Suppose I have a lot of data on conversations between humans and chatbots (human text, chatbot text, times, media used for chat, etc), and I want to be able to detect anomalies in these conversations. For example, an anomaly would be if the chatbot was supposed to help customers with navigating a website, but the website was down and displayed some sort of error. As a result, the chatbot gets a lot questions about that error in a "short window" of time. I want to be able to detect this quickly, so I can then make the chatbot be able to better handle questions about that error. I read "Incremental Tensor Analysis: Theory and Applications" by Sun, Tao, et al. but I was curious if there were other works that were more applicable to text and conversations that I should check out.

  • $\begingroup$ There's lots written on anomaly detection. Have you studied those available resources? What approaches have you considered? What's your analysis of them? $\endgroup$ – D.W. Mar 9 '18 at 0:18
  • $\begingroup$ There's lots, but I don't really see too much as applied to conversational data. That was what I was hoping to get some input on. $\endgroup$ – ilikecats Mar 9 '18 at 0:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.