Community Spring Cleaning week is here! Join your fellow Maveryx in digging through your old posts and marking comments on them as solved. Learn more here!

Engine Works

Under the hood of Alteryx: tips, tricks and how-tos.
NickJ
Alteryx Alumni (Retired)

For those who’ve followed my Community blogs in the past, you’ll know that I’m passionate about giving back some of my time and skills in analytics to charities and organisations that are really trying to change the world for the better. Thanks to Alteryx for Good, I get to contribute many hours of my working time to these projects, and I’d like to tell you more about my latest adventure! 

 

I was invited to be a Data Ambassador once again for DataKind UK – third time now! Quite an honour!

 

This time, the charity I was helping was The One Campaign – founded by U2’s Bono. Its primary aim is the eradication of poverty in Africa, and one of the means it does this is through empowering citizens of African countries to demand higher standards from their elected officials. Check out this short video to learn more  (the mention of being data-driven at 01:08 is particularly nice!)

 

As a background, the One Campaign (and its partners) have advocated for laws that would give African citizens the information they need to ‘follow the money’ in the extractives sector (Oil&Gas, Mining, etc.) Because of this, data is finally being published by hundreds of companies around the world. The DataDive focused on understanding what this data tells us, and to ensure that this increased transparency leads to real accountability.

 

Two goals for the weekend:

  • Find stories in the data, including ones that can help us show its value to policy makers.
  • Prototype the data structures and methodologies that will be needed to make the data actionable for citizens and activists.

 

Over the last two months, myself and two other data ambassadors got to work, refining the questions and the data – looking at mandatory filings, voluntary filings, open data, spatial data – a huge range of data sources. We cleaned, blended, fuzzy-matched and scraped as much as we could so that the data scientists over the weekend could be as productive as possible.

 

afg1.png

 

Data Dives are not about a single software vendor showing off and stealing the show – they’re about great data people coming together to do even greater things.

 

I’m always gratified to say that the Alteryx platform played its part in that process: getting to reuse custom tools from the last DataKind event allowed me to easily pull corporate filings, beneficial ownership and more from the OpenCorporates data service without having to write a line of code: fantastic when you’re dealing with thousands of organisations! Looking at a company’s portfolio of offshore oil rigs was also simple when you could load the geoJSON data straight into Alteryx and see that map and immediately understand the lie of the land.

 

On the weekend itself, there were actually three data dives happening simultaneously: the One Campaign, Christian Aid (looking at building data-driven measures of ‘resilience’ in Burkina Faso and Ethiopia) and FareShare (trying to optimise the logistics of delivering usable waste food to those who need it). The latter was particularly interesting as a previous data diver (and Alteryx customer) Matt was on-hand to help with his personal experience of using Alteryx to manage logistics and interpret that data!

 

afg2.png

 

The One Campaign had a team of around 20 data scientists split into a ‘data’ and ‘stories’ team – I ran the stories team: working on questions that focussed on a specific country (e.g. Nigeria), a specific company or looking for filings from subsidiaries from the American Petroleum Institute (API) who, as a consortium, are strongly resisting the mandatory release of transparent payments reporting.

 

We used a whole variety of technologies on the day: R, Python, Tableau, even plot.ly (which was nice to see – given our recent OEM deal). Some powerful visuals were produced, including ‘Sankey charts’ which show the flow of payments from Companies to Countries to Government Entities…

 

afg3.jpg

 

The API subsidiary analysis was very powerful – proving that their subsidiaries are already filing transparently is a strong case for advocating that their parent organisations do the same – the One Campaign have been looking for this evidence for a while now, and we found it! How did we achieve this? Through the Alteryx Fuzzy Matching tool – matching up company names from multiple sources and finding solid hits. Something like this has happened at every DataDive so far: in this case, data scientists struggling with the clunky code of a Python fuzzy matcher were astounded when we drop a single tool into a workflow and run Fuzzy Matching in seconds, as part of a train-of-thought analysis!

 

The Data Team also leveraged our fuzzy matching to improve the overall matching and aligning of the different data sets. One area of this team also discovered an unstructured text field showing ‘Payments in Kind’: a transfer of wealth that doesn’t hit the books in the same way as currency. Amazingly, through text analysis they found $78 BILLION of payments in kind around Africa, with some countries (e.g. Botswana) being paid almost exclusively in this way. (At this point, the charity spontaneously started clapping excitedly!!)

 

So, overall: we left the data in a better state than we found it – more connected, more complete and more dependable, and we left the charity with a huge number of stories to follow up, re-run the analysis for different countries and companies and start to use the story points for tougher advocacy – making these governments account for every dollar of their country’s natural resources.

 

If you’re interested in learning more about opportunities like this, please reach out via the Alteryx Community!

 

afg4.png

Nick Jewell

NEW! Find me on the Community at: @DataCurious_Nick

NEW! Find me on the Community at: @DataCurious_Nick

Comments