Identification of Safest Path using Spatial Analysis

Abstract

Safety is of critical importance to an individual while travelling. The goal of this project is to provide an effective solution to the masses to ensure their safety when travelling from one place to another. The proposed system is developed by building a robust model by using police FIR data and crime news articles scraped from the web. Techniques from Natural Language Processing(NLP) have been used to convert the unstructured to a form that can be applied to train a machine learning model. The location of a crime is extracted from these articles by using a technique called geo-tagging. The locations extracted from the articles are used map the location of the crime on a map and the occurrences of the crimes at there respective locations are stored. This problem can be now visualized as a graph search problem where the locations of crimes are the vertices and the crime intensities are the edges connecting the vertices. Finally, a shortest path algorithm is applied to find the safest route between two locations.

Keywords: Natural Language Processing, Web scraping, Naïve Bayes, Parts of Speech Tagging, Geo- tagging, Safest Path, Word Embeddings.

Cite this Article Aaditya Naik, Pritesh Parmar, Subhash Mudaliar, Monali Deshmukh. Identification of Safest Path using Spatial Analysis. Journal of Communication Engineering & Systems. 2020; 10(2): 17–24p.