Challenge 3 - Working with Cognitive Services

Prerequisites

  1. Challenge 2 - Working with Data in Power BI should be done successfully.

Introduction

Adventure Works has been getting great business value out of the new model published in Power BI. Now that users are gaining new insights into historical data they are asking for some advanced analytics features to get additional value and insights from the existing data. Adventure Works analytics team quickly realized that they don’t have the expertise to write artificial intelligence models. They would like to leverage some native features of the cloud to apply AI to their existing data. Presently they have a need to do the following:

The data warehouse team has the request to add reviews information to the data warehouse on their feature backlog, so you’ll have to acquire the file from the team that runs Adventure Works website. The team has been kind enough to extract the data for you as a delimited file in blob storage.

Success Criteria

  1. Extend the data flow to bring in data from the product reviews delimited file
  2. Implement cognitive services on the product reviews data to include language, sentiment, and key phrases
  3. In Power BI build one or more reports that demonstrates:
    • Review count by language
    • Which type of bikes have the lowest average sentiment
    • Key phrases for all reviews with a sentiment below .5

Hints

Learning resources

   
Description Links
Use Cognitive Services in Power BI https://docs.microsoft.com/en-us/power-bi/service-cognitive-services
Tutorial: Use Cognitive Services in Power BI https://docs.microsoft.com/en-us/power-bi/service-tutorial-build-machine-learning-model

Next challenge (Building Machine Learning in Power BI) >