Hi everyone,
I'm using TextiQ in my survey and I am wondering if there's any "machine learning" behind the scenes that improves it as we edit sentiment values.
I tried giving the same text response different times, and after loading it in TextiQ, I changed its score from 3 to 5. But everytime I do this, TextiQ still assigns a score of 3 to that response, even though the score of the previous ones has been changed to 5 and they have the same identical text.
Is this a normal thing or am I missing something?
Page 1 / 1
From what I understand, it's normal. There is machine learning but from I've been told, it's on a way bigger scale and is more tied to the way the words are used together for your company. I wish it would work the way you're describing but in my experience, I haven't seen it react that quickly.
I've noticed too that the sentiment score I assign to a statement differs from the "Positive" or "Negative" labels Qualtrics assigns it doesn't change the value, nor does the TextIQ tool seem to learn from similarly misattributed statements.
Leave a Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.