Having a hard time searching for relevant results in your data or spreadsheets? Sometimes it would be nice to just search through everything like a search engine—relevant results, ordered by rank.
The Tables: Find similar text action is an algorithm-powered action that acts just like a search engine. This action uses a powerful feature called fuzzy matching to find similar matches between data, not just exact matches. The action finds and ranks matches even if there is incorrect spelling, spelling variations, or slight differences.
To find exact matches, rather than similar matches, use the action Tables: Look up data in a column.
This action searches a column for a search term. If the action finds similar matches, it returns each result. Each result has a similarity score based on how similar it was to the search term. The action outputs five fields
- A table of all the matches (Matches Table)
- The number of matching rows (NumRowsMatched)
- The most similar result, based on your search term (Top Match)
- A similarity score for the top match, based on your search term (Top Match Similarity)
- The difference between the top match similarity and the average (Top Match Similarity Gap)
Check the full output field details for more information.
As an example, think about a table with data on different apple varieties. There are 3 columns:
- Deliver by
If we wanted to look up information on Honey Crisp and Crispin apples, we could search the table for
Crisp and return multiple results. The following illustration demonstrates this:
For technical details on the algorithm behind this action, see the Technical Details section of the find similar text action.
To imitate this functionality, just add the Tables: Find similar text action to your process. All you need is an existing data table to search through.
- First set the table to search through. In this example, the action searches through the “Quotes” table.
- Select the columns to search through. If the table has multiple columns, only pick the ones that you want to consider as factors.
- Select the search term. In this example, the term is a field reference.
- Select the columns to return. In other words, if you find a match in Column A, then return column B and C.
If you want to play with an example process that uses this technique, import and test the following Pushbot: Search sentences example. This example searches a database of quotes, and generates an inspirational PowerPoint slide based on the best match.
To learn how to import a Pushbot, check the How to import a Pushbot article.