Location obfuscation using smart meter readings

The World Wide Web has become a popular platform for the hosting of surveys. Such online surveys can more easily and efficiently collect data and are accessible using both desktop and mobile devices. In many cases personal information, such as postal addresses, is requested by these surveys. With the availability of location detection technology in mobile devices, some surveys even go further and request spatial coordinates (e.g. GPS information). This raises the issue of location privacy since this may lead to privacy attacks. One can prevent this by providing vague location information but, in many cases, this may defeat the purpose of having the location field since it may be required by the surveyors. Hence there is a delicate balance between user privacy and location accuracy. In this paper we address this trade-off by proposing a framework for location obfuscation in online surveys. In this framework we allow users to state their privacy requirement using a target probability which is the probability of they being correctly identified. We then provide the most accurate location information to the surveyors given the desired privacy level. This approach requires population density information and so we also provide an interesting approach to estimating this information from mining Smart Meter data.