RE: Important reminder: PROJECT HOPE against PLAGIARIZM and post spinning

You are viewing a single comment's thread from:

Important reminder: PROJECT HOPE against PLAGIARIZM and post spinning

in hive-175254 •  5 years ago 

You´re right.
I would consider a max of 20% -25%

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

Depending on the nature of the article being spun/written. Scientific write up can be very much similar and often times when an article will have up to 75% of similarity. However, the word will not follow, the depth of the knowledge will be different.

https://smallseotools.com/plagiarism-checker/

https://plagiarismdetector.net

These are some of the useful plagiarize checker for your perusal. I rely on them alot previously when I was still in the #esteem team.

It would be very interesting to implement the use of this type of verifier in our anti-plagiarism unit. Thanks for sharing.

As you have said, determining plagiarism can often be difficult. Except for vulgar plagiarism of those who make a full copy / paste (including images) of the article.

The nature of the article also influences the percentage of use of quotes.

Maybe you know any content verifier based on AI? I would consider it to be of a high level of reliability.

Sorry for late reply Juan. We have a set of algorithm to detect plagiarism, however nothing can be perfect. AI itself is AI because it's artificial, it need human to feed in the data so it can process. But when an AI can do it's own determination without human input, Houston we have a problem.

Can you imagine when a script can automatically scan for articles on the internet,

Automatically download pictures with the similar tag category from the internet, crop the pictures, apply shades and tilt the picture a little so the source code arrangement of the picture will also change arrangement

Automatically arrange the wordings like grammar checker, to make sure the meaning remains but the word being used become different.

And finally, automatically make a post with all the rearranged text, newly photoshopped photo, run a set of tags and posted in the frontend.


All these can be done automatically. So, imagine if you have an AI plagiarize checker, the next thing is someone use the AI checker algorithm reverse it, and you have an AI plagiariser automatically post spinned article everywhere everyday🤣

All these can be done automatically. So, imagine if you have an AI plagiarize checker, the next thing is someone use the AI checker algorithm reverse it, and you have an AI plagiariser automatically post spinned article everywhere everyday🤣

I wouldn't be surprised to know that this happens daily before our eyes.

A few months ago I read an article where large US newspapers (NYT, WP, Forbes) use AI to write articles.

For example, look at this article:
http://www.milb.com/milb/stats/stats.jsp?gid=2016_06_29_broasx_stcasx_1&t=g_rcp&sid=milb

It was automatically written using AI by the Associated Press, which has used for a couple of years the technology developed by the Automated Insights company, which is based on statistical information from the United States Minor Baseball League, published in MiLB. com.

This copywriting AI manages to consistently sort the large volume of statistical data, automatically interpret and present it in the form of a text article.

That being said, in the end we still need human to make the final touch, a FEEL of it. Just like what we have started, there's no perfect plagiarize checker. However, if there's an automated checker, it will be easier for us to filter which are the "potential" violator and we human to make a final decision if it's worth a vote or a kick in the ass. I wish you good day.

Have a nice day.
It was a pleasure chatting with you.