The purpose of this project lays at the intersection of image mining and virtual currencies for social networks: how to create virtual currencies backed by personal images meeting the specific marketing requirements of companies using the social networks. As a result, companies will reward users with virtual currency for the knowledge mined from their images shared on the social networks, and users in turn will use their personal images for trade in goods and services.

In the new Facebook news feed, the best ads will not look like traditional advertising. Instead, they’ll look like your friends’ photos. But it will be your friends’ photos which voluntarily celebrate their favorite brands that will become most valuable to advertisers.

The social web is turning visual, so that any picture of a person might give hints about the person, its family, work, and personal context that are of marketing interest. The work to be done is to extract the entities inside the information of the pictures, and then mine rules about what is behind these pictures, among the entities inside, (i.e., who is nurturing who, who is doing what, what is done in group, etc.) that is useful for companies in their segmentation and targeting activities (for example, targeting people who like drinking Coke in their daily jogging or those who might be willing to buy a Toyota SUV for the rocky weekends).

On the other hand, there is a growing usage of virtual currencies, as a sign of people accept using any new means for the trade exchange. The benefits are that the virtual currencies are available to everyone with computational resources for mining. The point is that the new virtual currencies like bitcoin or freicoin base their money supply by mining nonsense, with no utility and no benefit behind the mining activity.

The work to be done is to develop mining with a minimum sense of utility for visual content and exploit the properties of the virtual currencies.The project is a disruptive innovation in the field of image mining, marketing and virtual currencies, with diverse preliminary works such as the definition of Wits as a complementary currency.

In short, the project will develop a data mining tool for social networks and knowledge mining through image analysis. Taken together, these parts will provide the detection of brand images, extracted from the images displayed while rewarding users with digital currencies as a method of exchange. These functions will provide valuable “visual” information on public feedback to companies’ products.


Proyecto financiado con fondos del:


Referencia del proyecto: RTC-2014-2566-7