Chatterbot: Set a default response if no matching input is found

[datta90]

hi gunther,

I wanted to know if there is a way to set a default response like (i don't understand what you mean)
if no matching input is found

-Dattatreya

[datta90]

i am trying to set a threshold value in the closest match adapter for levenshtein distance
is this correct ?

[alekaizer]

hi @datta90 You use the process method of your logic adapter to get the confidence of the response then set the threshold on the confidence.

input = 'does linux rock'
bot = ChatBot('tux')
statement = bot.input.process_input(input)
confidence, response = bot.logic.process(statement)
if confidence > THRESHOLD:
   response = bot.output.process_response(response)
else
   response = 'I don't understand what you mean'

Maybe @gunthercox has a better way of doing it :)

[gunthercox]

@alekaizer That is an excellent example. There isn't a better way to provide a default response, I usually recommend something similar to that as a solution.

[datta90]

i am getting this error when i type the above code in python terminal

In [9]:  statement = bot.logic.input.process_input(input1)
----------------------------------------------------------------------
AttributeError                            Traceback (most recent call
<ipython-input-9-794f22aa52a2> in <module>()
----> 1 statement = bot.logic.input.process_input(input1)

AttributeError: 'MultiLogicAdapter' object has no attribute 'input'

[gunthercox]

@datta90 Are you sure you are typing it into the terminal _exactly_ as it was written above?
The error message is correct, there is no .input attribute after .logic. I think maybe you meant to type just bot.input.process_input.

[datta90]

thank you gunther
I am getting error UnicodeDecodeError: 'utf8' codec can't decode byte 0x93 in position 83: invalid start byte when i am adding my own training data in the English folder,
my file name is ending with .corpus.json and i have written this file properly
even when i am remove this file i am still getting this error

[datta90]

is it because of json adapter storage my strings are bit long in the file

[gunthercox]

A UnicodeDecodeError is a Python error that is triggered when a file is opened for reading without the encoding being properly specified. I'll look into this to make sure that ChatterBot's corpus reader is specifying it correctly.

[gunthercox]

@datta90 I've made a correction to the German corpus that should fix the issue you reported: https://github.com/gunthercox/ChatterBot/pull/394

[datta90]

No problem gunther actually mistake was from my side I used regex to remove the ASCII characters present in my JSON file

[lock[bot]]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.