Improving Tokenisation by Alternative Treatment of Spaces
This addresses tokenisation issues in NLP, particularly for handling complex words, but is incremental as it modifies existing algorithms rather than introducing a new paradigm.
The paper tackled problems in subword tokenisation algorithms, such as limited linguistic validity and inconsistent representations, by proposing an alternative approach where spaces are always treated as individual tokens, applied to BPE and Unigram algorithms, resulting in improved performance on downstream NLP tasks involving complex words without harming general natural language understanding tasks.
Tokenisation is the first step in almost all NLP tasks, and state-of-the-art transformer-based language models all use subword tokenisation algorithms to process input text. Existing algorithms have problems, often producing tokenisations of limited linguistic validity, and representing equivalent strings differently depending on their position within a word. We hypothesise that these problems hinder the ability of transformer-based models to handle complex words, and suggest that these problems are a result of allowing tokens to include spaces. We thus experiment with an alternative tokenisation approach where spaces are always treated as individual tokens. Specifically, we apply this modification to the BPE and Unigram algorithms. We find that our modified algorithms lead to improved performance on downstream NLP tasks that involve handling complex words, whilst having no detrimental effect on performance in general natural language understanding tasks. Intrinsically, we find our modified algorithms give more morphologically correct tokenisations, in particular when handling prefixes. Given the results of our experiments, we advocate for always treating spaces as individual tokens as an improved tokenisation method.