Hello folks, I am the intern who is working on the Mifos Chatbot project currently, I am willing to share my current state, which is the progress of this project. Here is the pull request that I sent https://github.com/openMF/mifos-chatbot/pull/5 <https://github.com/openMF/mifos-chatbot/pull/5> , basically the files changed in this PR are to build the basic structure.
After several discussions with my mentor Mr Aleks, we plan to divide the project into 5 components. 1. Core module: this module contains the interface for the other modules, it is like the controller of logic 2. Client module: this module handles the connection with Mifos main application through RESTful APIs 3. NLP module: this module implements the NLP engine, OpenNLP is implemented 4. Protocol module: this module implements the chatbot protocol, basically the Skype protocol is my choice 5. Server module: this module handles the requests and responses between user and server Basically another question that I would like to discuss with community is that how many commands that we plan to implement? If there are more commands that we want to achieve, more time of training the NLP model is required. 10 commands may be a reasonable choice. If you have any suggestions about the number of commands, feel free to reply to this email. Thanks in advance for any advice! Regards, Dingfan
