Challenge Overview
This F2F challenge is a part of IBM Cognitive ChatBot series, where we develop a chat bot based on IBM Cognitive Conversation Service, along with Angular 4 frontend and ExpressJS / NodeJS backend (ask in the challenge forum for authorization to access these repositories; conversation model for Cognitive service can be found in /docs/topcoder_chatbot.json in the backend repo). In this challenge you will do the following updates of the existing code:
- Right now all communications are treated by the frontend as purely textual. You should see however, that some conversation nodes assume get_option and get_multiple_options actions. You will implement the proper support for these. In the first case, the frontend should include into the quesion bubble buttons with valid options, assuming that user should click one of them to answer the question. In the second case, the frontend should include into the question bubble the list of options with checkboxes, and submit button, which sends the answer.
- At the welcome screen, right now clicking on the avatar, if all other data are introduced, immediately redirects to the chat. This should be updated, so that clicking on the avatar just selects (showing the same green border, visible when hovered), and activates Continue button. Only explicit click on the Continue button should take to the chat (if all other data are filled correctly). Also, please comment out the Plus Sign button, assumed to add custom avatars.
- When ChatBot does not understand something, it replies with I am just an example of what you can build with IBM Watson and Topcoder's Cognitive Community. If you would like to learn more, go here to launch live chat with Topcoder, where here is a link. Please (1) update the link's style to be underlined; (2) change the reference to point to www.topcoder.com.
- Right now in the frontend the backend URL is configured as a constant inside /src/app/chat/chat.component.ts. Please, create a proper configuration setup and move this url to config file. (e.g., look how it is done in the backend, where node config is used to handle configuration).