استفاده کلان داده از سیستم های اطلاعات مدیریت ارتباط با مشتری برای تعیین مشخصات مشتری در بخش هتل Using big data from Customer Relationship Management information systems to determine the client profile in the hotel sector

1. Introduction Customer knowledge is vital for the hospitality industry, and it plays a crucial role in improving the offer with better quality services (i.e., more adapted and customized), the relationship with customers, and the approach of marketing strategies (Adomavicius & Tuzhilin, 2001; Min, Min & Emam, 2002). All of them result in better customer satisfaction that increases the loyalty and ensures repeating customers, as well as higher profitability (Tseng & Wu, 2014). Over the last several years, this information has been mainly managed in many hotels by proactively gathering and recording customer preferences into the socalled Customer Relationship Management (CRM) systems (Sarmaniotis, Assimakopoulos, & Papaioannou, 2013). CRMs have become a key strategy for improving customer satisfaction and retention, especially in hotels (Padilla-Meléndez & Garrido-Moreno, 2013), and they are remarkably beneficial to those organizations by generating large amounts of valuable information about their customers (Chadha, 2015; Kotler, 2002; Nguyen, Sherif, & Newby, 2007). Nevertheless, it has been recently pointed out (Dursun & Caber, 2016) that even advanced analysis techniques, such as data mining, are not yet being adequately used in the hotel industry for the purpose of effectively profiling the customers by using the comprehensive data that are routinely collected with hotel CRM systems. A large amount of information is available nowadays in hotel companies, either internal and structured (from the Property Management and the CRM systems), or external and unstructured (such as opinion platforms, social networks, or geolocalization, among many others). This brings the need to consider powerful tools available from Big Data technologies, which have already been successfully used in other fields such as bioinformatics, healthcare, or finance (George, Haas, & Pentland, 2014), to name just a few. Big Data technologies are providing unprecedented opportunities for statistical inference on massive analysis, but they also bring new challenges to be addressed, especially when compared to the analysis of carefully collected smaller data sets. In Sivarajah, Kamal, Irani, and Weerakkody (2017), a systematic and illustrative review is presented on the state-of-art analysis of the literature on Big Data techniques and Big Data Analytics, which highlights the key challenges in terms of different data types, data processing, and data management. As pointed therein, descriptive statistics are the simplest form of Big Data analytic methods, and they involve the summarization and description of knowledge and patterns by using simple statistical tests, such as mean, median, mode, variance, or proportions. When scrutinizing the usefulness of Big Data technologies in a new application field, it is necessary to establish well the behavior and scope of basic statistics, before going into more sophisticated analytics such as data mining or advanced machine learning.