Partial Text and Approximate Matching
Match names that are not exactly the same.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
Help me match records that are almost the same. App: [SPREADSHEET_APP]. Two lists to match and how they differ (spelling, order, abbreviations, extra words): [LIST_DESCRIPTIONS]. Sample pairs that should match (fake): [SAMPLE_PAIRS]. Sample pairs that should not match: [NON_MATCH_PAIRS]. Provide a layered approach with formulas: 1. Normalize both lists: TRIM, LOWER, remove punctuation and common words (inc, ltd, the) using nested SUBSTITUTE, standardize abbreviations with a lookup table. 2. Exact match on the normalized key: XLOOKUP or INDEX/MATCH. 3. Partial match: wildcard lookups with "*"&text&"*", SEARCH/FIND with ISNUMBER, and the risk of short strings matching everything. 4. Token-based matching: split into words and count the common words (TEXTSPLIT and COUNTIF or an alternative). 5. Similarity scoring: describe how to compute a simple score (common words / total words) and a threshold; mention that fuzzy-matching add-ons or scripts exist for edit distance, with the pros and cons. 6. Review step: classify as auto-match, needs review, or no match, and keep an audit column showing the method used. Return a worked example with my sample pairs and results, warn about false matches (similar names for different entities), and provide a validation sample (check 50 random matches). Never merge records without review.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.