Search and Learning for Unsupervised Text Generation

TL;DR

This paper presents a novel unsupervised text generation method combining heuristic search and learning, reducing reliance on labeled data and improving efficiency for low-resource languages.

Contribution

Introduces a search and learning framework for unsupervised text generation that estimates quality via heuristics and refines with machine learning, bypassing the need for large labeled datasets.

Findings

Effective in low-resource language processing

Reduces human annotation labor

Improves generation quality through search and learning

Abstract

With the advances of deep learning techniques, text generation is attracting increasing interest in the artificial intelligence (AI) community, because of its wide applications and because it is an essential component of AI. Traditional text generation systems are trained in a supervised way, requiring massive labeled parallel corpora. In this paper, I will introduce our recent work on search and learning approaches to unsupervised text generation, where a heuristic objective function estimates the quality of a candidate sentence, and discrete search algorithms generate a sentence by maximizing the search objective. A machine learning model further learns from the search results to smooth out noise and improve efficiency. Our approach is important to the industry for building minimal viable products for a new task; it also has high social impacts for saving human annotation labor and for processing low-resource languages.