If one needs to get anything from Keap’s Pocket
Body: user’s request statement in natural language
To view sample body utterances, please glance through Brainstorming X-Bot
Also, try to reply to their same respective threads for consecutive get and put requests. This will help you maintain your and Keap’s inbox tidy 😉 Also, email’s subject is not the only but one of the important criteria that helps Keap understand which action to take on any particular email.
So, for development purposes for this segment we will be majorly using Python and libraries such as re, time, imaplib, email, datetime. We will be using IMAP (a standard protocol that stores email messages on a mail server), to login to Keap’s account every 5–10 minutes by defining a Cron Job and looking for emails that are in unseen state. After getting the list of all the unseen emails we filter them further by applying a relevant subject filter to it and extract date, subject, email, body from that email message. You can find full code here.
That’s it for this section. In the next blog we will try to generate sample dataset for training Keap’s brain in identifying relevant and irrelevant emails for his purposes and explore RasaNLU library.
read original article at https://hackernoon.com/brainstorming-x-bot-part-2-cc69a4372ef0?source=rss——artificial_intelligence-5