GrokItBot: A Python, AIML and Bayesian IM bot
How it all works
AIML
![GrokItBots' AIML files are small thanks to the Bayesian parser. GrokItBots' AIML](/archives/suttree.com/code/GrokItBot/images/AIML.jpg)
The ALICE bot uses AIML files that take advantage of wildcards, topics and templates to respond in an intelligent way. The downside of this is that, despite the availability of wildcards in the AIML specification, you soon need to have templated responses to a large number of anticipated statements.
Bayesian guessing
![GrokItBot can even tolerate speeling mistokes. GrokItBot in action](/archives/suttree.com/code/GrokItBot/images/how_uptime.jpg)
GrokItBot, however, guesses the topic of conversation and then extracts the relevant response from the AIML. This means that should be able to respond intelligently to a number of similar statements with only a fraction of the AIML code required.
GrokItBot is also very resilient when it comes to spelling mistakes and grammatical errors.
Read on to discover more about GrokItBots' features.
Suttree Beatniks with better clothing, brought to you by Duncan Gough. About Archives Casual Game Dev Code Disclaimer EcoLocal Music Knows PMOG Tumblee