A Supervised Approach To The Interpretation Of Imperative To-Do Lists

TL;DR

This paper presents a supervised method for interpreting electronic to-do lists by classifying user intentions and extracting relevant information, demonstrating strong performance across multiple datasets.

Contribution

The work introduces a novel supervised approach for classifying user intentions and extracting information from to-do lists, filling a gap in intelligent personal assistant research.

Findings

High accuracy in intention classification across datasets

Effective information extraction from to-do list entries

Method generalizes well to different to-do list domains

Abstract

To-do lists are a popular medium for personal information management. As to-do tasks are increasingly tracked in electronic form with mobile and desktop organizers, so does the potential for software support for the corresponding tasks by means of intelligent agents. While there has been work in the area of personal assistants for to-do tasks, no work has focused on classifying user intention and information extraction as we do. We show that our methods perform well across two corpora that span sub-domains, one of which we released.