About using a simple custom conversation format.

As mentioned here [1], we can create our own simple conversation data between 2 people. I created a file called simple_chat.txt and placed it in DeepQA/data/lightweight directory with content:

Hello
Hi
How are you?
I am good. How are you?
I am fine as well. How is the work?
work is going on as usual. Thanks for asking.
No problem.

Now, when I run

 python3 main.py --corpus lightweight --datasetTag simple_chat --test interactive

I do not see any answers from the bot. I get an output of something like:

Q: Hello
A:

Q: How are you?
A:

Is it because my sample data is small? Or can someone explain if I am doing things correctly.

[1] - https://github.com/Conchylicultor/DeepQA/tree/master/data/lightweight

[Conchylicultor]

Don’t forget to delete the pre-processed data inside data/samples/ to regenerate the vocabulary and training sentences.

[shank7485]

@Conchylicultor Thanks.

I am comparing your Chatbot and the one made here: [1]. In [1], it is sort of rule based bot which learns from the user input by linking response with input and builds its knowledge graph.

It would be amazing if you 2 had discussions with each other about making better Bots. It would be helpful for the rest of us. 😃

[1] - https://github.com/gunthercox/ChatterBot