To what extent TFIDF is legit for SEO purposes?

SirLouen

Jr. Executive VIP
Jr. VIP
Joined
Jan 17, 2015
Messages
3,680
Reaction score
2,669
Website
linktr.ee
Lately, I've been wondering if TFIDF formula is useful when optimising content.

Personally, I believe that it's partly useful but partly useless. I can't imagine G comparing sites for positioning... or taking this factor into account. But could be possible, especially considering that for Google, the most automation possible the easiest way to ponder niches that have never been considered before.

For those who are not familiarized with this formula
https://en.wikipedia.org/wiki/Tf–idf

TF-IDF is a numerical statistic that is intended to reflect how important a word is to a document in a collection or corpus

Somewhat controversial. Thoughts?
 

THUNDERELVI

Elite Member
Joined
Sep 12, 2009
Messages
2,547
Reaction score
2,297
I can't imagine G comparing sites for positioning... or taking this factor into account.

Google does a lot more things than you can possibly ever imagine, that is, unless you are a software engineer, know how search algorithm foundations work & are creative.
To give you a short answer: Yes, TF-IDF has been and still is used by Google for a very long time, along with LSI/LSA. Most people don't realize this, but these algorithms are really old and have been used for at least 1-2 decades now (BHW only started discussing them in the last few years). There are also newer algorithms that try to have a better accuracy than those.

So what does that tell you for 2018?
1) They are still used - not a lot of algorithms exist to extract meaning from words/text, so yes, they are still used.
2) They are not used alone - as it always happens with Google, they use these algorithms in conjunction with other ones (newer/more advanced) to produce a "relevance" score for text.
 

SirLouen

Jr. Executive VIP
Jr. VIP
Joined
Jan 17, 2015
Messages
3,680
Reaction score
2,669
Website
linktr.ee
Google does a lot more things than you can possibly ever imagine, that is, unless you are a software engineer, know how search algorithm foundations work & are creative.
To give you a short answer: Yes, TF-IDF has been and still is used by Google for a very long time, along with LSI/LSA. Most people don't realize this, but these algorithms are really old and have been used for at least 1-2 decades now (BHW only started discussing them in the last few years). There are also newer algorithms that try to have a better accuracy than those.

So what does that tell you for 2018?
1) They are still used - not a lot of algorithms exist to extract meaning from words/text, so yes, they are still used.
2) They are not used alone - as it always happens with Google, they use these algorithms in conjunction with other ones (newer/more advanced) to produce a "relevance" score for text.

Seems legit, this is one of the most complex quests out there. To know which algorithms could actually make sense.
 
Top