With tools, roles, and data architectures shifting all the time, it’s sometimes hard to keep up with the latest industry approaches to data science. But we’re here to help: this week, we’re highlighting several excellent posts that tackle a wide range of topics at the intersection of tech, data, and business. (Looking for great reads on other subjects? Just scroll down.) Here we go:
- Does the idea of effective data governance still make sense? The thought-provoking premise in Louise de Leyritz and Molly Vorwerck’s recent article is that data governance has failed to deliver on its promise to streamline data-focused workflows. But the authors also believe it’s not too late to revive and reshape the concept so that it works in harmony with business needs and modern data stacks.
- With great power, greater (corporate) responsibility. Recent advances in AI and algorithmic-driven decision-making are seeping into every facet of our lives. For Travis Greene, the massive social and economic footprint of data-guzzling tech giants must come with increased transparency and accountability—two areas where companies like Google, Meta, and others have a lot of room to grow.
- The promise of hybrid data organizations. After spending 15 years in data teams of various shapes, sizes, and configurations, Matthew Dawson-Paver has a strong sense of what works—and what doesn’t. His illustrated overview explains why a hybrid structure, which combines autonomous data teams and business or product teams with embedded data pros, is often the best approach.
- How can companies find the right data scientists? Before worrying about tools, frameworks, and team structure, there’s the need to hire data scientists whose skills align with company goals and mission. To do that, Rose Day recommends focusing a bit less on the technical side of the screening process, and more on conversations that can help identify shared interests and values.
If you’d like to stay up-to-date with recent posts about data and business, bookmark our Notes from Industry column.
Wait, there’s more! (But you knew that.) We also published fantastic articles in other areas recently, and we couldn’t just not share any of them. Here are several you definitely don’t want to miss:
- Nura Kawa introduced class maps, a novel visual tool that explains the results of classification algorithms (examples in R and Python included!).
- In his latest contribution, Michael Bronstein showed how feature propagation is an efficient, scalable approach for handling missing features in graph machine learning applications.
- If you’re in the market for a hands-on tutorial this week, you can’t go wrong with Rashida Nasrin Sucky‘s patient guide to chi-squared correlation tests.
- For those of you who are in the mood for theory, Gustav Šír‘s deep dive into the fascinating world of relational machine learning will absolutely do the trick.
- We never tire of Data Science projects that can help a greater good—like Himalaya Bir Shrestha‘s recent exploration of the learning curve effect, and how it can push renewable-energy technologies into a virtuous cycle.
We hope you had a good week, one full of learning and new ideas. Thank you, as always, for your support of the work we publish.
Until the next Variable,
TDS Editors
