Seashell language model reply generator and extended word filtering

Here’s a real quick reply generator I made. To expand it would require a serious plan. Either each word gets entered and you check each row OR you make a category list and then check each list to see if it contains the typed in word. then you could call that category from your table and it could pick out a reply. Just my quick thoughts. Not sure if thats what youre thinking or not.


This would work, it’s not the most efficient to enter 9 items into each row of a dataset but yes, it would work. The only thing is, we need a better language filter that can find any occurrence of any word marked N/NO in the dataset, not just the first. I’ll see what I can do about that, but yes, creating a language model is a lot of work. This is mainly a test for fun, learning and just to see what I can and can’t make with Thunkable, but I plan to expand it over time if it works. I’m willing to spend a lot of time on this project.

I’ll also need a minor spelling checker that can detect small spelling issues like tihs. But we’ll get to that part later. I was thinking of one more column with possible spelling errors, but don’t worry about that right now.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.