Stop Word Removal
A conventional text preprocessing operation that eliminates high-frequency function words such as 'the,' 'of,' and 'and' from text. Originating in information retrieval as a means of conserving storage and computational resources, the technique has been widely adopted in standard NLP pipelines.
Timeline (last 90 days)
A newly published arXiv paper argues that in interpretable legal text analysis, retaining stopwords is the correct choice, and that removing them may distort the jurisprudential and ideological signals the research is meant to capture.
The practice of stopword removal originated in information retrieval systems from decades ago, when removing high-frequency function words could save storage and computation costs and simplify indexing.
Experiments found that generic stopword lists fell below the 'no removal' baseline in every single test, actually dragging down classification performance.
All Facts (3)
A newly published arXiv paper argues that in interpretable legal text analysis, retaining stopwords is the correct choice, and that removing them may distort the jurisprudential and ideological signals the research is meant to capture.
50%UnverifiedThe practice of stopword removal originated in information retrieval systems from decades ago, when removing high-frequency function words could save storage and computation costs and simplify indexing.
50%UnverifiedExperiments found that generic stopword lists fell below the 'no removal' baseline in every single test, actually dragging down classification performance.
50%