Most cities have mechanisms for “measuring traffic and parking, but not social life” leading architect Jan Gehl famously said. The key to measuring what matters in neighbourhoods - like social life - is quantifying it. The process of measuring places, however, requires genuinely combining technical and non-technical expertise. This the intersection where Neighbourlytics sits. Our social analytics platform combines best practice urban planning knowledge with advanced analytics and mathematics to codify and quantify the value of places.
In the same way that we are continually thinking about advancing urban planning and social research, we are equally thinking about best practice analytics. In March, our Head of Analytics attended the 2019 Women in Analytics conference in Columbus, Ohio, USA. Touted as "An analytics conference. Featuring women. For everyone" the event brought together a wide variety of analytics professionals with entry level through to advanced technical talks, strategy talks, and a fantastic panel about ethics in algorithms.
In this blog you’ll find the crib notes summary of our key take aways, including:
Ethics in algorithms
Data and analytics in your organisation, from a strategic point of view.
Ethics relates to transparency and understanding. Mathematics is not something we need to intimidate people with, we need to learn how to convey difficult technical details to non-technical people and allow them to feel empowered to challenge algorithms when they are not doing the right thing.
Algorithms are opinion embedded in code. Therefore we need to focus on the context in which they are used, and to fully understand that context. one of the speakers Cathy O’Neil noted that "fairness comes from context". For Neighbourlytics this is a continual consideration as we create algorithms to help us understand cities, where the context is a) critical and b) always changing.
Nicole M. Alexander discussed how building algorithms needs to be an interdisciplinary process - interdisciplinary voices will help when making decisions around the use, deployment and upkeep of the algorithms.
Data and analytics in your organisation
Most organisations have datasets of various types, but they’re often inaccessible, particularly to non-technical teams. This makes it challenging for organisations to leverage the full benefits of data or to adopt an interdisciplinary process for building algorithms. Ursula Cottone spoke about her role at Huntington as the Chief Data Officer. She suggested we need to make it easy for people in our organisations to ‘shop’ for our data - the organisation’s data should be accessible and easy to understand (like for example, through an enterprise wide data dictionary). When you do give access to data, then think about: quality, understanding, and how to ensure it’s used in the way it was intended.
It can be hard for organisations to adopt data and analytics and to combine expertise across disciplines. It’s important to take time to empower people to adopt new processes and technology in order to leverage the benefits.
Yvonne Wang pushed for diverse teams. Dissenting opinions can create value. It’s important to create a culture where people feel comfortable speaking out and where different perspectives are allowed.
Data science isn’t just about creating new products, it can be used to improve and strengthen business as usual - to improve quality or optimise work pipelines. And this is best done with an inter-disciplinary process. Katie Malone and Annie Darmofal gave a joint presentation on how they combined Annie’s knowledge on product design and Katie’s knowledge on data science to create an internal product for the data science team that was user-friendly and practical.
If you are interested, you can watch the videos from the Women in Analytics conference here.