The successful completion of this project will lead to a clear go/no-go decision regarding the viability of commercializing the privacy-preserving location based services (LBS) techniques the PI developed in two NSF-sponsored research projects. Should a favorable decision be made, this project will also produce a transition plan for commercialization and a mobile application for demonstrating the privacy-preserving Location-Based-Services (LBS) technology to potential partners. LBS, are widely prevalent on the web but also raise privacy concerns, as evidenced by the negative publicity surrounding jigh-profile inadvertent collection of location data in mobile devices. The PI's privacy-preserving techniques enable LBS without violating a user?s privacy. Specifically, the techniques protect two types of privacy: (1) location privacy - i.e., the team prevents an LBS server from learning a user's real location; (2) query privacy - i.e., the team prevents the LBS server from learning what the user is searching for (e.g., bar, museum). At the same time, the echniques maintain the quality of LBS by providing the user with extremely accurate, if not precise, LBS query answers.
This project will has the potential to have broader impacts on the nation?s high-tech industries. The ability to control when (and whether) an employee's location can or cannot be disclosed to an LBS server is needed by a wide variety of security-sensitive corporations, governments, and security agencies. Similarly, empirically validated means to protect the location privacy of ordinary users of mobile devices will benefit the wireless user community at large, in addition to LBS providers and wireless carriers. This project will also have broader impacts on the market of mobile LBS. In particular, the team's offering of a privacy-preserving LBS application is expected to raise wireless users' awareness of privacy issues surrounding LBS, and to eventually convince most LBS providers to add privacy-preserving features to their systems (e.g., by integrating and/or licensing the proposed techniques). The broader impact of this project also extends to the research community and academia. In particular, the market studies conducted in the project will enhance scientific and technological understanding of the location privacy problem.
Since pivoted to the idea of deep-web analytics in November, 2011, the team interviewed a large number of potential customers in domains including government agencies, political parties, news media, public relations firms, accounting and business consulting, private equity and venture capital, education, etc. A finding from these interviews, which addresses the intellectual merits of this project, is that while the majority of customers being interviewed expressed significant interest in the technology, they were uncertain of how to take advantage of the analytical data the technology is able to provide. Specifically, many of them raised the 'so-what' question - i.e., while the analytical results produced by the technology was often eye-openning and sometimes surprising, many customers had difficulties figuring out how to deal with the results - e.g., which actions to take based on the new information provided by analytics. Another finding from the interviews, which is critical for addressing the broader impact of this project, is three possible approaches to address this critical problem: (1) to acquire domain expertise on what actions a customer can take based on the analytical answers. (2) to enable effective data visualization, and (3) to form partnership with software / web developers who already have the expertise to serve the end customers.
