I am working on an AI chat bot in Python. It works by comparing user input to a text file of possible responses. It finds the closes match to your question in the responce file, then uses the responce in the text file. This means you don’t have to spell words exactly, and there does not have to be a perfect string match.
This would be perfect for in game characters.
If you use this give me credit.
You can ask “How do you work?” for a simple summary of how the code is designed.
there are python implementations. i already successfully tested it. works quite well. depending on your chat-database you can have quite good conversations with it.
i planned to use it to make an in-game AI intelligent enough to talk to you (via alice and speech-synthesis). i didnt continue due to my lazyness to have a database which fits the in-game situation,tho
I might be able to help on this one - give me a shout if you feel so. I have recently had to develop a 3D Jabber Bot game for a client.
Based around “activities”/scenes. The player selects scene, logs in, bots appear in the scene (if online and if the scene configuration file dictates they appear in that scene). As noted, every bot is just a “jabber bot”, it signs on to an openfire server and if available - responds to chat.
User communicates … all 3D environment and animated characters. I wrote one generic “chat bot” that is driven by a configuration file itself - just run multiple instances (with different configurations) for each bot required (brain is powered by AIML).
Contact me via the FlatCoder link below if it sounds useful.
No, no lip syncing as of yet - that would imply parsing the bot response and mapping it… just animations for now, but I do like the idea of lip syncing ‘eventually’.
Meanwhile - as your project is a replacement for AIML… this project, the client in particular, is a completely generic 3d jabber bot environment. Bots could be those I have written (using AIML), human people, or ‘any other bot that can sign on to Jabber’.