In this speech, Silvia Semenzin will explore the topic of gender discrimination through well-known examples of algorithmic bias. The discussion will center on the importance of designing, envisioning, and promoting inclusiveness at every stage of technology development. She will also present the core principles of feminist AI and offer ideas on how to apply them to create better futures.

Our lives are comprised of moments shared with others or spent in solitude. Regardless of the circumstances, many of these moments occur within the confines of our household. The people with whom we live, as well as their absence, significantly influence who we are and the opportunities we have in life. Throughout our lifespan, our roles as parents, children, siblings, partners, or roommates, along with our well-being, are deeply affected by our living arrangements. These arrangements not only shape individual lives but also reflect societal norms on how we organize collectively. Despite the crucial role households play in social change, global data on the composition and evolution of living arrangements is limited. The CORESIDENCE Living Arrangements Database (CoLADB) fills this gap by providing a comprehensive dataset of over 782 million records from 107 countries over the period 1970-2020. Developed in collaboration with the Barcelona Supercomputing Centre, CoLADB brings together Social and Computational Sciences. This presentation explores the key challenges and breakthroughs behind building this innovative data-infrastructure.

In the growing domain of large language model (LLM) training and fine-tuning, the crucial aspect of tokenization is often overshadowed. Popular open-source language models frequently offer limited dictionaries for low-resource languages. This presentation illustrates a strategy for extending the token vocabulary and enhancing embeddings training specifically for the Ukrainian, Arabic and Georgian languages, using the Mistral model as a case study. Presentation will cover popular metrics like fertility and parity as well as custom metrics like code-switching that were created for the specific evaluation of quality at inference.

Our personal memories of a place are situated in fleeting sensory experiences – the waft of the petrichor, bird songs scoring the background, bakery aromas, sounds of traffic and footfalls whizzing by. This interplay between our senses structures and mediates our sense of a place. This talk centers on evocative experiences that highlight the rich, interconnected web of sensory information underlying our everyday encounters. Showcasing the project and process behind the piece, Atlas of Intangibles, Priti will examine alternate forms of data collection, such as visual ethnography, score-based data walks and the design of data experiences that encourage exploration and create a richer connection with the world we inhabit.

As machine learning becomes integral to business operations, the challenge lies in making these complex models understandable and actionable. In his talk, Karim Douieb, co-founder of Jetpack.ai, will showcase how effective data visualization can simplify machine learning insights, making them accessible to business leaders and driving adoption.  By sharing practical strategies and real-world examples, Karim will demonstrate how visualization can bridge the gap between data science and decision-making, ensuring that sophisticated analytics translate into impactful business outcomes. Whether you're a data visualization specialist or a business leader, discover how clear visual storytelling can enhance understanding and facilitate smarter decisions.

Data has become anything and everything. New techniques and explorations applied to data have created new forms of data interaction. Data is more than just informative; it is also a means of self-discovery, playfulness, interaction, and a creative space beyond the 2d dashboard screen. In this talk, Letícia Pozza reflects on new forms of data and technology interaction, bringing a source of inspiration, possibilities, and creativity. Whether visualizing information with spatial computing, going beyond chats for AI-human interaction, or using physical materials, we need to explore new ways of interaction to bring data closer to people.

We encourage you to meet new people or reconnect with old friends and grab a lunch at one of the nearby cafes.

In this talk, Nathan Coyle shares insights from his book, exploring the barriers identified through his research that have hindered the adoption of open data, particularly the entrenched governmental culture leading to ineffective policies. He also discusses how we can support peacebuilding efforts, addressing their needs and highlighting how we can influence data governance to make open data more accessible.  It's time to reframe open data as a tool for social good, capable of sparking positive change in our communities and empowering activists and the third sector to increase capacity, secure funding, and shape policy decisions.  You will also discover what peace tech is, how it is deployed, and what narratives we need to explore. Additionally, you will learn how the social tech community in Georgia can influence data providers to make their data more accessible for social activists and peacebuilding NGOs.

Imagine a world where each generation is happier than the one that’s gone before it. You aren’t living in this world. For some countries, the golden days belong to past generations. So, in an age of avocado toast and remote work, why are some of our younger generations the gloomiest yet?  In this talk, Kat will tell you a data story about happiness. She’ll then give you insight into the data storytelling process by showing you how it was created.