) and asymmetric impact-performance analysis (AIPA) are used to analyze 1,547,869 user-generated ratings collected form TripAdvisor posted by the travelers from 140 countries concerning 9,596 hotels from 75 capital cities around the world. The results suggest that the asymmetric effects of AP on CS may vary across different market segments, including different types of hotels, different types of travelers and travelers from different regions. In addition, the priorities of hotel attributes for each type of hotel with respect to different types of travelers and travelers from different regions are also analyzed by AIPA. The obtained results will be valuable for researchers to conduct further studies and hotel managers to formulate improvement strategies. • Whether the asymmetric effect of attribute performance (AP) on customer satisfaction (CS) varies across different types of hotels is examined. • Whether the asymmetric effect of AP on CS varies across different types of travelers is examined. • Whether the asymmetric effect of AP on CS varies across travelers from different regions is examined. • The priorities of hotel attributes for different types of hotels with respect to different types of travelers and travelers from different regions are analyzed. [ABSTRACT FROM AUTHOR]

Due to the severity of natural disasters, multiple areas can be affected simultaneously. Thus, the rescuers need to be assigned to multiple disaster areas to carry out the relief effort. In the process of the assignment of the rescuers, the rescuers usually have different professional skills and subjective preferences for the rescue tasks. Because of the differences of disaster situations, each disaster area has its rescue time requirement for the rescue tasks. This paper develops a model for the assignment of the rescuers considering multiple disaster areas in which the task fitness degrees and the time fitness degrees are introduced. Firstly, the satisfaction degree of each rescuer is calculated according to the intention list provided by each rescuer for the rescue tasks, and the competence degree of each rescuer is calculated according to the evaluation values of the rescuer with respect to the capability evaluation criteria. Then, according to satisfaction degrees and competence degrees of the rescuers, the task fitness degrees between the rescuers and the rescue tasks are obtained. Further, the time fitness degrees of the travel times are calculated based on the actual travel times and the time requirements of different disaster areas, and the matching degrees between each rescuer and the rescue tasks of different disaster areas are obtained by aggregating the task fitness degrees and the time fitness degrees. Afterwards, to maximize the sum of the matching degrees between the rescuers and the rescue tasks, an optimization model is constructed, and the optimal assignment result can be obtained by solving the model. Finally, an example on the assignment of the rescuers after earthquake is given to illustrate the use of the proposed model. • A novel rescuers assignment model considering multiple disaster areas is proposed. • The subjective preferences and competence degrees of rescuers for rescue tasks are considered comprehensively. • The task fitness degrees and the time fitness degrees are introduced. • The matching degrees between rescuers and the rescue tasks of different disaster areas are obtained. • A numerical example is used to illustrate the use of the proposed model. [ABSTRACT FROM AUTHOR]