Search

The rapid proliferation of social networking sites (SNSs) provides marketers with ample opportunities to explore advertising strategies based on electronic word-of-mouth (eWOM). With that in mind, this study aims to understand the influence of the eWOM diffusion process on SNSs. We propose a theoretical framework to investigate the roles of curiosity and influence in SNS users’ decisions to adopt eWOM and to pass it along to their contact groups. Two forms of curiosity, specific and diversive, are proposed to initiate specific exploration (consistency checking and knowledge-based validation) and diversive exploration (referral visit behavior), and eventually, adoption and pass-along behavior. In addition, SNS influence, formed by maven, persuasiveness, and connectivity, is proposed as a moderator affecting information diffusion. The proposed framework is empirically validated using the data collected from online surveys of 590 Facebook users. Implications for theory and practice are also discussed.

Quality contracting is critical and challenging due to the many unique issues related to quality. In this study, we analyze the first-mover right in quality contracting by considering two different strategies for the buyer: the quality requirement strategy (QR) where buyer moves first by posting quality requirement to suppliers and quality promise strategy (QP) where buyer voluntarily gives up the first-mover right to suppliers to ask them to promise quality. We study which strategy (1) better encourages suppliers' quality improvement efforts and (2) leads to a higher expected profit for the buyer. To analyze the drivers behind the buyer's choice between QR and QP, we start with the basic model where buyer faces only one supplier who has the opportunity to make quality improvements. We then gradually add other business features such as information asymmetry and supplier competition, analyzing how each feature adds/changes the driving forces and how they interact in the buyer's decision between QR and QP. We consider both the case where the wholesale price is fixed (when the buyer has the power to dictate price or price is set by the market) and the case where the wholesale price is included as a variable (when price is part of the negotiation). We find that QP always leads to the first-best quality efforts from the supplier(s) while QR limits their efforts. However, this does not guarantee higher expected profit for the buyer under QP. We provide insightful guidelines in buyer's choice between QP and QR. This research enriches the limited literature on quality contracting with quality improvement opportunity and asymmetric information

This article studies the effects of learning and forgetting on the design of maintenance outsourcing contracts. Consider a situation in which a manufacturer offers an outsourcing contract to an external contractor to maintain a manufacturing process. Under the contract, the contractor schedules and performs preventive maintenance and repairs the process whenever a breakdown occurs. Two types of learning effects on the cost and time of performing preventive maintenance are considered: learning from experience (natural) and learning by a costly effort/investment. It is assumed that forgetting occurs under each learning type. A model is developed for designing an optimal outsourcing contract to maximize the manufacturer's profit. An extensive numerical analysis is carried out to empirically demonstrate the effects of learning and forgetting on the optimal maintenance contract and the manufacturer's profit.

In traditional decision (classification) tree algorithms, the label is assumed to be a categorical (class) variable. When the label is a continuous variable in the data, two possible approaches based on existing decision tree algorithms can be used to handle the situations. The first uses a data discretization method in the preprocessing stage to convert the continuous label into a class label defined by a finite set of nonoverlapping intervals and then applies a decision tree algorithm. The second simply applies a regression tree algorithm, using the continuous label directly. These approaches have their own drawbacks. We propose an algorithm that dynamically discretizes the continuous label at each node during the tree induction process. Extensive experiments show that the proposed method outperforms the preprocessing approach, the regression tree approach, and several nontree-based algorithms.

In response to the thriving development in electronic commerce (EC), many on-line retailers have developed Web-based information systems to handle enormous amounts of transactions on the Internet. These systems can automatically capture data on the browsing histories and purchasing records of individual customers. This capability has motivated the development of data-mining applications. Sequential pattern mining (SPM) is a useful data-mining method to discover customers' purchasing patterns over time. We incorporate the recency, frequency, and monetary (RFM) concept presented in the marketing literature to define the RFM sequential pattern and develop a novel algorithm for generating all RFM sequential patterns from customers' purchasing data. Using the algorithm, we propose a pattern segmentation framework to generate valuable information on customer purchasing behavior for managerial decision-making. Extensive experiments are carried out, using synthetic datasets and a transactional dataset collected by a retail chain in Taiwan, to evaluate the proposed algorithm and empirically demonstrate the benefits of using RFM sequential patterns in analyzing customers' purchasing data.

We consider a repairable product with known market entry and departure times. A warranty policy is offered with product purchase, under which a customer can have a failed item repaired free of charge in the warranty period. It is assumed that customers are heterogeneous in their risk attitudes toward uncertain repair costs incurred after the warranty expires. The objective is to determine a joint dynamic pricing and warranty policy for the lifetime of the product, which maximizes the manufacturer’s expected profit. In the first part of the analysis, we consider a linearly decreasing price function and a constant warranty length. We first study customers’ purchase patterns under several different pricing strategies by the manufacturer and then discuss the optimal pricing and warranty strategy. In the second part, we assume that the warranty length can be altered once during the product lifetime in developing a joint pricing and warranty policy. Numerical studies show that a dynamic warranty policy can significantly outperform a fixed-length warranty policy.

An external contractor receives a payment from a manufacturer for periodically performing preventive maintenance and for performing minimal repairs whenever process failures occur. Suppose a new technology will be available in the future, but the timing is uncertain. We propose specifying one or multiple time points in a maintenance contract, at which the manufacturer can adopt (switch to) the new technology, if it becomes available, and change the preventive maintenance schedule for the remaining time in the contract period. A model is developed to determine the value of using these (switch) points in a maintenance contract to the manufacturer. The numerical results show that, with a higher arrival rate of the new technology, the value becomes greater, and thus, the optimal switch point should be set earlier. Furthermore, if the new technology reduces maintenance costs, using switch points adds more value to a cost-plus-margin method than either the fixed payment or cost-plus-fixed-fee methods. On the other hand, if adopting the new technology results in higher maintenance costs, the value using switch points is higher for the fixed payment method.

Most decision tree classifiers are designed to classify the data with categorical or Boolean class labels. Unfortunately, many practical classification problems concern data with class labels that are naturally organized as a hierarchical structure, such as test scores. In the hierarchy, the ranges in the upper levels are less specific but easier to predict, while the ranges in the lower levels are more specific but harder to predict. To build a decision tree from this kind of data, we must consider how to classify data so that the class label can be as specific as possible while also ensuring the highest possible accuracy of the prediction. To the best of our knowledge, no previous research has considered the induction of decision trees from data with hierarchical class labels. This paper proposes a novel classification algorithm for learning decision tree classifiers from data with hierarchical class labels. Empirical results show that the proposed method is efficient and effective in both prediction accuracy and prediction specificity.

Costs are often an important part of the classification process. Cost factors have been taken into consideration in many previous studies regarding decision tree models. In this study, we also consider a cost-sensitive decision tree construction problem. We assume that there are test costs that must be paid to obtain the values of the decision attribute and that a record must be classified without exceeding the spending cost threshold. Unlike previous studies, however, in which records were classified with only a single condition attribute, in this study, we are able to simultaneously classify records with multiple condition attributes. An algorithm is developed to build a cost-constrained decision tree, which allows us to simultaneously classify multiple condition attributes. The experimental results show that our algorithm satisfactorily handles data with multiple condition attributes under different cost constraints.

We studied time-based policies on pricing and leadtime for a build-to-order and direct sales manufacturer. It is assumed that the utility of the product varies among potential customers and decreases over time, and that a potential customer will place an order if his or her utility is higher than the manufacturer's posted price. Once an order is placed, it will be delivered to the customer after a length of time called “leadtime.” Because of the decrease in a customer's utility during leadtime, a customer will cancel the order if the utility falls below the ordering price before the order is received. The manufacturer may choose to offer discounted prices to customers who would otherwise cancel their orders. We discuss two price policies: common discounted price and customized discounted price. In the common discounted price policy, the manufacturer offers a single lower price to the customers; in the customized discounted price policy, the manufacturer offers the customers separately for individual new prices. Our analytical and numerical studies show that the discounted price policies results in higher revenue and that the customized discounted price policy significantly outperforms the common discounted price policy when product utility decreases rapidly. We also study two leadtime policies when production cost decreases over time. The first uses a fixed leadtime, and the second allows the leadtime to vary dynamically over time. We find that the dynamic leadtime policy significantly outperforms the fixed leadtime policy when the product cost decreases rapidly.

The objective of this paper is to study learning effects on maintenance outsourcing. We consider a situation in which a manufacturer offers a short-term outsourcing contract to an external contractor who is responsible for scheduling and performing preventive maintenance and carrying out minimal repairs when the process fails. The manufacturer’s payment to the contractor consists of a fixed amount along with cost subsidization for each maintenance operation performed. We assume learning occurs when the contractor performs preventive maintenance that reduces both time and cost. Two types of learning are considered: natural learning and learning by costly efforts. We demonstrate that a well-designed payment scheme can induce the contractor to adopt the maintenance schedule that maximizes the manufacturer’s profit.

We propose a new approach to performing market basket analysis in a multiple-store and multiple-period environment. In using the method, the user first defines a time concept hierarchy and a place (location) hierarchy, according to his or her application and needs. A set of contexts is systematically derived from the two hierarchies by combining the concept levels of the two hierarchies. We developed an efficient algorithm for extracting the association rules, which meet the support and confidence requirements for all the contexts. Using the approach, a decision maker can analyze purchasing patterns at very detailed concept levels of time and place, such as a combination of days and stores, at more general levels, such as a combination of quarters and states, and combinations of detailed levels of one with general level of the other, such as a combination of days and regions. In addition to this flexibility, the association rules are well organized, because they are generated according to the contexts derived from the time and place hierarchies. A numerical evaluation shows that the algorithm is efficient in running time and may generate more specific and richer information than the store-chain rules and the traditional rules.

Consider a manufacturer who has a process with an increasing failure rate over time. In order to improve the process performance, the following two types of maintenance activity are outsourced to an external contractor: (i) preventive maintenance is performed periodically to improve the reliability of the process when the process is functional; and (ii) corrective maintenance is used to restore the process to a specified condition when it fails. We consider the use of incentive contracts to induce the contractor to select the maintenance policy that optimizes the total profit of the manufacturer and contractor. It is demonstrated that an incentive contract based on a combination of a target uptime level and a bonus always leads to the desired win-win coordination, and provides flexibility in allocating the extra profit generated from coordination and, importantly, an incentive to the contractor to improve the efficiency of the maintenance operations. The incentive contract can also be used to select the m...

Consider a manufacturer with a manufacturing system that consists of multiple processes. The manufacturer's revenue is determined by the minimum of the uptimes among the processes. The maintenance functions of the processes are outsourced to independent contractors so that each contractor is responsible only for one process. A performance-based incentive contract is offered to each contractor, consisting of an uptime target level and a bonus rate for exceeding the uptime target. Under the incentive contract, a contractor receives a bonus only when the achieved uptime exceeds the target level specified in the contract. We develop a model for jointly determining the uptime target levels and bonus rates for the contractors which maximize system profit. We also demonstrate the financial benefits of coordination and added flexibility in allocating the additional profit to the contractors. In addition we show the impact of the variation in individual process maintenance times and costs on channel coordination a...

Market basket analysis (also known as association-rule mining) is a useful method of discovering customer purchasing patterns by extracting associations or co-occurrences from stores' transactional databases. Because the information obtained from the analysis can be used in forming marketing, sales, service, and operation strategies, it has drawn increased research interest. The existing methods, however, may fail to discover important purchasing patterns in a multi-store environment, because of an implicit assumption that products under consideration are on shelf all the time across all stores. In this paper, we propose a new method to overcome this weakness. Our empirical evaluation shows that the proposed method is computationally efficient, and that it has advantage over the traditional method when stores are diverse in size, product mix changes rapidly over time, and larger numbers of stores and periods are considered.

We consider the problem of finding the Kth shortest path for a time-schedule network, where each node in the network has a list of prespecified departure times, and departure from the node can take place only at one of these departure times. We develop a polynomial time algorithm independent of K for finding the Kth shortest path. The proposed algorithm constructs a map structure at each node in the network, using which we can directly find the Kth shortest path without having to enumerate the first K − 1 paths. Since the same map structure is used for different K values, it is not necessary to reconstruct the table for additional paths. Consequently, the algorithm is suitable for directly finding multiple shortest paths in the same network. Furthermore, the algorithm is modified slightly for enumerating the first K shortest paths and is shown to have the lowest possible time complexity under a condition that holds for most practical networks. © 2004 Wiley Periodicals, Inc. Naval Research Logistics, 2005.

A systematic method is proposed to generate time information on the paths and nodes on a time-window network for planning and selecting a path under a constraint on the latest entering time at the destination node. Specifically, three algorithms are proposed to generate six basic time characteristics of the nodes, including the earliest and latest times of arriving at, entering, and departing from each node on the network. Using the basic time characteristics, we identify inaccessible nodes that cannot be included in a feasible path and evaluate the accessible nodes’ flexibilities in the waiting time and staying time. We also propose a method for measuring adverse effects of including an arc. Finally, based on the time characteristics and the proposed analyses, we develop an algorithm that can find the most flexible path in a time-window network.

Quality experts have recognized two types of quality costs resulting from failure of control; namely, internal and external failure costs of control. Explicitly considering the difference between internal and external failure costs in designing statistical process control procedures is especially important in today’s business environment because (1) inventory holding time is strategically kept very short because of the just-in-time philosophy, and (2) product values decrease at a very fast rate over time because of advances in technology and keen competition. In this paper, we develop a process control model under a time-varying cost structure, based on internal and external failure costs. Using the model, we also study the process control policies when the value of the product may decrease over time. We consider two forms of product perishability. The first is an exponential product value drop, and the second is a fixed life product life.

In this paper, we study the efficacy of alternative process improvement strategies for a container-filling production process. Three types of improvement actions to modify process parameters are considered: reducing the process setup cost, reducing the arrival rate of the out-of-control state, and reducing the process variance. It is assumed that these process parameters can be changed with a one-time investment. The concept of a planning horizon is introduced as a means for modeling the investment decision and corresponding process improvement benefit. Models are formulated to determine the optimal process improvement and production parameters that minimize the unit time expected cost across a given planning horizon. Numerical analysis is used to examine relationships among the optimal investment strategy, production policy, and length of the planning horizon.

Selection of the mean (target value) for a container-filling process is an important decision to a producer especially when material cost is a significant portion of production cost. Because the process mean determines the process conforming rate, it affects other production decisions, including, in particular, production setup and raw material procurement policies. It is evident that these decisions should be made jointly in order to control the production, inventory and raw material costs. In this paper, we incorporate the issues associated with production setup and raw material procurement into the classical process mean problem. The product of interest is assumed to have a lower specification limit, and the items that do not conform to the specification limit are scrapped with no salvage value. The production cost of an item is a linear function of the amount of the raw material used in producing the item. A two-echelon model is formulated for a single-product production process, and an efficient algorithm is developed for finding the optimal solution.

The rapid evolution of sensor technology, using techniques such as lasers, machine vision and pattern recognition, provides the potential to greatly improve the Statistical Process Control (SPC) method for monitoring manufacturing processes. This paper studies the method of using on-line sensors to monitor manufacturing processes and compares that method with the control chart method, a widely used SPC tool. Two separate economic models are formulated for using either a sensor or a control chart to monitor a manufacturing process. Then, the two models are compared in a sensitivity analysis with lespect to several process parameters.

In the so-called two-attribute warranty policy, two types of warranty criteria, such as the age and mileage of an automobile, are employed simultaneously to determine the eligibility of a warranty claim. The authors propose in the paper several decision models that estimate the expected total cost incurred under various types of two-attribute warranty policies. They also perform a sensitivity analysis to study the effects of several model parameters, such as the discount rate, the product usage rate, and the warranty terms, on the total warranty cost.

The rapid evolution of sensor technology, using techniques such as lasers, machine vision and pattern recognition, provides the potential to greatly improve the Statistical Process Control (SPC) method for monitoring manufacturing processes. This paper studies the method of using on-line sensors to monitor manufacturing processes and compares that method with the control chart method, a widely used SPC tool. Two separate economic models are formulated for using either a sensor or a control chart to monitor a manufacturing process. Then, the two models are compared in a sensitivity analysis with respect to several process parameters.