5+ reasons to choose Python over NodeJS for your Dialogflow bot

This is NOT an article which gives you a comparison of Python and NodeJS when building your Dialogflow bot. This is very specific to my website, and explains how you can make the most of the materials you find on it.

Also, if you have already started using NodeJS, that is OK too. This article will be most useful for those who are just beginning their bot development. 

Python is easier to read

This is easy to observe.

You can compare the documentation for Python code and NodeJS code and see which one is easier to read.

Python is easier to maintain

If you hire a Python programmer to write some code for your Dialogflow bot, and they leave your team, it is usually quite easy for a new hire to maintain their code. This is not nearly as easy in the case of NodeJS

Python is the language of Natural Language Processing

The best NLP library spaCy is built using Python. The most famous open source chatbot framework RASA NLU is built with spaCy.

So someone with Python programming skills will be able to learn NLP more quickly than one without.

You can use spaCy to extract intents from chat logs

This follows from the previous point. You can use spaCy to get the dependency tree of chat log phrases and use them to design your intents. For example, that is what I do in my Autotrain tool.

You can use PyCharm to set up automated conversation testing

It is very easy to set up automated conversation testing for your Dialogflow bots using the PyCharm IDE. I explain why, and also provide some sample Python scripts, in my Improving Dialogflow ES accuracy course.

You can use spaCy to auto-generate test scripts

spaCy can be used to parse your existing training phrases and convert them into test scripts for conversation testing. I discuss the basic idea in the video below.

Use spaCy to understand Dialogflow’s intent matching

You can also use spaCy to analyze your user utterances so you can get a good picture of how Dialogflow’s intent mapping algorithm works

Migrate to RASA NLU more easily

If you ever need to migrate your bot to RASA NLU, you will also be able to do it much more easily for obvious reasons.


About this website

I created this website to provide training and tools for non-programmers who are building Dialogflow chatbots.

I have now changed my focus to Vertex AI Search, which I think is a natural evolution from chatbots.

Note

BotFlo was previously called MiningBusinessData. That is why you see that watermark in many of my previous videos.

Leave a Reply