News & blog

Read the latest news and blogs from The Data Lab and Scotland’s data science community.

Strata London: Day 1 highlights

We’re at Strata in London this week. Day 1 was a Data-Driven Business Day, packed with presentations on practical applications of analytics across domains including Government, health, internet and venture capital.

Top five stand out presentations of the day (in our opinion) were:

  1. Large-Scale Emotion Analytics by Daniel McDuff (@danmcduff) from Affectiva. Daniel walked through how Affectiva are using large volumes of facial video studies (3 million faces across 75 countries, generating 11 billion data points) to generate quantitive measures of facial recognition that can be used to generate valuable insights into consumers’ emotional engagement with digital content, brands, advertising etc as well as applications within healthcare. Some interesting stats:
  • Females are 30% more expressive than men.
  • Individuals in the US express outward emotions twice as much as those in China.
  • Affectiva have developed facial recognition algorithms that can measure heart rate more accurately than sensors.
  1. The Future of Machine Intelligence and Why It Matters by Shivon Zilis (@shivon) from Bloomberg Beta. Shivon described her approach to mapping the machine intelligence landscape and how she is using machine intelligence tools to improve her working life including algorithmic schedulers, VC growth scores, machine transcription and talent acquisition. Take a look at her post on Medium,plenty of inspiration for trying some new machine intelligence tools.
  1. Measuring and Understanding the Value of Social by Cory Levison (@kohwi) from SoundCloud Ltd. Cory presented SoundCloud’s approach to social media measurement across their social graph (currently with 2 billion edges). They have identified that users who are more active on social will engage more with their service, spending more hours using SoundCloud. The measurement approach focusses on three goals: user centric (i.e. down to a user level), intuitive and actionable, scalable. Specifically, SoundCloud analyse degree metrics including: in, out and mutual degrees, second degree networks and density.
  1. Situational Awareness: This is not the data your’re looking for by Simon Wardley (@swardley) from the Leading Edge Forum. Great presentation describing how may businesses and individuals operate with little situational awareness opting for the “tyranny of action”. Using inspiration from maps, Simon described the need to develop a strategic map modelling a situational position with movement dynamics.
  1. Getting the Public Sector to Act Data-Driven by Siim Sikkut (@sikkut) from the Estonian Government. Siim presented great anecdotes on the progress made in Estonia including:
  • The government has been paper fee since 2000. The only documents requiring written signatures are marriage certificates and property purchases (though they are looking to make these digital).
  • Each minister has responsibility for the digital component of their portfolio; digital isn’t a siloed department.
  • Estonians can vote in elections (from anywhere in the world). Currently a third of the population use this channel.
  • The country has a focus on skills development where children are taught coding from age 4.
  • Citizens can control their own digital footprint.

Also, “50 shades of dirty data”, a reference to data quality from smartphones during DataLions “Data Science as Art” presentation, generated the biggest laugh of the day.

Great start to the conference and we’re looking forward to day 2.

Article by Brian Hills, Head of Product Management, The Data Lab (@BrianHills)

Share this story:

Share on facebook
Facebook
Share on twitter
Twitter
Share on linkedin
LinkedIn
Share on email
Email