Mar 06, 2023 06:54 AM
Hi!
I am populating a table with plain text from user input via "Softr". I am now analyzing these entries via the Google Cloud Natural Language API to gain insights into Sentiment, Entities etc. and outputting this data into separate tables for each dimension I am analyzing.
How can I make sure to automatically link the analysis in the API output tables to the original entries in the "Entries" table? I would need to do this to identify the analysis data back to the user that has entered the text.
Thank you guys in advance!
Mar 06, 2023 09:34 PM
How are you analyzing said entries? Could you provide a screenshot of your tables with the relevant fields please?
Assuming the method you're using ot analyze the entires has access to the record IDs of the original records, is it possible to paste the record ID into the linked field as part of the API output update?
Mar 07, 2023 04:15 AM
You can see from my screenshots the input table "Entries" which I am taking the Entry text data from. I feed this information to the API and output the analysis to the table "Sentiment". The one thing that could identify the original record in the Input table "Entries" is the column "Text Content". This is what I mean by me needing to find this text content in the input table and link back to that record if that content is found in the column "Entry text" in the table "Entries".
Does this become more clear now?
Mar 10, 2023 07:12 AM
Yeap super helpful. To do this, you can use an automation that would trigger when a new record is created in "Sentiment", and it would have:
1. "Find record" action to look for a record in "Entries" that has the same "Entry-text" value as the triggering record's "Text Content (Senten...)" value
2. "Update record" action to update the triggering record's linked field to the "Entries" table with the found record's Airtable ID