A Simple yet Brisk and Efficient Active Learning Platform for Text Classification

TL;DR

This paper introduces a user-friendly, efficient active learning platform for text classification that enables business users to quickly build and deploy models using state-of-the-art text representations and simple incremental learning methods.

Contribution

It presents a fully managed active learning system that combines GPT2-based text representations with fast incremental linear models for efficient, accessible text classification.

Findings

Empirical results demonstrate high accuracy with simple models.

The platform enables rapid model building and deployment.

Effective on both public and real-world insurance datasets.

Abstract

In this work, we propose the use of a fully managed machine learning service, which utilizes active learning to directly build models from unstructured data. With this tool, business users can quickly and easily build machine learning models and then directly deploy them into a production ready hosted environment without much involvement from data scientists. Our approach leverages state-of-the-art text representation like OpenAI's GPT2 and a fast implementation of the active learning workflow that relies on a simple construction of incremental learning using linear models, thus providing a brisk and efficient labeling experience for the users. Experiments on both publicly available and real-life insurance datasets empirically show why our choices of simple and fast classification algorithms are ideal for the task at hand.