Image Analysis for Privacy Assessment in Social Networks
| Autor: | Elena del Val, Carlos Diez, Ramon Ruiz, Joaquín Taverner, José Alemany |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Service (systems architecture)
Social network Computer science business.industry Privacy policy Trust metric Cloud computing 02 engineering and technology Human–computer interaction 020204 information systems Metric (mathematics) 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing IBM business Set (psychology) |
| Zdroj: | Distributed Computing and Artificial Intelligence, Special Sessions II, 15th International Conference ISBN: 9783030005238 DCAI (Special Sessions II) |
| DOI: | 10.1007/978-3-030-00524-5_1 |
| Popis: | Nowadays, the concern about privacy in online social networks has increased. However, the definition of an appropriate privacy policy might be a complex task, especially when several users are involved and have different privacy preferences. This problem usually appears when a user publishes a photo. In this paper, we propose a tool to automatically define the audience of a photo based on a trust metric. This metric uses a set of features (i.e., distance between users, number of people, emotions, etc.) obtained by the image analysis provided by IBM Cloud Visual Recognition Service. In a preliminary experiment considering 40 photos of 4 users, the results show that the proposed trust metric approximates the real trust relationships between users. We plan to integrate the tool into a real online social network. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|