What if you had the ability to use common natural language to find answers to your own data without having to rely on complex sql statements, long winded code or cumbersome data scorecards and dashboards? What if interactive scorecards and dashboards could be accessible to you by simply posing questions and getting answers back?
One of the latest and greatest things to come out of O365 is Power BI Q&A and that is exactly what this new tool coming out of Power BI can do. As you type in the question, Power BI Q&A picks the best visual to display your answer; and the visual changes dynamically as you modify the question. You can also modify the display to best fit your interactive data report.
You and your marketing team need to strategize the next marketing campaign for the upcoming fiscal year. In prior scenarios, you would have a team of data analyst compiling and creating rigid scorecards and dashboards in a PowerPoint presentation. Questions that come up in meetings are tabled and teams would go back and “mine” the data again. Interactive reports is not an option. Strategic decisions are not so strategic among a swirl of data report change chaos.
With Power BI Q&A, the ability to go back into the data and pose questions and re-pivot reports is plausible.
There are two parts to setting up this feature on O365. The first part is the “modeling” and the second part is “consuming”. By modeling the data, you are able to join tables, sort and organize as well as define synonyms or keywords for your data. After modeling as been done, now your data is ready to be consumed. By consuming, we mean enable to be searched on Power BI Q&A.
For further information on setting up, defining and editing synonyms, click here.
After you have uploaded your Excel workbook up to the Power BI site, enabled Q&A to your workbook, you will now be able interact with it using Q&A.
The Q&A part of Power BI is using natural language (NL) models in Power Pivot to get the results back. Consider this technology working the same way that search engines mine data, except your data is your data report!
By leveraging natural language understanding engine based on traditional search and pattern matching techniques, you are able to bring data to the non-technical user. Business decisions can be made on demand without having to wait for results. This is a win-win for the business community as well as the IT professionals managing the back end reporting functions.
There are some things to consider when posing questions. For example a tip would be to use aggregation keywords in your questions. For example, Q&A understands words such as less than, maximum, number of, count of, average, between, before, and more.