Vaporetto: Efficient Japanese Tokenization Based on Improved Pointwise Linear Classification

TL;DR

This paper introduces Vaporetto, a method that significantly accelerates Japanese tokenization by optimizing the pointwise linear classification framework, achieving 5.7 times faster processing without losing accuracy.

Contribution

Vaporetto presents novel optimizations for PLC-based Japanese tokenization, including array operations, memory-efficient automata, and pre-processing techniques, enhancing speed while maintaining accuracy.

Findings

Tokenization speed increased by 5.7 times

Maintained the same accuracy as previous methods

Optimizations are orthogonal and broadly applicable

Abstract

This paper proposes an approach to improve the runtime efficiency of Japanese tokenization based on the pointwise linear classification (PLC) framework, which formulates the whole tokenization process as a sequence of linear classification problems. Our approach optimizes tokenization by leveraging the characteristics of the PLC framework and the task definition. Our approach involves (1) composing multiple classifications into array-based operations, (2) efficient feature lookup with memory-optimized automata, and (3) three orthogonal pre-processing methods for reducing actual score calculation. Thus, our approach makes the tokenization speed 5.7 times faster than the current approach based on the same model without decreasing tokenization accuracy. Our implementation is available at https://github.com/daac-tools/vaporetto under the MIT or Apache-2.0 license.