Spca2014 asking questions the natural way using q & a rapp
0 likes357 views
This document discusses using natural language questions (Q+A) in Power BI. It covers building effective data models, using synonyms, and how the system interprets questions and selects visualizations. The speaker's goals are to increase familiarity with Power BI capabilities in Office 365 and the cloud. Various tips are provided for data modeling best practices like fixing relationships, renaming for clarity, and setting default fields and labels.
![1. “medal count by country for 2012”
Search for interpretations of questions
Score and select best interpretation
Select best visual
from Olympics.xlxs
show medal[medal count], medal[country]
Where Olympics[year] = 2012
from Olympics.xlxs
show medal[medal count], Olympics[host country]
where olympics[year] = 2012
from Sales.xlxs
show customer[country]
where order[order year] = 2012](https://image.slidesharecdn.com/spca2014askingquestionsthenaturalwayusingqarapp-141125122857-conversion-gate01/85/Spca2014-asking-questions-the-natural-way-using-q-a-rapp-21-320.jpg)
![1. “medal count by country for 2012”
Search for interpretations of questions
Score and select best interpretation
Select best visual
from Olympics.xlxs
show medal[medal count], medal[country]
Where Olympics[year] = 2012
from Olympics.xlxs
show medal[medal count], Olympics[host country]
where olympics[year] = 2012
from Sales.xlxs
show customer[country]
where order[order year] = 2012
0.86
0.72
0.46](https://image.slidesharecdn.com/spca2014askingquestionsthenaturalwayusingqarapp-141125122857-conversion-gate01/85/Spca2014-asking-questions-the-natural-way-using-q-a-rapp-22-320.jpg)
![1. “medal count by country for 2012”
Search for interpretations of questions
Score and select best interpretation
Select best visual
from Olympics.xlxs
show medal[medal count], medal[country]
Where Olympics[year] = 2012
from Olympics.xlxs
show medal[medal count], Olympics[host country]
where olympics[year] = 2012
from Sales.xlxs
show customer[country]
where order[order year] = 2012
0.86
0.72
0.46](https://image.slidesharecdn.com/spca2014askingquestionsthenaturalwayusingqarapp-141125122857-conversion-gate01/85/Spca2014-asking-questions-the-natural-way-using-q-a-rapp-23-320.jpg)
Spca2014 asking questions the natural way using q & a rapp
- 2. Power BI: Askingquestions the natural way using Q + A Bjoern H Rapp
- 3. About me Business Developer at Steria Twitter: @bjoern_rapp SharePoint MVP and MCTBlog: www.sharepointviking.com 10 + years in system devEmail: safetox@hotmail.com
- 4. AGENDA •Power BI Primer •Q & A •Building Models •Synonyms •Behind the scenes
- 5. My Goals for the session…not like this.
- 7. POWER BI familiarity of Office and the power of the cloud EXCEL 2013 POWER BI for Office 365 Discover Analyze Visualize Share Find Mobile
- 8. Building Models -Modeling Tips •Fixing Missing Relationships •Renaming Tables and Columns •Choose a default Field Set for each table •Choose a default Label for each table •Fix incorrect data types •Mark year and id columns as “Do not Summarize” •Choose a data category for each data and geo col •Choose a sort by column for relevant column
- 9. Building Models –Missing Relationships POOR BETTER
- 10. Building Models –Renaming tables and columns POOR BETTER
- 11. Building Models–Default field set for each table
- 12. Building Models–Default label for each table
- 13. Building Models –Fix incorrect data types
- 14. Building Models –Do not summarize id and date columns
- 15. Building Models –Choose a Data Category for each date and geography column
- 16. Building Models –Choose a Sort By Column for relevant columns
- 17. Building Models –Normalize your model POOR BETTER
- 18. Synonyms •Unique for Q + A
- 19. 1. Sharing Questions •Sharing links •Featured Questions
- 20. 1. Behind the scenes –Choosing the best answer Search for interpretations of questions Score and select best interpretation Select best visual
- 21. 1. “medal count by country for 2012” Search for interpretations of questions Score and select best interpretation Select best visual from Olympics.xlxs show medal[medal count], medal[country] Where Olympics[year] = 2012 from Olympics.xlxs show medal[medal count], Olympics[host country] where olympics[year] = 2012 from Sales.xlxs show customer[country] where order[order year] = 2012
- 22. 1. “medal count by country for 2012” Search for interpretations of questions Score and select best interpretation Select best visual from Olympics.xlxs show medal[medal count], medal[country] Where Olympics[year] = 2012 from Olympics.xlxs show medal[medal count], Olympics[host country] where olympics[year] = 2012 from Sales.xlxs show customer[country] where order[order year] = 2012 0.86 0.72 0.46
- 23. 1. “medal count by country for 2012” Search for interpretations of questions Score and select best interpretation Select best visual from Olympics.xlxs show medal[medal count], medal[country] Where Olympics[year] = 2012 from Olympics.xlxs show medal[medal count], Olympics[host country] where olympics[year] = 2012 from Sales.xlxs show customer[country] where order[order year] = 2012 0.86 0.72 0.46
- 24. 1. Influencing the best answer Search for interpretations of questions Score and select best interpretation Select best visual Data modeling, synonyms Workbooks in site Data modeling (data types, data categories)
- 26. 1. THANK YOU!