Visualisation & Storytelling in Data Science & Analytics
6 likes2,380 views
The document provides an overview of data visualization and storytelling in data science and analytics. It discusses key concepts like what data visualization is, compelling reasons to visualize data like Anscombe's Quartet, visualization in the context of analytics workflows, components of effective storytelling, considerations for presentation, guidelines for data storytelling, and examples of interesting data visualizations. Throughout the document, the author emphasizes best practices like keeping visualizations clear, addressing the intended audience, and avoiding bias.
![@FelipeRego
feliperego.com.au
2
What is Data Visualisation
Many other definitions exist
“Visualization is the use of interactive visual representations of data to amplify
cognition.” (Card, 2008)
“The purpose of [information] visualization is to amplify cognitive
performance, not just to create interesting pictures. Information visualizations
should do for the mind what automobiles do for the feet.” (Card, 2008)
“[Information] visualization promises to help us speed our understanding and
action in a world of increasing information volumes.” (Card, 2008)
Sources: Wikipedia, http://www.infovis-wiki.net/ @FelipeRego](https://image.slidesharecdn.com/visualisationstorytellingindatascienceanalytics-180930134456/85/Visualisation-Storytelling-in-Data-Science-Analytics-2-320.jpg)
Visualisation & Storytelling in Data Science & Analytics
- 1. 1 feliperego.com.au Visualisation & Storytelling in Data Science & Analytics FELIPE REGO DATA SCIENCE & ANALYTICS PARTNER @FelipeRego
- 2. @FelipeRego feliperego.com.au 2 What is Data Visualisation Many other definitions exist “Visualization is the use of interactive visual representations of data to amplify cognition.” (Card, 2008) “The purpose of [information] visualization is to amplify cognitive performance, not just to create interesting pictures. Information visualizations should do for the mind what automobiles do for the feet.” (Card, 2008) “[Information] visualization promises to help us speed our understanding and action in a world of increasing information volumes.” (Card, 2008) Sources: Wikipedia, http://www.infovis-wiki.net/ @FelipeRego
- 3. @FelipeRego feliperego.com.au 3 Compelling reason to visualise data Anscombe’s Quartet Sources: Wikipedia @FelipeRego ● Importance of graphing data before analyzing it ● effect of outliers on statistical properties
- 4. @FelipeRego feliperego.com.au 4 Visualisation in the context of analytics Cognitive amplification flow Sources: General Assembly’s Analytical Workflow, Personal View @FelipeRego Identify Obtain Understand Prepare Analyse Present Identify Business Problem Opportunity for Analysis Brief from Business Meeting outcomes Internal Systems External Providers Social Media Spreadsheets Identify tools to work Internal client workshops Individual and group knowledge Industry benchmarks Peer review Any necessary aggregation Sampling Metric definitions and indicators Plan for delivery Answer questions Suggest novel ways to view the answers Summary Stats / Outliers Modelling and Advanced Analytics Narrative Caveats and Assumptions Static / Interactive Action-driven Future plans
- 5. @FelipeRego feliperego.com.au 5 What makes a good data visualisation Sources: Personal View, Knowledge is Beautiful @FelipeRego ● Clear problem definition ● Questions / Hypothesis ● Visualise to Analyse ○ Exploratory ○ Hypothesis testing ● Visualise to Communicate ○ Static ● Visualise to Empower ○ Interactive ○ Leads to Action
- 6. @FelipeRego feliperego.com.au Sources: Personal View, Edward Tufte, A. Abela, Slide Chooser http://extremepresentation.typepad.com/files/slide-chooser-11x17.pdf 6 Choosing the components of your story @FelipeRego Complex ideas communicated with clarity, precision and efficiency
- 7. @FelipeRego feliperego.com.au 7 Considerations during the ‘Present’ phase Cognitive amplification flow delivery Sources: General Assembly’s Analytical Workflow, Personal View, Fry, B. (2008). Visualizing data. Sebastopol, CA: O’Reilly Media, Inc @FelipeRego Identify Obtain Understand Prepare Analyse Present ● Filter: Remove all but the data of interest (caveat the exclusions though). ● Represent: Choose a basic visual model, such as a bar graph, list, or tree. ● Refine: Improve the basic representation to make it clearer and more visually engaging. ● Interactivity: Add methods for manipulating the data or controlling what features are visible.
- 8. @FelipeRego feliperego.com.au 8 Data Storytelling general guidelines Sources: Personal View, Tableau, HBR @FelipeRego ● Have the objectives clear ● Have underlying questions to be answered ● Draw up your storyline ● Consider Audience ● Be genuine / original ● Be visual but respect simplicity and clarity ● Avoid bias ● Consider purpose and expectations ● State your assumptions and caveats ● Don’t fall into the ‘it looks cool’ trap
- 9. @FelipeRego feliperego.com.au 9 Examples of interesting data vis Sources: http://bl.ocks.org/kerryrodden/7090426 @FelipeRego This example shows how it is possible to use a D3 sunburst visualization (partition layout) with data that describes sequences of events.
- 10. @FelipeRego feliperego.com.au 10 Examples of interesting data vis Sources: http://exposedata.com/parallel/ @FelipeRego
- 11. @FelipeRego feliperego.com.au 11 Examples of interesting data vis Sources: http://www.nytimes.com/interactive/2009/07/31/business/20080801-metrics-graphic.html?_r= 1& @FelipeRego The American Time Use Survey asks thousands of American residents to recall every minute of a day. Here is how people over age 15 spent their time in 2008.
- 12. @FelipeRego feliperego.com.au 12 Examples of interesting data vis Sources: http://bost.ocks.org/mike/sankey/ @FelipeRego Sankey diagrams visualize the magnitude of flow between nodes in a network.
- 13. @FelipeRego feliperego.com.au 13 Examples of interesting data vis Sources: http://euclid.psych.yorku.ca/SCS/Gallery/excellence.html @FelipeRego
- 14. @FelipeRego feliperego.com.au 14 Examples of interesting data vis Sources: David McCandless, Information is Beautiful @FelipeRego
- 15. @FelipeRego feliperego.com.au 15 Examples of interesting data vis Sources: http://flowingdata.com/2015/12/15/a-day-in-the-life-of-americans/ @FelipeRego
- 16. @FelipeRego feliperego.com.au 16 Examples of interesting data vis Sources: http://peterbeshai.com/buckets/app/#/playerView/201935_2015 @FelipeRego
- 17. @FelipeRego feliperego.com.au 17 Context is essential in graphical data Sources: The Visual Display of Quantitative Information by Edward Tufte @FelipeRego Graphics must not quote data out of context. “Compared to what?”
- 18. @FelipeRego feliperego.com.au 18 Graphical choice is important Sources: Visualize This, Nathan Yau @FelipeRego
- 19. @FelipeRego feliperego.com.au 19 Look for anomalies Sources: Online @FelipeRego Items, events or observations which do not conform to an expected pattern or other items in a dataset
- 20. @FelipeRego feliperego.com.au 20 Be extra careful with averages Sources: Personal View @FelipeRego Calculation House 1 House 2 House 3 House 4 House 5 Result Mean (Average) 200,000 1,200,000 1,300,000 1,500,000 18,000,000 4,440,000 Median 200,000 1,200,000 1,300,000 1,500,000 18,000,000 1,300,000 Most data points in real life are somewhat skewed
- 21. @FelipeRego feliperego.com.au 21 Correlation does not imply causation Sources: http://www.nejm.org/doi/full/10.1056/NEJMon1211064 @FelipeRego The principal finding of this study is a surprisingly powerful correlation between chocolate intake per capita and the number of Nobel laureates in various countries. Of course, a correlation between X and Y does not prove causation but indicates that either X influences Y, Y influences X, or X and Y are influenced by a common underlying mechanism. However, since chocolate consumption has been documented to improve cognitive function, it seems most likely that in a dose-dependent way, chocolate intake provides the abundant fertile ground needed for the sprouting of Nobel laureates. Obviously, these findings are hypothesis-generating only and will have to be tested in a prospective, randomised trial.
- 22. @FelipeRego feliperego.com.au 22 Pie charts can be tricky Sources: http://www.engadget.com/2008/01/15/live-from-macworld-2008-steve-jobs-keynote/ @FelipeRego
- 23. @FelipeRego feliperego.com.au 23 Good Data Vis Checklist Sources: Personal View @FelipeRego ● Goal: Define your business problem ● Hypothesis: Define questions that will help solve the problem ● Narrative: Draw up your storyline, almost like a film director ● Style / Format: Keep it clear, easy and clean (Less is More) ● Integrity: Be careful with bias, graphical choices and exclusions ● Audience: Think about who the viz is for and who is going to use it ● Expectation: Check whether the final product is aligned to objectives
- 24. “The greatest value of a picture is when it forces us to notice what we never expected to see.” John Tukey 24 @FelipeRego
- 25. 25 Thank you! feliperego.com.au DATA SCIENCE & ANALYTICS PARTNER FELIPE REGO @FelipeRego
- 26. feliperego.com.au @FelipeRego 26 With an extensive industry experience as well as a deep, technical analytical expertise, Felipe is often required by marketing, sales, finance, technology and strategy teams to provide support and to deliver robust analytical solutions that are easy to use, understand and implement. Felipe’s unique methodology focuses on a holistic organisational approach to using data and science to improve performance and reduce costs. Felipe helps organisations and teams with Data Science and Analytics Strategy, Predictive Analytics and Machine Learning solutions, Data Visualisation and Insights Automation and Analytical Training and Workshops. Felipe is also an analytics instructor helping disseminate practical, actionable analytics and data visualisation techniques in both classrooms and online settings. Organisations that partner with Felipe end up with a more engaged workforce and individuals feel more prepared to step up to their next challenge using data and analytics in their day-to-day jobs. When Felipe is not partnering with clients or helping students, he’s an M.Phil. research candidate in Learning Analytics at the University of Sydney making sense of students’ digital traces and the role learning analytics dashboards play in influencing learning outcomes. Felipe is also a blogger in predictive analytics, statistical learning and data visualisation. Last year alone, Felipe had over 62,000 visitors to his blog from more than 180+ countries. Some of his articles have been ranked #1 in google search and referenced by many sources and leading educational organisations including referrals / mentions from StackOverflow, Udacity, Western Michigan University, UC Santa Barbara, Edinburgh Napier University, among others. About Felipe Felipe is a leading advanced analytics and data science partner, helping teams build, manage and enhance their data science and visualisation solutions in a strategically-aligned, commercially-oriented and customer-centred way.
- 27. @FelipeRego feliperego.com.auCapabilities & Services 27 Felipe helps organisations make sense of their current analytics capability and provide them with a clear roadmap to find future growth opportunities with a distinctive balance between technical know-how and business experience. Data Science & Analytics Strategy From marketers requiring a robust segmentation model of their customers or a recommendation engine for their products to sales and finance leaders wanting to improve their forecasting ability, Felipe partners with teams to implement a variety of machine learning solutions. Predictive Analytics & Machine Learning Felipe partners with teams to help declutter their data assets and provide them with scalable, interactive and actionable data visualisation solutions. As a result, organisations become more transparent and can more seamlessly monitor their performance. Data Visualisation & Insights Automation Felipe partners with both educational institutions and corporate organisations to help disseminate practical, actionable and easy-to-understand data science, predictive analytics and data visualisation techniques to students and staff. Analytical Training & Corporate Workshops Sources: feliperego.com.au/about
- 28. @FelipeRego feliperego.com.auSome of the Brands I’ve Worked With 28 Long-term or ad-hoc consulting and permanent projects with leading brands Sources: feliperego.com.au/about
- 29. @FelipeRego feliperego.com.auExamples of Consulting Projects 29 Case Studies Sources: https://feliperego.github.io/cases/
- 30. @FelipeRego feliperego.com.auWhat Some of my Clients Say 30 Testimonials from consulting clients “We needed someone who could come in and help define and build a cutting-edge analytics solution for customer retention as part of a larger analytics investment. We at 3P have been laser-focused on our customers and further utilising advanced analytics for that was an easy decision. I was impressed by Felipe’s self-directed approach. Differently from other data science professionals we worked with, Felipe was really good at maintaining constant communication with the executive team and stakeholders. He also made sure that the results were easily interpretable for non-technical folks such as product, sales and design teams. But the most valuable thing for me was that he aligned technical solution to commercial benefits. At the end of the engagement, we had a clearer picture of the potential benefits of the solution he built which made it easier to adopt."- Simon Perry, Chief Information Officer, 3P Learning, 15 March 2018, Simon was a client of Felipe’s “Felipe was recommended to us as an experienced analytics expert who, in partnership, communicated technical terms in an easy-to-understand, business-friendly language. This was part of our recipe for success for winning the project. Felipe brought an impressive knowledge and wealth of experience which both impressed us and the client. He supported many conversations with the client during the process which helped to confirm why we were the team to deliver. Not only was the knowledge there but Felipe was hands on in translating the concept and in building the prototypes with us. He was a pleasure to work with and I would definitely recommend him and his expertise to anyone needing an analytics expert."- Tracy Voong, Data Planning & Strategy Director, The Works, 15 October 2017, Tracy was a client of Felipe’s Sources: Felipe Rego’s LinkedIn Profile
- 31. @FelipeRego feliperego.com.auWhat Some of my Clients Say 31 Testimonials from students “Data Science is one of 21st century digital wonders. Felipe onboards you to the data science with ease and makes you comfortable to start applying data science from the first day. It was a great workshop.” - Petras Janulevicius VR design & development at Diesel Immersive February 5, 2018, Petras was a client of Felipe’s “I went to Felipe's intro to data analytics & data science workshop recently at Academy Xi. Felipe teaches with passion and uses plain language so that students can follow. I picked up useful data analytics skills throughout the 3 hours workshop.” - Sandy Tsui CPA | Business Partner | Data Analytics | Data Visualisation | Passionate to make an impact January 28, 2018, Sandy was a client of Felipe’s “Felipe is very efficient at delivering highly technical knowledge. His teaching and mentoring methods are very engaging while also challenging. Felipe's real-life example approach made the Data Analytics course even more valuable; and provided me with the confidence to incorporate advanced data analysis into my current job.” - Jose Castell Technical Product Specialist at Roche (Haematology & Urinalysis) February 11, 2017, Jose was a client of Felipe’s “Felipe was my instructor for one of General Assembly’s Data Analytics courses in 2016. He was a highly effective and engaging teacher who was genuinely interested in his students. He routinely added more to the course (industry insights, best practices, things books don’t tell you, etc), which was incredibly valuable and helpful. I especially appreciated how encouraging and responsive he was when students asked unique or challenging questions. That was a very important point for me because that kind of dialogue keeps me motivated for what’s next. Felipe is a wonderful instructor. I highly recommend him!” - Byron Allen Data Analysis & Architecture June 30, 2016, Byron was a client of Felipe’s Sources: Felipe Rego’s LinkedIn Profile
- 32. @FelipeRego feliperego.com.auThought Leadership / Blogging Facts 32 Over +60,000 visitors / year from +180 countries - many content referrals Example of Facebook Referral by the Western Michigan University about a blog I wrote on Principal Component Analysis using R Example of one of my most popular blog posts which has been ranked #1 in google searches out of more than 9.2 million other similar results Sources: Google, Facebook
- 33. @FelipeRego feliperego.com.auVideo: Latest Talks + Workshop 33 Sources: Adapted from https://www.facebook.com/vicinitymarketing/videos/1877089065682753/
- 34. @FelipeRego feliperego.com.auPhotos: Latest Talks + Workshop 34
- 35. 35 I would love to hear your story, share my experience and learn about your challenges. feliperego.com.au @FelipeRego
- 36. 36 FELIPE REGO DATA SCIENCE & ANALYTICS PARTNER feliperego.com.au [email protected]@FelipeRego
- 37. 37 feliperego.com.au Visualisation & Storytelling in Data Science & Analytics FELIPE REGO DATA SCIENCE & ANALYTICS PARTNER @FelipeRego
- 38. Credits Slide Microsoft® and PowerPoint® are trademarks or registered trademarks of Microsoft Corporation. © 2015 Google Inc, used with permission. Google and the Google logo are registered trademarks of Google Inc. Google Drive® is a registered trademark of Google Inc. Photos in this template were downloaded from freedigitalphotos.net. Attribution is located in each slide notes and the Credits slide. The Template provides a theme with four basic colors: The backgrounds were created by Free Google Slides Templates. Vectorial Shapes in this Template were created by Free Google Slides Templates and downloaded from FreePik.com. Icons in this Template are part of Google® Material Icons and flaticons.com. Shapes & Icons Backgrounds Images Fonts Color Palette Trademarks The fonts used in this template are taken from Google fonts. ( Trebuchet, Arial) You can download the fonts from the following url: https://www.google.com/fonts/ #566579ff #666666ff #ee795bff #d8d8d8c7 38