In this State Changers meeting, the main topic of discussion revolved around utilizing Xano tables for storing and retrieving business data and how to effectively conduct a fuzzy search across these tables.
A participant explained his project using Xano tables, where one table hosts a form for collecting business information while another table defines businesses according to specific codes. He's looking for a method to conduct a fuzzy search across all this data, including data from the NAICS table. An issue discussed was the relationship between a business and its respective NAICS code, which identifies the type of business. The group discussed the possibility of a business having more than one NAICS code and how this might be handled in the database. They concluded that Xano, driven by Postgres, offers a solution where the structure type for NAICS codes can be changed to 'list', allowing for multiple codes to be associated with a single business. The discussion moved to the form where users input their respective NAICS codes. The initial expectation was that users would type or paste in their NAICS code, and the problem was how to connect this with the existing database. However, the participant clarified that adding the NAICS code to the database wasn't the issue, rather, he was looking for a method to allow a public user to search his database to find a particular business type. One of the suggested solutions was a fuzzy search performed via a sequence of queries rather than complex joins or fuzzy searches on multiple tables. The attendants also cautioned against using fuzzy searches across multiple tables as they tend to break down in terms of value for sort. Another suggested solution was to use something more generalized like Elastic Search but it was deemed as overly complicated for the requirement at hand. The discussion ended with a consensus to keep things simpler with a combination of conducting a few queries and a little bit of logic.
(Source: Office Hours 12/26 )
