Train Chatterbot on input

[GMT95]

hi, i want to do something like this

Me: what is the color of sky? bot: I dont know, can you tell me. Me: blue

and the bot saves the "blue" output on "what is the color of sky input".

Currently the bot saves my input "what is the color of sky?"

Please help.

[vkosuri]

@GMT95 chatterbot treas first statement and second as response. in your case

statement-1
response-1
....
....
statement-n
response-n

But you are trying in different fashion, for your use case it should be like this, Very good documentation on training process found here http://chatterbot.readthedocs.io/en/stable/training.html

-- what is the color of sky
-- blue

See Low confidence response selection filter out your responses with below confidence. http://chatterbot.readthedocs.io/en/stable/logic/index.html#low-confidence-response-adapter

[GMT95]

Thankyou @vkosuri, i have no problem training with corpus data or custom training file. I just wanted to ask that if I can train chatterbot while having a conversation with it.

For Example if chatterbot doesn't know something it says:

I dont know this can you tell me

and when i reply with the answer,chatterbot saves it in the database.

Me: what is the color of sky? bot: I dont know, can you tell me. Me: blue

and when i ask chatterbot again What is the color of sky? it replies with blue

Is it possible to do something like this?

[vkosuri]

yes, try this example https://github.com/gunthercox/ChatterBot/blob/master/examples/learning_new_response.py

[GMT95]

Thankyou @vkosuri, the last link is really helpful.

[GMT95]

@vkosuri can i set a default response when something is not in Chatterbot's Knowledge like

Bot replies: I don't know or Can you please tell me

[vkosuri]

Yes

[gunthercox]

It sounds like this question has been answered? @GMT95 I'm going to close this ticket off for now, feel free to reopen it if your original question wasn't resolved.

[AfrahAsif]

yes, try this example https://github.com/gunthercox/ChatterBot/blob/master/examples/learning_new_response.py

This link is not working.. How can I solve the same issue raised by @GMT95 ?

[vkosuri]

It's has been very long, I understood @GMT95 question, The Statements are sent through the console and the training process start on the fly, Bot return response

If that is the case I think this is not available right now.

[gunthercox]

@AfrahAsif See this example instead: https://github.com/gunthercox/ChatterBot/blob/master/examples/learning_feedback_example.py

[Augus2411]

@gunthercox Im also trying what @GMT95 and @AfrahAsif is trying can you help me how to implement it to the one im working on?

from chatterbot import ChatBot from chatterbot.trainers import ListTrainer

def get_response(usrText): bot = ChatBot('Bot', storage_adapter='chatterbot.storage.SQLStorageAdapter', logic_adapters=[ { 'import_path': 'chatterbot.logic.BestMatch' }, { 'import_path': 'chatterbot.logic.LowConfidenceAdapter', 'threshold': 0.70, 'default_response': 'I am sorry, but I do not understand.' } ], trainer='chatterbot.trainers.ListTrainer') bot.set_trainer(ListTrainer) while True: if usrText.strip()!= 'Bye': result = bot.get_response(usrText)
reply = str(result) return('Bot :'+reply) if usrText.strip() == 'Bye': return('Bot :'+reply) break

from chatterbot import ChatBot from chatterbot.trainers import ListTrainer import os

def setup(): chatbot = ChatBot( 'Bot', storage_adapter='chatterbot.storage.SQLStorageAdapter', trainer='chatterbot.trainers.ListTrainer') for file in os.listdir('C:/Users/hp/Chatbot/data/'): convData = open(r'C:/Users/hp/Chatbot/data/' + file,encoding='latin-1').readlines()

trainer = ListTrainer(chatbot)

    chatbot.set_trainer(ListTrainer)
    chatbot.train(convData)
    #print("Training completed")

setup()

[Augus2411]

@gunthercox also im having this kind of error Type something to begin... hi Traceback (most recent call last): File "feed.py", line 41, in input_statement TypeError: generate_response() missing 1 required positional argument: 'conversation_id'

[Augus2411]

@vkosuri can you help me with this sir?

[gunthercox]

@Augus2411 It looks like the code you are using is from an older version of ChatterBot. I would suggest that you look over the latest version of the examples.