This reference solution automates the process of analyzing the sentiment of customer reviews and visualizing the results in Amazon QuickSight. It also introduces parallel processing by leveraging AWS Step Functions Map state in Distributed mode and SDK service integrations.
Download AWS CloudFormation template reviewsentimentanalysistemplate.yaml and reviews file allreviews.csv to your local machine.
From the AWS CloudFormation console, create an AWS CloudFormation stack by going to Stacks, Create stack, With new resources (standard). Upload the template as shown below and click Next:
Enter a Stack name and click Next. Check the checkbox that says I acknowledge that AWS CloudFormation might create IAM resources with custom names. and click Next. Click Submit and wait until the stack creation is completed.
Alternatively, you can create the AWS CloudFormation stack with the AWS CLI by editing the following example:
aws cloudformation create-stack --stack-name <your-stack-name> --template-body file://reviewsentimentanalysistemplate.yaml --region <your-aws-region> --capabilities CAPABILITY_NAMED_IAM
Confirm that all the stack resources were created successfully, including Amazon S3 buckets reviewsinputbucket1-<AccountID> and reviewsoutputbucket1-<AccountID>, by checking the Resources tab in the stack details page.
Upload reviews file allreviews.csv to Amazon S3 bucket reviewsinputbucket1-<AccountID>. This should generate one JSON file per processed review in Amazon S3 bucket reviewsoutputbucket1-<AccountID>.
To automatically discover and catalog those JSON files as well as establish the integration with Amazon QuickSight, you will start by creating an AWS Glue crawler. From the AWS Glue console, click on Crawlers, then Create crawler, for Name enter scan-sentiment-output and click Next. Click Add a data source, for Data source select S3, for Location of S3 data select In this account, for S3 path enter s3://reviewsoutputbucket1-<AccountID>, click Add an S3 data source and click Next. Under IAM role click Create new IAM role, for Enter new IAM role enter AWSGlueServiceRole-sentimentgluecrawlerrole, click Create and click Next. Under Target database click Add database, for Database name enter sentiment-results. Under Crawler schedule set Frequency to Custom (cron expression), enter expression 1/10 * * * ? * which should translate to Every 10 minutes, starting at 1 minutes past the hour and click Next. Click Create crawler and wait for the crawler to complete its first run and create Database sentiment-results and Data Catalog table reviewsoutputbucket1-<AccountID>.
From the Amazon Athena console, go to Query editor, for Data source select AwsDataCatalog, for Catalog select None, for Database select sentiment-results and confirm that table reviewsoutputbucket1-<AccountID> is showing along with its two columns sentiment and reviewid.
Open the Amazon QuickSight console as an administrator user. Click on the user icon in the top right corner, set the region to N. Virginia and click Manage QuickSight. Go to Security & permissions and if Amazon Athena and Amazon S3 are not in the list of services that Amazon QuickSight is granted access to click on Manage and add them. Go to SPICE capacity and check that you haven't reached your available capacity; if you don't have enough unused capacity you can purchase more by clicking Purchase capacity. Go back to the Amazon QuickSight console by clicking on QuickSight in the top left corner. If your deployment is in a different region, switch to that region and go to Datasets, New dataset, under From new data source select Amazon Athena, for Data source name enter reviews-sentiment-analysis, click Create data source, for Catalog select AwsDataCatalog, for Database select sentiment-results, for Tables select reviewsoutputbucket1-<AccountID>, select Import to SPICE for quicker analytics and click Visualize. Go to Analyses, New Analysis, from Your Datasets choose dataset reviewsoutputbucket1-<AccountID> and click Create analysis. Select field sentiment, add a Donut chart visual and Horizontal bar chart visual and publish the visualization as a dashboard by clicking Publish.
Going forward, any additional reviews you upload to Amazon S3 bucket reviewsinputbucket1-<AccountID> will get automatically processed. The Amazon QuickSight dashboard will also account for those new reviews once the AWS Glue crawler discovers the newly generated JSON files in Amazon S3 bucket reviewsoutputbucket1-<AccountID> in 10-minute increments.
Open the Amazon QuickSight console as an administrator user and delete the dashboard, analysis and dataset.
Open the AWS Glue console and delete the crawler and database.
Open the AWS CloudFormation console and delete the stack.
Alternatively, you can delete the AWS CloudFormation stack with the AWS CLI by editing the following example:
aws cloudformation delete-stack --stack-name <your-stack-name>
See CONTRIBUTING for more information.
This library is licensed under the MIT-0 License. See the LICENSE file.