Costs to Consider in Adopting NLP for Your Business

TL;DR

This paper compares the performance and costs of classical versus deep learning NLP models in industrial tasks, highlighting that classical models often match deep models at lower costs and emphasizing the need for affordable NLP solutions for under-resourced languages.

Contribution

It provides a comparative analysis of classical and deep NLP models in terms of performance and cost, advocating for cost-effective NLP solutions especially for under-resourced languages.

Findings

Classical models often perform comparably to deep neural models in industrial datasets.

Deep models incur higher costs with marginal performance gains.

Cost-performance trade-offs are crucial for AI adoption in business.

Abstract

Recent advances in Natural Language Processing (NLP) have largely pushed deep transformer-based models as the go-to state-of-the-art technique without much regard to the production and utilization cost. Companies planning to adopt these methods into their business face difficulties because of the lack of machine, data, and human resources to build them. We compare both the performance and the cost of classical learning algorithms to the latest ones in common sequence and text labeling tasks. In our industrial datasets, we find that classical models often perform on par with deep neural ones despite the lower cost. We show the trade-off between performance gain and the cost across the models to give more insights for AI-pivoting business. Further, we call for more research into low-cost models, especially for under-resourced languages.