2020/02/19 by Sujoy Chatterjee, Chatterjee, Sujoy, Nicolas Pasquier +6
Business, Management and Accounting · Computer Science · Engineering · Mathematics · Social Sciences · #Artificial Intelligence (cs.AI) #Artificial intelligence #CURE data clustering algorithm #Cluster analysis #Computer science #Consensus clustering #Correlation clustering #Customer Service Quality and Loyalty #Data mining #Digital Marketing and Social Media #Engineering #FOS: Computer and information sciences #Information Retrieval (cs.IR) #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine learning #Market segmentation #Metric (unit) #Transportation Planning and Optimization #cs.AI #cs.IR #cs.LG #stat.ML
paper · pdf · doi:10.48550/arxiv.2002.10241
openalex publication_date 2020/02/19 · arxiv created 2020/02/26 · arxiv updated 2020/02/27 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
In the travel industry, online customers book their travel itinerary\naccording to several features, like cost and duration of the travel or the\nquality of amenities. To provide personalized recommendations for travel\nsearches, an appropriate segmentation of customers is required. Clustering\nensemble approaches were developed to overcome well-known problems of classical\nclustering approaches, that each rely on a different theoretical model and can\nthus identify in the data space only clusters corresponding to this model.\nClustering ensemble approaches combine multiple clustering results, each from a\ndifferent algorithmic configuration, for generating more robust consensus\nclusters corresponding to agreements between initial clusters. We present a new\nclustering ensemble multi-objective optimization-based framework developed for\nanalyzing Amadeus customer search data and improve personalized\nrecommendations. This framework optimizes diversity in the clustering ensemble\nsearch space and automatically determines an appropriate number of clusters\nwithout requiring user's input. Experimental results compare the efficiency of\nthis approach with other existing approaches on Amadeus customer search data in\nterms of internal (Adjusted Rand Index) and external (Amadeus business metric)\nvalidations.\n