Silent Night, Holy Night — The story behind the famous Christmas song has a lot in common with the content layer and the presentation layer in Data Science
August 2020 Mariapfarr, Salzburg, Austria.
Surrounded by a fantastic landscape, the nice village of Mariapfarr lies in the center of the Austrian Alps and offers a picturesque view into valleys and high mountains.
I return from a hiking tour and the bus drops me off in the center of this small village. Luckily I have to wait 20 more minutes for the next connection. to my hotel.
Luckily, because the place next to the bus station has been renamed to “Joseph Mohr” place it honor the famous son of the village of Mariapfarr. “Famous Joseph Mohr? Why that?
Believe me, you most likely heard of him and also re-cited some of this texts. He wrote the lyrics of the famous Christmas song “Silent Night, holy night”. And I had time to visit the exhibition on that famous song.
Memorable publications start with the content — Back in 1816 Joseph Mohr writes the lyrics of “Silent Night, Holy Night” in the parish of Mariapfarr, Salzburg
Josef Mohr worked as a priest here. Probably inspired by the altarpiece and the hard circumstances under which people had live and work here at the beginning of the 19th century, he was induced to write down his emotions in a text with six strophes. He created an important text which was close to his heart.
Similar to what we, the data scientists do, when we feel we have to solve a difficult analytical problem: we work hard, maybe even many nights to optimize the correctness of our predictive model, we summarize the findings from our data and we produce predictions and forecasts which we want others to understand and to use in their operational processes.
However it might have happened that Joseph Mohr’s important text would have never made it to the public.
This maybe also has happened to you and me already. Even after we built a great machine learning model and discussed the findings with others, the model did not really “take off” and did not get used or appreciated by others.
Content layer meets presentation layer — Joseph Mohr moves to Oberndorf in 1817 and works together with Franz Xaver Gruber
In winter Mariapfarr has an extremely cold climate and Josef Mohr suffered a lot from that. Thus he moved to the village of Oberndorf, Salzburg in 1817, to work there for a while. You can also call it the “first sabbatical in the 19th century” ;-) And it turned out to be a good move.
Also in our data science work environment we often find out that we can only solve a problem or are able to accomplish an important task after we changed our view on the project, paused our work for a while or after we spoke with peers from a different research area to broaden our perspective.
In Oberndorf Joseph Mohr met Franz-Xaver Gruber, who was a composer. They worked together on the song and Franz-Xaver Gruber composed the melody for the guitar and later also for the organ. On December 24th, 1818 they played the song together for the first time in the Christmas Eve service.
With his melody and his lyric, the song touched the people who attended that service a lot. And after that time the song was interpreted at many different places, translated into different languages and is now an integral part of many Christmas celebrations.
Make sure that your data science results do not end up at the “cemetery of forgotten models and ignored findings”
However it took the composer Franz-Xaver Gruber to bring this famous text into the mind of people. Without his work, who knows, the text might still reside completely unknown in the archives of the church of Mariapfarr.
And this is what we data scientists want to avoid. We do not want that our machine learning models end up in the cemetery of forgotten models. We feel unconformable if our important analytical findings are just a side note in a PDF or Powerpoint document. We want to publish our models to the business people and make them understand and re-use our analytical models.
Who is your personal Franz-Xaver Gruber?
When I look back to the year 2020, I want to ask myself “Who has been my Franz-Xaver Gruber this year”. What helped me this year to rise my machine learning from the Jupyter Notebook or the SAS Program Editor and to bring them to life. Where did I successfully mount my content findings to the presentation layer?
Who is your personal Franz-Xaver Gruber? Who and what helped you in 2020 to publish your models and make other colleagues use your insights?
I personally have three answers to this questions:
- The visual and interactive display of my the results of my machine learning models. I performed and lot of Monte Carlo Simulations for the health sector this year. And the visual and interactive capabilities of e.g. SAS Visual Analytics helped many of my clients to understand, to “touch” and use my models and my findings.
- When teaching at the university I also frequently meet many “Franz Xavers”. My student are a very important motivator for me to present and explain complicated content and machine learning results in very simple way.
e.g. by explaining and “experiencing” a lift chart while lining up according to scores on their card. - My clients who want to improve their business processes by being more safe, more profitable or more customer friendly. They only spend time with and my results if I manage to formulate my findings in the most appropriate form for them. And this might differ by persona roles:
I have to use a different presentation layer for the department head (maybe he prefers the sound of the organ)
compared to the IT administrator who wants to package my data science scoring solution into a lightweight container and deploy it in different environments (like carrying a guitar and being able to perform the song at different places).
Two days before Christmas my plan is to inspire you with this text a little bit and emphasize the importance of a good balance of both, content and presentation in our daily data science work.
Merry Christmas
Frohe Weihnachten
