6: Big Data Analytics in Predicting Tourist Preferences Based on Hotel Ratings Using Multiclass Multilabel Classification Algorithm

Multiclass multilabel classification algorithm for analyzing big data analytics in predicting tourist preferences based on hotel ratings in the growth of tourism, and it focuses primarily on behaviour of tourist travellers’ reviews.

Mirończuk and Protasiewicz (2018) have mentioned that a multiclass classification models the issue of classification with greater than one class. In multilabel classification, a classifier tries to allot numerous labels to every document. Multilabel classification is the task of allotting every given example to a predefined classis set in a domain where an example can belong to many classes simultaneously. Siblini et al. (2020) have stated that multilabel classification related every instance to numerous labels and has acquired huge attention in present years. Multilabel classification has acquired significance in the past decade, and it is nowadays confronted to present requirements to process huge raw information from heterogeneous sources. The multilabel classification objective is to find the proper label vector for any feature vector. During the phase of training in the feature matrix, a classifier is adjusted to fit its finding to label matrix. Pham et al. (2017) have mentioned that the multilabel classification is used to classify the new documents. The main task of multilabel classification is allotting every given example to a set of predefined classes in a domain where an instance can belong to many classes simultaneously. The Latent Dirichlet Allocation (LDA) based models were used in multilabel documents that relate separate tokens of words with varied labels. The multilabel classification is used to enrich the features for short documents by integrating different characteristics like LDA features, Term Frequency-Inverse Document Frequency (TF-IDF) bigrams and unigrams.