Is it possible to do Text Classification on unlabeled data? (Feat. Zero-Shot Classification) [Experiment]

Author(s): Ala Alam Falaki

Originally published on Towards AI.

Photo by Markus Spiske on Unsplash

A couple of months ago, I did an experiment to answer the question: Is it possible to do sentiment analysis if we do not have the target labels? The process was to use an unsupervised algorithm to generate labels that will be used to fine-tune a BERT model. The experiment resulted in a fine-tuned BERT model with 71.1% accuracy. (Click here to read the full article)

In this piece, we will try to get better accuracy by using the Zer-Shot Classification method. These models are extremely handy when we do not have enough training data. They… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI