Detect Product Names In Text

from a list of order line (list of json products) returns the most similar to input text based on fuzzy matching similarity

Multiple on-going orders can be linked to one customer, sometimes customers mention in their request the name of the item they have bought. This action helps link that customer request to one of the client's order based on words fuzzy matching.

This actions takes as input the customers' order line (list of all orders). For a more general fuzzy matching action, read here.

Example:
A customer asks "I still have not received my straw hat"
The on-going orders for that customer:

[
  {
    "itemDescription": "Straw hat with red ribbon",
    "status": "...",
    "createdAt": "..."
  },
  {
    "itemDescription": "Blue Jeans",
    "status": "...",
    "createdAt": "..."
  },
  {
    "itemDescription": "Black leather gloves",
    "status": "...",
    "createdAt": "..."
  },
]

In that example, the first item will be returned.

Inputs

InputTypeDescription
TextStringInput customer request
Product Name FieldsStringThe field containing item description, for the example above, we would put "itemDescription"
Order Line ItemsListList of user orders, list of jsons
Minimum ScoreIntegerMinimum number of letters you want to match. It helps avoiding the match on small common words like "a", "on", "and"... default is 4