A DEMONSTRATION OF APPLIED MACHINE LEARNING ON MULTI-SOURCE DATASETS FOR ANALYZING AND IMPROVING IN INCIDENTS MANAGEMENT AND RESPONSE TIME
– Auteurs: Pr. Patrick Mukala et Ass. Godwill Ilunga- Résumé: Various machine learning algorithms constitute vital solutions to big data processing. Data from various sources, in disparate formats can be integrated and seamlessly analyzed when the right tools and techniques are made use of. In this paper, we report on an application called Respond! reporting tool. This tool was realised by ICT students from Fontys University of Applied Sciences as part of their graduating project from the Applied Data Science minor. The aim of the project was to develop a tool based on advancedmachine learning techniques aimed at gathering information from a number of various different sources (news outlets and websites, social media and weather forecasting sources) in order to allow an automatic management of fire incidents in the Netherlands. The paper demonstrates the extensive research conducted on how to apply machine learning techniques on weather data. Making use of a combination of web scraping tools, social media mining and a number of machine learning techniques such as support vector machines and regression modeling, the tool provides some viable results that can be used to monitor weather information. – Mots-clés: Big Data, Data Mining, Clustering, Machine Learning, Text mining, Weather data, Web scraping. – Le lien: https://eoi.citefactor.org/10.11216/gsj.2021.08.53311