Privacy Preservation for Continuous Location-Based Misinformation Services
Abstract Several terms are given first before the research methods are explained:

• Location-based information (LBI), which is the information related to locations like a post of concerts,
• Continuous location-based information (CLBI), which is a kind of LBI and the information is continuously received according to the user’s current locations such as the ads of gas stations in a trip,
• Location-based misinformation (LBMI), which is a kind of LBI but is with incorrect or misleading information like a post of out-of-date concerts, and
• Continuous location-based misinformation (CLBMI), which is a kind of LBMI and the misinformation is continuously received according to the user’s current locations such as the ads of closed gas stations in a trip.

The LBMI is received based on the user’s current location; e.g., a movie theater is suggested but showing a different movie from the one sent. Whereas, the CLBMI is received according to the user’s continuous locations during a trip. For example, a user plans a flexible trip starting from source, restaurant, rest stops, beach, gas station, and destination. During the trip, the user may need to handle misinformation like out-of-date concerts or out-of-reach gas stations. Both LBMI and CLBMI are from location-based services (LBS), of which privacy is one of the critical issues. Without proper privacy preservation, users would be reluctant to use the services. The most common way to preserve privacy for the location-based information (LBI) is using dummy locations to conceal the true location. However, preserving privacy for CLBI is more difficult.

Instead of studying the generic LBS, this research will focus on privacy preservation for continuous location-based misinformation management (CLBMIM), which is to manage the misinformation, including misinformation identification, related to continuous locations. Other than the continuous locations of the true route, the locations of several dummy routes are also sent to the service provider for the requested information. To preserve the users’ privacy, the dummy routes include the following features:

• The dummy routes are randomly generated, but they have to be valid; e.g., the routes should be on roads.
• The dummy routes should not be too close to the true route, but they can not be too far away from the true route either.
• To facilitate the dummy-route generation, Google Directions, Places, Roads, and Routes APIs are used.

By doing this way, the service providers would not be able to tell the true routes taken by the users, and therefore the users’ privacy is preserved. Once the requested information is received by the user, ChatGPT will be used to identify whether the information is true or fake. If the information is found to be fake, next requested information will be checked before being used.

Keywords
Misinformation, misinformation identification, privacy, privacy preservation, location-based misinformation, LBMI, contniuous location-based misinformation, CLBMI, location-based service, LBS, ChatGPT

References
GPT models

Conference

The 63^rd ACM Southeast (ACMSE) Conference (ACMSE 2025), April 05, 2025.