BBVA DATA & ANALYTICS NEWSLETTER - The Monthly Briefing #6. Dec 2020.

The Monthly Briefing We have overcome this ~~year~~ stage December 23th, 2020 \| 8' reading time

The process of transforming a business need into a data-based product may in some cases look very much like the adventure of playing a video game. The long road is made up of several stages, -or screens, using the gaming terminology-. Screens that one must overcome one by one to finally achieve the final victory, which is none other than having the model productionised. Aspects such as the availability of data and its sources, the data ingestion & preprocessing stage or the model training are key to reach the final screen. However, other issues such as an accurate definition of the problem, a fluid communication between all the members of the team and a general knowledge of the tasks of all the teams involved are equally important to succeed. In this talk on December 10th at the SheStartup event, organized by AllWomen, our colleague Clara Higuera, explained how these stages have been overcome in a real use case using NLP techniques: a conversations classifier. Without a doubt, a masterly lesson on the daily work of a team that builds products based on data! Here you can watch the talks of the event. Thanks to AllWomen for inviting us!
Further reading +The annual Open Data Maturity Report is now available. (European Data Portal) The European Data Portal just released their annual Open Data Maturity Report that evaluates and compares over 30 countries in Europe on the impact open data creates and how it affects data policy and data strategy. +Better understanding of how NLP models work. (Adam Geitgey · Medium) This series of articles by Adam Geitgey explains in more detail how the NLP models work, step by step. Who said that Natural Language Processing is not fun? + The results of this year’s McKinsey Global Survey on Artificial Intelligence. (McKinsey) The results suggest that organizations are using AI as a tool for generating value and, increasingly, that value is coming in the form of revenues. Other relevant outcome: while companies overall are making some progress in mitigating the risks of AI, most still have a long way to go. Revenue increase from AI adoption, 2018 and 2019 \| Via Mckinsey +Another summary of AI top stories of 2020 (a16z) The article round up top stories on AI in business, in breakthroughs, in practice. Among the main topics addressed in the article, we highlight the Emerging Blueprint for Operational ML/AI, or the advances in AI in biology. + The EU wants to take control of its own data. (The Wall Street Journal) The EU’s executive arm proposed new legislation aimed at creating an EU-wide data marketplace to facilitate sharing of industrial and government information. This legislation is the result of a broader effort to wrest digital influence from large companies in the U.S. and China, among other issues. + A deep learning model to detect oil and gas infrastructure in aerial imagery. (Stanford ML Group) With this model, called OGNet, they've create an open dataset with 7,000+ images of nearly 150 oil refineries — including several facilities that were not yet included in existing public datasets — which they hope will make it easier to attribute satellite-detected methane emissions to their sources on the ground. Here the paper by Sheng & Irving et al. (2020). + How many different types of algorithms are trapping people in poverty (MIT Tech Review) A fascinating story about how a growing group of lawyers are uncovering, navigating, and fighting the automated systems that deny the poor housing, jobs, and basic services. By Karen Hao. For more information, read the report called Poverty Lawgorithms published by Data & Society. + Can we smell data? (Nightingale · Medium) According to Amy Cesal, we usually associate our sense of vision to how we interpret data, but we can consume data with other senses, too + The future of finance (Bloomberg) After a decade full of changes, some of the areas of the banking industry have seen major disruptions, and some of them, like retail trading and wealth management have experienced a boom. + What Spotify does to ensure raw data can be easily understood (Spotify) Spotify's users generate huge amounts of raw data. According to Sabrina Siu, raw data by itself is not that helpful though; "we need to be able to process, manage, and distill it into insights that can inform new features or improvements to the experience. And to do that, we need usable, well-designed tools that ensure these insights can be easily understood".
Quote of the month "AI is not fully ready to make the kind of decision-making corporates expect it to make and even if it were corporate teams and networks are not fully ready to implement and reap the full benefits of AI." Jennifer Schenker, Founder and Editor-in-Chief at The Innovator. The state of AI-decision making was the focus of an October 13 roundtable discussion moderated by The Innovator in partnership with DataSeries, a global network of data leaders led by venture capital firm OpenOcean. In this article, entitled AI Decision-Making: State Of Play And What’s Next, Jennifer Schenker writes a summary of the main ideas discussed in that session.
Happy Holidays! All of us at BBVA Data & Analytics wish you a happy holidays and a great start to the new year! 💙 See you folks! < Read previous issue

+The annual Open Data Maturity Report is now available. (European Data Portal)
The European Data Portal just released their annual Open Data Maturity Report that evaluates and compares over 30 countries in Europe on the impact open data creates and how it affects data policy and data strategy.

+Better understanding of how NLP models work. (Adam Geitgey · Medium)
This series of articles by Adam Geitgey explains in more detail how the NLP models work, step by step. Who said that Natural Language Processing is not fun?

+ The results of this year’s McKinsey Global Survey on Artificial Intelligence. (McKinsey)
The results suggest that organizations are using AI as a tool for generating value and, increasingly, that value is coming in the form of revenues. Other relevant outcome: while companies overall are making some progress in mitigating the risks of AI, most still have a long way to go.

Revenue increase from AI adoption, 2018 and 2019 | Via Mckinsey

+Another summary of AI top stories of 2020 (a16z)
The article round up top stories on AI in business, in breakthroughs, in practice. Among the main topics addressed in the article, we highlight the Emerging Blueprint for Operational ML/AI, or the advances in AI in biology.

+ The EU wants to take control of its own data. (The Wall Street Journal)
The EU’s executive arm proposed new legislation aimed at creating an EU-wide data marketplace to facilitate sharing of industrial and government information. This legislation is the result of a broader effort to wrest digital influence from large companies in the U.S. and China, among other issues.

+ A deep learning model to detect oil and gas infrastructure in aerial imagery. (Stanford ML Group)
With this model, called OGNet, they've create an open dataset with 7,000+ images of nearly 150 oil refineries — including several facilities that were not yet included in existing public datasets — which they hope will make it easier to attribute satellite-detected methane emissions to their sources on the ground. Here the paper by Sheng & Irving et al. (2020).

+ How many different types of algorithms are trapping people in poverty (MIT Tech Review)
A fascinating story about how a growing group of lawyers are uncovering, navigating, and fighting the automated systems that deny the poor housing, jobs, and basic services. By Karen Hao. For more information, read the report called Poverty Lawgorithms published by Data & Society.

+ Can we smell data? (Nightingale · Medium)
According to Amy Cesal, we usually associate our sense of vision to how we interpret data, but we can consume data with other senses, too

+ The future of finance (Bloomberg)
After a decade full of changes, some of the areas of the banking industry have seen major disruptions, and some of them, like retail trading and wealth management have experienced a boom.

+ What Spotify does to ensure raw data can be easily understood (Spotify)
Spotify's users generate huge amounts of raw data. According to Sabrina Siu, raw data by itself is not that helpful though; "we need to be able to process, manage, and distill it into insights that can inform new features or improvements to the experience. And to do that, we need usable, well-designed tools that ensure these insights can be easily understood".