User modelling and personalization are commonly used in multiple tasks, in which users are characterized only based on explicit information about their knowledge, behaviour, social relations or preferences, aiming at adapting generic systems to the particularities of each user. The ubiquitousness of social networking sites, and mobile and smart-devices offer new information sources, opportunities and challenges for changing the personalization paradigm. The analysis of these new data source offer new research opportunities across a wide variety of disciplines, including media and communication studies, linguistics, sociology, health, psychology, information and computer sciences, or education. This has important implications in the context of inclusive eGovernment and Smart Cities, which could leverage on the user’s models to design and tailor services according to the characteristics and needs of each particular citizen. This would allow to mine and analyse user behaviour aiming at better understanding users (and ourselves), and thereby create more accurate models and personalisation strategies.

The opportunities for advanced research are match by several challenges. First, the knowledge discovery process, i.e. how data could be collected and interpreted. Second, the long-term availability of data, the interpretation of user-generated information, and the need for qualitative and quantitative (as well as user-based and content-based) research approaches. This also leads to ethical and legal considerations. Third, due to the heterogeneous nature of smart devices it is necessary to develop strategies for representing users and their behaviour. Fourth, the processing and management of high volumes of generated multi-dimensional personal data hinders effective and efficient data management. Fifth, the creation of long-term user models, which should capture the particularities of users across long periods of time, as well as coping and adapting to dynamic changes in life patterns. Finally, these models should be made available for multiple applications, for example, they could be integrated in real world health systems.

Rapid urbanization in developing economies, coupled with reduced cost for information networks and data collection and transfer, are great drivers to the emergence of smart city markets. Vast amounts of data are becoming available from sensory and user mobility devices and other urban data collection sources. In this context, information retrieval and content analysis algorithms are required to make use of this large scale data to provide more intelligent urban services and to enhance user experience. This workshop targets academy and industrial practitioners leveraging on diverse data mining and machine learning techniques, including content aggregation, content analysis, predictive modeling, deep learning and user embedding for modeling user behavior and analyzing urban data.

Hence, a significant need arises for further development of innovative methods and approaches that are able to mine and deal with such new data sources. In this context, ECML-PKDD is the premier interdisciplinary conference for researchers and practitioners from data science, data mining, knowledge discovery, large scale data analytic, and big data.

The target audience of this workshop are professionals, researchers, and technologists who will benefit of a single forum where they can discuss and share the state-of-the-art of the development and applications related to knowledge discovery and user modelling for smart cities, present their ideas and contributions, and set future directions in innovative research in those areas. Particularly, at ECML-PKDD 2019 this workshop targets people who are interesting in sensing/mining/understanding data generated by citizens (including but not limited to social media data and data generated with smart devices) that can derive personalization models as well as to tackle challenges in cities and help better formulate the future.