Quickbase Sentiment Analysis

An AWS serverless app that provides a sentiment analysis capability to Quickbase through Pipelines.

Installation

View the end-to-end walkthrough to install and configure everything needed to add sentiment analysis to your Quickbase applications.

To install the Amazon services, sign in to the AWS Management Console and deploy the quickbase-sentiment-analysis app from the Serverless Application Repository. You may also take the steps detailed in the Development section to install the Amazon services using the SAM framework outside of the Serverless Application Repository model.

Quickbase requires basic schema and a Pipeline as detailed in the end-to-end walkthrough video.

Usage

The serverless app provides an API that Pipelines consumes to detect sentiment. Create a pipeline, and craft the json payload through the Body setting of the Webhooks Channel's Make Request action. An example of JSON payload that the API accepts is below:

{
    "text": "This is a simple but useful application.",
    "language": "en"
}

Let's say the Pipeline's Step A is the Quickbase channel's Record Created action. Assuming you have a field named Comment in the table in which the record was created, you might build the json above by configuring the Body setting in the Make Request action to {{ {'text': a.comment,'language':'en'}|to_json }}.

On success, the API will return a response like the one below:

{
    "Sentiment": "POSITIVE",
    "SentimentScore": {
        "Mixed": 0.001207727356813848,
        "Negative": 0.00012982868065591902,
        "Neutral": 0.000779111753217876,
        "Positive": 0.9978833794593811
    }
}

Assuming that the Make Request action happens in Step B, you can reference the sentiment in subsequent steps with {{b.json.Sentiment}}.

Development

Suppose you wish to develop additional features or modify and host the application outside of the Serverless Application Repository workflow. In that case, this project contains tools to make it easy to do so.

First, install the prerequisites below:

• An AWS account
• Docker
• Node.js
• AWS Command Line Interface
• SAM CLI
• jq (optional)

The project includes a Makefile to set up, deploy, and teardown the application. Reference the targets in the Makefile as documentation for underlying commands.

First, clone the repository.

git clone [email protected]:cpliakas/quickbase-sentiment-analysis.git
cd ./quickbase-sentiment-analysis

Then, create the bucket that will host your code:

make bucket

You are now ready to deploy the application.

make deploy

Once the stack is created, run the settings target to download the settings.json file that contains the Pipelines configuration.

make settings

Tear down the application after you are finished working with it.

make teardown

The defaults are suitable for setting up a development environment. You can configure the Makefile by setting the following environment variables:

• APP_NAME: The CloudFormation stack name prefix, defaults to quickbase-sentiment-analysis
• APP_STAGE: The CloudFormation stack name suffix, defaults to the USER environment variable
• AWS_REGION: Defaults to the output of the aws configure get region command