CrowdSPaFE: A Crowd-sourced Multimodal Recommendation System for Urban Route Safety

In the age of information technology, location-based services such as Google Maps, Bing Maps, and Apple Maps have become popular for navigation and traffic information. These services usually consider the shortest path, traffic information, nearby places, and multi-modal alternative route suggestions based on a user's constraints. Nevertheless, these services do not always provide the best choice in terms of "user safety". Recently, some research and mobile applications have considered safety issues. Notably, none of these are capable of adapting to dynamic, conflicting safety features and real-time user feedback. Recently, a population-based algorithm called "SPaFE" has been introduced, which deals with crowd-sourced data along with historic data. This population-based approach, however, does not give more weight to recent feedback than to earlier feedback and does not consider updated crime reports. Furthermore, this approach does not consider distance and performs poorly in the area with insignificant or zero feedback. Considering the above background, we have introduced the population based algorithm "CrowdSPaFE" to adapt with dynamic crime reports, latest feedback, navigation in areas with insignificant feedback, and a trade-off between distance and safety factors. Finally, our empirical results of the "CrowdSPaFE" algorithm depict that it significantly outperforms the state of the art.