Hybrid machine learning model for a next-generation ecofriendly travelling and guides to reduce carbon emissions

In this modern era, travelling has become an inevitable activity for medium-high income people and students to energise their mental strength. As a result, they tried to have a budget tour, such as finding a cheap place to stay or purchasing a cheap ticket. Although, numerous systems have been developed to target travel, yet, innumerable gaps exist between them. For instance, reducing carbon emissions and identifying human needs automatically have not been applied to date. Hence, it is crucial to comprehend the needs of individuals through opinion mining and to design the system accordingly. In this study, we analysed human sentiment through Natural Language Processing (NLP) and Machine Learning (ML) to propose a hybrid software framework (HSF) that will automatically fulfil people’s needs and reduce carbon emissions. Latent Dirichlet Allocation (LDA) techniques are used to analyse the sentiment regarding tourism, including locations, costs, hotels, the nearest food, etc. Positive and negative opinions were classified using the Naive Bayes (NB), Logistic Regression (LR), and Gradient Boosting (GB) models. In contrast, the HSF describes a hybrid system designed as the next generation of travelling software. The accuracy of LR and GB models achieved 95.98% and 95.88%, respectively. These models are further evaluated with the performance of the Receiver Operating Characteristic Curve (ROC) and Average Precision (AP). They enumerated the value of 0.91 and 0.40 in Area-Under the Curve-Precision Recall (AUC-PR). This study even utilised the most dominant words in the sentiment analysis from the hotel review. Therefore, the proposed model can be one of the possible solutions to reduce carbon emissions, minimise plastic use, and be a game-changer for green-ecofriendly living.