Hollaka Hollala
IRC bot. Hollaka Hollala is a chatbot engine and IRC bot written in Java. The chatbot engine uses a knowledge base organized in an XML file and WordNet for dictionary lookup. The IRC Bot has some WordNet- and some Web-based functionality and is hot-extensible through Ruby programs (new commands and functionality can be defined in Ruby scripts which are loaded at runtime.
Hollaka Hollala is a learning chatbot engine and IRC bot. The chatbot engine uses XML (using JDOM) to store the knowledge base and WordNet (using JWNL) as an external resource. The knowledge base is organized in topics, which consist of keywords and possible answers. These are organized in a simple XML format. The program tries to find the most suitable topic for a user input (e.g. using semantic relations from WordNet and word counting). On the selected topic the answer is then chosen randomly, where answers with higher frequencies (meaning things often said by users) occur more often. This is a place for possible extensions (e.g. speech act theory, pragmatics in general, any comparison between possible answers and input). When no suitable topic is found new topics are made up from the user input, using information from WordNet and the user input for initial content. The chatbot can therefore start with an empty knowledge base and learn new topics and answers for these topics from user conversation.
Hollaka Hollala is a learning chatbot engine and IRC bot. The chatbot engine uses XML (using JDOM) to store the knowledge base and WordNet (using JWNL) as an external resource. The knowledge base is organized in topics, which consist of keywords and possible answers. These are organized in a simple XML format. The program tries to find the most suitable topic for a user input (e.g. using semantic relations from WordNet and word counting). On the selected topic the answer is then chosen randomly, where answers with higher frequencies (meaning things often said by users) occur more often. This is a place for possible extensions (e.g. speech act theory, pragmatics in general, any comparison between possible answers and input). When no suitable topic is found new topics are made up from the user input, using information from WordNet and the user input for initial content. The chatbot can therefore start with an empty knowledge base and learn new topics and answers for these topics from user conversation.
| Category : | Chatbots - English |
| Submitted : | 6th, August 2008 |
Report Broken Link
