Goodlookup is a smart function for spreadsheet users. It’s a pre-trained model that has the intuition of GPT-3 and the join capabilities of fuzzy matching. Use it like vlookup or index match to speed up your topic clustering work in google sheets!

Yearly Subscription US $15 per year

Here’s how to start:

1. Subscribe! You must be subscribed.
2. Install goodlookup from the google workspace maketplace
3. Use this example and activate the function in the sheet menu > extensions > manage add-ons > goodlookup

The main limitation of traditional fuzzy matching is that it doesn’t take into consideration similarities outside of the strings. Topic clustering requires semantic understanding. Fortunately, recent advances in NLP technology have unlocked new possibilities with text data. Goodlookup is a smart function for spreadsheet users that gets very close to semantic understanding.

Modern spread sheet users have a join problem. Their data lives in multiple places without uniform naming conventions. This makes it hard to get a clear and unified view of said data. Goodlookup helps people solve these text-to-text record linking problems.

Goodlookup can match similar text the way a human would. Goodlookup can match semantic relationships, synonyms, and even cultural similarities between text strings. It doesn’t try to replace fuzzy matching, it’s simply a new tool for your data ops repertoire.

Goodlookup can match “Ronaldo” to “Ronaldo Luís Nazário de Lima” as well as match “Ronaldo” to “Soccer” or “Football.” The Score next to the match indicates the intensity of the neighboring words in the vector space. Longer strings matching identically to longer strings will have the highest scores.