Demo: MoCA+: Incorporating user modeling into mobile contextual advertising

In-app advertising has become a signifcant source of revenue for mobile apps. Mobile contextual advertising is one of the recent approaches to improve the effectiveness of inapp advertising, which seeks to target an app page content that a user is viewing. Typically, mobile contextual advertising is based on the cloud-based architecture, which may cause many privacy concerns, because in-device user data inevitably sends to ad servers. In our previous work [3], we developed a novel mobile contextual advertising platform, called MoCA, which was designed to improve the relevance of in-app ads in a privacy protecting manner. However, MoCA does not explicitly model user interests. In this demo, we present yet another mobile contextual advertising platform, called MoCA+, which incorporates user modeling into MoCA. It is designed to provide contextual in-app ads to third-party apps through its well-defned APIs. MoCA+ collects a variety of user data inside a mobile device to model user interests. It then matches contextual ads considering both the user interests and an app page content based on the semantic technique [2]. Since the proposed platform explicitly targets user interests, it is expected to satisfy the user's information needs, resulting in a better user experience on in-app advertising. As opposed to typical mobile contextual advertising that is based on big data analytics on ad servers, MoCA+ performs all the key essential tasks locally. It, therefore, protects user privacy without sending out any in-device data. To the best of our knowledge, this is one of few works to implement the mobile contextual advertising platform without resort to servers.