KB442148: Natural Language Query in A Nutshell in MicroStrategy Web

User Input Contains	Sample Query Pattern	Sample Question	Widget Used	Visualization and Insights
Only 1 Metric	What is the [metric]	What is the Number of Flights?	KPI
1 Metric + 1 Time Attribute + 0~N Attributes	What is the [metric] by "time attribute"	Please show Profit by Month.	Line Graph
1 Metric + 1 Geo Attribute + 0~N Attribute	Distribution of [metric] by "geo attribute" A map of [metric] by "geo attribute" and "attribute 2"	What is the Revenue by City?	Map
2 Metrics + 1 Attribute	Correlation between "metric 1" and "metric 2" by [attribute]	Can you tell me the correlation between Cost and Revenue by Customer Age?	Scatter-plot
0~N Metric + 1~N Attribute Or 2~N Metric only	What is [metric] by "attribute 1" and by "attribute 2"	What is the Profit by ‌Item and by ‌Customer Gender‌?	Bar Graph

• Grid
• Heat map, Heatmap
• Tree map
• Bar, Bar Chart, Bar Graph
• Line, Line Chart, Line Graph
• Pie, Pie Chart, Pie Graph
• Network, Network Visualization
• KPI
• Map

• sort by + [attribute]/[metric]
• sorted by + [attribute]/[metric]

• [metric] + bigger than + xxx (a number)
• [metric] + larger than + xxx (a number)
• [metric] + greater than + xxx (a number)
• [metric] + smaller than + xxx (a number)
• [metric] + less than + xxx (a number)
• [metric] + is + xxx (a number)
• [metric] + are + xxx (a number)

Within the filtering condition of the created visualization, you can see that an advanced filter condition, “[Flights Cancelled] Greater than [1500]”, was created. This was translated from natural human language to a filter on a MicroStrategy visualization.

• Top ‘N’ + [attribute]/[metric]

With this query, an advanced filter was created:

• Average + [attribute]/[metric] -> avg (attribute/metric)
• Maximum + [attribute]/[metric] -> max (attribute/metric)
• Max + [attribute]/[metric] -> max (attribute/metric)
• Minimum + [attribute]/[metric] -> min (attribute/metric)
• Min + [attribute]/[metric] -> min (attribute/metric)
• Sum + [attribute]/[metric] -> sum (attribute/metric)
• Sum of [attribute]/[metric] -> sum (attribute/metric)
• Count + [attribute]/[metric] -> count (attribute/metric)
• Count of + [attribute]/[metric] -> count (attribute/metric)

• The keyword “Sort by” only supports ascending sorting.
• Numbers in the query can only be numeric, not words (e.g., 20 not twenty).
• NLQ filtering capabilities are currently only available with metric-based queries, NOT attributes.
• NLQ ranking capabilities do not include the following keywords: “highest”, “lowest” “largest”, or “smallest”.
• With NLQ-generated derived metrics, the user is not able to specify aggregation function parameters (e.g., “count distinct”). However, NLQ-generated derived metrics can be edited and manually adjusted to set more advanced parameters.
• When a user specifies KPI visualizations in the NLQ, all attributes will be ignored.
• If there are attributes/metrics with the same name from different datasets in a Dossier, there is no guarantee which dataset the object/data is coming from (as you cannot specify which dataset to use).
• Currently, “Sort-by” aggregation is unavailable (e.g., “Show me country sort by count of state as a grid”).
• You must type exact attribute/metric name in your query (e.g., no plurals).
• Changing the order of attributes/metrics in your query may lead to slightly different visualizations, as the order of objects within the query is used to decide how the visualization is built.