Build A Smart Chat Bot Using Python & Machine Learning

Build A Smart Chat Bot Using Python & Machine Learning

⭐Please Subscribe !⭐

⭐Support the channel and/or get the code by becoming a supporter on Patreon:

⭐Website:

⭐Helpful Programming Books
► Python (Hands-Machine-Learning-Scikit-Learn-TensorFlow):

► Learning Python:

►Head First Python:

► C-Programming :

► Head First Java:

Promising new research on kidneys:

Mayo Clinic Website on Chronic Kidney Disease:

[TheSewDew]

That's freaking genius. appending the users sentence to the sentence list to keep info is a smart way to do it. nice one.

[stefanzander]

Some clarifications: This chatbot is not able to answer any questions about a given text corpus. It uses a comparably simple approach (cosine similarity) to compute the relative distance between a query (the request) and sentences contained in a corpus. Responses like greetings are chosen randomly from a predefined list. The bot has no memory (state tracking) nor does it use ML or AI techniques. It can also not identify user "intents".
It is therefore neither smart nor does it use ML/AI techniques.

[sanketgadge]

Sometimes you do same thing like index thing was really simple ...you just sorted it but you made it really confusing i got alot...good stuff

[jamesbondisamonkey]

Really great tutorial, good starting point for this project, thank you

[sanketkotkar]

can you tell which NLP model u have used to build the chatbot ?

[soumojitchowdhury]

by the way in the index_sort() function why are we sorting them manually instead of using in build sort()?
can anyone explain?

[fajard]

is cosine similarity has an alternative ? such as machine learning algorithm, knn / svm ?

[MsAora]

Thank you so much for this tutorial! It was very educational, and the best part is that it works. Even on vscode, just tried it there!

[zradbaamine]

everythink is good, but at the very end i got this error:

in <module>()
13 print('doc Bot:
14 else:
---> 15 print('doc Bot: '+bot_response(user_input))

TypeError: must be str, not NoneType


when i add str like this :


the doc bot answer my question with : none
help me please!

[peterkhomchenko]

Hey, I have a question. When I type a question to the bot, it does not output the scraped data of the website, instead it just duplicates the message.
So if I type: What is chronic kidney disease, the bot outputs: What is chronic kidney diseaseWhat is chronic kidney diseaseWhat is chronic kidney disease
If you could help that would be great.

[anujshaan]

hi there i am working on college project i.e. Chatbot which helps with knowing disease when user provide with some symptoms.How can i train my chat bot inorder to get perfect precaution tips. Help me in this please i am so confused.

[guilhermecfialho]

Your channel is awesome, congrats! There is a continuation about this video?

[srikanthmore]

hey bro it is telling no Could not find a version that satisfies the requirement newpaper3k (from versions: none) please help me out

[shaktinishad]

print('Doc Bot: '+bot_response(user_input))
in this line am getting error of
TypeError: can only concatenate str (not "NoneType") to str

alll working great but when i ask question this error comes

[prodevmahi]

are these packages installed via pip install, are installed in our computer or are installed in cloud temporarily

[nuritwitch]

Thank u pls keep going .I love ur channel <3

[dionisnavarro]

Congratulation, you presented this tutorial in an easy way to follow and understand.
I might be missing somehing becuase the bot does search the website one set at the article = Article() cell, however when I or someone ask a question or give a key word about that Article, the bot doesn't bring back any answer ; it just says None. ..
Maybe you can give me some clue as why it's doing that?
Thanks, :)

[sollekram]

Hello, thanks for your contribution on this one!
So I have a couple of questions.
So I am using a different approach to talk the the chatbot, by have a JSON data file with patterns(user inputs) and responses and calculated inside which are close so it will print and appropriate response...
Though the thing I wanna ask you is, how to implement a whole wiki_en_articles.txt corpus to all this scenario, so I will get back info on what the user put.
The file is a huge txt.
Let's say i input how many planets exists in the solar system?
So I am expecting the chatbot to reply to me with the correct answer.
Is there a way to implement such a thing with my current way of doing it?
I don't want an API, I have the whole data to text, each line an article

[TheBornfreetofly]

any suggestion on how to improve the answers ? or how to improve the model that we create out of article ?

[justhink]

Our big-hearted software developer brother who fits 5 lines of code into 100 lines.