Forum Discussion
Very slow after Key phrase extraction
I thought I would trial using the Text Analytics insights within Power Query on data that includes user comments. The step itself does not take long however any further steps within Power Query or trying to load the data to continue with my report takes an age.
Is this likely just because of an increase in rows and duplicated information following key phrase extraction? For example,
Before Key phrase extraction:
| 1 | This is my comment |
Might become:
| 1 | This is my comment | This | ||
| 1 | This is my comment | is | ||
| 1 | This is my comment | my | ||
| 1 | This is my comment | comment |
My real data set is 133,559 rows before key phrase extraction. I tried to find out the number of rows after key phrase extraction but got bored of waiting for the result! Is this service meant for very small datasets?
Anonymous Yeah, if it blows out your table to several million rows or more then there is going to be some slowness involved. Also, using the AI stuff is going to mean that query folding is broken at that point so any steps after that will not be query folded and run locally. Better to do as many steps with query folding as possible.
2 Replies
- Greg_Deckler
Community Champion
Anonymous Yeah, if it blows out your table to several million rows or more then there is going to be some slowness involved. Also, using the AI stuff is going to mean that query folding is broken at that point so any steps after that will not be query folded and run locally. Better to do as many steps with query folding as possible.