Analyzing Text Data with Google Sheets and Cloud Natural Language (Cloud Next '18)

Some of the most valuable insights for businesses come from free-form user feedback, but text data can be difficult to process and summarize in a scalable way. This session will show how to use Cloud Natural Language to open up opportunities for analyzing qualitative feedback alongside quantitative data. I’ll show you how to use Google Forms to collect feedback, Google Sheets and Cloud Natural Language to analyze it, and Data Studio to visualize the insights; a powerful yet lightweight solution!
Event schedule → g.co/next18
Watch more Collaboration & Productivity sessions here → bit.ly/2LldTsw
Next ‘18 All Sessions playlist → bit.ly/Allsessions
Subscribe to the Google Cloud channel! → bit.ly/NextSub
#CustomizingGoogleWorkspace
re_ty: Publish; product: Cloud - AI and Machine Learning - Cloud Natural Language, Workspace - Applications - Sheets, Cloud - Data Analytics - Dataprep, Workspace - Applications - Apps Script; fullname: Alicia Williams; event: Google Cloud Next 2018;

[@lazarinastoy]

Amazing tutorial, Alicia!

[@nafizhaq4554]

there is a tab called "Sample data" - is that tab is used for tagging and trainning the model ?

[@aliciawilliams5520]

The sample data tab was used to populate new form responses during the demo. I essentially saved back some data to send through the form during the demo.

[@nafizhaq4554]

Alicia Williams thanks Alicia, can the same exercise could be done without using API but a actual GCP console where data could be uploaded ?

[@aliciawilliams5520]

@@nafizhaq4554 There is not a UI for the Natural Language API that allows you to make predictions in the console. In the AutoML NL UI, there is a possiblity to make one prediction at a time. I have sent this feedback to the product team so hopefully it could be an option in the future.

[@johng5295]

Thanks in a million. You have an inventive mind. You got the best stuff on earth. I have just subscribed. Do you teach these stuffs. If yes, how can I join? Any book?

[@aliciawilliams5520]

Thanks John! Our team just launch a solutions gallery that have example Apps Script to help start off projects: cloud.google.com/blog/products/g-suite/new-solution-gallery-helps-you-customize-and-extend-gsuite-apps. I also do a bit of blogging on Medium and tweet @presactlyalicia. Ben Collins has great learning material on Sheets/Apps Script on his website benlcollins.com. Hope that helps!

[@simoverse]

Thanks Alicia

[@danielmarcellino1360]

Excelente Alicia!!

[@sumansourabh5913]

Great Video

Can you share the spreadsheet?

[@aliciawilliams5520]

Thanks for watching! All of the code, including a spreadsheet template is available in the Github repo: github.com/aliciawilliams/analyzingtext

[@daveojeda]

At 37:10 are the links to all the code.

[@studybronzwik2404]

@@aliciawilliams5520 You have made me quit film-making and hooked to something I new I could do. Thanks

[@nafizhaq4554]

@@aliciawilliams5520 Hi Alicia, that's a great content it might help me with my MSc Dissertation in NLP sentiment analysis of user content. Could you help me with my project if possible ? I just need some basic understanding of this function and apps script to attribute sentiment analysis of around 20k reviews in a google sheet.

[@aliciawilliams5520]

@@nafizhaq4554 I wish I could help, though I can point you to some resources. I recommend looking at the demo to see how the different features work for your text data (cloud.google.com/natural-language#natural-language-api-demo) and then consider AutoML if you find that the pre-trained model doesn't quite get you what you need. I wrote a set of two blog posts on how to use AutoML NL and then process predictions using Apps Script (medium.com/google-cloud/analyzing-sentiment-of-text-with-domain-specific-vocabulary-and-topics-726b8f287aef). Hope this helps!