Authors - Govind Kumar, Amresh Kumar, Ajeet Singh Abstract - The process of selecting the right Indian city to live in is an extremely crucial one, which can have a huge impact on One’s life, safety, work and happiness every day. However, the tools available today, The kind of websites that tell about a property, or simple map applications, aren’t smart enough. They Do not know what each member of a neighbourhood really wants. This paper introduces Neighbor- Fit, an innovative AI-driven solution that suggests neighborhoods. Based on the actual need of the user. The system has three new ideas, the first of which is: A composite neighborhood suitability score (CNSS) as a six-part score that perates safety, facilities in the area, travel time, cost of living, green areas, and community life; (2) a smart algorithm called Preference-Adaptive Cascade Hybrid (PACH) which alters its style of recommendation according to the amount of recommendation it already has knows about the user; and (3) an explanation system based on LIME which explains to the user in simple words why a neighborhood was suggested. Tests done on 250 PIN codes In three major cities of India, namely, Delhi, Mumbai and Bengaluru, Preci- shows across. sion@10 of 87.3%, Recall@10 of 84.1%, and F1-Score of 85.7% — better than all There were five methods of comparison (p ¡ 0.05). The system reacts in an average of 340ms time even for 50 users using simultaneously.
Open Zoom