We will discuss some of the main tasks bots must perform. How is machine learning used to improve performance in these tasks? Some tasks to be discussed will involve aspects of natural language processing and forms of classification or inference such as intent recognition. What kinds of data are available to learn from and what kinds of models are used? Are context and conversation important and to what extent are they handled and how are they handed currently?
To what extent are bots taking action and learning to improve their actions? Where are the challenges and where will things go next?