An Intelligent Home Service Recommendation System using Geolocation and user Preferences

Urban services are growing digitalized and thus the need to have an intelligent and location aware real time home service systems is on the rise. Most of the existing applications have been based on static directories and rule-of-thumb filtering, which is not sufficient to deliver tailored recommendations and active tracking of availability. To enhance services discovery efficiency and accuracy, the current paper introduces Craft-Con- nect, a hybrid intelligent home service platform, a hybrid that integrates machine learning with geolocation awareness and is built on the foundations of the Node.js and Python platforms. It is a system with an 89% accuracy in ranking and appropriateness prediction of professionals using a Node.js Express.js backend to manage RESTful API, a MongoDB database to store data in a NoSQL, and a prediction machine using Python-based machine learning Scikit-learn and XGBoost. The hybrid recommendation engine considers many user-user-professional interaction fea- tures, including availability, price range, location, service ratings, and category match to bring about the most contextually relevant recommendations. It is also based on Twilio and Firebase Cloud Messaging API to ensure the safety of the communication process, OTP validation, and the delivery of real-time notifications. The findings confirm Craft-Connect as a smart, scalable and data- oriented solution towards customized home service discovery and recommendation in urban digital ecosystems.