A heuristic line piloting method to disclose malicious taxicab driver's privacy over GPS big data

Abstract While privacy preservation is important, there are occasions when an individual’s privacy should not be preserved (e.g., those involved in the case of a terrorist attack). Existing works do not generally make such a distinction. We posit the importance of classifying an individual’s privacy as positive and negative, say in the case of a misbehaving driver (e.g., a driver involved in a hit-and-run or terrorist attack). This will allow us to revoke the right of the misbehaving driver’s right to privacy to facilitate investigation. Hence, we propose a heuristic line piloting method, hereafter referred to as HelpMe. Using taxi services as a case study, we explain how the proposed method constantly accumulates the knowledge of taxi routes from related historical GPS datasets using machine-learning techniques. Hence, a taxi deviating from the typical route could be detected in real-time, which may be used to raise an alert (e.g., the taxi may be hijacked by criminals). We also evaluate the utility of our method on real-life GPS datasets.