Systems and methods for data insight generation and display

Patent metadata

Field	Value
Publication number	US-11989174-B2
Application number	US-202017026749-A
Country	US
Kind code	B2
Filing date	Sep 21, 2020
Priority date	Feb 5, 2020
Publication date	May 21, 2024
Grant date	May 21, 2024

Abstract

Disclosed herein are systems and methods for intelligent generation and display of insights using information in a data repository. For example, disclosed herein are methods for generating and displaying insights using initial data from a data repository, and intelligently/automatically proposing, generating, and displaying further insights using previously-generated insights.

First claim

What is claimed is:

1. A method of generating and displaying insights using repository data, the method comprising:
receiving, at a processor, a request from a user to analyze a metric generated using data from a repository;
accessing a semantic graph to determine a plurality of attributes for analyzing the metric;
using the plurality of attributes associated with the data, creating a plurality of analyses of the metric, wherein each analysis includes a graph or chart, and each analysis of the plurality of analyses is based on at least one attribute;
testing each analysis from the plurality of analyses for at least one of: a statistically significant trend, a pattern, or an outlier;
selecting an analysis from the plurality of analyses based on the testing each analysis;
generating an insight based on the selected analysis;
based on the generated insight and the selected analysis, generating a plurality of recommended actions, the plurality of recommended actions including replacing a first attribute of the data used in the selected analysis with a child of the first attribute;
outputting, to a user device, a display of the selected analysis, the generated insight, and the plurality of recommended actions;
receiving, from the user device, an instruction to replace the first attribute with the child of the first attribute in response to the output display;
based on the received instruction to replace the first attribute with the child of the first attribute, generating a further analysis, a further insight, and a plurality of further recommended actions that are based on the further analysis and the further insight; and
outputting, to the user device, a display of the further analysis, the further insight, and the plurality of further recommended actions.

2. The method of claim 1 , further comprising:
recursively performing the steps of:
receiving, from the user device, an instruction to take an action in response to the output display;
based on the received instruction, generating an additional analysis, an additional insight, and a plurality of additional recommended actions that are based on the further analysis and the further insight; and
outputting, to the user device, a display of the additional analysis, the additional insight, and the plurality of additional recommended actions.

3. The method of claim 1 , wherein the step of outputting, to the user device, the display of the further analysis, the further insight, and the plurality of further recommended actions comprises updating a previously output display of the selected analysis, the generated insight, and the plurality of recommended actions.

4. The method of claim 1 , wherein the received instruction to take an action in response to the output display includes one of:
a selection of a recommended action; or
an instruction input into a user device in spoken or written natural language.

5. The method of claim 1 , wherein each of the plurality of attributes that describe the data includes a categorization or value assigned to data points in the data, wherein the categorization or value is not reflected in the metric.

6. The method of claim 1 , further comprising:
using repository data and/or user preference data, training a machine learning model to choose attributes suitable for creating an analysis of a metric; and
using the trained machine learning model to choose the plurality of attributes for creating the plurality of analyses of the metric.

7. The method of claim 1 , wherein the step of generating a plurality of recommended actions comprises:
accessing a database of potential recommended actions and conditions for recommending each potential recommended action, wherein the potential recommended actions are assigned a priority order;
identifying whether the conditions for recommending each potential recommended action have been met; and
selecting a predetermined number of potential recommended actions for which the conditions have been met.

8. The method of claim 1 , further comprising:
receiving, from a user device, an instruction to add a generated insight to a primary visual interface; and
adding the generated insight to the primary visual interface.

9. The method of claim 1 , further comprising:
receiving, from a user device, a request to share a generated insight; and
creating a copy of the generated insight for attachment to a digital message.

10. A method of generating and displaying insights using repository data, the method comprising:
receiving, at a processor from a user device, a request to analyze a metric generated using data from a repository;
accessing a semantic graph to determine a plurality of attributes for analyzing the metric;
using the plurality of attributes that describe the data, creating a plurality of analyses of the metric, wherein each analysis includes a ranked list or map, and each analysis of the plurality of analyses is based on at least one attribute;
reviewing the generated analyses to select a predetermined number of analyses, each of which (1) shows an outlier, a pattern, or a trend, or (2) is based on an attribute of interest to a user;
generating an insight based on each selected analysis;
for each of the generated insights, generating a plurality of recommended actions, the plurality of recommended actions including replacing a first attribute of the data used in the selected analysis with a child of the first attribute;
outputting, to the user device, an interactive display including each selected analysis, the insight generated based on each selected analysis, and the plurality of recommended actions generated for each insight;
receiving, from the user device, an instruction to replace the first attribute with the child of the first attribute in response to the output display;
based on the received instruction to replace the first attribute with the child of the first attribute, generating a further analysis, a further insight, and a plurality of further recommended actions that are based on the further analysis and the further insight; and
outputting, to the user device, an updated interactive display including the further analysis, the further insight, and the plurality of further recommended actions.

11. The method of claim 10 , wherein the step of receiving, from the user device, an instruction to take an action in response to the interactive display includes receiving, from a user device, text input by a user, and wherein the method further comprises:
analyzing the text to identify a query or command; and
performing an action in response to the query or command.

12. The method of claim 11 , wherein the action includes one of:
generating an additional analysis or insight;
comparing two or more analyses; or
outputting an answer to a question.

13. The method of claim 10 , wherein each generated insight is a summary of a conclusion based on an analysis, and wherein the step of generating an insight based on each selected analysis includes:
identifying whether the analysis includes an outlier;
identifying whether the analysis exhibits a pattern; or
identifying whether the analysis exhibits a trend in data over time.

14. The method of claim 10 , further comprising:
receiving an instruction to make a prediction based on an analysis;
identifying a trend over time exhibited by the analysis; and
predicting a future data point using the identified trend over time.

15. The method of claim 14 , further comprising:
receiving an instruction to output a suggested method for changing a predicted future data point;
identifying one or more attrib

Assignees

• Microstrategy Inc

Inventors

• Mei Yingchun
• Zhong Xiaodi
• Li Jiacheng
• Jiang Wei
• Liu Shu
• Zhang Lina
• Li En

Classifications

• G06F16/2423 Primary

  Interactive query statement specification based on a database schema · CPC title

• G06F16/24564

  Applying rules; Deductive queries · CPC title

• G06F16/24578

  using ranking · CPC title

• G06N5/045

  Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence · CPC title

• G06N20/00

  Machine learning · CPC title