Will GDPR redefine consent?
How is data going to change over the next three years? Can blending datasets improve healthcare? How can companies use and share customer data without damaging trust? What unexpected things can we learn from the CIA leak?
The Information Commissioner's Office (ICO) has published draft guidance for consultation on the interpretation of "consent", as a part of the GDPR. Companies must offer individuals genuine choice and control, so consent must be a positive opt-in. 📃📃
Martin Sloan, data protection law guru, outlines his interpretation of the proposed changes. Key takeaways: consent must be “unambiguous” and be given “by a statement or clear affirmative action.” Furthermore, "consent cannot be a precondition of a service", such as in an employer/ employee relationship "where it is arguable that the individual is not in a position to freely give consent". 🙌
Need an overview on the digital economy and GDPR? Excellent write-up by Stephen Cameron, in particular the key CTA's for businesses on what they NEED to be doing in preparation for the upcoming changes.
How is big data going to change over the next three years?
"As data leaders, we need to make sure we fulfil the obligations of GDPR" whilst showing customers how they can benefit from handling over their personal data.
Watch out for talent raids - companies need to “keep investing in their employees to foster improved retention”.
Joining up data will be KEY to winning big - blending disparate data sources will redefine the "user experience of data to create a much more interactive connection". This will "drive information to insights and actions". ✅
Adl Gaskell discusses blending user-generated data with electronic patient records and the implications for healthcare. 🤔 Validic has entered a partnership with The Salford Royal NHS Foundation Trust to test how richer datasets, which include user-generated data, can improve patient care and engagement. 🏥 Definitely one to watch.
Google’s DeepMind announced plans to implement "a new technology loosely based on bitcoin to let hospitals, the NHS and eventually even patients track what happens to personal data in real-time." 😎 DeepMind has previously received criticism for how they protect personal data, so now plan to incorporate an audit trail to increase transparency.
An open-data contest "unearthed scientific gems" - but dramatically reduced the ROI for the investigators who collected the data. ⚖ However - the contest was based on incomplete data, so could potentially generate hypotheses for the investigators to test and analyse on the full dataset. 🤔
How can companies use and share customer data without damaging trust? 👍 Excellent piece by Steve Shoaff. Trust can be difficult to gain but easy to lose - so protecting customer privacy is more vital than ever for ensuring customer retention. Summary of top tips for nurturing trust:
Go beyond the regulations
Put users in control
Be careful with third parties
Use security best practices
Joan Gamell discusses the unexpected learnings from the CIA leak. They search answers on StackOverflow too. 😅
Google is acquiring Kaggle. Will this help Google get "more entrenched within the data science community" whilst it explores "more perpendicular approaches to ensure its dominance in AI"? Interesting.
Researcher/ data scientist Evan Warfel explains how to make easily understandable node-link graphs. Interesting read - with lots of nice visualisations. For example, differences between Apple and Google's patent collaborations:
Left: Apple Patent Collaboration. Right: Google Patent Collaboration. From Fast Company.
Reverse engineering the Hacker News ranking algorithms. Intriguing. 💥
Szabolcs Horvát used Mathematica to measure the fractal dimension of a photo of tree branches. Neat. 😍
Measuring the fractal dimension of a tree photo - alongside scaling plot, by Szabolcs Horvát