Transformer Models for Mining Intents and Predicting Activities from Emails in Knowledge-Intensive Processes

Faria Khandaker, Arik Senderovich, Junda Zhao, Eldan Cohen, Eric Yu, Sebastian Carbajales, Allen Chan. Engineering Applications of Artificial Intelligence, 2024.

In knowledge-intensive processes, much of the work happens in email, which makes it hard to connect messages to the process activities they belong to. This work treats the business intents that users express in emails as semantic representations of their goals, and uses them to bridge emails and activities. It leverages user actions as a source of weak supervision, in addition to a limited set of annotated examples, to train transformer models that detect intents in emails and predict activities.

• Paper