We caught up with Dr Chong to learn more about the research paper.
Q. Why did you choose to this topic for your research?
Dr Chong: Effective promotions of destinations, attractions and services require careful consideration of the target markets. Clustering techniques have sought to identify target markets but are widely criticised for the biases they induce. We found that Persistent Homology (PH) can identify key tourist groupings with similar behaviours without the prejudice of the functional forms inherent in most regression models. It further produces more focused, and therefore easily promoted to, markets. Consequently, PH can highlight obtainable promotion opportunities that otherwise would be missed. This research provides an example of its application to identify the highest, and lowest, spenders amongst tourists visiting the United Kingdom. We further provide an intuitive theoretical background highlighting the inherent value of the methodology for tourism research. Potential for impact in applications to other aspects of tourism practice is also great as we signpost therein.
Q. Why is Persistent Homology considered to be effective for customer survey data commonplace in tourism?
Dr Chong: It is the ability of topographical data analysis techniques like PH to break down the complexity of data structure without being reliant on imposed relationships which offers most to developing theory in tourism management. PH offers a laboratory in which data changes may be trialled and evaluated by perturbations to the cloud. Such a test bed is a valuable tool for testing policy or business strategy constructs. Unique filtration processes ensure that any correspondences identified have the requisite robustness for inductive research. Our example uses simplified expenditure data to show patterns therein that indicate a lack of adherence to established simple relationships.
Q. Does Persistent Homology help in marketing tourism?
Dr Chong: Practical value in the application of PH can help practitioners/government implement better marketing activities/policies which can effectively eliminate factors that could hinder tourist activities, and which can encourage tourism expenditure. In addition, the more sophisticated representation of spending groupings can provide a more accurate assessment of the tourist market. This assessment can assist in the design and implementation of marketing strategies for both practitioners and government. Our results exploit to date uncharted, intrinsic topological features about tourist spending patterns helping to formulate better marketing strategies.