(Embarrassing LOL) - Quillbot bug "On August 1, 2010, this entry was published"

N

Nattsurfaren

Power Member

Apr 12, 2010

593

82

• Sep 26, 2022

• #1

So I noticed that Quillbot loves to add nonsense into my text when I do a grammar check.

This is inserted in many places.
On August 1, 2010, this entry was published

So I checked if anyone more than me have this problem.


Have anyone more than I noticed it? Maybe you've automated this and don't have this under control.
Just saying.

 

Steptoe

Elite Member

Jr. Executive VIP

Jr. VIP

Aug 9, 2017

5,592

6,878

• Sep 26, 2022

• #2

Interesting! Can be used as a footprint to identify sites using automated content I suppose

*Edit: scrolling through some of those sites I've already got a few ideas - nice!

 

N

Nattsurfaren

Power Member

Apr 12, 2010

593

82

• Sep 26, 2022
• Thread Starter Thread Starter

• #3

Steptoe said:

Interesting! Can be used as a footprint to identify sites using automated content I suppose

*Edit: scrolling through some of those sites I've already got a few ideas - nice!

Click to expand...

Did some more research and this happens when you fail to filter out garbage characters...
This particular character will cause problems '�'

 

terrycody

Elite Member

Sep 29, 2012

6,423

3,324

• Sep 26, 2022

• #4

Smart lol, this is a hilarious footprint.

 

Steptoe

Elite Member

Jr. Executive VIP

Jr. VIP

Aug 9, 2017

5,592

6,878

• Sep 26, 2022

• #5

Nattsurfaren said:

This particular character will cause problems '�'

Click to expand...

Yep, from my early days of scraping SERPs and then putting the resultant mess through Markov chains, I've learned that you should just remove all non ascii characters from your content sources programmatically (as long as you're using English of course!)