[Tutorial] How search engines understad wich content is unique

naskootbg

Senior Member
Joined
Nov 8, 2010
Messages
847
Reaction score
280
I write this post mostly to help people that rewrite or spun content for their backlinks and have limited copyscape access. Because most of the content spinner tools create articles from paras from different articles and very regular the final result do not make sence, but if we can spin/rewrite single article wich will pass copyscape will be great correct!?

I think the best plagiarism checker tool is copyscape. I think it is the tool that works very similar to search engine bots and even I think it works in the same way. I thought a lot for this and I consider, that this is the only algoritm to find if the content is unique in deep.

Why?
After I tried a lot of plagiarism checker tools (have tons free), I noticed that most of them just checking sentences or parts from content on g00gle search. Than they do simple math:
numberUniqueSentences/totalNumberSentence*100
In most cases when such tools return 90%+ uniquiness, the content passed copyscape. But, sometimes not! This is the reason in my head come the questions:
How copyscape algoritm catching the dupe content?
When they do it so fast (copyscape and search engines) it must be some simple thing, correct?

So to answer these questions I tryed to make a programme, that will try to find the original source from spun content (like copyscape do). And consider to only way to do this is in 3 modules:
1. split the content and chek each part of content against g00gle search (,;:!?. are parts' separators)
2. if some part return results compire words popularity and words count between the checked content and the content of returned URL's content. If popular words and number of words are near the same, than most possibly the content is spun/rewritten and go to module 3.
3. calculate how many words exact matches and if the % of the mached words is higher than e.g. 20%, mark as duplicate.

I don't thing, there have other way to check fast.

So how to be sure search engine algorithms and copyscape will not understand if we spin/rewrite an article (well copyscape at least)?
Just be sure each part of the content returns no results from g00gle search. You can use some plagiarism checker from the download section on the forum.

FAQ
Q. Why it will pass copyscape?
A. Copyscape will have nothing to use to move on the next steps from the algorithms.

Q. Well, ofcourse if each part of the article returns no results from G00gle it is unique. Then why you wrote this?
A. Because most common practice is to rewrite few sentences if copyscape returns dupe or even to skip/remove the paragraph. Now if you understand my post, with only free tools and very little rewrite - only the part of the article wich returns results, you can produce hq copyscape passed articles.
 
in which "sizes" you spilt the content to check ?
 
can you please give some live example, I feel little confuse - you want to put the the above symbol between two paragraphs and check it on Google ?

PART1: can you please give some live example
PART2: I feel little confuse - you want to put the the above symbol between two paragraphs and check it on Google

"can you please give some live example" returns 1 (or2) results from g00gle

So if "can you please give some live example" is part from the 'article to check', than copyscape will move on second module (words compire)
 
Let say our article is:

Image_44.png

When copyscape or your favorite tool split on parts and check on g00gle search the result will be similar to:
Image_45.png


In this example the copyscape algorithm will not start compire words, because no results matching to compire.

Ofcourse the 'article to check' can be spit on different way, but my tests shows that splitting it on some of these symbols: [,:!?.] works.

What you can't understand? Please, be more concrete.
 
Emm, I believe it's more complicated that that. The same concept but might be taking a number of words together.

Cuz, if it's checking sentence by sentence, then even if there is duplicate words it will pass. The only way it would be marked as duplicate if the two sentences are 100% exact
 
Emm, I believe it's more complicated that that. The same concept but might be taking a number of words together.

Cuz, if it's checking sentence by sentence, then even if there is duplicate words it will pass. The only way it would be marked as duplicate if the two sentences are 100% exact
Maybe the way it find the possible dupe targets is more complex, but my tests shows, that when each part of an article passed g00gle search, then the copyscape returns unique even it is single spun article.
Other way to look up over the possible duplicate targets is some LSI engine like g00gle suggest, but it will require much more time to check each variant. There will have millions of variants for each article.

Keep in mind, that I'm not claim copyscape use g00gle to check - even think they have own spider, becasue creating spider is really simple - just require resourses.
 
I write this post mostly to help people that rewrite or spun content for their backlinks and have limited copyscape access. Because most of the content spinner tools create articles from paras from different articles and very regular the final result do not make sence, but if we can spin/rewrite single article wich will pass copyscape will be great correct!?

I think the best plagiarism checker tool is copyscape. I think it is the tool that works very similar to search engine bots and even I think it works in the same way. I thought a lot for this and I consider, that this is the only algoritm to find if the content is unique in deep.

Why?
After I tried a lot of plagiarism checker tools (have tons free), I noticed that most of them just checking sentences or parts from content on g00gle search. Than they do simple math:
numberUniqueSentences/totalNumberSentence*100
In most cases when such tools return 90%+ uniquiness, the content passed copyscape. But, sometimes not! This is the reason in my head come the questions:
How copyscape algoritm catching the dupe content?
When they do it so fast (copyscape and search engines) it must be some simple thing, correct?

So to answer these questions I tryed to make a programme, that will try to find the original source from spun content (like copyscape do). And consider to only way to do this is in 3 modules:
1. split the content and chek each part of content against g00gle search (,;:!?. are parts' separators)
2. if some part return results compire words popularity and words count between the checked content and the content of returned URL's content. If popular words and number of words are near the same, than most possibly the content is spun/rewritten and go to module 3.
3. calculate how many words exact matches and if the % of the mached words is higher than e.g. 20%, mark as duplicate.

I don't thing, there have other way to check fast.

So how to be sure search engine algorithms and copyscape will not understand if we spin/rewrite an article (well copyscape at least)?
Just be sure each part of the content returns no results from g00gle search. You can use some plagiarism checker from the download section on the forum.

FAQ
Q. Why it will pass copyscape?
A. Copyscape will have nothing to use to move on the next steps from the algorithms.

Q. Well, ofcourse if each part of the article returns no results from G00gle it is unique. Then why you wrote this?
A. Because most common practice is to rewrite few sentences if copyscape returns dupe or even to skip/remove the paragraph. Now if you understand my post, with only free tools and very little rewrite - only the part of the article wich returns results, you can produce hq copyscape passed articles.
If I were you I'd invest in a spell checker first lol
 
Well, I am using CopyScape to check articles and It's a familiar and great tool as well. ;) I feel you make it complicated by explaining brief about CopyScape.Anyways, Thanks for sharing the info.
 
Well, I am using CopyScape to check articles and It's a familiar and great tool as well. ;) I feel you make it complicated by explaining brief about CopyScape.Anyways, Thanks for sharing the info.

As I wrote, I think copyscape is the best.
Becasue copyscape is limited or paid.
 
Now this will help all you understand better (even my bad grammar :) :
PROOF
This is paragraph from ezinearticles:
1e5b0ffc1bc64204b9ab178c7acd88f7.png

When I chek agains google search I got this:
44fea4d848e44ec48128818348a2b45b.png

Which means only 2 parts from the paragraphs returns results from g00gle search. If you check this content with copyscape will see it is duplicate.

Now I rewrite the parts that returned results. They are actually in 1 sentence.
You can easy see what I changed. When check aginst g00gle search it returned no results for all parts of the paragraph:
2bbbff1a9166445196cf1eb4b7e528fc.png

When I checked this on free copyscape (you can check too), it is returned as unique.
;)
 
thanks for your example. did you use copyscape service to check the sentence like the image below here?i never use copyscape before.

44fea4d848e44ec48128818348a2b45b.png



i'm using the sample on grammarly. without editing. and it shows no error on plagiarsm. did it means this ezine article is not indexed on google?
 
Last edited:
thanks for your example. did you use copyscape service to check the sentence like the image below here?i never use copyscape before.

44fea4d848e44ec48128818348a2b45b.png



i'm using the sample on grammarly. without editing. and it shows no error on plagiarsm. did it means this ezine article is not indexed on google?

This is from self developed on php plagiarism checker, wich can't be shared. There have free plagiarism checker tools on the download section and I post a link early on this thread.

The ezine article is indexed - it's ID is 9457999 .
 
Last edited:
Back
Top