Text Ranking and Classification using Data Compression
This work addresses text categorization challenges for users needing simpler, language-agnostic tools, but it is incremental as it builds on existing compression-based approaches.
The paper tackled text classification and ranking by enhancing data compression-based methods with the Zstandard compressor, resulting in a language-agnostic technique called Zest that simplifies configuration and competes with language-specific embeddings in production, though it does not outperform other counting methods on public datasets.
A well-known but rarely used approach to text categorization uses conditional entropy estimates computed using data compression tools. Text affinity scores derived from compressed sizes can be used for classification and ranking tasks, but their success depends on the compression tools used. We use the Zstandard compressor and strengthen these ideas in several ways, calling the resulting language-agnostic technique Zest. In applications, this approach simplifies configuration, avoiding careful feature extraction and large ML models. Our ablation studies confirm the value of individual enhancements we introduce. We show that Zest complements and can compete with language-specific multidimensional content embeddings in production, but cannot outperform other counting methods on public datasets.