The Internet, globalization, and the emergence of global digital platforms have transformed our lives and transcended geographical and cultural boundaries. Thanks to the web and social media, it is easy to connect to anyone in the world, who lives in a different culture, in a different social context, with different social norms, with different habits, and who speaks a different language. The amount of diversity we have access to has increased exponentially. We are exposed to a seemingly unlimited amount of diversity on a daily basis, most of which is unexpected and represents the ‘unknown unknowns’. The overall objective of WeNet is to develop a technology that enables to empower richer and deeper social interactions through diversity-aware Artificial Intelligence (AI). Diversity-awareness is the ability to cope with and capitalize on this multidimensional variability of humans, by being robust to changes across human characteristics that are not consequential for the problem at hand, while being also sensitive to the changes of those human characteristics that are so.
The overall objectives of WeNet are organized along three main lines of work. The first objective focuses on how to develop an AI technology and a platform which empowers online social relationships. This is interdisciplinary work where we integrate competencies and work from many different areas. The first is the social sciences, where we have developed a sociological model of diversity built around the notion of communities of practice. The second is AI, where we are developing a set of algorithms that allow the AI to learn the profile of the user, with specific reference to her situational and social context. The third is ethics as an underlying methodology and concrete approach providing requirements in the development of WeNet. The second objective is focused on an intense piloting activity. The main idea is that technological innovation should be motivated and validated by a clear and positive societal innovation. We have therefore developed a highly innovative diversity-aware pilot design research, and we are running pilots in Europe, India, China, Mongolia, Paraguay, and Mexico. These pilots have involved thousands of students in synchronic studies (performed via online questionnaires) and hundreds of students in diachronic studies, collecting both user responses and sensor data (via a dedicated app developed as part of the project). The third is the bootstrap of a community around the WeNet idea, involving users, developers and innovators, and scientists. The main tools for this are the WeNet platform, which is open-source and available to third parties, and a research infrastructure whose main goal is to make the data collected available to third parties. The bootstrap of the WeNet idea is further pushed by a set of accompanying measures including an intensive dissemination activity, several online meetings, and webinars, a set of online courses, an open call for third parties, and, at the end of the project, a two-day event whose main goal will be to bootstrap the post-project development of the WeNet.
During the first phase, which lasted about twelve months, the work focused on developing the building blocks of the three lines of work mentioned above. A lot of emphasis was placed on planning and organizing the pilot activities and ensuring that the pilots would comply with the data protection (GDPR) and ethics requirements, also taking into account that some of them would have to be carried outside the European Union. In a second phase, approximately lasting another twelve months, a first version of the technology and the WeNet platform was produced, deployed and used in the first round of pilots. In the current, third phase, i.e. the last twenty-four months, the data collected during the pilots have been prepared and processed, including a complex and intensive anonymisation phase, and made available to the partners. This data is now being analyzed and studied with the main goal of designing the second and third rounds of pilots, which were executed within the third phase. The ultimate goal is to have a deeper understanding of how diversity appears in the pilot regions and countries, and the implications of this for the WeNet technology and platform.
