If you are an existing customer of Chatbot or new, you can make use of our Classification API to extend your Chatbot.
You can inlcude sentiment analysis or general text classification usign the same API. We have provided steps to implement sentiment analysis with our API below.
If you are new to Chatbot check out their site at Chatbot
We will be using the Webhook integration method provided by Chatbot
This repo contains a sample NodeJS Express server which provides a webhook endpoint for sentiment analysis.
Clone the repo in your system and run below command to install all dependencies
npm install
Once above is done, you need to open 'webhook.js' file and replace 'CHATBOT_VERIFICATION_TOKEN' with the verification code you entered while creating the webhook in Chatbot console. This is required to ensure your webhook is authentic and working. In this code, we’re checking if your verification token is correct with the one sent in the request.
You also need to replace 'FIRSTLANGUAGE_API_KEY' with your apiKey to access FirstLanguage APIs. If you do not have one, head to FirstLanguage to get one for free.
Once the Chatbot is published in chatbot console, you need to copy the JS embed code and replace the line "TODO: Copy the Chatbot published code here " with the script code.
Once that is done, just run
node webhook
to start serving responses
You can use NGROK to expose your localhost server to Chatbot. Run the below command if you are running on port 3000
ngrok http 3000
We are assuming the steps to create Chatbot stories and other fucntions are known. If not please refer their documentation page.
You can check below for a short demo {% include youtubePlayer.html %}
Below are few screenshots for your reference:
PFB a sample Story for our Chatbot to work
Select the webhook you created in the Integrations window.
User input can be configured like below
Final output of our Chatbot app
