LION-APP Conference, Trento, Italy
Applications of Machine Learning and Intelligent Optimization to Tourism and Hospitality
Summer School and Conference
Applications, business cases, software
For Students, Practitioners and Business Participants
(in alphabetical order)
In this short course I will review how hotel reservation systems work, including explanation of the reservation channels. These reservation-based data provides large amounts of information that allow us to optimize the hotel processes, especially the optimization of revenue. I will explain the analytics behind hotel revenue optimization. For successful revenue optimization, an accurate forecast of incoming reservations and occupancy is needed. I will review the major time series forecasting approaches, as applied to hotel data.
Hotel reviews on reservation sites are very informative for the guest, and guide him towards better hotel selection. Moreover, they give the hotel feedback that helps him to improve his services. Machine learning models have been introduced for automatic analysis of the reviews, leading to the so-called opinion mining concept. I will explain the concept, and provide an overview of machine learning models for opinion mining.
Tourism is a complex phenomenon and tourism systems are complex systems, which needs a more “systemic” approach to the study of the different aspects, a more realistic, organic approach, in which the system is considered to have ‘life-like’ characteristics. After initial “logical” and qualitative studies, the literature has started to produce more rigorous and quantitative analyses which we will present in this course. In particular, the field of network science has attracted most efforts, focusing on emergent properties caused by the interaction of the different actors.
Complexity and network science acknowledge a growing usage in a number of different settings. Having put the bases and achieved a recognition of its validity for the theoretical appeal as well as for the capability to provide good “practical” insights, we now face a further phase of development, since more refined and sophisticated techniques were made available in the last years by scholars of many disciplines. There is now a need to apply these to the study of tourism destinations and their components, and a need for better simulation tools for a more profound understanding of the structure and the dynamic behavior of the actors involved. I will discuss the most challenging and promising areas for research and applications.
Like most businesses, tourism and hospitality are undergoing a phase of disruptive innovation caused by the wider adoption of computers, networks and, above all, sophisticated and powerful algorithms. Algorithms automate business processes in a partial or total manner, by starting from repetitive and simple tasks but progressively reaching also more complex and “creative” tasks, traditionally associated with human decision making. The concepts of “automated creativity” or “automated business innovation” sound like contradictions. We like to think that only human people can discover truly innovative ways of solving problems and radically improving business performance. In this course we summarize two theoretical advancements in the past years which permit this disruptive innovation: machine learning and intelligent optimization.
Managers and decision makers reach decision by some level of anticipation (expectation, prediction) of the effects of different choices. These decisions are based on a series of “What if?” questions, with answers given by expertise, gut feeling, or some level of logical and mathematical modelling. Machine learning or learning from data is a theory for deriving flexible models by starting only from the data produced by the business. After a model is available, computers can simulate the effects of zillions of possible decisions, by predicting the output (for example the total profit of the hotel), and by creating and selecting one among the best decisions. Intelligent optimization is this automated process of creating in an intelligent manner a large series of possible decisions, aiming at improving the current way of doing business.
Through machine learning and intelligent, optimization hotel managers have extremely powerful tools in their pockets to improve total profitability and customer satisfaction. It is up to them to understand the new possibilities (the overall vision), decide which possible changes they are considering (e.g., acting on prices, availability of different types, reservation rules, kind of offer, etc.), collect and organize the relevant data about the past performance, deliver them to ML tools to build models and run zillions of software experiments via intelligent optimization (IO) to identify improving solutions.
In this course we will highlight some fundamental tools and use simulation-based optimization software for exercises with the participants about managing a test hotel and measuring the improvement in profitability that can be obtained by LION techniques in realistic contexts.
Professor Dimitrios Buhalis of Bournemouth University will explain that Smart Tourism revolutionizes tourism and hospitality and change market conditions and industry structure. This leads to tourism and hospitality organisations to readdress the sources of competitiveness and repositioning their strategy and operations in their marketplace. Network economics and strategies suggest that organization need to reengineer their processes to take advantage of their ecosystem. Technology has emerged as the pervasive and robust platform for the tourism organisation and destination distribution and management. The Web 2.0 and consumer generated content based social media engagement are revolutionising global tourism. New developments such as Augmented Reality provide incredible opportunities for tourism organisations to develop their competitiveness. Only tourism organisations and destinations that can take full advantage of the opportunities will be able to capitalise on the benefits in the future and enhance their competitiveness. This seminar will challenge participants to think of their use of technology and their digital foot print to maximise their visibility, engagement, conversion and loyalty. Participants will be encouraged to think of how they can cocreate tourism experiences and how they can develop benefits for all participants in the marketplace.
Smartness takes advantage of interconnectivity and interoperability of integrated technologies to reengineer processes and data in order to produce innovative services, products and procedures towards maximising value for all stakeholders. This reengineering enables shaping products, actions, processes and services in real-time, by engaging different stakeholders simultaneously to optimise the collective performance and competitiveness and generate agile solutions and value for all involved in the value system. Looking into the future this seminar will emphasise the importance ofnetwork competitiveness and how to maximise the benefits for all stakeholders. Smartness is the glue of interconnected and mutually beneficial systems and stakeholders and provides the infostructure for the value creation for all. Participants will be encouraged to consider how they can optimise their competitiveness based on optimising the performance of their networks in smart destinations and smart tourism and hospitality ecosystems.
Over the past four decades many researchers have examined various components of the tourism system. This work along with other advances in science and technology delineates four essential advances which now enable tourism planners. First, the development of a considerable body of research in a variety of disciplines and areas of application ranging from psychology, social psychology, environmental psychology, geography, landscape architecture, urban and regional planning, economics, marketing, and communications provides a reasonably comprehensive understanding of the touristic experience and the factors influencing these experiences. Second, the development of the Internet and related technologies now enables researchers to collect and analyze traveler-related data almost anywhere and in real time; this new capability affords new opportunities to understand how travelers respond to various stimuli while in situ, thereby overcoming a number of important limitations of previous methods. Third, the coalescence of the basic theories and new technologies gives rise to a new understanding of design, which argues that it (i.e., design) is not simply a property of the artefact (i.e., event or place which supports the traveler experience), but rather it is a way of thinking. As such, design thinking is a basic process driving innovation and new ways for supporting the creation of customer value, i.e., the tourism experience. Finally, the development of new, highly sophisticated systems (including the Internet of Things (IoT) and the Quantified Traveler) for seamlessly tracking and communicating with visitors enables the tourism industry to manage the visitor experience in much more personal and innovative ways.
These developments in theory, methodology, and application provide the foundation for a new paradigm which can be characterized as Design Science
in Tourism (DST) and supports a framework for designing systems and artefacts to improve travel experiences. DST is explicitly focused on the
development of new artifacts and, as such, it provides the foundation for enabling tourism managers to develop innovative processes, systems and places.
The tourism design system is comprised of six key components: (1) Themes, (2) Stories, (3) Atmospherics; (4) Affordances; (5) Co-creation; and, (6) Technology.
As illustrated, each of these components represent a specific aspect of the system which determines which sensations are received and how they are interpreted
and communicated so as to create memorable visitor experiences. Thus, one of the most important findings of this research over the past forty years is
the clear linkage between environmental stimuli, sensation, emotions and decision making and the nature of tourism experiences.
Persuasive technologies are those that elicit specific behaviors, manage to change attitudes or encourage users into forming habits by using fundamental principles of persuasion, such as social influence, scarcity or authority. Websites, recommender systems, mobile apps, robots, online games, social media platforms, etc. all rely on their persuasive capacity to keep users engaged and encourage particular responses or behaviors. This requires intricate knowledge of the social psychology of users and an understanding of how persuasion principles can be integrated into interfaces and algorithms. It also demands a basic understanding of the business models of the systems or tools to determine specific persuasion goals.
This course will provide an overview of persuasion principles and their applicability to different aspects of technology design. Specifically, it will discuss the persuasive potential of various technologies in the context of particular use scenarios and will point out specific challenges to persuasion for technology use/design in tourism and hospitality settings. It will further highlight the importance of persuasive technologies from a business perspective and debate benefits for users. It will also explore the incredible potential of persuasive technology to initiate positive behavior change (e.g. in the context of health or environmental sustainability) and will discuss ethical implications of persuasion, as the implementation of persuasive designs might lead to unwanted consequences such as technology addiction.
The hospitality industry is a highly competitive space and inventory management plays a critical part in overall hotel success. The overall objective of revenue management is not to maximize occupancy but to optimize revenue and profit. When it comes to rate setting and inventory allocation decisions, data should be at the heart of strategic decision.
In this session, I will explain the principles of optimal inventory allocation and introduce analytical tools for hotels. In particular, this session will focus on the unique optimization issues in hotels such as length of stay controls and multiplier effects. It will further highlight the limitations of current approaches used in the industry and discuss how machine learning can help to solve the issues we face today.
The next 15-20 years will witness the massive introduction of robots – both as consumer robots (including companion robots) and industrial robots as result of the advances in robotics, artificial intelligence and automation. Economists expect this with mixed feelings. While some extort the benefits artificial intelligence and robotics will bring to societies, others predict a darker scenario. The massive introduction of robots and the transition of the economic system to robonomics (robot-based economy) will cause many people to lose their jobs, new jobs would be created, production facilities will scale down and change their geographic location, and the sources of employees’, companies’ and countries’ competitive advantages will change drastically. This will have profound implications on the nature of work, level and sources of incomes, leisure time, politics, international trade and relations, ownership rights, etc., hence leading to major social, economic and political challenges and tension. Societies will be forced to find unconventional solutions to these challenges – birth right patents, universal basic income, constant and fluid free life-long education of population, robot-based tax system, redefinition of human rights, etc. In this course, Professor Stanislav Ivanov will elaborate on the economic principles and drivers of robonomics, will pinpoint its benefits and challenges, and sketch some of the solutions to its challenges. Furthermore, the course will outline the implications of robonomics for the travel, tourism and hospitality companies – robots as tourists and service providers, impacts on servicescape, marketing, operations, human resource managements, performance, competitiveness, etc.
The recent development of the Internet supports practitioners in tourism and hospitality to disseminate information, reach potential customers worldwide, and facilitate bookings through business websites. From the perspective of suppliers, a website is a worldwide distribution channel of products or service to consumers. From the perspective of consumers, the Internet enables them to access websites and make reservations anytime and anywhere. Therefore, a functional and easy-to-use website is crucial for managers to facilitate the process of reservation, thereby meeting the needs of consumers, and increasing revenue.
Website usefulness is of great importance when consumers evaluate a website. Website usefulness consists of website functionality and website usability. Website functionality refers to information provision (i.e. website contents and features), whereas website usability denotes information use and processing (i.e. design). Overall, a good website should be useful and easy to use. Easily accessible information, consistent appearance of content, and a good navigation system all contribute to the enhanced usefulness and ease-of-use of a website. Besides basic information about reservation, the information available on websites can be extended to a relational level, such as by integrating social networking sites, providing customized service and information on loyalty program, which will enhance customer relationships.
In this presentation, the historical background of the Internet will be introduced. Then, some recent research findings related to website, such as website visibility, website evaluation approaches, and website performance measurements will be illustrated with some real-life examples. After that, an overall picture of the progress and future development directions of websites will be presented. Specifically, terminologies used for website evaluation, measurements of website functionality and usability, chronological development of the mainstream website evaluation models from 1990s to present will be reviewed. The adoption of digital footprints in tourism will also be delivered in this presentation.
The ultimate goal of this presentation is to provide industry practitioners, post-graduate students, and other tourism professionals the insights on improving websites, such as exerting effort in achieving high-level information communication in order to bring more convenience and personalized service to consumers, and in the meantime, increase business revenue.
Recommender systems are software applications that attempt to reduce information overload. Their goal is to recommend items of interest to the end users based on their preferences. To achieve that, most Recommender Systems exploit the Collaborative Filtering approach. In parallel, Multiple Criteria Decision Analysis (MCDA) is a well-established field of Decision Science that aims at analyzing and modeling decision maker’s value system, in order to support him/her in the decision making process.
In this course, we will present:
The basic concepts of Multiple Criteria Decision Analysis (MCDA) and Aggregation - Disaggragation approach.
Two recommender systems. Initially, a Multicriteria Recommender Systems whose purpose is to recommend items of interest to users based on their preferences will be presented. To achieve that, most Recommender Systems apply a widely used algorithm, named the Collaborative Filtering algorithm. In parallel, Multiple Criteria Decision Analysis (MCDA) is a well-established field of Decision Support Systems that aims at analyzing and modeling a user’s value system. In this system, a hybrid framework that incorporates techniques from the field of MCDA together with the collaborative filtering algorithm is proved to enhance the performance of existing Recommender Systems. More specifically, the Disaggregation-Aggregation approach of MCDA is exploited that builds user’s value system through iterative interactive procedures, where the attributes of the problem and the user’s global judgment policy are analyzed and then aggregated into a value system. Subsequently, system’s users are clustered into groups of similar preferences and the collaborative filtering algorithm adapted to multiple attributes is applied, to successfully propose items of interest to these users. The proposed methodology improves the performance of Recommender Systems as a result of two main causes. First the creation of user profiles prior to the application of collaborative filtering algorithm and second, the integration of multiple criteria in the recommendation process. Next, we will present the methodology and results, of a new hybrid multi-criteria hotel recommendation system. The problem of hotel recommendations using multi-criteria methods, as there are many parameters that users consider important and which should be taken into account for an accurate and efficient final recommendation. Within the methodology, we combine three different methods of analysis (MUSA, Sentiment Analysis, Filtering). A variant of WAP method is also used to create a preferential user profile for the system. We end up producing personalized product recommendations to system users, which are commensurate with their preferences. Additionally, the user is able to filter the available alternatives with a selection from a set of standard criteria. The use of the minimum satisfaction threshold, that is calculated using sentiment analysis in customer reviews, guarantees the quality of the recommendations. The recommendation system uses real reviews and ratings for hotels, as well as static hotel features that have been extracted, using data mining methods, from online hotel reservation platform. Inputs of the system are user choices, based on standard criteria, as well as classification of specific criteria in order to create her preferential model. The evaluation of the recommendation system is done by measuring the accuracy of forecasting of evaluations in a real-user experiment. For the case study, we used data for hotels in the prefecture of Chania, Crete.
In this tutorial we will cover recent approaches and advances in the optimization and estimation and forecasting aspects of Dynamic Pricing and Revenue Management. The availability of massively parallel computers and theoretical advances have led to new algorithms and approaches with both practical and theoretical interest in the area of network optimization.
Likewise, new machine learning approaches and the availability of new data sources is also influencing the area of estimation and forecasting of pricing models.