Kin-Nam Lau

Text Mining for the Hotel Industry

With the availability of huge volumes of text-based information freely available on the Internet, text mining can be used by hoteliers to develop competitive and strategic intelligence. Although the software needed to analyze online text files remains comparatively expensive, the cost will inevitably fall. A demonstration of text mining compiled information about Hong Kong hotels and about would-be travelers. Such relatively invariant information as statistics on competitors’ facilities is fairly easy to assemble, but variant information, such as room rates, can be elusive. Customers’ demographics and attitudes can be mined with reasonable accuracy from newsgroup postings and the like.

Web-site marketing for the travel-and-tourism industry.

Abstract Marketers may learn much about their targets from examining personal web sites by means of automated web-crawling tools usually called bots.

A mathematical programming approach to clusterwise regression model and its extensions

Abstract The clusterwise regression model is used to perform cluster analysis within a regression framework. While the traditional regression model assumes the regression coefficient (β) to be identical for all subjects in the sample, the clusterwise regression model allows β to vary with subjects of different clusters. Since the cluster membership is unknown, the estimation of the clusterwise regression is a tough combinatorial optimization problem. In this research, we propose a “Generalized Clusterwise Regression Model” which is formulated as a mathematical programming (MP) problem. A nonlinear programming procedure (with linear constraints) is proposed to solve the combinatorial problem and to estimate the cluster membership and β simultaneously. Moreover, by integrating the cluster analysis with the discriminant analysis, a clusterwise discriminant model is developed to incorporate parameter heterogeneity into the traditional discriminant analysis. The cluster membership and discriminant parameters are estimated simultaneously by another nonlinear programming model.

A Modeling Approach to Evaluating Strategic Uses of Information Technology

Abstract:Traditional static benefit-cost methods were useful when evaluating transaction processing systems. Strategic benefits are more difficult to evaluate, since they involve dynamic interactions between customers, suppliers, and rivals. In an attempt to gain a competitive advantage, there is a strong incentive to be the first implementor of new technology. However, information technology (IT) costs decline over time, so there is an incentive to delay implementation. A model is developed that enables managers to evaluate this trade-off and choose the best implementation time. The model emphasizes competition between large firms in a regional (or national) market, interacting with firms in a local market. The model is illustrated with an application to the banking industry. It compares the implementation times of larger regional banks vis-a-vis smaller local banks, and shows how the banks might use technology to respond to various changes in the banking industry.

Estimating the city-block two-dimensional scaling model with simulated annealing

Abstract Two-dimensional Scaling is a technique to represent dissimilarities among n objects in a two-dimensional space so that the interpoint distances can best approximate the observed dissimilarities between pairs of objects. The coordinates are found by minimizing the STRESS function. It is well known that the number of local minima of the STRESS function increase with n. In this paper, we present a new approach for finding the global minimum of the STRESS function for the city-block two-dimensional scaling model. The proposed method consists of two stages. While the least square regression is used to obtain the local minimum of the STRESS function in stage 1, simulated annealing is applied to search for the global minimum in stage 2. Real and simulated examples (n=30, 50, 70) are used to assess the performance of the proposed algorithm. Results show that the coordinates can be quite accurately recovered by the proposed method.

Economic freedom ranking of 161 countries in year 2000: a minimum disagreement approach

Since 1995, the Heritage Foundation has published the economic freedom index for countries throughout the world. The fundamental challenge is how to objectively measure economic freedom so that the index can be fair, credible and has the least disagreements among all countries. In this paper, we propose a new approach to construct economic freedom ranking which minimizes disagreements among countries. The mathematical model consists of two stages: (1) calculate the set of weights for each country which would give each country the best ranking, and (2) aggregation of ranks from all countries to minimize their disagreements. We apply the model to the data set of year 2000 on Economic Freedom from the Heritage Foundation, and find that the minimum disagreements from the proposed ranking improves by 9%.

Combining ordinal forecasts with an application in a financial market

The literature on combining forecasts has almost exclusively focused on combining point forecasts. The issues and methods of combining ordinal forecasts have not yet been fully explored, even though ordinal forecasting has many practical applications in business and social research. In this paper, we consider the case of forecasting the movement of the stock market which has three possible states (bullish, bearish and sluggish). Given the sample of states predicted by different forecasters, several statistical and operation research methods can be applied to determine the optimal weight assigned to each forecaster in combining the ordinal forecasts. The performance of these methods is examined using Hong Kong stock market forecasting data, and their accuracies are found to be better than the consensus method and individual forecasts.

Stochastic preference modeling within a switching regression framework

The multinominal logit framework is extended by allowing the deterministic utility to be a switching function of the explanatory variable. This can be regarded as a piecewise linear approximation to the unknown specification of the deterministic utility or as a model to incorporate the threshold effect into the consumer behavior. A new estimation procedure is proposed to estimate the utility parameters and the threshold points simultaneously. This new procedure can be formulated as a linear integer programming model solvable by the standard mathematical programming package and the resulting estimates are maximum likelihood estimates. The proposed framework can be applied to the modeling and the estimation of reservation price and reference price.

Market share modeling within a switching regression framework

Marketers often need to analyze and predict market share for a brand or a firm. These predictions can be based on product attributes and marketing variables. Two analytical techniques have been shown to be better at this task than other models: Multinomial Logit (MNL) and the Multiplicative Competitive Interaction (MCI) model. This study shows that the MNL and MCI models can be derived from a common foundation. It also extends the MNL model by integrating it with switching regression techniques. This new model, switching multinomial logit (SMNL), provides several theoretical advantages over both the basic MCI and MNL models. With a sample data set, the SMNL model outperforms the simple MNL model with a 10% reduction in errors.